Split (1)

train · ~9.11M rows (showing the first 93.5k)

identifier stringlengths 11 32	pdf_url stringlengths 17 4.62k ⌀	lang stringclasses 120 values	error stringclasses 1 value	title stringlengths 2 500 ⌀	source_name stringlengths 1 435 ⌀	publication_year float64 1.9k 2.02k	license stringclasses 3 values	word_count int64 0 1.64M	text stringlengths 1 9.75M
https://openalex.org/W4296006871	https://findresearcher.sdu.dk/ws/files/215078532/1_s2.0_S0002939422003518_main.pdf	English	null	Peripapillary Hyperreflective Ovoid Mass–like Structures (PHOMS) in Children: The Copenhagen Child Cohort 2000 Eye Study	American journal of ophthalmology	2,023	cc-by	9,013	Download date: 24. Oct. 2024 University of Southern Denmark Citation for pulished version (APA): Behrens, C. M., Malmqvist, L., Jørgensen, M., Sibony, P. A., Munch, I. C., Skovgaard, A. M., Larsen, M., & Hamann, S. (2023). Peripapillary Hyperreflective Ovoid Mass–like Structures (PHOMS) in Children: The Copenhagen Child Cohort 2000 Eye Study. American Journal of Ophthalmology, 245, 212-221. https://doi.org/10.1016/j.ajo.2022.09.003 Go to publication entry in University of Southern Denmark's Research Portal Terms of use This work is brought to you by the University of Southern Denmark. Unless otherwise specified it has been shared according to the terms for self-archiving. If no other license is stated, these terms apply: Citation for pulished version (APA): Behrens, C. M., Malmqvist, L., Jørgensen, M., Sibony, P. A., Munch, I. C., Skovgaard, A. M., Larsen, M., & Hamann, S. (2023). Peripapillary Hyperreflective Ovoid Mass–like Structures (PHOMS) in Children: The Copenhagen Child Cohort 2000 Eye Study. American Journal of Ophthalmology, 245, 212-221. https://doi.org/10.1016/j.ajo.2022.09.003 Terms of use This work is brought to you by the University of Southern Denmark. Unless otherwise specified it has been shared according to the terms for self-archiving. If no other license is stated these terms apply: • You may download this work for personal use only. y p y • You may not further distribute the material or use it for any profit-making activity or commercial Y f l di t ib t th URL id tif i thi i y p y You may not further distribute the material or use it fo y p y • You may not further distribute the material or use it for any profit-makin at this document breaches copyright please contact us providing details and we will investigate your claim. enquiries to puresupport@bib.sdu.dk If you believe that this document breaches copyright please contact us providing details and we will investigate your claim. Please direct all enquiries to puresupport@bib.sdu.dk If you believe that this document breaches copyright please contact us providing details and we will investigate your claim. Please direct all enquiries to puresupport@bib.sdu.dk Download date: 24. Oct. 2024 CHRISTOPHER MAXIMILIAN BEHRENS, LASSE MALMQVIST, MORTEN JØRGENSEN, PATRICK A. SIBONY, INGER CHRISTINE MUNCH, ANNE METTE SKOVGAARD, MICHAEL LARSEN, AND STEFFEN HAMANN Given the high prevalence of PHOMS, they should not unreservedly be taken as evidence of optic neu- ropathy. (Am J Ophthalmol 2023;245: 212–221. © 2022 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )) • PURPOSE: To determine the prevalence of peripapillary hyperreflective ovoid mass–like structures (PHOMS) in a population-based child cohort and to study their associ- ation with other optic nerve head features and myopia. • DESIGN: Observational, population-based cohort study of 1407 children aged 11-12 years. • METHODS: Optical coherence tomography scans of op- tic nerve heads were graded for PHOMS, disc tilt, prelam- inar hyperreflective lines, and scleral canal diameter and investigated for associated prenatal and ocular parame- ters. Children with optic disc drusen or optic disc edema were excluded. • METHODS: Optical coherence tomography scans of op- tic nerve heads were graded for PHOMS, disc tilt, prelam- inar hyperreflective lines, and scleral canal diameter and investigated for associated prenatal and ocular parame- ters. Children with optic disc drusen or optic disc edema were excluded. A A peripapillary hyperreflective ovoid mass–like structure (PHOMS) is commonly seen on optical coherence tomography (OCT) of the optic nerve head (ONH). As such, it is a unique, dynamic in vivo rep- resentation of the corresponding localized peripapillary ax- onal distension (LPAD) seen histopathologically. 1 , 2 Early OCT studies, mostly based on low-resolution time domain OCT, mistakenly interpreted PHOMS as optic disc drusen (ODD). 3-6 Newer OCT equipment has revealed that a PHOMS is a separate structural entity with its own unique OCT appearance as described in a recent consensus re- port from the Optic Disc Drusen Studies (ODDS) Consor- tium. 7 , 8 • RESULTS: PHOMS were found in 8.9% of children. The location of PHOMS was predominantly in the super- onasal section of the optic disc. Myopia and optic nerve head tilt were more common in children with PHOMS than in children without PHOMS ( P < .001 and P < .001, respectively). Prelaminar hyperreflective lines were found in 17.9% of children with PHOMS compared to 7.3% of children without PHOMS ( P < .001). Prelami- nar hyperreflective lines with and without PHOMS were associated with a shorter axial length of the eye ( P < .001). There were no prenatal factors associated with PHOMS. Supplemental Material available at AJO.com . Accepted for publication September 1, 2022. Department of Ophthalmology, Rigshospitalet(C.M.B., L.M., M.J., M.L., S.H.), Glostrup, Denmark; Department of Ophthalmology, State University of New York at Stony Brook(P.A.S.), Stony Brook, New York, USA; Center for Clinical Research and Prevention, Bispebjerg and Fred- eriksberg Hospital(I.C.M.), Copenhagen, Denmark; National Institute of Public Health, Faculty of Health Sciences, University of Southern Den- mark(A.M.S.), Copenhagen, Denmark; Faculty of Health and Medical Sciences, University of Copenhagen(M.L., S.H.), Copenhagen, Denmark Inquiries to Steffen Hamann, Department of Ophthalmology, Rigshos- pitalet, University of Copenhagen, Glostrup, Denmark.; e-mail: steffen.hamann@regionh.dk © 2022 THE AUTHOR(S). PUBLISHED BY ELSEVIER INC. THIS IS AN OPEN ACCESS ARTICLE UNDER THE CC BY LICENSE ( HTTP://CREATIVECOMMONS.ORG/LICENSES/BY/4.0/ ) 0002-9394/$36.00 https://doi.org/10.1016/j.ajo.2022.09.003 CHRISTOPHER MAXIMILIAN BEHRENS, LASSE MALMQVIST, MORTEN JØRGENSEN, PATRICK A. SIBONY, INGER CHRISTINE MUNCH, ANNE METTE SKOVGAARD, MICHAEL LARSEN, AND STEFFEN HAMANN Prelaminar hyperreflective lines were associ- ated with higher birth weight and continued maternal smoking during pregnancy ( P = .01 and P = .02, respec- tively). The defining morphologic features of PHOMS are their (1) peripapillary location, abutting on the retina, (2) hy- perreflectivity on OCT, (3) ovoid shape on linear OCT scans through the center of the optic disc, and (4) mass- like, space-filling structural characteristic of displacing the adjacent retina away from the disc often expanding behind the retina, inducing a “ski-slope” sign. 2 , 7 , 8 Unlike ODD, where more than 1 can be present per eye, 3-dimensional analyses have shown that PHOMS are contiguous singular elements with the shape of a torus (doughnut) or toroid (a partial torus) that wholly or partially encircles the prelam- inar part of the optic nerve. 2 , 9 • CONCLUSIONS: PHOMS had a prevalence of 8.9% in healthy children without optic disc drusen or op- tic disc edema and was associated with increasing myopic refraction and the presence of a tilted op- tic nerve head and prelaminar hyperreflective lines. PHOMS are associated with several disorders of the ONH: (1) ODD-associated PHOMS, (2) anomalous disc– associated PHOMS as seen in association with myopic tilted discs, and (3) disc edema–associated PHOMS 2 —the third one being a consequence of increased retrolaminar pressure secondary to intracranial pressure elevation, and to axoplasmic stasis with distended prelaminar axons. The expanded axons herniate away from the disc and displace the peripapillary retina. 10 A similar process of axoplasmic stasis and nerve fiber herniation may play a role in the de- velopment of PHOMS in other types of optic disc edema, S A d f Supplemental Material available at AJO.com . Accepted for publication September 1, 2022. METHODS • STUDY POPULATION: The Copenhagen Child Cohort 2000 Eye Study is an addendum to a study of mental health in children that includes both registry studies and physical examinations. 12 , 13 Eye examinations were made in 2011- 2012, when the children, all born in the year 2000, were aged 11-12 years and again in 2016-2017. The study was approved by the Medical Ethics Committee of the Capital Region of Denmark and performed in accordance with the Helsinki Declaration. Informed consent was obtained from the children’s parents or legal guardians before the exam- inations. The present analyses are based on data from the 2011-2012 eye examination, where 1407 children partici- pated. After the initial assessment by the primary grader (C.M.B.), the presence of PHOMS was confirmed by 3 experienced graders (L.M., M.J., and S.H.). Afterward, the primary grader (C.M.B.) reassessed the 20 ° radial OCT scans to map the precise distribution and frequen- cies of PHOMS out of 12 possible peripapillary locations ( Figure 2 ). l Prelaminar hyperreflective lines were graded in all chil- dren by a single grader (C.M.B.) based on established cri- teria 16 requiring (1) a location in the prelaminar portion of the optic nerve, (2) horizontal extent longer than 50 µm, and (3) lack of interruption by more anterior structural el- ements ( Figure 1 ). • DATA ACQUISITION: Collection of data and perfor- mance of eye examinations have previously been de- scribed. 14 , 15 In brief, the children, who were accompanied by their parents or legal guardians, were questioned about medial history, ophthalmic history, and use of medication. Visual acuity was measured using Early Treatment Diabetic Retinopathy Study Charts (4m original series; Precision- Vision). An interferometric device was used to measure oc- ular axial length (IOL-MASTER, version 3.01.0294; Carl Zeiss Meditec). Objective refraction was measured by an au- torefractor (Retinomax K-plus 2; Right MG Co, Ltd). Assessment of external oblique border tissue (EOBT) length and ONH tilt was made by a single grader (C.M.B.) in all children. The presence of ONH tilt was identified by a white half-moon or C-shaped halo on the color fundus pho- tograph being present together with a corresponding cross- sectional EDI-OCT revealing a corresponding extension of the border tissue of Elschnig. Department of Ophthalmology, Rigshospitalet(C.M.B., L.M., M.J., M.L., S.H.), Glostrup, Denmark; Department of Ophthalmology, State University of New York at Stony Brook(P.A.S.), Stony Brook, New York, USA; Center for Clinical Research and Prevention, Bispebjerg and Fred- eriksberg Hospital(I.C.M.), Copenhagen, Denmark; National Institute of Public Health, Faculty of Health Sciences, University of Southern Den- mark(A.M.S.), Copenhagen, Denmark; Faculty of Health and Medical Sciences, University of Copenhagen(M.L., S.H.), Copenhagen, Denmark Inquiries to Steffen Hamann, Department of Ophthalmology, Rigshos- pitalet, University of Copenhagen, Glostrup, Denmark.; e-mail: steffen.hamann@regionh.dk 212 ONH OCT scans from both eyes of all children (2814 eyes of 1407 participants). The 20 ° 6-line radial OCT scan was used as the gold standard in detecting PHOMS. In case of this scan missing, a 7-line horizontal dense ONH OCT-scan was assessed if it proved coverage of the entire peripapillary circumference to detect all possible locations of PHOMS; otherwise, the eye was excluded. i in eyes with ODD, and in eyes with tilted discs. 2 PHOMS have been shown to be the most common cause of pseu- dopapilledema in children. 11 The prevalence of PHOMS in children has not previ- ously been assessed in a population-based cohort. The aim of this study was to examine the prevalence of PHOMS in children aged 11-12 years without ODD or optic disc edema and to investigate its association with prenatal fac- tors, refraction, and ONH characteristics such as tilted discs, hyperreflective lines, and variations in Bruch mem- brane opening (BMO) diameter. Furthermore, because of its 3-dimensional toroidal nature, we examined the peripapil- lary sectorial distribution of PHOMS. PHOMS were identified in accordance with the OCT characteristics defined by the ODDS Consortium and a recent state-of-the-art review on PHOMS 2 , 7 , 8 as ho- mogenous hyperreflective mass-like structures, appearing ovoid in cross section, always located in the peripapil- lary area immediately on top of the Bruch membrane, of- ten displacing the overlying retinal layers upward in a characteristic ski-slope sign. Mimics of PHOMS include ODDs that were excluded according to previous analy- ses of the data set, 15 , 16 and ONH blood vessels that, like PHOMS, are circular in cross section but differ in their superficial location and by casting a prominent shadow ( Figure 1 ). 2 METHODS Right eye with PHOMS (blue arrow) located superonasally. The BMO is measured as the distance between the 2 yellow arrows. C. PHOMS on a left eye located nasally (blue arrow). Furthermore, the nasal side is slightly lifted, and on the opposite (temporal) side, the border tissue of Elschnig is externally oblique and extends down in the optic nerve head (red arrows) as seen in myopic tilted nerve heads. D. PHOMS in a right eye located superiorly in association with several hyperreflective lines (green arrows). FIGURE 2. PHOMS locations by 20 ° radial B scan of the peri- papillary circumference. En face OCT of the optic nerve head of the right eye of a participant with PHOMS and associated 12- o’clock hour peripapillary radial B scan cross sections. On the en face OCT, the optic disc has a blurry margin on the superior, nasal, and nasoinferior section of the circumference. The OCT B scans reveal PHOMS in 8 of 12 possible locations (blue ar- rows) in the peripapillary space, in this case a 240 ° continuous torus. On the right en face OCT, the partial toroidal distribution of PHOMS is highlighted in transparent blue. (SE) refraction was calculated by adding the sum of the sphere to half the power of the cylinder. In children without PHOMS, only the right eye was used for the analysis. In children with unilateral PHOMS, only the affected eye was measured. In children with bilat- eral PHOMS, the eye with the subjectively assessed largest PHOMS was used. In children without PHOMS but with ONH tilt, the eye with the largest EOBT length and ONH tilt was used. Student t tests and the Wilcoxon rank test for skewed distributions were used to compare the children with PHOMS and subgroups of PHOMS to the group of children without PHOMS. χ2 and Fisher exact test were used for categorical data such as Tanner pubertal develop- ment scores and maternal tobacco consumption. FIGURE 2. PHOMS locations by 20 ° radial B scan of the peri- papillary circumference. En face OCT of the optic nerve head of the right eye of a participant with PHOMS and associated 12- o’clock hour peripapillary radial B scan cross sections. On the en face OCT, the optic disc has a blurry margin on the superior, nasal, and nasoinferior section of the circumference. METHODS Reis and associates 17 named the border tissue EOBT, and Han and associates 18 defined its associated ONH characteristics as an oblique optic canal and a tilted ONH on SD-OCT. To minimize uncertainty, only individuals with an EOBT length of at least 150 µm were included in the analysis of ONH tilt. The length of the EOBT and ONH tilt were mea- sured as described by Han and associates 18 ( Figure 3 ). The tilt angle was measured using GeoGebra software (version 5.0.16; International Geogebra Institute). Subjective refraction was measured with the optimal spherical correction, as found by adjustment of values from objective refractioning. Nondilated color fundus photogra- phy was made (Topcon Medical Systems). Spectral-domain O CT (Spectralis HRA + O CT; Heidelberg Engineering) scans were acquired in the form of one 20 ° 6-line enhanced depth imaging radial ONH scan, one 12 ° diameter peripap- illary scan, and one 30 ° transfoveal 7-line horizontal scan per eye. The scleral canal diameter or BMO diameter was mea- sured by a single grader (C.M.B.) as the mean of 6 ra- dial measurements. 15 , 16 Each radial measurement was done by connecting the 2 endpoints of the Bruch membrane on opposing sides. The scleral canal diameter was only measured in children with PHOMS without ONH tilt ( Figure 1 ). • IMAGE ANALYSIS: Using the manufacturer’s viewing software (Heidelberg Eye Explorer, version 1.6.1.0; Hei- delberg Engineering), a single grader (C.M.B) assessed all 213 VOL. 245 PHOMS IN A POPULATION-BASED CHILD COHORT FIGURE 1. Morphology of PHOMS and associated optic nerve head findings on EDI-OCT. A. Normal optic nerve head and no PHOMS. A blood vessel and its shadow (white arrows) is seen. B. Right eye with PHOMS (blue arrow) located superonasally. The BMO is measured as the distance between the 2 yellow arrows. C. PHOMS on a left eye located nasally (blue arrow). Furthermore, the nasal side is slightly lifted, and on the opposite (temporal) side, the border tissue of Elschnig is externally oblique and extends down in the optic nerve head (red arrows) as seen in myopic tilted nerve heads. D. PHOMS in a right eye located superiorly in association with several hyperreflective lines (green arrows). FIGURE 1. Morphology of PHOMS and associated optic nerve head findings on EDI-OCT. A. Normal optic nerve head and no PHOMS. A blood vessel and its shadow (white arrows) is seen. B. METHODS The OCT B scans reveal PHOMS in 8 of 12 possible locations (blue ar- rows) in the peripapillary space, in this case a 240 ° continuous torus. On the right en face OCT, the partial toroidal distribution of PHOMS is highlighted in transparent blue. Correlations between the measured ONH tilt and other ocular parameters such as axial length and spherical equiv- alent were evaluated by Pearson correlation. Univariable and multivariable logistic regression analysis were used to examine the risk factors associated with PHOMS and prelaminar hyperreflective lines. Odds ratios (ORs) for PHOMS in relation to myopia was additionally calculated using a 2 × 2 contingency table. The level of statistical sig- nificance was set at P < .05. • STATISTICS: Statistical analyses were performed using R studio software (version 4.0.5). Examination of histograms and performance of Kolmogorov-Smirnov and Shapiro- Wilk tests were used to evaluate normality of distribution and homogeneity of variance. For continuous variables, mean values and SDs were calculated for data of normal distribution, whereas medians and interquartile ranges were calculated for skewed distributions. Spherical equivalent RESULTS • STUDY POPULATION: The 2011-2012 visit of the CCC2000 Eye Study was attended by 1407 children aged 11-12 years. Of these, 93 were excluded after inspection of their OCT records, because they had ODD (n = 14) or were 214 AMERICAN JOURNAL OF OPHTHALMOLOGY MONTH 2023 FIGURE 3. Optic nerve head (ONH) tilt measurements. In accordance with the deep ONH morphology, the external oblique border tissue (EOBT) length is defined as the distance between the termination of the Bruch membrane at the neural canal opening anteriorly and the end of the EOBT (yellow line). The Bruch membrane opening (BMO) plane was defined as the length between the end points of the BMO (red line, between 2 red arrows). The ONH tilt angle was defined as the angle between the BMO and the optic canal plane, the latter defined as the line connecting the nasal BMO with the central end of the EOBT (blue dotted line). A and B. Two right eye examples of nasal PHOMS (blue arrows) and temporal EOBT of different lengths. C and D. Measurement of EOBT length. E and F. Measurement of the ONH tilt angle, illustrating larger EOBT length resulting in greater ONH tilt angle. FIGURE 3. Optic nerve head (ONH) tilt measurements. In accordance with the deep ONH morphology, the external oblique border tissue (EOBT) length is defined as the distance between the termination of the Bruch membrane at the neural canal opening anteriorly and the end of the EOBT (yellow line). The Bruch membrane opening (BMO) plane was defined as the length between the end points of the BMO (red line, between 2 red arrows). The ONH tilt angle was defined as the angle between the BMO and the optic canal plane, the latter defined as the line connecting the nasal BMO with the central end of the EOBT (blue dotted line). A and B. Two right eye examples of nasal PHOMS (blue arrows) and temporal EOBT of different lengths. C and D. Measurement of EOBT length. E and F. Measurement of the ONH tilt angle, illustrating larger EOBT length resulting in greater ONH tilt angle. body mass index, or pubertal development. The 117 partic- ipants with PHOMS had a median SE of –0.125 diopters (D) (interquartile range 0.75) and the 1197 children with- out PHOMS had a median SE of 0.00 D (interquartile range 0.625), P < .001 ( Table 1 ). RESULTS Inspection of the histogram dis- tribution of SE showed that the PHOMS group with ONH tilt had a large myopic tail. suspicious of having ODD (n = 5), OCT scans were missing for both eyes (n = 24), or OCT scan quality was insufficient (n = 50). This left 1314 children with gradable OCT data from at least 1 eye for analysis. Optic disc edema did not occur in the study population. • PREVALENCE OF PHOMS: Data from both eyes were ex- amined in all children ( Table 1 ). Of the 1314 children that were eligible for analysis, 117 (8.9%) had PHOMS in at least 1 eye. Of these 117 children, 73 (63%) had bilateral PHOMS, 24 (20%) had right-eye-only PHOMS, and 17 (14%) had left-eye-only PHOMS. In 3 additional partici- pants (2.6%) with PHOMS in one eye, symmetry could not be evaluated because scans from the contralateral eye were unavailable, which left 190 eyes with PHOMS. Of the 1197 children without PHOMS, 23 (1.9%) children were miss- ing OCT scans from the contralateral eye. • ANATOMICAL LOCATION OF PHOMS: In the 190 eyes with PHOMS, the 20 ° radial OCT scans were assessed to map the precise distribution and frequencies of PHOMS sections out of 12 possible peripapillary locations ( Figure 4 ). The most frequent cross sections with PHOMS were the nasosuperior (89%, 169 of 190 eyes), nasal (74%, 141 of 190 eyes), and superonasal (57%, 109 of 190 eyes). The least common cross sections with PHOMS were the tem- poral (1.1%, 2 of 190 eyes), the tempoinferior (1.1%, 2 of 190 eyes), and the temposuperior (2.6%, 5 of 190 eyes). Of the 167 eyes with PHOMS sections in more than 1 cross section, 153 had sections that were all adjacent and connected. • ANATOMICAL LOCATION OF PHOMS: In the 190 eyes with PHOMS, the 20 ° radial OCT scans were assessed to map the precise distribution and frequencies of PHOMS sections out of 12 possible peripapillary locations ( Figure 4 ). The most frequent cross sections with PHOMS were the nasosuperior (89%, 169 of 190 eyes), nasal (74%, 141 of 190 eyes), and superonasal (57%, 109 of 190 eyes). The least common cross sections with PHOMS were the tem- poral (1.1%, 2 of 190 eyes), the tempoinferior (1.1%, 2 of 190 eyes), and the temposuperior (2.6%, 5 of 190 eyes). RESULTS This was due to either large blood vessels obscur- ing the view or otherwise poor scan quality of selected ra- dial scans. The frequencies of PHOMS based on quadrants were as follows: 60.8% were located nasally, 31.7% supe- riorly, 1.5% temporally, and 6.0% inferiorly (n = 190). The distribution was identical in the right and left eyes. FIGURE 5. Toroidal coverage of PHOMS from 1 up to 12 cross sections (n = 190). 1 cross section corresponds to a 30 ° partial torus, whereas 12 cross sections correspond to a com- plete 360 ° torus. PHOMS ranged from a 30 ° partial torus to a complete 360 ° torus, and in both eyes PHOMS were on average found to cover 25% (3 of 12 cross sections) of the peripapillary circumference, corresponding to a 90 ° torus ( Figure 5 ). • ASSOCIATION OF PHOMS WITH OPTIC NERVE HEAD TILT AND HYPERREFLECTIVE LINES: Of the 117 participants with PHOMS, 60 (51.3%) had an ONH tilt and 7 of these had both prelaminar hyperreflective lines and ONH tilt. FIGURE 4. Total PHOMS section distribution and frequen- cies by 12-hour clock peripapillary circumference mapping (n = 190). I = inferior, IN = inferonasal, IT = inferotemporal, N = nasal, NI = nasoinferior, NS = nasosuperior, S = supe- i SN l ST t l T t l FIGURE 5. Toroidal coverage of PHOMS from 1 up to 12 cross sections (n = 190). 1 cross section corresponds to a 30 ° partial torus, whereas 12 cross sections correspond to a com- plete 360 ° torus. FIGURE 5. Toroidal coverage of PHOMS from 1 up to 12 cross sections (n = 190). 1 cross section corresponds to a 30 ° partial torus, whereas 12 cross sections correspond to a com- plete 360 ° torus. FIGURE 5. Toroidal coverage of PHOMS from 1 up to 12 cross sections (n = 190). 1 cross section corresponds to a 30 ° partial torus, whereas 12 cross sections correspond to a com- plete 360 ° torus. FIGURE 4. Total PHOMS section distribution and frequen- FIGURE 4. Total PHOMS section distribution and frequen- cies by 12-hour clock peripapillary circumference mapping (n = 190). RESULTS Of the 167 eyes with PHOMS sections in more than 1 cross section, 153 had sections that were all adjacent and connected. • DEMOGRAPHICS: There were no significant differences between children with PHOMS (n = 117) and without PHOMS (n = 1197) in gestational age, sex, birth weight, and maternal tobacco consumption ( Table 1 ). There was no difference at the time of examination in age, visual acuity, In 14 eyes of 10 participants with PHOMS, we were un- able to confirm or reject whether PHOMS sections were 215 VOL. 245 PHOMS IN A POPULATION-BASED CHILD COHORT TABLE 1. Characteristics of 11-12-Year-Old Children With or Without PHOMS PHOMS (n = 117; 8.9%) No PHOMS (n = 1197; 91.1%) P Value ∗ Boys/girls, n (%) 60 (51) / 57 (49) 560 (47) / 635 (53) .47 a Age, y 11.64 ± 0.38 11.67 ± 0.40 .41 b Spherical equivalent (sphere + 0.5 × cylinder, D) –0.125 (0.75) 0.00 (0.625) < .001 c Visual acuity (ETDRS letters) 88.38 ± 3.06 88.78 ± 3.10 .18 b Axial length (mm) 23.28 ± 0.91 23.18 ± 0.76 .22 b Tanner pubertal development stage, n (%) 1 25 (23.6) 247 (21.8) .95 d 2 54 (50.9) 597 (52.7) 3 24 (22.6) 258 (22.8) 4 3 (2.8) 31 (2.7) Birth weight (g) 3625 ± 609 3532 ± 595 .13 b Gestational age (d) 278 ± 12.6 278 ± 11.7 .90 b Maternal smoke during pregnancy, n (%) None 87 (81.3) 943 (81.1) .88 d Ceased 1 (0.9) 22 (1.9) Continued 19 (17.8) 197 (17.0) BMI 18.99 ± 3.42 18.35 ± 2.92 .053 b BMI = body mass index, D = diopters; ETDRS = Early Treatment of Diabetic Retinopathy Score, PHOMS = peripapillary hyperreﬂective ovoid mass–like structures. Data are presented as mean ± SD unless otherwise noted. In variables with skewed distribution, median and interquartile range (IQR) are presented. a Chi squared test. b Student t test. c Wilcoxon rank sum test. d Fisher exact t test. TABLE 1. Characteristics of 11-12-Year-Old Children With or Without PHOMS FIGURE 4. Total PHOMS section distribution and frequen- cies by 12-hour clock peripapillary circumference mapping (n = 190). I = inferior, IN = inferonasal, IT = inferotemporal, N = nasal, NI = nasoinferior, NS = nasosuperior, S = supe- rior, SN = superonasal, ST = superotemporal, T = temporal, TI = tempoinferior, TS = temposuperior. adjacent. RESULTS I = inferior, IN = inferonasal, IT = inferotemporal, N = nasal, NI = nasoinferior, NS = nasosuperior, S = supe- rior, SN = superonasal, ST = superotemporal, T = temporal, TI = tempoinferior, TS = temposuperior. PHOMS ranged from a 30 ° partial torus to a complete 360 ° torus, and in both eyes PHOMS were on average found to cover 25% (3 of 12 cross sections) of the peripapillary circumference, corresponding to a 90 ° torus ( Figure 5 ). adjacent. This was due to either large blood vessels obscur- ing the view or otherwise poor scan quality of selected ra- dial scans. The frequencies of PHOMS based on quadrants were as follows: 60.8% were located nasally, 31.7% supe- riorly, 1.5% temporally, and 6.0% inferiorly (n = 190). The distribution was identical in the right and left eyes. • ASSOCIATION OF PHOMS WITH OPTIC NERVE HEAD TILT AND HYPERREFLECTIVE LINES: Of the 117 participants with PHOMS, 60 (51.3%) had an ONH tilt and 7 of these had both prelaminar hyperreflective lines and ONH tilt. • ASSOCIATION OF PHOMS WITH OPTIC NERVE HEAD TILT AND HYPERREFLECTIVE LINES: Of the 117 participants with PHOMS, 60 (51.3%) had an ONH tilt and 7 of these had both prelaminar hyperreflective lines and ONH tilt. • ASSOCIATION OF PHOMS WITH OPTIC NERVE HEAD TILT AND HYPERREFLECTIVE LINES: Of the 117 participants with PHOMS, 60 (51.3%) had an ONH tilt and 7 of these had both prelaminar hyperreflective lines and ONH tilt. 216 AMERICAN JOURNAL OF OPHTHALMOLOGY 216 MONTH 2023 216 AMERICAN JOURNAL OF OPHTHALMOLOGY AMERICAN JOURNAL OF OPHTHALMOLOGY TABLE 2. RESULTS Risk Factors Associated With PHOMS (n = 117) Variables Univariable Model Multivariable Model a OR 95% CI P Value OR 95% CI P Value Presence of ONH tilt vs not 5.70 3.84-8.47 < .001 5.96 3.92-9.08 < .001 ONH tilt angle per 1 ° increase 1.37 1.14-1.65 < .001 1.38 1.13-1.69 .002 SE per 1-D increase 0.57 0.46-0.71 < .001 0.51 0.39-0.68 < .001 AL per 1-mm increase 1.19 0.93-1.52 .16 0.92 0.71-1.19 .52 VA per 1-letter increase 0.97 0.71-1.31 .25 Presence of prelaminar hyperreﬂective lines vs not 2.85 1.69-4.80 < .001 2.34 1.92 0.62-8.86 0.51-7.28 .006 .34 b Age per 1-y increase 0.82 0.51-1.33 .43 Sex female vs male 0.84 0.58-1.23 .37 GA per 1-wk increase 1.00 0.98-1.02 .89 BW per 100-g increase 1.03 0.99-1.06 .12 1.03 0.99-1.07 .14 Maternal smoking ceased during pregnancy vs no smoking 0.49 0.07-3.70 .49 Maternal smoking continued during pregnancy vs no smoking 1.04 0.62-1.75 .88 Birth complications vs none 1.14 0.61-2.14 .63 AL = axial length, BW = birth weight, D = diopter, GA = gestational age, ONH = optic nerve head, OR = odds ratio, PHOMS = peri- papillary hyperreﬂective ovoid mass–like structures, SE = spherical equivalent, VA = visual acuity. Boldface indicates signiﬁcance. a Adjusted for AL, age, gender, and birth weight. b Further adjusted for ONH tilt angle per 1 ° increase. TABLE 2. Risk Factors Associated With PHOMS (n = 117) TABLE 2. Risk Factors Associated With PHOMS (n = 117) Overall, 14 participants (12%) had hyperreflective lines only, and 43 (37%) had none of the above-mentioned find- ings. Of the 1197 children without PHOMS, significantly fewer (14.5%, n = 174) had an ONH tilt, P < .001. In ad- dition, 87 (7.3%) of the children without PHOMS had prelaminar hyperreflective lines, and 7 of these had both ONH tilt and prelaminar hyperreflective lines. sex, and birthweight (OR = 0.49, 95% CI = 0.36-0.66, P < .001, per 1-mm increase in length). The presence of PHOMS was associated with prelaminar hyperreflective in the same statistical model (OR = 2.25, 95% CI = 1.22- 4.13, P = .009) but not when further adjusting for ONH tilt angle (OR = 1.99, 95% CI = 0.51-7.69, P = .32) ( Table 3 ). RESULTS The following prenatal factors were associated with prelaminar hyperreflective lines in a multivariable logis- tic regression model: continued maternal smoking during pregnancy (OR = 1.88, 95% CI = 1.11-3.20, P = .02) and higher birth weight (OR = 1.05, 95% CI = 1.01-1.09, P = .01, per 100-g increase) ( Table 3 ). • RISK FACTORS ASSOCIATED WITH PHOMS: Presence of ONH tilt increased the risk of PHOMS (OR = 5.96, 95% CI = 3.92-9.08, P < .001). A 1 ° increase of ONH tilt increased the risk (OR = 1.38, 95% CI = 1.13-1.69, P = .002), and a 1-D increase of SE lowered the risk of PHOMS (OR = 0.51, 95% CI = 0.39-0.68, P < .001). Pres- ence of prelaminar hyperreflective lines increased the risk of PHOMS (OR = 2.34, 95% CI = 0.62-8.86, P = .006) in the first multivariable regression model, but the association was not significant in the second model further adjusted for increasing ONH tilt angle (OR = 1.92, 95% CI = 0.51- 7.28, P = .34) ( Table 2 ). • RISK FACTORS ASSOCIATED WITH PHOMS: Presence of ONH tilt increased the risk of PHOMS (OR = 5.96, 95% CI = 3.92-9.08, P < .001). A 1 ° increase of ONH tilt increased the risk (OR = 1.38, 95% CI = 1.13-1.69, P = .002), and a 1-D increase of SE lowered the risk of PHOMS (OR = 0.51, 95% CI = 0.39-0.68, P < .001). Pres- ence of prelaminar hyperreflective lines increased the risk of PHOMS (OR = 2.34, 95% CI = 0.62-8.86, P = .006) in the first multivariable regression model, but the association was not significant in the second model further adjusted for increasing ONH tilt angle (OR = 1.92, 95% CI = 0.51- 7.28, P = .34) ( Table 2 ). • BMO DIAMETER IN CONTROLS AND IN CHILDREN WITH PHOMS WITH OR WITHOUT PRELAMINAR HYPERREFLEC- TIVE LINES: In the group of participants with PHOMS and prelaminar hyperreflective lines without ONH tilt (n = 14), 10 children (71.4%) had a scleral canal diameter in the low- est quartile. In participants with PHOMS without supple- mental findings (n = 43), 13 children (30.2%) had a scleral canal diameter in the lowest quartile ( Table 4 ). RESULTS TABLE 3. Risk Factors Associated With Prelaminar Hyperreﬂective Lines (n = 108) TABLE 4. Distribution of Bruch Membrane Opening Diameter in Controls a and in Children With PHOMS With or Without Hyperreﬂective Lines Q1: 1182-1399 µm, n (%) Q2: 1400-1500 µm, n (%) Q3: 1501-1605 µm, n (%) Q4: 1602-2034 µm, n (%) Total Median (IQR), µm Controls 23 (17.7) 35 (26.9) 36 (27.7) 36 (27.7) 130 1508 (196) Children with PHOMS and hyperreﬂective lines but without optic nerve head tilt 10 (71.4) 2 (14.3) 1 (7.1) 1 (7.1) 14 1354 (93) Children with PHOMS but without hyperreﬂective lines and without optic nerve head tilt 13 (30.2) 15 (34.9) 7 (16.3) 8 (18.6) 43 1461 (159) IQR = interquartile range, PHOMS = peripapillary hyperreﬂective ovoid mass–like structures. a From Malmqvist and associates. 14 tribution of Bruch Membrane Opening Diameter in Controls a and in Children With PHOMS With or Without Hyperreﬂective Lines sections were identified in all of the 12 cross sections. In- terestingly, in the 168 of 190 eyes where PHOMS sections were seen in more than 1 cross section, the sections were all adjacent and connected. This supports the notion that it is not possible for an individual to have more than 1 PHOMS as it is one contiguous and united structure. population-based, birth cohort study using EDI-OCT scans of 1314 children. This is, to our knowledge, the first study to assess the prevalence of PHOMS in children in a population-based cohort. i PHOMS were first observed and described as “dome-like hyperreflective” structures seen in cross-sectional spectral domain OCT B scans of patients with ODD 3 and children with tilted disc syndrome. 19 This study supports the recent finding that a PHOMS is a single coherent mass whose true 3-dimensional shape is that of a partial or full, continuous torus along the peripapillary circumference. 2 , 9 A very consistent finding was that the PHOMS were preferentially located in the nasal and superior peripapil- lary region, where 60.8% and 31.7%, respectively, of all PHOMS were located. A recent review by Fraser and as- sociates 2 discussed the pathogenesis of PHOMS in differ- ent conditions, including tilted ONH. RESULTS • BMO DIAMETER IN CONTROLS AND IN CHILDREN WITH PHOMS WITH OR WITHOUT PRELAMINAR HYPERREFLEC- TIVE LINES: In the group of participants with PHOMS and prelaminar hyperreflective lines without ONH tilt (n = 14), 10 children (71.4%) had a scleral canal diameter in the low- est quartile. In participants with PHOMS without supple- mental findings (n = 43), 13 children (30.2%) had a scleral canal diameter in the lowest quartile ( Table 4 ). In a 2 × 2 contingency table, the odds of having PHOMS with a refraction of –1.0 to below –2.0 D, –2.0 to below – 3.0 D, and –3.0 D or below, were 1.5:1, 5.2:1, and 10.6:1, respectively. • RISK FACTORS ASSOCIATED WITH PRELAMINAR HYPER- REFLECTIVE LINES: The presence of prelaminar hyper- reflective lines was associated with a shorter axial length in a multivariable logistic regression model adjusting for age, We found PHOMS in 8.9% of 11-12-year-old children without ODD or optic disc edema in a cross-sectional, VOL. 245 217 PHOMS IN A POPULATION-BASED CHILD COHORT 217 PHOMS IN A POPULATION-BASED CHILD COHORT TABLE 3. Risk Factors Associated With Prelaminar Hyperreﬂective Lines (n = 108) Variables Univariable Model Multivariable Model a OR 95% CI P Value OR 95% CI P Value Presence of PHOMS vs not 2.85 1.69-4.80 < .001 2.25 1.99 1.22-4.13 0.51-7.69 .009 .32 b Presence of ONH tilt vs not 0.67 0.37-1.20 .18 0.69 0.36-1.34 .27 ONH tilt angle per 1 ° increase 1.29 0.94-1.75 .11 1.46 0.97-2.18 .07 SE per 1-D increase 1.04 0.82-1.34 .73 AL per 1-mm increase 0.54 0.42-0.71 < .001 0.49 0.36-0.66 < .001 VA per 1-letter increase 0.96 0.91-1.02 .22 Age per 1-y increase 0.93 0.56-1.52 .76 Sex female vs male 1.09 0.73-1.61 .68 GA per 1-wk increase 1.00 0.98-1.02 .76 BW per 100-g increase 1.04 1.00-1.08 .04 1.05 1.01-1.09 .01 Ceased smoking during pregnancy vs no smoking 2.93 1.61-5.33 .06 2.89 0.92-9.11 .07 Continued smoking during pregnancy vs no smoking 1.57 1.20-2.06 .08 1.88 1.11-3.20 .02 Birth complications vs none 1.18 0.61-2.27 .63 AL = axial length, BW = birth weight, D = diopter, GA = gestational age, ONH = optic nerve head, OR = odds ratio, PHOMS = peri- papillary hyperreﬂective ovoid mass–like structures, SE = spherical equivalent, VA = visual acuity. Boldface indicates signiﬁcance. a Adjusted for AL, age, gender, and birth weight. b Further adjusted for ONH tilt angle per 1 ° increase. RESULTS Comparing myopic children with tilted disc–associated PHOMS to nonmyopic controls, a signif- icant correlation between myopia and tilt angle was dis- covered. They also found that children with PHOMS have larger disc tilts than healthy controls; however, myopia seemed to be the sole reason , as the PHOMS group was highly myopic and the control group was not. It shows the same pattern: that the PHOMS are over- represented in the nasosuperior peripapillary circumference and no children have isolated temporal PHOMS. An ex- planation could be that the nasal and superior peripapillary spaces are the most vulnerable for nerve fiber herniation ir- respective of etiology. Interestingly, there are other cases in literature describing the nasosuperior peripapillary retina as particularly exposed, supporting our hypothesis. We did the same comparison in our study where we evaluated ONH tilt and SE as risk factors associated with PHOMS on a multivariable logistic regression model. We found that the presence of ONH tilt as well as the mag- nitude of ONH tilt was associated with elevated odds of PHOMS. ln regard to refraction, the model showed that a myopic shift was highly associated with PHOMS. These findings support the fact that refraction seems to be the uni- fying cause of ONH tilt causing PHOMS as previously dis- cussed. 2 For simplicity and comparability, we treated refrac- tion as a continuous variable in the multivariable regression model and not as a categorical variable, which is a limita- tion of our study. The adduction-induced phosphenes originally described by Purkinje located temporally of the blind spot have been proposed to be caused by mechanical strains of the nasal peripapillary axons. 22 Furthermore, a recent study by Sibony and associates 23 described 10 individuals with crowded or tilted discs who developed asymptomatic peri- papillary subretinal hemorrhages, all confined to the nasal or superonasal peripapillary area. The proposed vulnerabil- ity of this particular area could also explain why ODDs are mainly located in the nasal region. 24 Seven children with PHOMS had ONH tilt and prelam- inar hyperreflective lines whereas 14 children had hyper- reflective lines only. We measured the BMO diameter in these latter 14 children and found that 10 of these were in the lowest quartile based on a nested cohort of 130 ran- dom participants measured by Malmqvist and associates 15 (see Table 4 ). RESULTS It is believed that the protrusion of Bruch membrane and the nasal drag- ging of the lamina cribrosa relative to the BMO as well as the stretching of the temporal sclera cause impinge- The distribution of PHOMS in our study, however, var- ied from individual to individual; some children only had a small PHOMS seen in 1 of 12 cross sections, whereas the largest PHOMS was a complete 360 ° torus, where PHOMS AMERICAN JOURNAL OF OPHTHALMOLOGY 218 MONTH 2023 ONH tilt group and controls (0.00 D in both groups) is most likely caused by a large high myopic tail in the ONH tilt group. ment and stress on optic nerve axons, which lead to the occurrence of nasal PHOMS. 19-21 Our findings of nasosu- perior overrepresentation support this theory of PHOMS pathogenesis, but the nasal overrepresentation was not lim- ited to children with ONH tilt/anomalous disc-associated PHOMS. As myopia is first fully developed after puberty, 19-21 it is likely that several children with tilted ONHs were pre- myopic. As they age, their axial length, tilt, and myopia would increase, possibly making the difference in refrac- tion between the groups even greater. Indeed, children with PHOMS were largely overrepresenting myopia in the co- hort, and especially high myopia. The ORs of myopic eyes having PHOMS reveal that not only is myopia a risk factor for having PHOMS, but increasing myopia causes an expo- nentially increasing risk of having PHOMS, likely through the severity of axial elongation and disc tilt. 5 The participants with PHOMS and hyperreflective lines as well as the group with neither hyperreflective lines nor disc tilt also mainly had PHOMS in the nasal and supe- rior peripapillary circumference. Surprisingly, only 7 eyes of 4 participants had PHOMS in the temporal peripapillary space, totaling 9 cross sections of PHOMS ( Figure 4 C). In none of these cases, the PHOMS were exclusively located temporally, but they were also observed in the superior and nasal regions and were subjectively larger in volume. If we look at the ODD-associated PHOMS previously identified in the CCC2000 Eye Study cohort, 15 11 of 14 children had ODD-associated PHOMS that were distributed 35% nasally, 36% superiorly, 10% temporally, and 18% inferi- orly. A recent case-control study by Lyu and associates 25 sup- ports these findings. RESULTS A recent study of PHOMS in children 26 found that the BMO in children with PHOMS was signifi- cantly smaller than that of controls. Of a total of 45 eyes with PHOMS, 16 eyes had coexisting buried ODD, and there was no report of presence of prelaminar hyperreflec- tive lines. In contrast, there were no coexisting ODD in the control group of 39 eyes. As axoplasmic stasis of the ONH increases, per se, be- cause of enlargement of ODD or increased optic disc edema, the lateral bulging not only increases the volume of nasal and superior PHOMS but also spreads to the inferior and temporal peripapillary spaces en route to a complete 360 ° torus. More research is needed to fully understand the step- wise development of PHOMS. p The 117 children with PHOMS without ODD were sub- divided according to associated papillary findings. More than half (51.3%) had a tilted disc, which makes this the commonest ONH-associated finding in children with PHOMS aged 11-12 years. We might have slightly under- estimated the number as we only included disc tilts in which the EOBT length was at least 150 µm. Subdividing the par- ticipants with PHOMS also revealed that it was indeed the group with ONH tilt that statistically varied in refraction from the rest of the cohort. The significant difference in refraction combined with identical median values in the In the CCC2000 Eye Study cohort, a significant associa- tion between ODD and a small BMO was found, 15 and we reevaluated the OCT scans and found that 11 of 14 of these children had PHOMS. In the same cohort, it was also found that hyperreflective lines are precursors of ODD. 16 Interest- ingly, hyperreflective lines were not a significant risk factor for PHOMS when adjusting for the magnitude of disc tilt. We suggest that a small BMO and a large ONH tilt are 2 219 VOL. 245 VOL. 245 PHOMS IN A POPULATION-BASED CHILD COHORT independent risk factors for axonal stasis causing PHOMS and hyperreflective lines. genesis and nervous system development can cause damage to the optic nerve, causing prelaminar hyperreflective lines to occur. We found that a low axial length was an associated risk factor for the presence of prelaminar hyperreflective lines, independently of PHOMS. We did not find any similar studies assessing the prevalence of these lines in association to axial length. RESULTS 29-31 It seems likely that a continued toxic exposure of tobacco during fetal organo- PHOMS will likely be an increasingly common finding in clinical practice. tral domain optical coherence tomography. Ophthalmology . 2014;121(4):959–963 . Funding/Support: This work was supported by Horizon 2020, the European Union’s Framework Programme for Research and Innovation, under grant agreements no. 732613 (GALAHAD) and no. 780989 (MERLIN) and by the Bagenkop-Nielsen foundation, and was performed within the framework of ERN-EYE (authors M.L. and S.H.). The funding organizations had no role in the design or conduct of this research. Financial Disclosures: The authors indicate no financial support or conflicts of interest. All authors attest that they meet the current ICMJE criteria for authorship. Funding/Support: This work was supported by Horizon 2020, the European Union’s Framework Programme for Research and Innovation, under grant agreements no. 732613 (GALAHAD) and no. 780989 (MERLIN) and by the Bagenkop-Nielsen foundation, and was performed within the framework of ERN-EYE (authors M.L. and S.H.). The funding organizations had no role in the design or conduct of this research. Financial Disclosures: The authors indicate no financial support or conflicts of interest. All authors attest that they meet the current ICMJE criteria for authorship. Author Contributions: Conceptualization: C.M.B., S.H.; Writing – original draft: C.M.B.; Writing – editing: C.M.B., S.H.; Methodology: C.M.B., S.H.; Investigation: C.M.B., L.M., M.J., P.A.S., I.C.M., A.M.S., M.L., S.H.; Data collection: C.M.B.; Data analysis: C.M.B., L.M., M.J., P.A.S., I.C.M.; Data curation: A.M.S., M.L.; Validation: S.H.; Supervision & Project administration: S.H. p Author Contributions: Conceptualization: C.M.B., S.H.; Writing – original draft: C.M.B.; Writing – editing: C.M.B., S.H.; Methodology: C.M.B., S.H.; Investigation: C.M.B., L.M., M.J., P.A.S., I.C.M., A.M.S., M.L., S.H.; Data collection: C.M.B.; Data analysis: C.M.B., L.M., M.J., P.A.S., I.C.M.; Data curation: A.M.S., M.L.; Validation: S.H.; Supervision & Project administration: S.H. RESULTS This association, however, seems plausible given the fact that these lines in some cases were found to be precursors of ODD in the CCC2000 Eye Cohort follow- up study, and that these children with ODD had an em- metropic shift in refraction that is associated with a low ax- ial length. 16 i Regarding the 43 children with PHOMS without supple- mental findings, we were limited by only having the OCT scans of 2011-2012 to assess and subdivide these further. Reassessing these 43 cases of PHOMS, 1 participant had an unknown macular pathology on the contralateral eye, 1 had tortuous retinal blood vessels, and 13 (30.2%) had BMO diameters in the lowest quartile. In the remaining 28 participants (65.1%), however, it was not possible to find any abnormal finding on the OCT. Another interesting finding was the association between continued maternal smoking during pregnancy and pres- ence of prelaminar hyperreflective lines. Tobacco smok- ing is a known independent risk factor of many common eye diseases such as age-related macular degeneration, glau- coma, and cataracts, and on a biochemical level, smoking enhances free radical generation and oxidative stress in the ocular tissue including the optic nerve. 27 , 28 This raises questions as to whether PHOMS, in some children, may be a nonpathologic finding, which resolves as the eye evolves and the child goes into puberty. We hope to answer this question in an assessment of the Copenhagen Child Cohort Eye Study follow-up of 2016-2017 to inves- tigate how the PHOMS identified in this study will change over a 5-year period, and whether new PHOMS will arise. Our study suggests that PHOMS are common in chil- dren without ODD or optic disc edema. Tilted discs and high myopia are risk factors for having PHOMS. With the increasing incidence of myopia in children worldwide, 19-21 PHOMS will likely be an increasingly common finding in clinical practice. Our study suggests that PHOMS are common in chil- dren without ODD or optic disc edema. Tilted discs and high myopia are risk factors for having PHOMS. With the increasing incidence of myopia in children worldwide, 19-21 Furthermore, maternal smoking during pregnancy has been associated with large cilioretinal arteries, increased risk of severe retinopathy of prematurity, and a reduced reti- nal nerve fiber layer thickness. REFERENCES Invest Ophthalmol Vis Sci . 2016;57(11):4979–4987 . 13. Olsen EM, Rask CU, Elberling H, et al. Cohort profile: the Copenhagen Child Cohort Study (CCC2000). Int J Epidemiol . 2020;49(2):370–371 . 23. Sibony P, Fourman S, Honkanen R, El Baba F. Asymptomatic peripapillary subretinal hemorrhage: a study of 10 cases. J Neu- roophthalmol . 2008;28(2):114–119 . 14. Li XQ, Munkholm A, Larsen M, Munch IC. Choroidal thick- ness in relation to birth parameters in 11- to 12-year-old chil- dren: the Copenhagen Child Cohort 2000 Eye Study. Invest Ophthalmol Vis Sci . 2014;56(1):617–624 . 24. Lorentzen SE. Drusen of the optic disk. A clinical and genetic study. Acta Ophthalmol (Copenh) . 1966(Suppl 90):1–180 . 25. Lyu IJ, Park KA, Oh SY. Association between myopia and peripapillary hyperreflective ovoid mass-like structures in children. Sci Rep . 2020;10(1):2238 . l 15. Malmqvist L, Li XQ, Eckmann CL, et al. Optic disc drusen in children: the Copenhagen Child Cohort 2000 Eye Study. J Neuroophthalmol . 2018;38(2):140–146 . 26. Ahn YJ, Park YY, Shin SY. Peripapillary hyperreflective ovoid mass-like structures (PHOMS) in children. Eye (Lond) . 2022;36(3):533–539 . 16. Malmqvist L, Li XQ, Hansen MH, et al. Progression over 5 years of prelaminar hyperreflective lines to optic disc drusen in the Copenhagen Child Cohort 2000 Eye Study. J Neurooph- thalmol . 2020;40(3):315–321 . 27. Solberg Y, Rosner M, Belkin M. The association between cigarette smoking and ocular diseases. Surv Ophthalmol . 1998;42(6):535–547 . 17. Reis AS, Sharpe GP, Yang H, Nicolela MT, Burgoyne CF, Chauhan BC. Optic disc margin anatomy in patients with glaucoma and normal controls with spectral do- main optical coherence tomography. Ophthalmology . 2012;119(4):738–747 . 28. Sanz-Morello B, Ahmadi H, Vohra R, et al. Oxidative stress in optic neuropathies. Antioxidants (Basel) . 2021;10(10):1538 . 29. Jønsson LH, Larsen M, Olsen EM, Skovgaard AM, Munch IC. Incidence of cilioretinal arteries in 11- to 12-year-old children and association with maternal smoking during pregnancy: the Copenhagen Child Cohort 2000 Eye Study. Acta Ophthalmol . 2021;99(7):e1162–e1167 . 18. Han JC, Lee EJ, Kim SB, Kee C. The characteristics of deep optic nerve head morphology in myopic normal tension glau- coma. Invest Ophthalmol Vis Sci . 2017;58(5):2695–2704 . 19. Pichi F, Romano S, Villani E, et al. Spectral-domain optical coherence tomography findings in pediatric tilted disc syndrome. Graefes Arch Clin Exp Ophthalmol . 2014;252(10):1661–1667 . 30. Hudalla H, Bruckner T, Pöschl J, Strowitzki T, Kuon RJ. REFERENCES 6. Bassi ST, Mohana KP. Optical coherence tomogra- phy in papilledema and pseudopapilledema with and without optic nerve head drusen. Indian J Ophthalmol . 2014;62(12):1146–1151 . 1. Skougaard M, Heegaard S, Malmqvist L, Hamann S. Preva- lence and histopathological signatures of optic disc drusen based on microscopy of 1713 enucleated eyes. Acta Ophthal- mol . 2020;98(2):195–200 . 7. Malmqvist L, Bursztyn L, Costello F, et al. The Optic Disc Drusen Studies Consortium recommendations for diagnosis of optic disc drusen using optical coherence tomography. J Neu- roophthalmol . 2018;38(3):299–307 . 2. Fraser JA, Sibony PA, Petzold A, Thaung C, Hamann S. Peripapillary hyper-reflective ovoid mass-like structure (PHOMS): an optical coherence tomography marker of axoplasmic stasis in the optic nerve head. J Neuroophthalmol . 2021;41(4):431–441 . 8. Malmqvist L, Bursztyn L, Costello F, et al. Peripapillary hy- perreflective ovoid mass-like structures: is it optic disc drusen or not?: Response. J Neuroophthalmol . 2018;38(4):568–570 . l 3. Lee KM, Woo SJ, Hwang JM. Differentiation of op- tic nerve head drusen and optic disc edema with spec- tral-domain optical coherence tomography. Ophthalmology . 2011;118(5):971–977 . 9. Fraser JA, Hamann SA. 360-degree peripapillary hyper-reflec- tive ovoid mass-like structure (PHOMS). Can J Ophthalmol . 2021;56(2):146 . 4. Traber GL, Weber KP, Sabah M, Keane PA, Plant GT. En- hanced depth imaging optical coherence tomography of optic nerve head drusen: a comparison of cases with and without visual field loss. Ophthalmology . 2017;124(1):66–73 . 10. Paton L, Holmes G. The pathology of papillœdema: a histo- logical study of sixty eyes. Brain . 1911;33(4):389–432 . 11. Mezad-Koursh D, Klein A, Rosenblatt A, et al. Peripapillary hyperreflective ovoid mass-like structures—a novel entity as frequent cause of pseudopapilloedema in children. Eye (Lond) . 2021;35(4):1228–1234 . 5. Kulkarni KM, Pasol J, Rosa PR, Lam BL. Differentiating mild papilledema and buried optic nerve head drusen using spec- 5. Kulkarni KM, Pasol J, Rosa PR, Lam BL. Differentiating mild papilledema and buried optic nerve head drusen using spec- 220 MONTH 2023 AMERICAN JOURNAL OF OPHTHALMOLOGY 12. Skovgaard AM, Olsen EM, Houmann T, et al. The Copenhagen County child cohort: design of a longitudi- nal study of child mental health. Scand J Public Health . 2005;33(3):197–202 . Boramae Myopia Cohort Study Report 2. Ophthalmology . 2018;125(8):1224–1233 . Boramae Myopia Cohort Study Report 2. Ophthalmology . 2018;125(8):1224–1233 . 22. Sibony PA. Gaze evoked deformations of the peripapillary retina in papilledema and ischemic optic neuropathy. REFERENCES Ma- ternal smoking as an independent risk factor for the devel- opment of severe retinopathy of prematurity in very preterm infants. Eye (Lond) . 2021;35(3):799–804 . 31. Ashina H, Li XQ, Olsen EM, Skovgaard AM, Larsen M, Munch IC. Association of maternal smoking during preg- nancy and birth weight with retinal nerve fiber layer thickness in children aged 11 or 12 years: the Copenhagen Child Cohort 2000 Eye Study. JAMA Ophthalmol . 2017;135(4):331–337 . 20. Kim TW, Kim M, Weinreb RN, Woo SJ, Park KH, Hwang JM. Optic disc change with incipient myopia of childhood. Oph- thalmology . 2012;119(1):21–26 e1-3 . 21. Lee KM, Choung HK, Kim M, Oh S, Kim SH. Po- sitional change of optic nerve head vasculature during axial elongation as evidence of lamina cribrosa shifting: VOL. 245 221 VOL. 245 PHOMS IN A POPULATION-BASED CHILD COHORT
https://openalex.org/W4205238719	https://curationis.org.za/index.php/curationis/article/download/481/420	English	null	Orientation	Curationis	1,985	cc-by	2,772	Mrs E. Anderson, RN, RM, OTT, DNA, DNE, CHN, BA (Unisa). Deputy Chief Nursing Officer, Natal Provincial Admi nistration. Elizabeth Anderson Elizabeth Anderson To settle you in the job. To settle you in the job. To the writer orientation means the first thing to be done for a new employee. Holmes and Rabe’s re search found that a new job gave a score of 36 in a rating scale of stress ranging from 100 to 11. At the first opportunity we rushed over and, fortunately, the list was always made out using the same format. Your eyes and index finger went straight to the theatre spot — your name wasn’t there, escape, you lived another month. This is no exaggeration, as we dreaded going to theatre. I’m not sure why except that we were all frightened of it. Looking back they were the sis ters who took no nonsense, disci plined you, took time to teach you — and to orientate you to their ward. It did not take long to know what she wanted and what she did not want you to do. Admittedly, that sort of orientation is not as ac ceptable as it should be, but it was better than none. Introduction The Oxford Dictionary defines the verb orientate as bring into clearly understood relations, or determine how one stands. On asking others what orientation meant to them, there were several replies: Oriëntasie van ’n nuwe werknemer is noodsaaklik om angs te verlig en om horn/ haar van die inligting wat nodig is vir bevredigende werksverrigting te voorsien. Orientasieprogramme moet aangepas word by die spesifieke behoeftes van elke nuwe werknemer. Daar kan onderskei word tussen makro- en mikro-oriëntasie. Eersgenoemde behels inligting oor die hospitaal as geheel en behoort deur die streeksmatrone be- hartig te word. I think it means to know where you are — left from right, East from West. I think it means to know where you are — left from right, East from West. Mikro-oriëntasie vind in die saaleenheid plaas en is die suster-in-bevel se ver- antwoordelikheid. Die skrywer gee verskeie riglyne vir geslaagde oriëntasie. Finding your whereabouts in a new situation. Finding your whereabouts in a new situation. Something for new people. Something for new people. that I am, in all probability, a good deal older than those same Sisters were at the time, I feel a tinge of bathos. Now one tends to see that type of sister as the efficient ward sisters! change list went up at the end of the month and the cry went out: the change list is up! The new employee Orient is derived from the Latin word oriens, meaning East. It is in teresting to know that the East window in a church is placed so that the rising sun will shine on the altar. Either way, orientation means starting early, rising sun, new em ployee, (probably scrubbed and dressed in her best uniform with highly polished shoes and badges). Orient is derived from the Latin word oriens, meaning East. It is in teresting to know that the East window in a church is placed so that the rising sun will shine on the altar. Either way, orientation means starting early, rising sun, new em ployee, (probably scrubbed and dressed in her best uniform with highly polished shoes and badges). I don’t know about you, but every time I have started a new job there was feverish activity in the house the night before, best uni form got ready, badges on epaul- lettes polished, shoes polished, hair set, alarm clock set and more. From that point on, we checked the various wards where the known battle-axes reigned supreme. Now .. AND WE RE VERY PROUD OF OUR NEW OPERATING THEATRE!” CURATIONIS M A A R T1985 I don’t know about you, but every time I have started a new job there was feverish activity in the house the night before, best uni form got ready, badges on epaul- lettes polished, shoes polished, hair set, alarm clock set and more. Why does one do this? — Be cause we want to make a good im pression on our new boss; we want to present a nice clean, groomed appearance for we all know that first impressions are lasting ones. Do you remember when you were a student nurse? Certainly in the hospital where I trained, the .. AND WE RE VERY PROUD OF OUR NEW OPERATING THEATRE!” CURATIONIS M A A R T19 18 • Most employees like to be chal lenged. A competitive spirit will result in increased productivity and increase personal capabili ties. • Most employees like to be chal lenged. A competitive spirit will result in increased productivity and increase personal capabili ties. The regional matron should, prior to the tour, read the appli cant’s application for appointment- /transfer, in order to gain some background knowledge of the new employee. The orientation process If you don’t organise a good orientation programme, this may easily be tapped dry. The second aspect to look at is the areas of orientation. These will be, firstly, the hospital as a whole (macro-orientation) and secondly, the ward/department (gýcro-orien- tation). It has been shown in a hundred studies over the last 20 years that what workers want most is to become masters of their environ ment and to feel that they them selves are important — the twin in gredients for a high self-esteem. New employees thrust into an alien environment are rarely receptive to anything more than the simplest concepts. Emphasise W H Y the job is done rather than HOW . Establish a good emotional relationship first and teach the technical skills later. Confusion and enthusiasm are the two dominant emotions during one’s first day at work. The ward sister can diminish the former and foster the latter. We all know what chaos there is on change-over day, and how everybody hates the first of the month. This is the time when the highest number of mistakes occur, in fact, it is a loaded day — one way and the other. Orientation programmes should be practical and m eet the real needs. What does a new employee expect from orientation? — A review of duties — what is ex pected? — Introduction to fellow workers. — View of the ward/areas of hospi tal facilities. — Explain hours, pay arrange ments, overtime, and so on. There cannot be a fixed orienta tion programme, standardised and dished out for you to implement. You may be given guides, but they must be tailored to meet your needs. — How the job fits. — Cordial greetings. The new employee Is she married or single, does she have children or other de pendants, has she found a house or flat, how old is she, what experi ence and qualifications does she have? By taking time to gain know ledge of the new person she may be able to help her with problems that arise when one moves to a new town. Memories of a first day can bring a flood of emotions, fear, excite ment, embarrassment and humility. It depends on how the supervisor had tailored the orientation pro gramme in her ward to meet all the levels of the variety of new staff. An orientation programme should reduce anxiety and provide the in formation required for reasonable performance. Most im portant, it should develop an atmosphere for effective communication so that new employees feel that are free to ask questions. One nurse did not go to the toilet all day because she felt too shy and didn’t have the courage to ask where it was! Her first day, needless to say, was not a happy one. Micro-orientation — Discussion of reports. Micro orientation (small) takes place in the actual limited area of the ward/department. The sister-in- charge will take over from the regional matron. In order to protect yourself and your new employees from the nega tive effects of poor orientation, here are four reasonable assump tions to use as guides: Macro-orientation The term macro (long, large, the great world) is a good description for that part of orientation which the regional matron should carry out. Who better than this matron to introduce the sister to the ward sister to whom she will be assigned, and then to take her on a fact-find- ing introductory tour of the hospi tal. Do not delegate the job of wel coming a new employee (senior staff should greet the new em ployee). Hello, welcome aboard, or Show sister the ropes and find some job for her to do is out. This shows you are harried and she is a nui sance. The newcomer is looking for guidance and reassurance. • Most individuals take the job be cause they believe it is important. • Most individuals want to do a good job. • Most people like to work for a ward sister who sets high stan dards. 19 CURATIONIS The orientation process Orientation should be seen as having two aspects. The first is from a broader view and incorporates various person diversities. Is the new person a registered nurse moving from one ward to another, or is she coming from another hos pital, or is she returning to nursing, having not nursed, for many years, or is she perhaps, a brand new re gistered nurse having just qualified? Orientation should be seen as having two aspects. The first is from a broader view and incorporates various person diversities. Is the new person a registered nurse moving from one ward to another, or is she coming from another hos pital, or is she returning to nursing, having not nursed, for many years, or is she perhaps, a brand new re gistered nurse having just qualified? Macro-orientation will include visits to: the medical superinten dent, the M atron’s Secretary, the staff office, and facilities that may be required at the commencement of the job — such as uniforms, parking, identification ticket. The problem of induction or ori entation is one common to all wards. Its most crucial period is the first day — it can be good or bad. For you, the person in charge, it is a unique opportunity to make a good impression. It begins with one dis tinct advantage which is that a person starting a new job has a re serve of goodwill towards his new employer. If you don’t organise a good orientation programme, this may easily be tapped dry. The problem of induction or ori entation is one common to all wards. Its most crucial period is the first day — it can be good or bad. The hospital as a whole is shown to the new employee which may in clude: library, college, theatre, X- Ray Departem ent, tuck shop, dif ferent wards, OPD, dental depart ment, laundry, pharmacy, blood transfusion, nurses’ home, hairdres ser, clinical department, sick bay. (You make a list of what your hos pital has!) You will see immediately that each of these examples would re quire a different orientation pro gramme to make it effective. For you, the person in charge, it is a unique opportunity to make a good impression. It begins with one dis tinct advantage which is that a person starting a new job has a re serve of goodwill towards his new employer. V O L 8 N 0 1 V O L 8 N 0 1 Figure 1: Questions in the new employee/employer relationship Few people can remember the names of everyone they are intro duced to and it is important for an orientation programme that the staff wear their name badges. This helps in the mental orientation. EMPLOYEE ADMINISTRATOR What is really expected of me? A , / Will she perform 1 up to my standard? How do I gain acceptance around here? \ / Will my other staff support her? How do I get ahead in this ward? \ / Is she ambitious? What is the sister really like? s ' X What is this person really like? I know the policies and procedures, but what are the real rules of the game? \| Unknown ] Will she understand the informal laws, will she conform? How do I fit into the total picture? / \ Will she be a real asset to the ward? Just how much security do I have here? / \ Is security an important issue with her? What is this ward really like? / ' Elementary information such as the location of lavatories, change rooms, fire escape, tea room and in struction on how to use telephones, are essential to help the new recruit to orientate herself physically. An other help is to assign the new person to a fellow worker (be sure to choose a suitable one). Here it has been shown that it is not the oldest sister or a sister who has been in the department for a long time that is the best. Persons of the same age group of the employee or newer staff are found to be more flexible. In orientation programmes it has been found to be beneficial if the new person is given a small task on the first day. There should be a high probability that it can be handled successfully to get her off to a confi dent start. One cannot specify that the ori entation programme must be one day, one week, one month or even, as read in an American Journal, six months. It must tailored according to your work load, staff available, type of nursing the person has a preference for and so on. The new sister must be assessed and be put in charge when ready to cope alone. —group integration factors. —group integration factors. Conclusion —group integration factors. • Who should tell them? — The regional matron — the immediate boss (ward sister). Remember that a good orientation programme is more than getting to know you, it’s finding reasons to stay. Orientation to the ward is a must because each employee inevi tably learns about his employer — som etim es rightly, som etim es wrongly, sometimes slowly, some times at great expense to himself and his management. • When should they be told? — Pre-employment — day one of employment — during the first fortnight — by the end of week six. When she is first placed in charge arrange that there are good back-up staff. y p y — during the first fortnight — by the end of week six. This last point holds happy me mories for me, and I would like to share them. As a newly qualified sister, I found myself in the operat ing theatre in a non-White hospital. After two weeks, my turn came to be on call to cope with the emergen cies after 17h00. A n e ffe c tiv e o rie n ta tio n p r o gramme starts the employer- em ployee relationship off on a positive note and facilitates learning by re ducing the personal anxiety com monly caused by a new or changed w ork setting. R em em ber that V.I.P. stands for Very Important Person (new employee) in a Very Important Programme. A planned programme helps to make learning pleasant, quick and accurate. It highlights those activi ties and services which are likely to win sister’s/nurse’s goodwill, confi dence and co-operation. Orienta tion can, and should be, seen as a necessary extension of the selection process. Imagine my chagrin on being called out of the local cinema to scrub for a patient with a compound fractured femur. Never will I forget the orderly Ambrose. The trolly was set, with all the correct instru ments in the correct order for me, and Ambrose stood by with the cheatles and Pass that one next, get the needles ready, etc. The elements of a good orienta tion programme can be put into a nutshell by asking three ques tions:— V O L 8 N 0 1 This may be after one month or three months, or never. REFERENCES 1. Arbose J. (1976) How to greet a new employee. Inter national Management. August 1976. 2. Ellis: Nowlis (1980) Nursing: A Hum an Needs Approach U .S.A . Houghton Mifflin. 2nd Ed. • What should they be told? — Job factors — personal factors • What should they be told? — Job factors — personal factors 3. Kanhouse D .N .; W arihay P I. (1980) A new look at em ployee orientation. Training and D evelopm ent Journal. July 1980. C llRATlON lS 20 MARCH 1985
https://openalex.org/W2100852058	https://europepmc.org/articles/pmc3224354?pdf=render	English	null	Theoretical investigation of malaria prevalence in two Indian cities using the response surface method	Malaria journal	2,011	cc-by	9,929	© 2011 Roy et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 * Correspondence: ramrup@ccmb.res.in Centre for Cellular and Molecular Biology (CSIR), Uppal Road, Hyderabad, India Theoretical investigation of malaria prevalence in two Indian cities using the response surface method Sayantani Basu Roy, Ram Rup Sarkar* and Somdatta Sinha Sayantani Basu Roy, Ram Rup Sarkar* and Somdatta Sinha Abstract Background: Elucidation of the relationships between malaria incidence and climatic and non-climatic factors in a region is of utmost importance in understanding the causative factors of disease spread and design of control strategies. Very often malaria prevalence data is restricted to short time scales (months to few years). This demands application of rigorous statistical modelling techniques for analysis and prediction. The monthly malaria prevalence data for three to five years from two cities in southern India, situated in two different climatic zones, are studied to capture their dependence on climatic factors. Methods: The statistical technique of response surface method (RSM) is applied for the first time to study any epidemiological data. A new step-by-step model reduction technique is proposed to refine the initial model obtained from RSM. This provides a simpler structure and gives better fit. This combined approach is applied to two types of epidemiological data (Slide Positivity Rates values and Total Malaria cases), for two cities in India with varying strengths of disease prevalence and environmental conditions. Results: The study on these data sets reveals that RSM can be used successfully to elucidate the important environmental factors influencing the transmission of the disease by analysing short epidemiological time series. The proposed approach has high predictive ability over relatively long time horizons. Conclusions: This method promises to provide reliable forecast of malaria incidence across varying environmental conditions, which may help in designing useful control programmes for malaria. Background Data-based statistical modeling, based on the available disease prevalence data for different cate- gorizations, is an integrative approach that uses the past data to predict the future trend. The applicability of such models becomes important while assessing their numeri- cal outputs, estimating parameters, and predicting future values from past observations. Much research has gone in the study of statistical models for malaria to describe the relationship between disease incidence and climatic as well as non-climatic factors [10,14-17]. Several techni- ques, such as, Logistic regression modeling [18], Poisson regression modeling [19], and Binary logistic regression modeling with fractional polynomial transformations [8] have been used to study the disease transmission process under different environmental factors. A combined mathematical-statistical approach (the Liverpool Malaria model), that uses the dynamic transmission models of the SIRS-type framework and statistical methods for cor- relating environmental dependencies, was developed recently to successfully forecast the evolution of malaria epidemiology in western Africa [20,21]. specific models with a large number of linear and non- linear terms are typical features of these models. Epide- miologists developing models to understand the underlying causes, their relative importance, and relation- ship with the pattern of the disease have not been able to reach a consensus as to which of the several processes in existence would work the best. In this paper, the incidence of malaria in two cities in southern India - Chennai and Mangalore - in two different climatic zones, are studied using the Response Surface Method (RSM), which was first introduced to obtain opti- mum conditions in a chemical investigation [29] and then in various optimization problems such as, structural relia- bility and biochemical processes [30-32]. It is known that malaria incidence in any region is modulated in response to several environmental factors, and these two cities have considerably different levels of disease severity, population density, rainfalls and vegetation patterns. The RSM mod- els, used for the first time to study any epidemiological data, is applied to develop a general framework to forecast malaria incidences by considering the relative effects of environmental factors without a priori assumptions on independent or response variables (discrete or continu- ous). Two types of epidemiological data for malaria cases are considered - i) a three-year time series data of Slide Positivity Rates (SPR) values of malaria for Chennai, and ii) a five-year time series of Total Malaria cases (TMC) for Mangalore. Background biology. However, factors such as the complexities in the life cycle of the parasite, environmental interactions, evolu- tionary pressure of drugs and control measures contribut- ing to drug resistance of parasite, and migration of population between endemic and non-endemic areas, con- tinue to contribute to the huge burden of morbidity and mortality accompanying the disease. These also present new challenges to researchers and public health profes- sionals to combat the disease. Relating epidemiological and clinical data, collected by researchers and public health professionals using different methods and analysis tools, to the associated biotic and abiotic factors is an extremely daunting task - particularly in a large and, eco- logically as well as demographically diverse country like India. Malaria is one of the world’s major micro-parasitic infec- tions in humans killing nearly two million (mostly chil- dren) each year, in addition to more than 500 million people living in the epidemic regions. It has been esti- mated that almost half the world’s population is at a risk of the disease [1,2]. Malaria was declared endemic and a major public health problem in 109 countries in 2008 [2]. It is known to be spreading rapidly in different tropical countries of Central and South America, Asia and Africa, with the sub-Saharan Africa credited for 85-90% of deaths [2-6]. In India, malaria is highly endemic in most regions [5,7]. Despite the introduction of control programs in many parts of the world over the past few decades, the impact of malaria on human population continues to increase. Scientific research has improved our understand- ing of the host-parasite-vector interactions and their Epidemiological research on micro-parasitic infections such as malaria, is often based on two separate measures of parasite distribution among hosts (humans, in this case) - the incidence of the disease, and the infection pre- valence. The incidence or prevalence measures are * Correspondence: ramrup@ccmb.res.in Centre for Cellular and Molecular Biology (CSIR), Uppal Road, Hyderabad, India Page 2 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 primarily based on classifying the population under study with respect to a range of factors such as age, sex, social factors, environmental variability etc. The climatic and environmental factors affect production and survival of the vector (mosquito), and the speed of parasitic life cycle [8-12]. Such studies involve a host of modelling approaches based on mathematical and data-based statis- tical methods [13]. Background The RSM models delivered good predictive ability over long time horizons. In addition, a step-by-step model reduction technique is proposed to refine the initial model obtained from RSM, which not only provides a sim- pler structure but also gives better fit as supported by Akaike’s Information Criterion [33]. The study on two data sets with different strengths of malaria incidence, reveals that this method can be used successfully to ana- lyse the behavior of epidemiological time series, and the forecasts obtained from the simpler yet improved models show high predictive ability. Hence, this approach leads to detection of crucial environmental factors influencing the transmission of the disease while offering a coherent and integrated understanding of the disease process in any area. Given the fact that data collection is mostly restricted to short time scales (months to few years), in recent studies, researchers have used elaborate time series analysis mod- els to show seasonality pattern in malaria incidence, and the Monte-Carlo Markov Chain methods with Bayesian techniques of a-priori probability assignment, to estimate the risk factors [22-24]. However, there exist some draw- backs, which affect the suitability of these models being fitted into the incidence pattern of the disease. For instance, discontinuity is observed in the time series at high temporal resolution while studying extensive dataset for elucidating climatic role in the transmission of malaria in Africa and Europe [25,26]. Quantitative modeling was applied in Punjab, India but its operation at low temporal resolution renders its wider applicability questionable [27]. The induction of linearity between malaria deaths and temperature via fractional polynomial algorithm in this methodology, too, is not obvious for modeling malaria risk with climatic factors. A recent study [28] attempted to deal with these drawbacks by developing a simple non-lin- ear regression methodology in modeling and forecasting malaria incidence in Chennai city, India. This method introduces successively higher powers of the chosen inde- pendent variables and leads to a complex model that does not ensure a trade-off between the number and numerical order of terms and, the goodness of fit of the model. Data- Methods Data Malaria incidence data, along with different demographic and environmental factors, were collected from literature and different other sources for two cities in India - Chen- nai and Mangalore [28,34-38]. Both cities are situated in southern part of India with Chennai in the coast of the Bay of Bengal and Mangalore near the Arabian Sea (see Figure 1). Even though both the cities have tropical cli- mate, their rainfall pattern is quite different, thereby put- ting them in two different climate zones - tropical wet for Roy et al. Malaria Journal 2011, 10:301 Page 3 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 http://www.malariajournal.com/content/10/1/301 Figure 1 Geographical location [34 ]and climatic regions for Mangalore and Chennai cities. Map within the box is not to the scale. Figure 1 Geographical location [34 ]and climatic regions for Mangalore and Chennai cities. Map within the box i cal location [34 ]and climatic regions for Mangalore and Chennai cities. Map within the box is not to the scale. dependent variable under study, were obtained from the Corporation of Chennai, Tamil Nadu, India for two types of malaria, P. vivax and P. falciparum, for the period Jan- uary 2002 - December 2004. The temporal scale of SPR is the monthly values collected from the Corporation. The threshold defining the positive slide is calculated based on JSB Stain for thick and thin films, which is stan- dard method used by the laboratories under the National Malaria Eradication Programme in India [37]. A thin blood smear film with parasitaemia value of 2-3% or above is considered to be a positive slide. The number of blood smears collected lies in the range of 16306-63717 depending on the seasons. The number of cases that tested positive was between 1184 and 4275, of the col- lected smears. Since the number of people that are exposed to the disease is directly proportional to the inci- dence, the population of the city is considered as an inde- pendent variable for this study. Monthly values for population were obtained from a third order polynomial fitting approach described in [28]. The environmental variables, such as, average temperature, average humidity and rainfall values, were also used as other independent variables. Average monthly values for temperature and humidity were taken as the arithmetic mean of minimum and maximum values available. Further details are avail- able in [28]. Methods Data Mangalore and tropical wet-dry climate for Chennai (Figure 1). Malaria is endemic in both the cities for the last few decades, though Chennai has lower endemicity compared to Mangalore. They are victims of rapid industrialization, witnessing an unprecedented spurt in construction activ- ities in recent years, thus facing the problem of mosquito breeding in man-made clear water sources like wells, overhead tanks, sumps, cisterns as well as other defective and illegal drainage systems. Yet there are several differ- ences between them, specifically in their demographic and climatic factors. Mangalore has less population (398,745) as compared to Chennai (5.6 million, third lar- gest city) according to the 2001 census of India. In addi- tion, the urban area in Mangalore has 32 recognised slums, and nearly 22,000 migrant labours live in slums within the city limits, whereas, Chennai has the fourth largest population of slum dwellers among major cities in India, with about 820,000 people (18.6% of its popula- tion) living in slum conditions [35]. Two types of data for malaria incidence, Slide Positivity Rate (SPR) and Total Malaria Cases (TMC), along with different demographic and environmental factors were collected from Chennai and Mangalore. The data for two types of malaria (due to Plasmodium vivax and Plasmo- dium falciparum), show similar temporal patterns of dis- ease incidence, and therefore are combined together for further analysis. The SPR values (bars) and the environmental variables, including population (lines), for Chennai are presented in Figure 2(a and b). Chennai gets most of its seasonal rain- fall from the North-East monsoon winds, from mid-Octo- ber to mid-December, followed by light rainfalls during Chennai city Slide Positivity Rate (SPR = number of cases found posi- tive/number of blood smears examined) values, the Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Roy et al. Malaria Journal 2011, 10:301 Page 4 of 12 y http://www.malariajournal.com/content/10/1/301 Figure 2 Malaria incidence and climate data. (a) SPR (%) values in Chennai, during January 2002 - December 2004, and (c) TMC values in Mangalore, during January 2003 - December 2007. The climate variables, Rainfall (cm), Temperature (°C), Humidity (%), and Population (in millions) are appropriately scaled to fit in the same plot for: (b) Chennai, and (d) Mangalore. Figure 2 Malaria incidence and climate data. (a) SPR (%) values in Chennai, during January 2002 - December 2004, and (c) TMC values in Mangalore, during January 2003 - December 2007. The climate variables, Rainfall (cm), Temperature (°C), Humidity (%), and Population (in millions) are appropriately scaled to fit in the same plot for: (b) Chennai, and (d) Mangalore. March to June of the South-West monsoon season show- ing two clear peaks for rainfall, with late May to early June being the hottest part of the year. This results in two rain- fall peaks each year as seen in Figure 2(b). The variation in the levels of the peaks can be attributed to climatic sto- chasticity. Average humidity variations follow the rainfall pattern and reach a temporary peak just before the start of the first monsoon in May - July, though its variance is low. And, the average temperature oscillates within a fixed range, and increases when the rainfall reaches the high peaks. Population remained relatively invariant during the period of study. Some influence of climatic variations on the malaria incidence pattern can be seen in Figure 2(a), where the SPR values increase with higher rainfall and subsequently decrease during the seasons with sparse rainfall. cases) are the measures of malaria incidence, which are different from the numerical nature of SPR values. The total malaria cases include both symptomatic and asymp- tomatic cases. Also, average monthly values of tempera- ture, humidity, and rainfall for the same interval were computed from a weather database that gives the daily minimum and maximum values for these variables [38]. Mangalore is under the direct influence of the Arabian Sea branch of the South-West monsoon and has a very high annual precipitation [36]. Chennai city It receives about 90% of its total annual rainfall between May to September, with occasional rains in October, while remains extremely dry from December to March. The TMC values and the climate variables are pre- sented respectively as bars and line graphs in Figure 2(c and d). Here, one can observe a more regular pattern of rainfall with a yearly peak (June-August), preceded by a dry season, and followed by gradual decrease. Also, the rainfall pattern is mimicked by humidity, albeit with a smaller variance. Owing to a tropical climate, the tem- perature shows limited variation. The TMC values clearly Autocorrelation function Autocorrelation of any time series data describes the correlation and relation in general between values of the series at different points in time [40]. Intuitively for infectious diseases, the autocorrelation at lag one is expected to be important among other lags since values at the previous time point tend to correlate closely with those of the current time. The first step in RSM is to find the First Order model with appropriate estimation of model parameters. If this preliminary analysis indicates non-linearity in the relation between independent variables and the response, along with the linear terms, a quadratic polynomial is used to include the second order terms of the independent vari- ables individually and all possible interaction terms of sec- ond order so as to propose the Second Order model (see Additional File 1, Section C for details). The method of least squares is then used to obtain an initial estimate of the parameters for the First or Second Order model depending on the cases. The response surface analysis is then performed, using the fitted surface to obtain the values of the parameters that optimize the response value [42]. In this study, this is attained in a complementary manner, by obtaining better, successive forecasts as newer (reliable forecasts) climate variable values are used in the model, thus, optimizing our future estimates of incidences. For obtaining the optimum parameter values following RSM, 80% of the data have been used (28 time points for SPR, 48 time points for TMC), and the remaining 20% of the data are used for validating the model predictions. The hypothesis of using a part of the data for model-building and retaining the rest for validation is common practice in Presence of autocorrelation was studied for the inde- pendent variables (SPR and TMC) (Additional File 1, Section A, Figure S1). The value at a particular lag was considered significant if it extends beyond the 95% con- fidence limits. In addition, Ljung-Box Q statistic (LB) is used to obtain the significance for autocorrelation values [40]. For autocorrelation value, rk at lag k, for sample size n and testing for cumulative autocorrelation for h lags, LB is given by Q = n(n + 2) h k=1 (ρ2 k /(n −k)). (1) (1) Variables with significant cumulative autocorrelations at a particular lag (mostly one) are also considered as independent variables in the model. Autocorrelation function For SPR and TMC values, autocorrelation values at lag-one were found to be high with significant LB values (Additional File 1, Section A, Table S1). Mangalore city Monthly malaria cases from January 2003 through December 2007 (60 time points), were obtained from Mangalore city, Karnataka [37]. These Total Malaria Case (TMC) values (sum of P. vivax and P. falciparum Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Page 5 of 12 Page 5 of 12 follow the rainfall pattern, though there is a consistent lag between the yearly maxima for TMC values and that of rainfall. tα 2 , n−2 = r (n −2) (1 −r2) (2) (2) It may be noted here that the data for climate vari- ables were obtained from daily weather sources, and they were averaged to monthly scale to ensure compat- ibility with the monthly parasitological data (SPR and TMC). Being in a tropical climatic region, the climatic variables for both the cities show small daily variations and hence the choice of arithmetic mean for calculating the monthly averages is appropriate and easy to apply. Also, none of the data are measured in logarithmic scale, or are highly skewed, or show high variations, hence geometric mean is not considered here [39]. Correlations at different lags (Additional File 1, Section B, Tables S2-S4) were considered to investigate the varia- tion or consistency in the different segments of the data [41]. Further, when correlation was found to be low or non-significant, perhaps indicating non-linearity, Resi- dual Plots were employed for further confirmation. Such plots show non-random distribution of points around the mean axis if the relationship between the variables under consideration is nonlinear. High correlations were observed between SPR and SPR-at-lag-one, and TMC and TMC-at-lag-one. For other combinations of inde- pendent and dependent variables, where correlation was found to be low, non-linearity was confirmed by residual plots (Additional File 1, Section B, Figures S2 and S3). Statistical methods Before the model development is discussed, statistical techniques for exploring the relationships between dependent variables (SPR and TMC values) and inde- pendent variables (climatic factors, previous incidence of the disease etc.) are presented. The details are given in Additional File 1 (Sections A-E). A t l ti f ti Model development using Response Surface Method Response Surface Method (RSM) is a classical optimiza- tion technique that integrates mathematical and statisti- cal approaches to the problem of finding the values of several independent variables (factors) affecting the opti- mum response in the variable of interest - the dependent (response) variable [42]. This approach has mostly been employed when experiments are conducted by varying the levels of the different factors and observing the effect on the response, training the values in a manner so as to reach the optimum point on the response surface. AIC = (−2)max(log(likelihood) + 2k AIC has been used to substantiate the claim that con- sidering R2 for reducing the number of terms in the model does provide us with a model that attains a trade-off between the variation explained and the num- ber of parameters estimated. Prediction Intervals in Forecasting Method Model reduction process A systematic model reduction process is developed to derive a simpler model with fewer but important terms, while ensuring efficiency (coefficient of determination) similar to that obtained in the initial model fitting through RSM. Figure S4 in Additional File 1 (Section C) shows the flow chart of the systematic process followed for model development using RSM and the model reduction technique. Model validation and forecast To assess the goodness of fit for the initial models using RSM and the reduced models using model reduction technique, four diagnostics tests are carried out: (a) Con- fidence Intervals for the data points, which are used for fitting the model, (b) Akaike’s Information Criterion to determine the better fit model, (c) Prediction Intervals in Forecasting Method for the data points estimated from the reduced model but not used for fitting the model, and (d) Collinearity Analysis for testing homoscedasticity of the residuals. The foremost task in this process is to enlist the resul- tant orders of magnitude of each term. In a model, each term consists of two components: the variable - linear or quadratic and the corresponding estimated coefficient or parameter value. The resultant order of a term is defined as the sum of the orders of the variable and the parameter in scientific notation. This exercise asserts its importance in cases when there is a great disparity between the orders of the variable and the parameter of a single term. Even though the parameter has very low order, a high order of the variable component may render the resultant order to become higher than that of the response, which may finally lead to its exclusion from consideration for the next step and vice-versa. Confidence Intervals Using Gauss Markov Theorem [41], it is tested whether each estimated value of the data points used for fitting the model (both for the optimized and reduced models) lies within the 95% confidence intervals or not (Addi- tional File 1, Section D). Correlation and residual plots Linear dependencies were quantified using correlation, and t-statistic was used for testing of significance. The two-sided t-statistic for a correlation, r at a level of sig- nificance is given by Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Page 6 of 12 statistical methods and the specific ratio may vary with the size of data at hand, as well as, the nature of data. removing the terms in the model since it is a statistic tra- ditionally used for providing the measure of the propor- tion of the variation explained by the model [43,44]. The final estimates of the parameters in a model give optimum goodness of fit realized using RSM. Such a model, if Second Order, will have a large number of terms, all of which may not have equal contribution to the varia- tion in response. The presence of more terms also renders the model complicated, and decreases its efficiency statisti- cally as more parameters are estimated from limited data. Hence, there is a requirement for a method to reduce the number of terms in a model without reducing the good- ness of fit. The outcome from this method remains unchanged to changes in the unit of the variables. Such a change in a variable is proportionately adjusted in the parameter component of the terms which contain that variable, pre- serving their resultant orders of magnitude. This may imply re-arrangement of the resultant orders of magni- tude of the terms in the model, in turn, leading to a dif- ferent reduced model. Also, the criterion (that is, the order of decimal place differences in the coefficient of determination considered significant) described above to arrive at the decision of retaining or removal of terms in the model will vary according to the nature of response being studied and variation in the allowed amount of loss in accuracy owing to model reduction. Akaike’s Information Criterion (AIC) Following the above process, the coefficient of determi- nation (R2) is examined at each step of the model by removing each term, beginning with the term with the smallest resultant order (Additional File 1, Section C, Tables S5 and S6). A stepwise process is followed, consid- ering removal of terms with successively higher resultant orders, and observing the change in the coefficient of determination (R2) at each step, to decide about the reten- tion or removal of the term under consideration. This pro- cess is continued till the terms with resultant orders equal to that of the response or higher are reached. Here, the underlying assumption is that the contribution of a term in a particular model equation is proportional to its resul- tant order. This is similar to backward regression, since the removal of terms from the model is performed once all the independent variables as well as corresponding pos- sible terms up to second order have been included in the model. R2 is employed for the decision of retaining or This statistic measures the appropriateness of forecasts of the estimated statistical models and selects the better model from the given models [33,40]. The model with the lower (or lowest) AIC is the better (or best) model. For a model based on n observations, having k para- meters and the residual errors denoted as Îi,i = 1,2,..,n, the value of the criterion is given by formula (3). AIC = (−2)max(log(likelihood) + 2k (3) (3) Tools and software Several tools and software are used for different statistical analysis and model fitting to the data. MS Excel was used for data handling. The model computation and verifica- tion of results was performed using MATLAB [45] and the R-package [46]. SPR = 115.25 −3.53 (SPR−1) −4.78 × 10−03T −5.51 × 10−05P + 6.75 × 10−03R + 3.64 × 10−1T−1 + 10.07(SPR−1)2 −3.61 × 10−04T2 + 6.52 × 10−12P2 −3.48 × 10−07R2 −1.63 × 10−03(T−1)2 −8.35 × 10−02SPR−1T −3.48 × 10−07SPR−1P + 1.09 × 10−03SPR−1R + 8.83 × 10−02SPR−1T−1 + 1.41 × 10−08T P + 2.34 × 10−05T R + 4.18 × 10−04T (T−1) −1.57 × 10−09P R −6.65 × 10−08P (T−1) −2.02 × 10−05R (T−1) . (4) SPR = 115.25 −3.53 (SPR−1) −4.78 × 10−03T −5.51 × 10−05P + 6.75 × 10−03R + 3.64 × 10−1T−1 + 10.07(SPR−1)2 −3.61 × 10−04T2 + 6.52 × 10−12P2 −3.48 × 10−07R2 −1.63 × 10−03(T−1)2 −8.35 × 10−02SPR−1T −3.48 × 10−07SPR−1P + 1.09 × 10−03SPR−1R + 8.83 × 10−02SPR−1T−1 + 1.41 × 10−08T P + 2.34 × 10−05T R + 4.18 × 10−04T (T−1) −1.57 × 10−09P R −6.65 × 10−08P (T−1) −2.02 × 10−05R (T−1) . (4) Results The RSM models and their modified forms are discussed below for two types of malaria incidence data - SPR and TMC values - in two cities of India - Chennai and Man- galore. The important independent variables (climate variables, population size, and previous incidence of the disease) are assessed using the step-wise model reduction process (described in Methods section), which offer bet- ter models as well as reliable predictions for these two cities. (4) Figure 3(a) shows the actual time series data and the simulated SPR values from the initial model Eqn. (4). The lag variables ensure that every successive simulation (or estimation) provides an input to the model to obtain the next SPR value, capturing the autoregressive nature of the time series. The 95% confidence interval bars are shown in the plot of fit, which serve as error limits and validate the range of estimations. The estimations are considered reasonable when within the 95% confidence Prediction Intervals in Forecasting Method Once the reduced model is obtained, using 80% of the available data, the next step to validate the model is to Page 7 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 sporogyny (15 to 23 days at 20°C of Plasmodium develop- ment inside the mosquito) [47]. This justifies the consid- eration of ‘SPR values at lag-one’ as an independent variable in the model. The other independent variables, on which the response (SPR values) depends, are envir- onmental factors, such as average temperature, average temperature-at-lag-one and rainfall. On Examination of the pair-wise linear correlations and corresponding t-sta- tistics (Additional file 1, Section B, Tables S2 and S3) between these independent variables and SPR did not give any significant relationship, indicating the presence of non-linearity. Since rainfall is strongly related to humidity, the latter was not included in the model. Thus the independent variables selected for model formulation were: SPR-at-lag-one (SPR-1), average temperature (T), population (P), rainfall (R) and average temperature-at- lag-one (T-1). The non-linear relationship shared between the independent variables (except SPR-at-lag-one), and the SPR values, were confirmed using residual plots (Additional file 1, Section B, Figure S2). The correlation and residual plots of these indicate non-linear relation- ships, and the Second Order RSM model gave better initial fit than the First Order. test the forecasting ability. This is performed by employ- ing the model to predict values for the remaining 20% time points. Since the model incorporates auto-regres- sive terms (SPR-at-lag-one for SPR values, TMC-at-lag- one for TMC values), this prediction ensures that each estimated response serves as an input for the subsequent estimate. This means that the predictions use the model prediction as the input for successive estimates, in addi- tion to reliable climate forecasts and/or population pro- jections. Further, this feature leads to training the data on hitherto unforeseen terrain, making it more condu- cive to unknown data. Using ancillary statistics [41], the 95% prediction intervals are calculated for each pre- dicted value from the reduced models for the data points which are not used for fitting the model (Addi- tional File 1, Section D). Collinearity Analysis For a least square regression approach, it is imperative for the residuals to be able to adhere to the assumption of constant variance. This property is also known as homoscedasticity. Two different measures are considered, namely, the Variance Inflation Factor and Breusch-Pagan test [40], for analyzing this property shared between dif- ferent cofactors in the models (Additional File 1, Section E). This also ascertains that there is no significant evi- dence that the assumptions underlying the fitting and estimation methods are violated. The model formulation using the Second Order model is then performed. The initial values of the parameters are optimized using response surface analysis. A salient feature of this system is the existence of a unique solu- tion, which implies that each parameter can assume only one value, and hence, the initial solution space and optimized values of the estimated parameters overlap. The initial model with parameter estimates, based on the 80% of the SPR time series data of Chennai, is given by, SPR time series in Chennai The SPR data for Chennai city covers the period, January 2002 - December 2004 (Figure 2(a)). Study of the auto- correlation of SPR values, at various lags ranging from 1 through 11, shows significant relationship for lag-one (Additional file 1, Section A, Figure S1(a) and Table S1). This observation is a reflection on the period of Roy et al. Malaria Journal 2011, 10:301 Page 8 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 http://www.malariajournal.com/content/10/1/301 SPR = 75.96 −20.40 (SPR−1) −1.31 × 10−01T −3.71 × 10−05P + 5.30 × 10−1 (T−1) + 8.474(SPR−1)2 −1.07 × 10−04T2 + 4.46 × 10−12P2 −1.21 × 10−03(T−1)2 + 4.4 × 10−06SPR−1P + 3.28 × 10−08T P + 2.56 × 10−05T R + 2.402 × 10−04T (T−1) −1.62 × 10−10P R −1.04 × 10−07P (T−1) . (5) Figure 3 The observed data and the model fit for SPR values in Chennai. (a) Initial model using RSM; (b) Final model using model reduction techniques (Error bars show the 95% confidence intervals); and (c) Model validation using the final reduced model for SPR values in Chennai from June 2004 to October 2004 (Error bars show the 95% prediction intervals). (5) The reduced model (Eqn.5) has R2 = 89.75%. It is important to note that through this process it is possible to remove six terms from the initial model (about 30% reduction in coefficients) compromising only less than 2% decrease in the value of R2. Figure 3(b) shows the reduced model and fitted values along with the 95% confidence intervals computed for each time point. The estimated values are found to be consistently lying within the confidence intervals. This exercises a second level check on the efficiency of the model, in addition to the R2. The AIC for the initial and final models for Chennai are - 3827.692 and 3410.028. It is clear that the reduced final model is indeed better owing to its lower AIC. For model validation the 20% of the SPR data (June to Octo- ber 2004), which were not utilized for building the model, were estimated from the final model. Figure 3(c) shows the observed and predicted SPR values estimated from the final model of Chennai (Eqn. 5) along with the 95% prediction intervals. TMC time series in Mangalore Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Page 9 of 12 optimized estimates from response surface analysis. The initial model with parameter estimates is given by, Figure 4 The observed data and the model fit for TMC values in Mangalore. (a) Initial model using RSM; (b) Final model using model reduction techniques (Error bars show the 95% confidence intervals); and (c) Model validation using the final reduced model for TMC values in Mangalore from February 2007 to December 2007 (Error bars show the 95% prediction intervals). TMC = 6995.79 −1087.87T + 4.03R −1.44TMC−1 + 516.88 (T−1) + 40.03T2 −1.9 × 10−4R2 −2.1 × 10−4 (TMC−1)2 + 6.08(T−1)2 + 5.68 × 10−2T R −5.64 × 10−2TMC−1T −35.24T (T−1) −2.6 × 10−04TMC−1R −1.59 × 10−1R (T−1) + 1.67 × 10−1TMC−1T−1. (6) (6) −1.59 × 10−1R (T−1) + 1.67 × 10−1TMC−1T−1. Figure 4(a) gives the initial model estimates of the response, TMC, and its observed values along with the 95% confidence limits. On applying the method of model reduction to equation (6), the final model with corresponding parameter estimates is given by, TMC = 9800.71 −1077.33T + 3.43R −2.30TMC−1 + 348.61 (T−1) + 18.64T2 −2.2 × 10−4(TMC−1)2 + 9.958(T−1)2 −4.378T (T−1) −1.1004R (T−1) + 1.40 × 10−1TMC−1T−1. (7) (7) This process reduces the number of terms from 15 to 11 (~27% reduction), and hardly reduces the R2 from 84.61% to 84.28% (only by 0.4%). Further attempts to remove terms (Additional file 1, Section C, Table S6) resulted in drastic reductions in the value of R2. Figure 4(b) shows the observed and estimated values for the final reduced model. In this case also, the AIC can be computed for the initial as well as the final model, which are - 3812.048 and 3359.722. The lower AIC value for the final model proves that this parsimonious representation of the same information with lower num- ber of terms is a useful approach to model epidemiolo- gical data. For model validation TMC for Mangalore from February to December 2007 were considered. Fig- ure 4(c) shows the observed and predicted TMC values estimated from the final model of Mangalore (Eqn. 7) along with the 95% prediction intervals. TMC time series in Mangalore Moreover, to assess the linear accuracy, homoscedasticity and to observe the differences after removing lower order variables, residual plots for the initial and reduced models have been plotted for both SPR and TMC values (Addi- tional file 1, Section C, Figure S5), which show uniform spread of the residual cloud indicating homoscedasticity. Figure 4 The observed data and the model fit for TMC values in Mangalore. (a) Initial model using RSM; (b) Final model using model reduction techniques (Error bars show the 95% confidence intervals); and (c) Model validation using the final reduced model for TMC values in Mangalore from February 2007 to December 2007 (Error bars show the 95% prediction intervals). TMC time series in Mangalore The TMC data for Mangalore city covers 60 months, from January 2003 through December 2007 (Figure 2 (c)). This dataset has an inherently different numerical nature from the SPR data. Based on the autocorrelation and correlation studies (Additional file 1, Section A, Table S1 and Figure S1(b)), the independent variables selected were: TMC-at-lag-one (TMC-1), average tem- perature (T), rainfall (R) and average temperature-at- lag-one (T-1). Also, even though the lag-two TMC vari- able (TMC-2) showed significant autocorrelation, it is not considered as an additional independent variable since the number of cases in the present month is con- sidered to be directly influenced by the previous month only [47]. Figure 3 The observed data and the model fit for SPR values in Chennai. (a) Initial model using RSM; (b) Final model using model reduction techniques (Error bars show the 95% confidence intervals); and (c) Model validation using the final reduced model for SPR values in Chennai from June 2004 to October 2004 (Error bars show the 95% prediction intervals). Figure 3 The observed data and the model fit for SPR values in Chennai. (a) Initial model using RSM; (b) Final model using model reduction techniques (Error bars show the 95% confidence intervals); and (c) Model validation using the final reduced model for SPR values in Chennai from June 2004 to October 2004 (Error bars show the 95% prediction intervals). intervals. This model gives a coefficient of determination (R2) of 91.53%. The linear correlation between TMC and the climate variables (Additional file 1, Section B, Table S4) were found to be significant. Here, 48 time points (80% of the data) were used to obtain the initial parameter estimates. The Second Order model gave better results, in contrast to the first order equation, owing to the non-linear rela- tionships (Additional file 1, Section B, Figure S3). This system, too, has a unique solution which leads to an intersection between the initial estimates and the Next, the method of model reduction is followed, cal- culating the changed R2 at each step and making the decision for removing or retaining a term accordingly (Additional file 1, Section C, Table S5). Being a manual and rigorous process, it is always possible to induct a term already removed if epidemiologically more impor- tant, though computationally it may have already been removed. The final model obtained using the above- mentioned process is given by, Roy et al. Discussion This reinstates that the residuals show no erratic behavior as a function of the value of the responses, and hence it can be assumed that there is no significant level of heteroscedasticity. This provides a comprehensive reason for the use of our model fitting approach for the available data. and Mangalore city, it is clear from the final model equa- tions that previous incidence of the disease, Temperature as well as Temperature-lag-one and Rainfall have strong influence on the disease incidence and play important roles to shape the disease curve. The predictions for Chen- nai city do not show any clear trend for the disease to decrease but show more monthly periodic variations. But for Mangalore city, the prediction show initial decay in the total malaria cases from February to May, then show an upward trend till August, and finally decreasing again till December following the pattern of rainfall distribution of that region. RSM is a global approximation method which is ideally suited for solving problems with a relatively noisy response, where a gradient based method would lead to a local optimum instead of a global one [32]. The Second Order model given by the RSM approach, in spite of not having terms higher than degree 2, still represents the temporal variations of both datasets very well. Further, there is no linearity assumption in the analysis. The pre- vious incidence of the disease and the influence of pre- vious temperature as independent variables are also introduced. This is an ‘auto correlative concept’ that ensures training of the model on past values and leads to autoregressive forecasts by the model, prediction of future incidence and, thus, establishing the applicability of RSM to statistical modeling of epidemiological data. A model reduction method is introduced, which ensures filtering of terms that do not make valuable contribution to the fit of the model. It is a simple but useful approach, and has no underlying assumption except that the contri- bution of a term in the model is directly proportional to its order of magnitude. Based on this, a criterion to remove terms from the model without a significant change in the coefficient of determination is successfully outlined. At each step of the model reduction, it is ensured that the predictions remain within the confidence limits that mini- mize error variance. Discussion and Total Malaria cases (TMC) for Mangalore. This approach shows the applicability of the algorithm to differ- ent measures of incidence of malaria. The analyses of the parasitological data for the two Indian cities using the RSM approach not only capture the essential dynamics of the disease incidence, but also show the influence of differ- ent climatic and non-climatic factors. For both Chennai In this study, a combination of statistical modelling approach (the RSM) and a simple model reduction method is applied to describe the incidence of malaria in two geographically, ecologically and demographically dif- ferent cities in India (Chennai, Tamil Nadu and Manga- lore, Karnataka) for two types of epidemiological data - Slide Positivity Rates (SPR) values of malaria for Chennai Page 10 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 Page 10 of 12 Page 10 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 lie within the 95% confidence limits. Also, the forecasts lie within the 95% prediction intervals (Figures 3(c) and 4 (c)). Thus, the reduced models perform well and are use- ful for forecasting purposes, provided reliable climate variable estimates and population projections are avail- able. The chi-square goodness of fit is also found to be highly statistically significant (p-value < < 0.001), which gives another affirmative test, in addition to the confi- dence limits, AIC and prediction limits, thereby render- ing significantly good comparability between the observed and predicted values. From the residual plots for both SPR as well as TMC values (Additional File 1, Figure S5), it is observed that there is slight compression of the residual cloud from the initial to the reduced model for SPR as well as TMC, but the spread in general is uniform on either side of the x-axis confirming homo- scedasticity. To explore the nature of the residuals and further to measure the collinearity imposed by different cofactors in the models, criteria such as, Variance Infla- tion Factor and Breusch-Pagan test have been used. It is observed that the chosen independent variables in both data, do not lead to violation of the assumption of con- stant variance of the residuals (Additional File 1, Table S7 (a) and (b)) and the homoscedastic null hypothesis is not rejected at 99% level of significance (Additional File 1, Table S8). Discussion It is worth noting here that the pro- posed model reduction technique simplifies the model structure by reducing number of coefficients (terms) by almost 30% compared to the initial model by RSM without reducing the coefficient of determinant values significantly (around 0.4-2% decrease in R2). The comparison of initial and final model formulations (Eqns. 4 and 5) for the SPR values in Chennai city reveals some interesting features. The final model not only con- tains less number of dependent terms but also shows reduction of direct dependence of both linear and higher order terms for rainfall. The dependence of SPR on rainfall in this final model (Eqn. 5) is only depicted by the interac- tion terms associated with other factors, such as tempera- ture and population size. This feature is also observed in the original data (Figure 2(a) and 2(b)) as the variation in SPR values do not follow exactly the rainfall distribution in this region, which clearly has two peaks every year due to North-East and South-West monsoons. Moreover, the ups and downs of the SPR values, whatever was observed in the real data, follow the temperature distribution of this city broadly. This pattern is also captured by the RSM model, which shows more dependence of SPR values on temperature, either through direct presence of linear and higher order terms, or through interactions with other associated factors including rainfall. The presence of inter- action terms and non-linear terms enables proper repre- sentation of the nature of SPR values of Chennai. On the other hand, the Total Malaria cases for Mangalore follows The results based on the initial model by RSM, and final model through the proposed model reduction tech- nique, show good fits for both the data types in two dif- ferent cities (Figures 3 and 4). It may be worth mentioning that the R2 in this study for the SPR values in Chennai city is much higher (89.75%) compared to pre- viously reported values (63.85%) observed on the same data using a different regression method [28]. Using Akaike Information criterion (AIC), it is successfully established that the reduced models for both cities are the best fit models. The observed and estimated values Page 11 of 12 Page 11 of 12 Page 11 of 12 Roy et al. Discussion Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 of the different environmental factors, previous incidence of the disease and their influence on the prevalence pat- tern, a reduced model with fewer terms is helpful for clearer understanding and reliable predictions. a cyclic pattern each year with gradual increase leading to the peak during June - August, followed by a gradual decrease, quite similar to the rainfall distribution in this region due to the direct influence of the Arabian Sea branch of the South-West monsoon (Figures 2(c) and 2 (d)). This specific trend is also mimicked by the model proposed here using the combined RSM and model reduc- tion technique. The model equation (Eqn. 7) shows direct dependence on rainfall term and captures the essential dynamics of the disease along with other factors. Another interesting observation in this study is the presence of less direct terms (linear or higher-order) and more cross inter- action terms in Chennai model compared to Mangalore model. This also highlights the complex influence of envir- onmental variables on the epidemiological status of these regions. Chennai being in the tropical wet-dry climatic region shows sustained prevalence of malaria through-out the year with less periodic variations, and hence the varia- tions in the SPR values are associated with more cross interaction terms. Whereas, Mangalore being in tropical dry climatic region shows more periodic variations in TMC values following the climatic changes in the region, which are observed to be directly influencing the model formulation. The interesting effects of climatic and non-climatic factors, such as previous disease incidence, population, are clearly visible from the analysis. The model predic- tions capture the climatic variations, mainly rainfall and temperature, for both the regions under study and resembles well with the observed disease incidence. This approach leads to detection of the most crucial environ- mental factors influencing the transmission of the dis- ease while offering a coherent and integrated understanding of the disease process in any area. Thus, the proposed combined method gives a simple, but highly predictive model for malaria incidence without compromising on the proportion of variation repre- sented and may be useful for public health professionals for adopting better strategy not only to control malaria but for other infectious diseases, if suitable climatic information and disease prevalence data are available. Additional material Since the observed and estimated values do compare well, we would like to infer that these techniques of mod- elling (the RSM and the model refinement/reduction method) perform well for analyzing epidemiological data and may be useful for forecasting purpose, provided reli- able climate variable estimates and population projections are available. Additional file 1: Statistical techniques and Model Reduction Process. The additional information provided describe the statistical techniques used for exploring the relationships between dependent variables (SPR and TMC values) and independent variables (environmental factors, previous incidence of the disease etc.) and the details of the model reduction process (also includes Tables S1-S8 and Figures S1-S5). [40-42]. Authors’ contributions SBR carried out the theoretical analysis, and contributed in writing the manuscript. RRS contributed in theoretical analysis and formulating the approach, and writing of the manuscript. SS conceived the idea for the article, and contributed to designing the study and writing of the manuscript. All authors read and approved the final manuscript. Conclusions g The authors are grateful to the anonymous referee for valuable comments and suggestions in improving the content. We acknowledge the help rendered by Sandip Mandal and Aridaman Pandit. We thank Department of Science and Technology, Government of India for financial support (SR/SO/ AS-25/2008). Mathematical analysis of the models may not always cap- ture the essential features of the disease transmission pro- cess, as it is sometimes difficult to understand the direct and interacting effects of temperature, seasonal forcing, excessive rainfall, correlation between different variables and other model parameters. Also, among the innumer- able statistical models based on malaria incidence data, only a few approaches have shown promising results. The results of these models are mostly data specific, and applicable primarily to the particular data set studied. Therefore, there is always a need for suitable models and methods to understand the important features of the epi- demiological data for providing better predictions. Competing interests The authors declare that they have no competing interests. Received: 13 June 2011 Accepted: 14 October 2011 Published: 14 October 2011 Received: 13 June 2011 Accepted: 14 October 2011 Published: 14 October 2011 In this paper the Response Surface Method (RSM) is applied for the first time in analyzing epidemiological data, specifically for malaria. It has been shown here that this method models the existing time series data well. The model is subjected to further analysis and develop- ment of a simpler, but statistically good, reduced model, which can offer better forecast of the disease. Since the major challenge is to ascertain the detailed understanding References 33. Akaike H: A new look at the statistical model identification. IEEE T Automat Contr 1974, 19:716-723. 34. The Survey of India. [http://www.surveyofindia.gov.in] y 9. Anderson RM, May RM: Infectious diseases of humans - dynamics and control London: Oxford University Press; 1991. 35. Urban Areas by Population Density. World Urban Areas (World Agglomerations). Demographia. March 2007. p. 77. [http://www. demographia.com/db-worldua.pdf]. gg g demographia.com/db-worldua.pdf]. 10. Chattopadhyay J, Sarkar RR, Chaki S, Bhattacharya S: Effects of environmental fluctuations on the occurrence of malignant malaria - a model-based study. Ecol Model 2004, 177:179-192. g p p 36. Indian Institute of Tropical Meteorology. [http g p p 36. Indian Institute of Tropical Meteorology. [http://www.tropmet.res.in/]. 36. Indian Institute of Tropical Meteorology. [http://www.tropmet.res.in/]. 37. Malaria Web Site. [http://www.malariasite.com/index.htm]. 11. Mwangi TW, Fegan G, Williams TN, Kinyanjui SM, Snow RW, Marsh K: Evidence for over-dispersion in the distribution of clinical malaria episodes in children. PLoS One 2008, 3:e2196. 38. TuTiempo.net - El Tiempo. [http://www.tutiempo.net/en]. 39. Briet OJT, Gunawardena DM, van der Hoek W, Amerasinghe FP: Sri Lanka malaria maps. Malar J 2003, 2:22. 40. Shumway RH, Stoffer DS: Time Series Analysis and Its Applications New York, Inc: Springer-Verlag; 2005. 12. Craig MH, Sharp BL, Mabaso MLH, Kleinschmidt I: Developing a spatial- statistical model and map of historical malaria prevalence in Botswana using a staged variable selection procedure. Int J Health Geogr 2007, 6:44. 41. Kirk RE: Statistics: An Introduction. 5 edition. Belmont, CA: Thomson Learning, Inc; 2008. 13. Mandal S, Sarkar RR, Sinha S: Mathematical models of malaria - a review. Malar J 2011, 10:202. 42. Montogomery DC: Design and Analysis of Experiments Wiley India (P.) Ltd; 2007. 14. Abeku TA: Forecasting malaria incidence from historical morbidity patterns in epidemic-prone areas of Ethiopia - simple seasonal adjustment method performs best. Trop Med Int Health 2002, 7:851-857. 43. Kunimitsu A: The accuracy of clinical malaria case reporting at primary health care facilities in Honiara, Solomon Islands. Malar J 2009, 8:80. 15. Lindblade KA, Walker ED, Wilson ML: Early warning of malaria epidemics in African highlands using Anopheles (Diptera: culicidae) - indoor resting density. J Med Entomol 2002, 37:664-674. 44. Shaman J, Stieglitz M, Stark C, Le Blancq S, Cane M: Using a dynamic hydrology model to predict mosquito abundances in food and swamp water. Emerg Infect Dis 2002, 8:6-12. g y 16. References 1. Gratz NG: Emerging and resurging vector-borne diseases. Annu Rev Entomol 1999 44 51 75 1. Gratz NG: Emerging and resurging vector-borne diseases. Annu Rev l Entomol 1999, 44:51-75. 2. World Health Organization (WHO) and WHO Global Malaria Programme: [http://www.who.int/topics/malaria/en/], and [http://www.who.int/malaria/ about_us/en/index.html].. 3. Sabatinelli G, Majori G, D’Ancona F, Romi R: Malaria epidemiological trends in Italy 1994. Eur J Epidemiol 1994, 10:399-403. Page 12 of 12 Page 12 of 12 Roy et al. Malaria Journal 2011, 10:301 http://www.malariajournal.com/content/10/1/301 29. Box GEP, Wilson KB: On the experimental attainment of optimum conditions. J R Stat Soc B (Methodological) 1951, 13:1-45. 4. Mert A, Ozaras R, Tabak F, Bilir M, Ozturk R, Aktuglu Y: Malaria in Turkey: A review of 33 cases, 2003. Eur J Epidemiol 2003, 18:579-582. 5. Dev V, Sangma BM, Dash AP: Persistent transmission of malaria in Garo 5. Dev V, Sangma BM, Dash AP: Persistent transmission of malaria in Garo hills of Meghalaya bordering Bangladesh north-east India Malar J 2010 30. Kim S, Na S: Response surface method using vector projected sampling points. Structural Safety 1997, 19:3-19. 5. Dev V, Sangma BM, Dash AP: Persistent transmission of malaria in Garo hills of Meghalaya bordering Bangladesh, north-east India. Malar J 2010, 9:263. points. Structural Safety 1997, 19:3-19. hills of Meghalaya bordering Bangladesh, north-east India. Malar J 2010, 9:263. 31. Kaymaz I, MacMahon CA: A response surface method based on weighted regression for structural reliability analysis. Probabilist Eng Mech 2005, 20:11-17. 6. Hay S, Guerra C, Tatem A, Noor A, Snow R: The global distribution and population at risk of malaria: past, present and future. The Lancet Infectious Diseases 2004, 4:327-336. 6. Hay S, Guerra C, Tatem A, Noor A, Snow R: The global distribution and population at risk of malaria: past present and future The Lancet 32. Vasconcelos AFD, Barbosa AM, Dekker RFH, Scarminio IS, Rezende MI: Optimization of laccase production by Botryosphaeria sp. in the presence of veratryl alcohol by the response-surface method. Process Biochem 2000, 35:1131-1138. Infectious Diseases 2004, 4:327-336. 7. Dash AP, Valecha N, Anvikar AR, Kumar A: Malaria in India: challenges and opportunities. J Biosci 2008, 33:583-92. 8. Ye Y, Louis VR, Simboro S, Sauerborn R: Effect of meteorological factors 8. Ye Y, Louis VR, Simboro S, Sauerborn R: Effect of meteorological factors on clinical malaria risk among children: an assessment using village- based meteorological stations and community-based parasitological survey. BMC Public Health 2007, 7:101. References Sarkar J, Murhekar MV, Shah NK, van Hutin Y: Risk factors for malaria deaths in Jalpaiguri district, West Bengal, India: evidence for further action. Malar J 2009, 8:133. 45. The Math Works Software. [http://www.mathworks.com]. 46. R foundation for Statistical Computing, Version R 2.11.1 Vienna, Austria [http://www.R-project.org], ISBN 3-900051-07-0.. 17. Khanum S, Singh A: Health, poverty and human security: illustrations from malaria in south-east Asia. Regional Health Forum 2007, s11:33-44. 47. Manual on Practical Entomology in Malaria: Part II Geneva: World Health Organization; 1975. 18. Kleinschmidt J, Bagayoko M, Clarke GPY, Craig M, Le D: A spatial statistical 18. Kleinschmidt J, Bagayoko M, Clarke GPY, Craig M, Le D: A spatial statistica approach to malaria mapping. Sauer-International Epidemiological Association 2000, 29:355-361. doi:10.1186/1475-2875-10-301 Cite this article as: Roy et al.: Theoretical investigation of malaria prevalence in two Indian cities using the response surface method. Malaria Journal 2011 10:301. approach to malaria mapping. Sauer-International Epidemiological Association 2000, 29:355-361. 19. Ruru Y, Barrios EB: Poisson regression models of malaria incidence in Jayapura, Indonesia. The Philippine Statistician 2003, 52:27-38. 20. Gaudart J, Touré O, Dessay N, Dicko A, Ranque S, Forest L, Demongeot J, Doumbo OK: Modelling malaria incidence with environmental dependency in a locality of Sudanese savannah area, Mali. Malar J 2009, 8:61. 21. Ermert V, Fink AH, Jones AE, Morse AP: Development of a new version of the Liverpool Malaria Model. II. calibration and validation for West Africa. Malar J 2011, 10:62. 22. Briet O, Vounatsou P, Gunawardene DM, Galppaththy GNL, Amerasinghe PH: Models for short-term malaria prediction in Sri Lanka. Malar J 2008, 7:76. 22. Briet O, Vounatsou P, Gunawardene DM, Galppaththy GNL, Amerasinghe PH: Models for short-term malaria prediction in Sri Lanka. Malar J 2008, 7:76. 23. Abellana R, Ascaso C, Aponte J, Saute F, Nhalungo D, Nhacolo A, Alonso P: Spatio-seasonal modelling of the incidence rate of malaria in Mozambique. Malar J 2008, 7:228. 24. Cancre N, Tall A, Rogier C, Faye J, Sarr O, Trape JF, Spiegel A, Bois F: Bayesian analysis of an epidemiological model of Plasmodium falciparum malarial infection in Ndiop, Senegal. North American Journal of Epidemiology 2000, 152:760-70. References Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: p gy 25. Malaria risk in Africa (MARA) project. [http://www.mara.org.za/]. 26. Kuhn KG, Campbell-Lendrum DH, Davies CR: A continental risk map of malaria mosquito (Diptera: culicidae) vectors in Europe. J Med Entomol 2002, 39:621-630. 27. Bourma MJ, Dye C, Van Der Kaay HJ: The el nino southern oscillation and historic malaria epidemics on Indian subcontinent and Sri Lanka: an early warning system for future epidemics. Trop Med Int Health 1996, 1:86-96. 28. Chatterjee C, Sarkar RR: Multi-step polynomial regression method to model and forecast malaria incidence. PLoS One 2009, 4:e4726.
https://openalex.org/W2549594739	http://www.scielo.br/pdf/remhu/v24n48/1980-8585-REMHU-24-48-011.pdf, https://ri.conicet.gov.ar/bitstream/11336/107106/2/CONICET_Digital_Nro.3574f0a9-e00c-4492-9684-74538a909d2b_A.pdf, https://www.scielo.br/j/remhu/a/JkqWQYPcXnBcqkDXNK4fDQN/?lang=es&format=pdf, https://www.redalyc.org/pdf/4070/407048610002.pdf	es		Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia	REMHU	2,016	cc-by	9,022	Valentine Le Borgne de Boisriou Dossiê: “Movimentos sociais e mobilizações de migrantes e refugiados” EXPONER LO DISIMULADO. LOS ALCANCES DE LAS MOVILIZACIONES DE PERSONAS INDOCUMENTADAS EN FRANCIA Exhibiting the undisguised. The reaches of the mobilizations of undocumented people in France Valentine Le Borgne de Boisriou1 Resumen. La cuestión de la participación política, en el espacio público francés, de los inmigrados cobra actualmente una gran relevancia, en el contexto de la crisis migratoria que atraviesa el conjunto del continente europeo. Frente a las interrogaciones que plantea esta situación, en torno a la capacidad de las sociedades a recibir y sumar a sus sociedades los que suelen ser considerados como su “excedente”, este artículo, basado sobre un trabajo de campo realizado entre 2007 y 2014, propone una reflexión acerca de la historia y los desafíos actuales de los movimientos sociales llevados a cabo por personas indocumentadas en Francia. Palabras clave: inmigrados, sin-papeles, espacio público, acción colectiva, movimientos sociales. Abstract. The matter of political participation, in the French public space, of immigrants currently takes great relevance, in the context of the immigration crisis facing the European continent. Faced with the questions posed by this situation, about the capacity of societies to receive and add to their societies those who are often seen as their “surplus”, this article, based on a fieldwork conducted between 2007 and 2014, proposes an investigation about the history and the current challenges of social movements carried out by undocumented people in France. Keywords: immigrants, undocumented-people, public space, collective action, social movements. 1 Universidad de Buenos Aires/Conicet. Buenos Aires, Argentina. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 11 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia La cuestión de los movimientos de personas indocumentadas, los llamados “sin-papeles” atraviesa tanto la filosofía política francesa como las ciencias sociales, desde hace varios años. La evolución en su aproximación por parte de distintas disciplinas da cuenta de sus transformaciones. Desde trabajos cómo La cause des sans-papiers2, hasta los trabajos de Etienne Balibar, Didier Fassin y Gerard Noiriel3, entre otros, se operó un movimiento que, a partir de las acciones llevadas a cabo en el espacio público por personas sin-papeles, desplazó el eje problemático del tema, desde una perspectiva humanitaria hacia un cuestionamiento del sentido de la ciudadanía y de los contornos del espacio político que esta diseña. Creemos que este desplazamiento se debe no solo a una evolución de los movimientos, de sus formas de movilización y de los desafíos que plantean, de los objetivos que persiguen o de los contenidos de sus discursos, sino también a una cierta aproximación teórica que buscó, en este sentido, acompañar la evolución del movimiento. Aquí se trata entonces, de un desafío en cuanto al pensamiento crítico, que impone además, una reflexión acerca del estatuto del investigador y sobre los fines de una corriente de pensamiento situada en la encrucijada entre teoría social y política, filosofía política y un pensamiento políticamente comprometido. Este artículo busca entonces, en paralelo al análisis de las luchas de sin-papeles, reflexionar acerca de los aportes hechos por los trabajos teóricos que se acercaron a este problema. Esta arista reflexiva nos parece más necesaria aún, en el trastornado contexto actual que atraviesa Europa, en un momento donde muchos buscan enfrentar las dos facetas de la cuestión migratoria en Europa, constituida por una parte, por la denominada “crisis de los refugiados” que interroga la voluntad de los estados europeos a hacerse cargo de las consecuencias de sus políticas exteriores y por otra parte, el temor instalado en las sociedades por los repetitivos atentados que ocurren en distintos países que, más allá del terror y del espanto que logran instalar, deberían poner en cuestión las formas en las cuales se desarrolló, o se obstaculizó, la perspectiva de la igualdad y de la ciudadanía plena de los inmigrados y sus hijos en los estados europeos y en particular, en Francia. Este artículo está dividido en tres secciones. En un primer momento, reconstruimos la evolución de las luchas de los inmigrantes sin-papeles, prestando atención a la historia en la que se insertan y presentando las distintas luchas llevadas a cabo por las organizaciones de indocumentados en las que 2 3 SIMÉANT Johanna. La cause des sans-papiers. El libro se basa en una tesis de doctorado, y se le considera como uno de los trabajos pioneros sobre los sin-papeles. En particular, ver BALIBAR, Etienne et alii. Sans-papiers, l’archaïsme fatal; FASSIN, Didier et alii (dirs.). Les lois de l’inhospitalité. Les politiques de l’immigration à l’épreuve des sans-papiers; FASSIN, Didier. The biopolitics of otherness. Undocumented foreigners and racial discrimination in french public debate; NOIRIEL, Gérard. Réfugiés et sans-papiers, la France face au droit d’asile, XIXe - XXe siècles. 12 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou hemos realizado un trabajo de campo4, realizado durante varios años, entre 2007 y 2014. Los movimientos presentados aquí dan cuenta de la diversidad de los movimientos de sin-papeles. Surgen en la urgencia de una situación, se construyen “así como vienen”, y siempre pareciera que interrumpen brutalmente el orden mediático para luego disolverse, sin que se pueda saber lo que fue de aquellas y aquellos. Sin embargo, cada una de estas disrupciones construyen una historia cuya origen suele identificarse con la toma de la iglesia parisina de Saint Bernard, en el verano 1996, mientras autores como Gerard Noiriel y Johanna Siméant las vinculan a las movilizaciones llevadas a cabo por trabajadores inmigrados en los años setenta. Luego, nos detendremos en la relación de los movimientos de sin-papeles y la violencia, desde dos perspectivas: por un lado, la violencia de Estado y la violencia institucional que sufren las personas indocumentadas, y por el otro, el uso y la instrumentalización de la violencia por los colectivos de sin-papeles, proponiendo así la idea de una fuerza política de la debilidad. Se tratará de dar cuenta de las prácticas de exposición de su lucha, por parte de las organizaciones y los movimientos de indocumentados, como así también de la manifestación de la clandestinidad como respuesta a una forma de violencia administrativa. En tercer lugar, analizaremos el impacto de la movilización de los sin-papeles, tanto en los propios movilizados como en las sociedades. Intentaremos demostrar que el acceso de los sin-papeles a la acción pública no solo permite la irrupción, en el escenario político, de temas que los gobiernos prefieren esconder, sino que implica una reconfiguración de la propia definición del espacio público y del ejercicio de la ciudadanía. I. Breve recorrido por la historia de la construcción de una categoría: los sin-papeles en Francia Gerard Noiriel, en el prólogo a la segunda edición de Réfugiés et sans-papiers, la France face au droit d’asile, XIXe - XXe siècles5, expone su 4 5 El trabajo de campo mencionado fue desarrollado en Paris entre 2007 y 2014, en el marco de una tesis doctoral. Más allá del trabajo documental y de la observación, en este plazo, de las diferentes ocurrencias de los movimientos de sin-papeles, en particular consistió en la observación participante de las movilizaciones callejeras y de las reuniones de un colectivo de sin-papeles, el 9ème collectif de sans-papiers (9º colectivo de sin-papeles). Asimismo, realizamos sucesivas series de entrevistas en profundidad con los integrantes del colectivo, que en ciertos casos, repetimos a lo largo del tiempo. Anteriormente, diferentes contactos habían sido realizados en otros grupos, y durante el trabajo, las observaciones se extendieron a otros colectivos y grupos, como el caso particular de la movilización de los refugiados tunecinos en Paris, en 2011 (sobre este tema, ver: BOISRIOU, Valentine de, YAHYA, Hamza Ben. Errances révolutionnaires. Entretien avec H. Variations [En linea], n. 16, 2012. Disponible en: <http:// variations.revues.org/135>), o la movilización alrededor de la cuestión del trabajo de los sinpapeles, organizada por la Confederación General del Trabajo francesa, a partir de 2008. NOIRIEL, op. cit. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 13 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia propósito con este libro. Se trataba, para el historiador, de proponer una historia de la inmigración “desde abajo”. De esta manera, el autor buscaba desprenderse de otros trabajos sobre las migraciones que, entrando en otra relación entre lo sabio y lo político, adoptaban o la posición de expertos, o la de militantes. Sobre la cuestión de los sin-papeles, los trabajos de Noiriel aportan una fuente de conocimiento particularmente importante, en lo que aclaran la historia de las migraciones en Francia que precede a los movimientos de sin-papeles que analizamos. En particular, nos referimos aquí a las investigaciones del historiador sobre la producción de sin-papeles por parte del Estado, a partir de los cambios legislativos, y a la construcción de la idea de Nación a partir o a pesar de la inmigración. Noiriel escribe Réfugiés et sans-papiers a fines de los años ochenta, en un contexto marcado por un debate creciente en Francia alrededor de “la integración de los inmigrados” que suponía que, en la situación de la época, la integración se revelaba dramáticamente incompleta, mientras había sido exitosa en las olas de migración anteriores. En efecto, para Noiriel, la inmigración se revela estrechamente ligada a la emergencia de los EstadosNación. El autor subraya “el proceso histórico de asimilación social por el cual los diversos componentes de la población del país ‘se funden’ dentro de una nueva entidad, a la vez social y política: la nación juega un papel esencial, difundiendo sus normas en el conjunto de la sociedad”6. Eso se revela particularmente importante para nuestro propósito, en la medida en que Noiriel postula que “la ruptura política que se produce a fines del siglo XIX tiene por principal consecuencia imposibilitar toda vida social individual o colectiva, que no esté fundada sobre el principio de la nacionalidad”7. Surge la cuestión entonces, a partir de la aparición de la categoría jurídica de “refugiado”, la “diferenciación y la identificación de lo nacional y lo extranjero, el inmigrante y el refugiado“8. La cuestión de la gestión política de los refugiados también se revela clave en torno a nuestro objeto de reflexión. Para el autor, se fundamenta sobre una forma elemental de la dominación: controlar por medio de la asistencia. Por otra parte, los trabajos de Johanna Siméant, que mencionaremos más adelante, demuestran de manera muy detallada como el estatuto de debilidad en el que están mantenidos los sin-papeles va a dirigir su relación con el Estado hasta volverse una de las herramientas de su lucha. En los trabajos de Noiriel, los primeros movimientos de sin-papeles surgen a partir de 1972, en consecuencia directa de la modificación de la legislación 6 7 8 Ibidem, p. 19. Ibidem, p. 22. Ibidem, p. 23. 14 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou que, instalando la obligación de poseer un contrato de trabajo, creaba de repente una nueva categoría: el extranjero sin-papeles. En respuesta a esta brutal modificación de la ley, múltiples huelgas de hambre se realizan en Francia, entre 1972 y 1975. El año 1981 constituye a su vez, una bisagra en la constitución de los movimientos de sin-papeles. En primer lugar, porque el periodo que anticipa la llegada de la izquierda al poder, en Francia, ve la multiplicación de controles de identidad, dirigidos “de modo preventivo”, a los extranjeros, permitiendo así la confusión entre inmigración y clandestinidad, lo que constituirá más adelante un eje de las luchas de sin-papeles. Por otra parte, porque la efímera “ventana”, que representa 1981 y la elección de François Mitterand, con la suspensión de todas las expulsiones, la apertura de las condiciones de las migraciones con propósito de reunión familiar, la regularización de 130 000 personas y la supresión del régimen derogativo que imponía a las asociaciones extranjeros que reciban la autorización del Ministerio del Interior, se cierra pronto. Entre 1986 y 1996, se abre una década que modificará las disposiciones de 1981, hasta que las restricciones aportadas al derecho de estadía y el control de los extranjeros culminen en los movimientos de sin-papeles de los años 90, que describiremos más adelante. Posteriormente, si bien estas movilizaciones permitirán obtener regularizaciones, estas siempre quedaran insuficientes para responder a la presencia en el territorio de decenas de miles de extranjeros sin-papeles. Los movimientos de sin-papeles se transformarán entonces, evolucionando a raíz de nuevas ocurrencias ligadas a las modificaciones legislativas o a la precipitación de los acontecimientos políticos en los países de origen. En fin, a fines de los años 2000, la cuestión del trabajo de los extranjeros sin-papeles tiende a imponerse al centro de las reivindicaciones llevadas por las organizaciones. El plural de las movilizaciones de sin-papeles En trabajos recientes, Guillaume Garcia9, volviendo sobre cuarenta años de historia de los movimientos de sin-papeles, subraya a la vez el carácter momentáneo de sus luchas – es decir que estas luchas están siempre vinculadas a una secuencia política, y se deshacen, para renacer en otra oportunidad – y la diversidad de registros que movilizan. Desde la cuestión del trabajo a la de la reunión familiar y el derecho de asilo, hasta temas vinculados a una nacionalidad de origen en particular, estos movimientos se caracterizan por la heterogeneidad de sus actores, de los motivos de su movilización, de sus apoyos y de sus modos de intervención pública. Johanna Siméant determina, en La cause des sans-papiers, que, menos aun que toda otra colectividad “inmigrada”, los sin-papeles no constituyen 9 GARCIA Guillaume. La cause des sans. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 15 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia una “comunidad”10. Los sin-papeles, así definidos desde afuera, por las leyes y los cambios de reglamentación, solo se hacen de esta nominación para tratar se desprenderse de ella. Su agrupación, hecha de múltiples nacionalidades y situaciones administrativas, difícilmente permite la reunión alrededor de una identidad positiva. Sin embargo, tanto los trabajos de campo que realizamos con colectivos de sin-papeles en Paris, como las más recientes luchas llevadas a cabo por los sin-papeles en conjunto con la Confédération Générale du Travail (CGT), acerca de la cuestión de los trabajadores sin-papeles, tienden a demostrar cómo, alrededor de la lucha política, tiende a surgir la construcción de una identidad, pudiendo superar la negatividad. La sospecha “comunitarista” a la que se enfrentan las movilizaciones de inmigrados constituye, según Siméant, una constante de la reflexión en sociología. Por la denuncia de la comunidad detrás de la movilización, se trata de descalificar las luchas de sin-papeles, como por otra parte, se suelen descalificar las luchas de sectores populares con argumentos cercanos. ¿Debemos entonces resolvernos a limitar las movilizaciones inmigradas a cuestiones de orden comunitarista, o al contrario, ignorar este aspecto? Johanna Siméant propone substituir a esta disyuntiva la cuestión de las condiciones políticas de la movilización, prefiriendo el análisis de la entrada en la movilización al presupuesto de comunidades previas a la acción. Según la autora, en los años 90, pocos trabajos planteaban la cuestión inmigrada desde el enfoque de la acción colectiva. Es que, según Siméant, el estatuto de inmigrado está considerado como fuerza de trabajo y no como fuerza política. Hablar de “inmigrados” conlleva ambivalencias: si bien disimula la heterogeneidad de las movilizaciones inmigradas, no distinguir las acciones llevadas a cabo por “extranjeros”, “exiliados”, “refugiados”, “sin-papeles”, permite a su vez no entrar en el reparto de las administraciones. Antes del desarrollo de los primeros trabajos sobre las movilizaciones de sin-papeles, los estudios realizados enfocándose en la acción colectiva de los inmigrados se dividen en dos épocas: en los años 60 y 70, plantean el tema del trabajador inmigrado, de la explotación laboral; mientras en los años 80, el “joven” reemplaza el trabajador, y se vuelve dominante el “problema de la integración”. II. Violencia de Estado y formas de acción colectiva en la precariedad Hace poco más de una década, se consideraba la huelga de hambre como el principal modo de protesta al alcance de los sin-papeles. Era, por ejemplo, la posición defendida por Johanna Siméant en La cause des sans-papiers11. 10 11 SIMÉANT, op. cit., p. 44. Ibidem. 16 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou En los años siguientes, su repertorio de acción se diversificó nítidamente y ya no se limita al ejercicio público de la violencia hacia sí-mismo12. En este mismo periodo parecen haber logrado adquirir cierta legitimidad en el espacio público. Siguiendo el objetivo de determinar lo que hizo posible esta evolución, partimos de la situación en la cual la huelga de hambre constituía el único modo de acción colectiva de los sin-papeles, analizándolo en su relación con la violencia, para llegar a una perspectiva que pueda trazar la evolución del movimiento y su desplazamiento, desde el registro humanitario de la compasión hacia el campo político. Considerar los movimientos de sin-papeles a partir de la violencia moviliza diferentes registros de reflexión: se trata, en primer término, de interrogar la violencia ejercida contra los sin-papeles, abriendo una serie de preguntas: ¿cuál es el tipo de violencia empleado, por quién, y con qué justificación? En segundo término, tenemos que examinar el uso de la violencia por los sin-papeles, dentro de la cual debemos distinguir la violencia de los individuos hacia ellos, de la violencia ejercida por los sin-papeles como movimiento social. En esta perspectiva, debemos prestar atención a la transformación que los movimientos de sin-papeles supieron operar, desde un movimiento de protesta cuyos modos de acción eran limitados, hasta volverse una figura de de movilización política auto-legitimada. Cuando se inició el trabajo de campo realizado en el marco de nuestra investigación, en 2007, la actualidad mediática reportaba cotidianamente las detenciones violentas, las redadas y las expulsiones del territorio francés que sufrían los sin-papeles. La cara más visible de esta violencia es la violencia policial que se manifiesta en los casos de detenciones y las expulsiones, cuando el expulsado o los pasajeros del avión en el que se debe ejecutar la expulsión intentar oponerse a la “medida de alejamiento”. Se trata aquí de una violencia de Estado, estando gran parte de sus operaciones enmarcadas dentro de la ley francesa. Las detenciones resultan, con frecuencia, marcadas por el ejercicio de la violencia, que en la mayoría de los casos, se resuelve unilateralmente bajo la acusación de ofensa a la policía y rebelión. Escogimos aquí dos figuras para el análisis de la violencia de Estado: el centro de retención y el método de detención a gran escala que lleva el nombre polémico de “redada”. No pretendemos que la violencia sufrida por los sin-papeles se limite a estas dos situaciones, pero creemos que ambas permiten elaborar las premisas de una comprensión de la violencia que caracteriza el vínculo del Estado hacia los sin-papeles. En esta relación atravesada por la violencia, consideramos que es necesario fijar puntos de observación, que 12 Esto, sin embargo, sigue siendo válido en el caso extremo del centro de retención administrativa, como lo veremos más adelante. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 17 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia deben permitir orientarnos en la larga lista de las violencias ligadas a la cuestión de trato policial de los sin-papeles. 2.1. Redadas y detenciones Durante la presidencia de Nicolas Sarkozy (de 2007 a 2012) se implementó una política de control masivo de identidad en la calle, operativo policial que fue denominado bajo el término de “redada”. En un primer momento, el uso del término –hasta entonces utilizado para las persecuciones de los judíos durante la segunda guerra mundial– se hizo de forma polémica, pero con el tiempo su uso se extendió y normalizó, más allá de los sectores militantes, para finalmente imponerse como el paradigma de la persecución de las personas indocumentadas. Contrariamente a las detenciones a domicilio después de una denuncia o a los controles de identidad en la calle, que tienen que ser motivados por la sospecha de un delito, la redada ocurre sin generar resistencia física, este último punto pudiendo justificarse por el gran número de policías que participan de los operativos. Sin embargo, suele ser sufrida por los habitantes de los barrios donde estos operativos ocurren con frecuencia como un modo de acción particularmente violento, y este es el punto que nos proponemos analizar aquí. En la mayoría de los casos, se desarrollan según el mismo procedimiento: se bloquea una salida de la red de subterráneo, en general en un barrio popular, donde se supone que viven y transitan muchas personas indocumentadas, luego se controlan a todas las personas que van saliendo. Sin embargo, no sería del todo exacto decir que todos son revisados, en la medida en que se trata de un control de caras. Así, se ven dos filas estableciéndose, una, saliendo sin ser demorada, constituida por “individuos de tipo europeo”, y la otra en la cual se exige la presentación del documento de identidad para poder salir. Al terminar la redada, los móviles cargados de detenidos se dirigen hacia las comisarias. Poco tiempo después de la aparición de este tipo de controles sistemáticos, que corresponden a la implementación de políticas de cupos de expulsiones del territorio por año, vecinos y asociaciones decidieron resistir a este procedimiento. Así surgieron los números de teléfono de emergencia a los que se podía llamar para avisar de un operativo y difundir la información a los listados de contacto. Una vez emitida la alerta, los indocumentados podían evitar el lugar, mientras los grupos de apoyo acudían al mismo para mantener una mirada sobre el operativo, buscando así molestarlo y volverlo más breve. Desde ese momento, la violencia física se hizo presente en las redadas, cuando las personas detenidas buscaban resistirse y los apoyos interponerse, física y verbalmente. Esta táctica de resistencia puede considerarse exitosa, en la medida en que las redadas se hicieron más cortas y sobre todo, los operativos invistieron con más dificultad los barrios, cuando esto daba lugar a reacciones 18 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou abruptas por parte del vecindario, como se pudo ver en el barrio parisino de Belleville, durante el verano 2007. Según los testimonios que recogimos en aquella época, la fuente del malestar residía en el procedimiento. La redada molesta porque funciona bajo el modelo de la trampa: la estación de subterráneo, lugar banal y cotidiano, se vuelve trampa cuya salida está bloqueada. Provoca una brutal irrupción del orden policial en lo cotidiano, y de repente, la vuelta a casa se vuelve peligrosa. Las personas que salen de la redada nos decían haber sentido vergüenza por encontrarse fuera de las camionetas, como si el operativo revelaba así su injusticia, su incursión dentro de un momento de igualdad forzada, el transporte público: de repente, el vecino en el subterráneo ya no es igual, porque a él, lo detienen. Podemos distinguir en el traumatismo que provoca la redada otro motivo, que se reúne con la propaganda alrededor de la figura del invasor: si se separa la gente al salir del subterráneo, se revela la presencia, invisible de otra forma, de los sin-papeles en la población. Para los sin-papeles, se trata de un peligro cotidiano: es preferible evitar determinados barrios, en ciertos horarios (que coinciden con el horario de salida del trabajo). Contribuye, con las detenciones a la salida de las escuelas, y cerca de los registros civiles, a difundir una amenaza difusa en el conjunto de los desplazamientos en la ciudad y las actividades de la persona sin-papeles. El cotidiano se vuelve peligroso, y el miedo a ser detenido dirige la conducta, diseñando los contornos de existencias invadidas por las “tácticas” cotidianas para evitar a la policía. 2.2. El centro de retención Sin bien no concentra la totalidad de la violencia sufrida por los sin-papeles, el centro de retención administrativa puede ser considerado como un arquetipo de la violencia ante los sin-papeles, porque lleva en sí-mismo la significación del nexo entre la violencia de Estado y la “inmigración ilegal”. Este nexo se caracteriza por su carácter defensivo, y su repertorio utiliza todas las categorías de la amenaza y la invasión, hecho de prácticas administrativas que, a pesar de que se mantengan dentro del marco de la ley, prefieren no publicitarse. El centro de retención es el lugar donde las personas sin-papeles que fueron detenidas, esperan su expulsión del territorio o un juicio, si apelaron su expulsión. Es un lugar cerrado, donde el régimen de visita está restringido. En su apariencia, se acerca a una cárcel, pero puede tratarse de un simple local requisicionado en caso de necesidad. Las organizaciones de defensa de los derechos de los migrantes y de derechos humanos suelen denunciar la multiplicación de los centros de retención en Francia y la extensión de la REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 19 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia posible duración de la retención13, criticando una forma de sistematización del encierro de los extranjeros. Según los informes de la Cimade14, que realizó numerosas entrevistas con los retenidos –en el marco de la misión de apoyo que hasta los últimos cambios en la legislación15, era la única asociación autorizada a realizar entrevistas dentro de los centros de retención– el periodo dentro del centro es sufrido como un traumatismo profundo por los retenidos. Los sin-papeles suelen auto-mutilarse. Intentos de suicidio, huelgas de hambre, ingestión de filos de afeitadoras, son prácticas comunes dentro de los centros. La mutilación de sí, en el contexto cerrado del centro y con la expulsión por único horizonte, aparece como la ultima respuesta posible, traduciendo la imposibilidad de otros recursos. Si nos adelantamos aquí sobre el análisis propuesto y nos centramos en el trabajo de Johanna Siméant y su análisis de las condiciones y las significaciones de uso de la huelga de hambre por los sin-papeles, observamos que desde la publicación de su La cause des sans-papiers, la situación de los indocumentados se modificó: Siméant escribe en 1998 que el repertorio alcanzable por los sin-papeles está principalmente concentrado en la huelga de hambre, siendo este modo de acción el único medio para la acción política al que puedan acceder. Sin embargo, 10 años más tarde, constatamos que este repertorio se amplifico de manera significativa, lo cual analizaremos más adelante. Vemos entonces, en la práctica de la huelga y la auto-mutilación en el centro de retención una respuesta extrema a una situación de violencia exacerbada, en la cual la desproporción de las fuerzas es tanta que estas prácticas se imponen como los únicos y últimos remedios contra la amenaza de la expulsión. De esta manera, el centro de retención adopta la forma de una zona límite del derecho, a la frontera de la jurisdicción nacional, siendo que las leyes que lo enmarcan dejan un lugar significativo a la apreciación subjetiva de las situaciones. Intimidaciones, maltratos y humillaciones son casos frecuentes en el centro. Los casos de rebelión ante estos tratos existen, como lo demostró el incendio del centro de retención parisino de Vincennes, en junio 2008, pero son poco frecuentes, en particular por que las entradas y salidas rápidas de los En 2016, el límite de la retención es de 45 días. Una vez pasado este plazo, si la administración no pudo efectivizar la expulsión, el extranjero debe ser liberado. El número de personas retenidas por año no deja de aumentar. Según los informes anuales de la Cimade, en 2010, se constataban 33692 casos. En 2014, fueron 49537. 14 La Cimade, o Comité Inter-Mouvements Auprès Des Evacués (Comite inter-movimiento acerca de los evacuados), es una de las principales organizaciones de apoyo a los migrantes, los refugiados y los pedidores de asilo, fundada en 1939 en el contexto de la 2da guerra mundial, en Francia. 15 Desde 2011, la Cimade comparte esta misión con 4 asociaciones: France Terre d’Asile, l’Ordre de Malte, l’ASSFAM, Forum Réfugiés. 13 20 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou retenidos no dejan a los migrantes la posibilidad de organizarse ante la sucesión de violencias dentro del centro. Dentro del centro de retención, la vida está encerrada, atrapada en una trampa de la que difícilmente se sale. La liberación, después de haberse transcurrido los 45 días del actual periodo máximo de encierro, tampoco aparece como una liberación, porque los extranjeros en situación irregular siguen siéndolo una vez liberados y los periodos de encierro y de libertad, pueden sucederse. En este caso, es la propia naturaleza del centro de retención que se debe interrogar, ¿cuándo deja de constituir un lugar de tránsito, donde el extranjero está retenido hasta que se pueda efectivizar su expulsión y se vuelve una cárcel para extranjeros, a la que se llega por haber cometido el delito de presencia indeseable? En este marco, el informe de 2007 de la Cimade subraya el deterioro de las condiciones de encierro y condena la extensión del periodo de retención posible (que pasa por esta fecha de 32 a 45 días). Este último punto, junto a la multiplicación de la cantidad de centros de retención y del numero de retenidos, permite pensar un desplazamiento de este dispositivo, desde la figura de la excepción hasta la de la banalización de un procedimiento administrativo. Entonces, tanto la redada como el centro de retención, aparecen ahí como el motivo ambiguo de la expresión de la violencia del nexo represivo entre Estado y sin-papeles. Se tratan de procedimientos violentos, cuya repercusión en las vidas tiene efectos desastrosos. Sin embargo, están inscriptos dentro de la ley, definidos por circulares oficiales, programados, preparados y legitimados. El derecho francés inscribe la presencia en su territorio de zonas de infra-derecho, zonas de sombra, suspensivas, habitadas por la violencia física e institucional. En numerosos casos, esas zonas son desconocidas para la mayoría y las personas que viven cerca de los centros de retención, aun cuando saben que ahí se ubica un centro de retención, suelen ignorar su función y lo que allí ocurre. Por su parte, la redada, practicada de manera repetitiva en ciertos barrios, modifica por algunas horas el paisaje de un barrio. Se forma una “nasa” a la salida de un subterráneo, donde se ordenan los individuos según su apariencia. Allí, la violencia simbólica se deja ver, pero lo hace de forma fugaz, como avergonzada: por la periodicidad de las redadas en determinados barrios del norte de París, habiendo fastidiado a sus habitantes, los operativos se transformaron en “operaciones relámpago”, para preservar el personal policial de la agresividad que se generaba en contra de ellos y a su vez, para que esta práctica se vuelva menos ostensible. Podemos decir, frente a estas figuras, que el Estado busca esconder lo que el ojo ciudadano no debe ver. Si el centro de retención se disimula detrás de sus paredes, si las manifestaciones en sus alrededores están prohibidas por ley, y si la redada busca volverse invisible, es porque no pueden aparecer sin revelar REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 21 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia su carácter inhumano y provocar la ira. Disimulados detrás de un silencio demasiado organizado para no asemejarse al secreto, solo llaman la atención de vecinos atentos y organizaciones de apoyo. 2.3. Formas de la acción colectiva en los movimientos de sinpapeles Tratándose de caracterizar la violencia de los sin-papeles, nos encontramos ante una situación compleja. Violencia, sin-papeles: por un lado, la articulación parece ser evidente. Una gran cantidad de imágenes, transmitidas por los medios, se presentan ante nosotros: son casos de fallecimientos en circunstancias de persecución, de suicidios en los centros de retención o ante una detención, vuelos que no pueden despegar por el revuelo causado por la presencia a bordo de un sin-papeles en proceso de expulsión. Sin embargo, si buscamos interrogar la violencia que pueden utilizar los movimientos de sin-papeles, o las personas sin-papeles de forma individual, la pregunta parece más complicada, porque la reciprocidad de la violencia se encuentra más difícil de demostrar. ¿Podría tratarse de una situación de monopolio de la violencia, en la cual los sin-papeles se constituirían como victimas absolutas? Analizaremos aquí la práctica de la huelga de hambre como violencia utilizada en contra de sí-mismo con un objetivo político, y las ocupaciones, estas dos practicas formando, junto a las manifestaciones en la calle, los modos de acción más representativos de acción de los colectivos de sin-papeles. Intentaremos caracterizar estas acciones y sus modos de justificación junto a un análisis de la construcción de los movimientos de sin-papeles y su capacidad para forjar esta imagen de víctima. 2.3.1. La huelga de hambre La práctica de la huelga de hambre es un hecho recurrente de las movilizaciones alrededor de la cuestión de los migrantes desde los años 70. Buscaremos en este apartado aclarar lo que su utilización puede enseñarnos acerca del tipo de justificación a las violencias contra sí-mismo, así como la trama que construyen con la violencia del poder. Esta práctica está necesariamente puesta en escena y espectacularizada para lograr su objetivo, siendo así mediante su puesta en escena como puede construir una relación de fuerza política, al poder establecer un enfrentamiento con el Estado. Por eso, llaman a los medios de comunicación. Sin su presencia, la huelga es inútil porque no se trata de morir de hambre, sino de mostrar que se está dispuesta a llegar hasta esta última instancia. Por lo tanto, los sin-papeles en huelga de hambre consideran la posibilidad de la muerte, pero buscan atrasar hasta su máximo punto su término. En todo el transcurso de la huelga, se diseña una relación de fuerza que adopta la forma de un equilibrio tenso 22 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou entre las amenazas de los huelguistas, los informes de salud alarmistas de sus apoyos, y la inflexibilidad del poder, la cual no duda en enviar a sus propios médicos para auscultar la determinación y el estado de salud de los huelguistas, cuando no alimenta a la fuerza los más debilitados. En esta configuración, el papel de los sin-papeles es el de la debilidad expuesta. Los cuerpos se muestran en las pantallas cada vez más enflaquecidos y cuanto más debilitados aparecen los huelguistas, más cruel se vuelve el poder para la opinión pública. En este contexto, podemos plantear la huelga de hambre como una manifestación de la fuerza política de la debilidad. Sin embargo, esta debilidad se vuelve fuerza solamente si está transmitida por los medios de comunicación, expuesta y espectacularizada, y es ahí donde los apoyos en la sociedad toman una relevancia capital. Así, lo que diferencia la ocupación de la iglesia Saint-Ambroise de la de Saint-Bernard16, que la sigue de pocos meses, es el hecho que en la segunda ocupación, el grupo que lidera la acción ya pudo constituir una base de apoyo consistente, que luego se abrirá a personalidades mediáticas y políticas. El contexto de la huelga, el verano en Paris, particularmente vacío de informaciones que pudieran ocupar los medios, al que se suma las condiciones extremas de la huelga, que sobrepasa cuarenta días, crean el tejido explicativo del éxito sin precedente de la ocupación de Saint-Bernard, de la que podemos decir que funda el movimiento de los sin-papeles, lo instituye en tanto movimiento político con una fuerza propia y modos de acción, superando en este momento el estatuto de movimiento de protesta de rasgo humanitario. En la medida en que la huelga de hambre aparece, en este contexto y al menos por un periodo determinado, como el modo de acción política privilegiado por los movimientos de sin-papeles, proponemos caracterizar la relación de fuerza que allí se expone. El mensaje enviado al Estado por los sin-papeles puede ser resumido como lo planteó Johanna Siméant: “que se atrevan a violentarnos, a nosotros que ya nos hacemos sufrir y nos hemos adelantado”17. Este punto debe permitirnos discernir la justificación de las violencias hacia sí-mismo, cuando están perpetradas con un horizonte político. En primer lugar, debemos tomar en cuenta que la violencia dirigida hacia cualquier otra persona o institución es un modo de acción prohibido a los sin-papeles: parece imposible reivindicar ser integrado en una comunidad política utilizando la Detallamos más adelante estas dos ocurrencias, en el apartado dedicado a las ocupaciones llevadas a cabo por los sin-papeles organizados. En pocas palabras, podemos mencionar que ambas movilizaciones tuvieron lugar en Paris, en 1996. La segunda duró todo un verano, y marcó de manera contundente la entrada en la agenda social y política de la problemática de los sin-papeles. 17 SIMÉANT, Johanna. Violence d’un répertoire. Les sans-papiers en grève de la faim. 16 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 23 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia violencia, más allá del hecho que usar este medio expondría a una expulsión inmediata, o al encarcelamiento, cuando el objetivo principal de los sin-papeles es evitar todo contacto con la policía. La violencia hacia sí-mismo constituye entonces el último recurso para unos militantes que no disponen sino de pocos medios para ejercer algún tipo de presión. Se trata de crear una situación que pueda incrementar el estatuto de víctima. Si está bien desarrollada, la huelga de hambre debe tener por efecto dar vuelta a la violencia del Estado contra sí-mismo. La visión de los cuerpos sufridos tiene por efecto la producción de un sentimiento de empatía, acentuado por la determinación y como en espejo, reforzando así la inflexibilidad del Estado. 2.3.2. Las ocupaciones Las ocupaciones constituyen, dentro del repertorio actual de acciones utilizado por los movimientos de sin-papeles, una de las formas de acción colectiva más frecuentes. Desde los años 90, diferentes organizaciones de sin-papeles ocuparon, en distintas ocasiones (campañas de lucha política y/o sindical, resistencia ante desalojos, oposición a las reformas legislativas) una gran variedad de edificios: iglesias, locales de campaña de partidos, ministerios, etc. Buscaremos aquí analizar cuáles son los tipos de edificios ocupados, en qué circunstancias y con qué legitimación. La ocupación de un edificio público, al contrario de la huelga de hambre, es una forma de violencia explicita: haya o no secuestración de personas. En todo caso, esta violencia suele ser minimizada en comparación con las violencias sufridas por los sin-papeles, pero no entra dentro del registro de los derechos humanos, de lo humanitario o de la debilidad ligado a la huelga de hambre. Esta práctica incisiva persigue un doble objetivo: expandir la visibilidad del conflicto, y radicalizarlo por la implementación de una estrategia del enfrentamiento. Observamos una incrementación del valor estratégico de los lugares tomados. Si bien en un primer tiempo, las iglesias están priorizadas, porque constituyen una meta simbólica y a su vez, porque implican menos riesgo de ser detenidos, en una etapa siguiente, los objetivos se desplazan hacia terrenos más políticos y a desafíos más arriesgados: así, en 2007, la sede de campaña presidencial del candidato Nicolas Sarkozy fue ocupada por un colectivo de sin-papeles, con el que realizamos un trabajo de campo. 10 años antes, las primeras ocupaciones significativas realizadas por grupos de sin-papeles habían tenido lugar en dos iglesias parisinas, Saint-Ambroise en marzo de 1996 y Saint-Bernard en junio del mismo año. Estos acontecimientos, que podemos caracterizar como los actos de nacimiento de los movimientos de sin-papeles, habían surgido bajo la iniciativa de un pequeño grupo de conocidos, sobre la 24 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou base de una protesta frente al endurecimiento de las leyes migratorias por el gobierno de Charles Pasqua. A posteriori, se construye una reflexividad de la ocupación, con la construcción de una relación de fuerza. Esta relación se establece, en un primer momento, con las autoridades eclesiásticas: poco tiempo después del inicio de la ocupación de la iglesia Saint-Ambroise, proponen desplazar los ocupantes hacia un lugar más adecuado. En este momento se construye la idea según la cual el lugar ocupado es elegido en función del desafío que representa, y el grupo decide permanecer en la iglesia. La ocupación de Saint-Ambroise no duró más de algunas horas, pero la ocurrencia siguiente de las luchas de sin-papeles, la toma de la iglesia Saint-Bernard, será mucho más larga y sus consecuencias fundamentales para la construcción de la identidad de los movimientos de sin-papeles. Las ocupaciones de las iglesias, si bien tienen por objetivo la puesta en escena de la condición de víctimas de los sin-papeles, conllevan otras víctimas –los fieles, a los que se negaba la posibilidad de asistir al culto– como argumentaron las autoridades católicas cuando llamaron a la policía para desalojar la iglesia, después de dos meses de presencia de los sin-papeles. La construcción de un imaginario victimizado alrededor de los ocupantes revela un esquema que aparece explícitamente en las estrategias de ocupación, pero más aún en el caso de las huelgas de hambre: al poner en escena la debilidad, al demostrar que estos últimos recursos son los únicos alcanzables por individuos carentes de todo, los sin-papeles construyeron una fuerza de presión que encuentran precisamente en su posición de debilidad y utilizan diversos modos de representación de sí que les permiten, por un juego de espejo, tomar a su cuenta a la violencia de Estado. III. Superar la experiencia silenciosa de la violencia. El colectivo como fuerza política La situación inicial que pudimos observar es la de la experiencia silenciosa de la violencia institucional por los sin-papeles. Esta experiencia no es solamente silenciosa, sino también solitaria y vergonzosa. La persona sin-papeles vive aislada del conjunto de la sociedad y se ve afectada por diversas dificultades con las que se enfrenta su existencia en varios puntos: en el ámbito laboral, está limitada al trabajo informal, así expuesta a la arbitrariedad en sus relaciones de trabajo; su vida familiar está marcada por el riesgo permanente de la expulsión; en un sentido general, la persona sin-papeles debe permanecer desaparecida en todo momento, dirigiendo su vida según lo que podría resumirse de esta manera: hacer siempre el menor ruido posible, no llamar nunca la atención. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 25 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia La necesidad de disimularse implica diferentes consecuencias: incita a callar las violencias sufridas, por una parte, porque levantar la voz expone a ser arrestado. Por otra parte, constatamos en las entrevistas realizadas que llevaría una tendencia, en un largo plazo, a justificar las violencias. Con gran parte de los sin-papeles, una palabra se repite a lo largo de las entrevistas: la vergüenza. Vergüenza al deber esconderse en todo momento, por trabajar en negro, tener que mentir, disimular sus condiciones de vida. Así, cuando durante una entrevista, Ahmed –sin-papeles argelino– habla de la dificultad de vivir con toda su familia en un monoambiente, no lo hace reclamando mejores condiciones de vida, sino para mencionar su temor a ser denunciado por sus vecinos, si el ruido llegara a molestarlos. En este sentido, los sin-papeles no organizados en colectivos o que no se acercaron a ninguna asociación de apoyo, se asemejan a átomos dispersos, sufriendo una violencia institucional que ataca a sus vidas en su individualidad y en cada una de sus actividades, encerrando a los individuos en una relación de fuerza cuya desproporcionalidad tiene efectos petrificantes. Cada faceta de su existencia puede ser judicializada, sometida al arbitrario, y los individuos se enfrentan a la necesidad permanente de justificarse, ante sus propios juicios y los de la sociedad, más aun cuando deben afrontar procedimientos policiales que los consideran como delincuentes por no llevar papeles. En este conjunto, la conjunción de un discurso oficial estigmatizante, de prácticas policiales traumatizantes, con la mantención forzada en una situación por fuera, o por debajo, del derecho, produce una disminución de la capacidad de resistencia de los individuos. Por una acumulación de prácticas y de discursos, los sin-papeles interiorizan la imagen negativa de sí-mismos que se les transmite. Siguiendo a esta línea, podemos afirmar que los mecanismos de asignación identitaria por los que cada uno permanece en el lugar que le corresponde y en las condiciones que se le permiten funcionan así plenamente. La aparición de los colectivos de sin-papeles, que suele ser vinculada a la fecha del 28 de junio de 1996, cuando inició la ocupación de la iglesia Saint-Bernard, transforma la situación anterior. Desde entonces, opone a la experiencia silenciosa de la violencia, la reivindicación de los derechos humanos y la denuncia de las condiciones de vida de los sin-papeles, mientras, por otra parte, revela la presencia disimulada de los sin-papeles en la ciudad, otorgándoles una visibilidad reivindicada y haciéndose de modos de expresión que se les negaban. Así, a quince años de la toma de la iglesia Saint-Bernard, después de su remarcada irrupción en las escenas políticas y mediáticas, los sin-papeles se constituyeron como fuerza política. Los colectivos de sin-papeles se difundieron en todo el territorio y se desarrollaron en profundidad, ampliando tanto sus redes de apoyo como sus repertorios de acción. El hecho que los apoyos estén, ahora, apoyando y no liderando las movilizaciones es un punto 26 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou clave de la constitución de los colectivos de sin-papeles como actor político en su propio nombre. Estos apoyos, movilizados en función del contexto, ya no tienen una función tan esencial. Por otra parte, si Johanna Siméant escribía, de manera muy justa, en 1993, que la huelga de hambre constituye el modo de acción privilegiado, sino el único, la situación actual cambió en profundidad. La huelga de hambre y las mutilaciones hacia sí-mismo que la acompañan se siguen utilizando, pero solo se observa dentro de los centros de retención que, como lo hemos visto, representan el paroxismo de la violencia sufrida por los sin-papeles. Los colectivos supieron desarrollar nuevos modos de acción, tales como las marchas y las ocupaciones, ya no solo de iglesias sino también de edificios oficiales. El punto donde relevamos un cambio en la apropiación de la acción colectiva sobre la escena política por los colectivos de sin-papeles yace no tanto en los modos de acción escogidos, sino por el hecho que estos modos de acción están apropiados por una categoría de actores a la que esto parecía le era imposible unos años atrás. Los colectivos demostraron entonces su capacidad de adquisición de los recursos simbólicos y sociales necesarios a la acción política. Relevamos numerosos impactos de esta evolución. En particular, se destaca las marchas y protestas callejeras organizadas por colectivos de sin-papeles. Hasta los últimos años, las manifestaciones públicas de sin-papeles eran consideradas imposibles, por el riesgo muy alto que conllevaban, y tal vez, porque no se las podía imaginar. Ahora tales protestas son corrientes. Los sin-papeles, organizados en colectivos, explican a su vez que el hecho de pertenecer a un colectivo permite una más rápida liberación cuando están arrestados. Asimismo, las huelgas de trabajadores sin-papeles se multiplican, y con ello la cuestión del trabajo de los indocumentados se vuelve pública, permitiendo romper con estas formas de explotación laboral. En fin, cuando una persona sin-papeles está arrestada y condenada a la expulsión, su pertenencia a un colectivo complica o imposibilita el cumplimiento de la decisión. Sin embargo, es al nivel de los individuos y de su percepción de sus condiciones que la diferencia entre ambas temporalidades se vuelve más concisa. Los sin-papeles que entrevistamos mencionaron, con gran recurrencia, como la pertenencia a un grupo solidario y activo constituye el primer logro de la organización. Si bien pueden tener diferentes expectativas en cuando a la posibilidad de lograr su regularización, afirman que “el colectivo protege”, “te permite sentirte más fuerte”, “hace bien”. A nivel individual, podemos decir que el colectivo de sin-papeles funciona como una barrera ante la violencia arbitraria a la que los sin-papeles están expuestos. Mientras al nivel de la acción política colectiva, la organización de los sin-papeles en colectivos permitió que estos grupos puedan afirmarse en tanto detentores de una identidad REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 27 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia reivindicativa. Afirmarse sin-papeles y marchar como tal en la calle, implica deshacerse de la asignación peyorativa a la que se sometían anteriormente y reivindicarse un derecho a vivir en el lugar donde, de hecho ya se vive, se relaciona y se trabaja. Por último, queremos mencionar aquí como la ilustración más clara del proceso por el cual los colectivos de sin-papeles lograron transformar el estatuto de los indocumentados, desde la postura de fantasmas avergonzados a la de militantes determinados, el cambio semántico que logro su organización. Así, el término sin-papeles pudo instalarse mediante las luchas. Este término, si bien expresa una falta, está dirigido a la reivindicación de lo que hace falta, y al mismo tiempo, prepara el juego estratégico de espectacularización de la debilidad que hemos analizado. Reemplaza el término “clandestino” que, utilizado anteriormente o ahora, con cierta carga ideológica, corresponde a una identidad disimulada de un individuo que no puede ser más que un cuerpo expuesto a todo tipo de violencia física, psicológica e institucional. Referencias bibliográficas ANDRES, Hervé. Les enjeux théoriques du droit de vote des étrangers : la démocratie contre la souveraineté. Migrations et sociétés, v. 19, n. 114, 2007, p. 47-64. ARIES, Paul. 1973, les “sans-papiers” du bidonville de Feyzin. Hommes et migrations, n. 1177, 1994, p. 43-47. BALIBAR, Etienne; CHEMILLIER-GENDREAU, Monique; COSTA-LASCOUX, Jacqueline; TERRAY, Emmanuel. Sans-papiers, l’archaïsme fatal. Paris: La Découverte, 1999. BOUAMAMA, Saïd (dir.). La République à l’école des sans-papiers. Paris: L’Harmattan, 2009. BOUZIRI, Saïd. La participation des étrangers à la vie associative. Migrations et Sociétés, n. 38, 1995, p. 47-52. BROSSAT, Alain. Autochtone imaginaire, étranger imaginé: Retours sur la xénophobie ambiante. Paris: Editions du souffle, 2013. CHAUVIN, Sébastien. En attendant les papiers, l’affiliation bridée des sans-papiers aux Etats-Unis. Politix, n. 87, 2009, p. 47-69. CHOUKRI, Hmed. Sonacotra cèdera! La construction collective d’une identité collective à l’occasion de la grève des loyers dans les foyers de travailleurs migrants (1973-1981). Agone, n. 40, 2008, p. 7-14. DUBET, François. Immigrations, qu’en savons-nous? Paris: La documentation, 1989. FASSIN, Didier; MORICE, Alain; QUIMINAL, Catherine (dirs.). Les lois de l’inhospitalité. Les politiques de l’immigration à l’épreuve des sans-papiers. Paris: La découverte, 1997. FASSIN, Didier. The biopolitics of otherness. Undocumented foreigners and racial discrimination in french public debate. Anthorpology Today, v. 17, n. 1, 2001, p. 3-7. 28 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 Valentine Le Borgne de Boisriou GARCIA, Guillaume. La cause des sans. Rennes: Presses Universitaires de Rennes, 2012. GRIMSON, Alejandro. La vida política de la etnicidad migrante: hipótesis en transformación. Estudios Migratorios Latinoamericanos, año 17, n. 50, Buenos Aires, 2003, p. 143-159. JAZOULI, Adil. L’action collective des jeunes Maghrébins de France. Paris: L’Harmattan, 1986. LABELLE, Micheline; ROCHER, François. Contestation transnationale, diversité et citoyenneté dans l’espace québécois. Montréal: PUQ, 2004. LAPEYRONNIE, Didier. Assimilation, mobilisation et action chez les jeunes de la seconde génération de l’immigration maghrébine. Revue française de sociologie, v. 28, n. 2, 1987, p. 287-318. MASSON, Paul; BALARELLO, José. Rapport de la commission d’enquête sur les régularisations d’étrangers en situation irrégulière. Journal officiel du 3 juin 1998. NOIRIEL, Gérard. Réfugiés et sans-papiers, la France face au droit d’asile, XIXe - XXe siècles. Paris: Hachette, 1998. NOIRIEL, Gérard. Le creuset français. Paris: Editions du Seuil, 1988. NOIRIEL, Gérard. Color blindness et construction des identités dans l’espace public français. In FASSIN, Eric; FASSIN, Didier (dir.). De la question sociale à la question raciale, Représenter la société française. Paris: La Découverte, 2006. PACECCA, Maria Ines. El fantasma en la maquina: la praxis política de los extranjeros. In VILLAVICENCIO, Susana (dir.). Los contornos de la ciudadanía. Buenos Aires: Eudeba, 2003. SAYAD, Abdelmalek. Qu’est-ce qu’un immigré. In IDEM. L’mmigration ou les paradoxes de l’altérité. Paris-Bruxelles: De Boeck Université, 1997, p. 53-54. SAYAD, Abdelmalek. Le phénomène migratoire, une relation de domination ou les conditions de possibilité d´un science de l´émigration, Maghrébins en France, émigres ou immigres ? Annuaire de l´Afrique du nord. Paris: CRESM-CNRS, 1983. SAYAD, Abdelmalek. Tendances et courants dans les publications en sciences sociales. sur l´immigration en France depuis 1960. Current Sociology, ISA, v. 32, n. 3, Sage Publication, 1984. SAYAD, Abdelmalek. A land of no return. Exile. UNESCO Courier, 1996, p. 10-12. SAYAD, Abdelmalek. L´immigration ou les paradoxes de l´altérité. Paris: De boeck &Larcier S.A.,1997. SAYAD, Abdelmalek. A imigração. San Pablo: Ed. Usp, 1998. SAYAD, Abdelmalek. La doublé Absence. Des illusions aux souffrances de l’immigré. Paris: Seuil,1999. SHEPARD, Todd. Une République française “postcoloniale”, la fin de la guerre d’Algérie et la place des enfants des colonies dans la Ve République. Contretemps, n. 16, mai 2006, p. 54-53. SIMÉANT, Johanna. La cause des sans-papiers. Paris: Presses de Sciences – Po, 1998. REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016 29 Exponer lo disimulado. Los alcances de las movilizaciones de personas indocumentadas en Francia SIMÉANT, Johanna. Violence d’un répertoire. les sans-papiers en grève de la faim. Cultures et Conflits, n. 9-10, 1993, p. 13-42. TASSIN, Etienne. La cosmopolítica a prueba: la cuestión del extranjero. Revista Posdata. Reflexión y análisis Político, n. 9, Septiembre, 2003. WITHOL DE WENDEN, Catherine. Les immigrés et la politique, cinquante ans d’évolution. Paris: Presses de la SNSP, 1988. WITHOL DE WENDEN, Catherine; LEVEAU, Remy. La deuxième generation. Pouvoirs, n. 47, 1988, p. 61-73. Recibido para publicación en 09.09.2016 Aceptado para publicación en 07.11.2016 Received for publication in September 09th, 2016 Accepted for publication in November 07th, 2016 ISSN impresso: 1980-8585 ISSN eletrônico: 2237-9843 http://dx.doi.org/10.1590/1980-85852503880004802 30 REMHU - Rev. Interdiscip. Mobil. Hum., Brasília, Ano XXIV, n. 48, p. 11-30, set./dez. 2016
W3154027813.txt	https://www.mdpi.com/2079-6382/10/4/457/pdf?version=1618883361	en		Comparative Studies on Different Extraction Methods of Centella asiatica and Extracts Bioactive Compounds Effects on Antimicrobial Activities	Antibiotics	2,021	cc-by	13,918	antibiotics Review Comparative Studies on Different Extraction Methods of Centella asiatica and Extracts Bioactive Compounds Effects on Antimicrobial Activities Farhana Nazira Idris and Masrina Mohd Nadzir * School of Chemical Engineering, Engineering Campus, Universiti Sains Malaysia, Nibong Tebal, Pulau Pinang, Seberang Perai Selatan 14300, Malaysia; farhana_nazira@yahoo.com * Correspondence: chmasrina@usm.my Abstract: The interest of consumers in using products containing phytochemicals derived from plants is growing day by day due to the shift of consumers’ preferences from convenience to environmental sustainability. One plant utilized in many products is Centella asiatica, a herb commonly used in folk medicine, health supplements, and beauty products. Extraction of bioactive compounds from C. asiatica was performed using conventional methods and modern methods (e.g., microwave or ultrasound-assisted and subcritical water extraction). This review summarizes the variety of methods used to extract active compounds from C. asiatica, their influence on the bioactive compounds and antimicrobial activity in vitro and in vivo, and the safety and toxicology of C. asiatica extract. Keywords: antimicrobial; bioactive compounds; Centella asiatica; extraction; in vitro; in vivo Citation: Idris, F.N.; Mohd Nadzir, M. Comparative Studies on Different Extraction Methods of Centella asiatica and Extracts Bioactive Compounds Effects on Antimicrobial Activities. Antibiotics 2021, 10, 457. https:// doi.org/10.3390/antibiotics10040457 Academic Editor: William N. Setzer Received: 18 March 2021 Accepted: 16 April 2021 Published: 17 April 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1. Introduction Medicinal plants have been used widely in nutraceuticals and cosmeceuticals in which phytochemicals from plants represent natural sources of compounds with several biological benefits. Extracts from plants have been used as additives in different drug formulations to enhance their use as basic health care by 80% of the world’s population [1]. Nowadays, more and more developed countries have started to grow more herbs to turn into plant-based products, therefore, these activities are no longer associated with third-world countries. Centella asiatica is one of the most popular herbs that can be found abundantly in China, Japan, Italy, Sri Lanka, Iran, India, Madagascar, America, Australia, South Africa, Indonesia, and Malaysia [2–5]. It is a perennial herb in the Umbelliferae family and presents medicinal properties such as anti-inflammatory [6–8], anti-ulcer [9], antimicrobial [10–12], and memory-enhancing properties [13,14]. It is also used for healing skin diseases such as leprosy, psoriasis, eczema, and itching [9,15,16]. This plant is known as pegaga in Malaysia, codagem in Brazil, tsubo-kusa in Japan, and mandukaparni in India. Other common names are pennywort, gotu kola, and brahmi [5,17]. Different studies published worldwide have reported that up to seventy compounds have been extracted from C. asiatica [5]. The most abundant bioactive compounds found in C. asiatica are represented by asiaticoside, madecassoside, asiatic, and madecassic acid from the triterpene class [5,6,18,19]. Triterpenes are used in cosmeceuticals mainly for woundhealing, anti-wrinkle, and anti-cellulite effects since they increase the synthesis of collagen and improve the production of fibronectin in human skin fibroblasts [6,20]. Specifically, asiaticoside increases the hydroxyproline content and tensile strength in wound tissue, raises the collagen Type I and III content, expedites the wound-healing process, and induces glycosaminoglycan synthesis [6,20–22]. Madecassoside can also stimulate the production of collagen Type III [6]. Other compounds from the saponins group that can be found in a smaller amount in C. asiatica are brahmoside, centelloside, glycosides, and alkaloids [5,15]. Antibiotics 2021, 10, 457. https://doi.org/10.3390/antibiotics10040457 https://www.mdpi.com/journal/antibiotics Antibiotics 2021, 10, 457 2 of 24 Flavonoids such as kaempferol, quercetin, rutin, catechin, naringin, and apigenin that contribute to the total phenolic content can also be found in C. asiatica [14]. The market for C. asiatica is primarily in the Asia Pacific, but keeps rising in the U.S. and Europe due to its wide range of medicinal purposes [23]. Many researchers have studied this plant due to its potential benefits like antioxidant, immune system enhancers, and antimicrobial effects, which could be useful traits when applied in cosmeceutical, pharmaceutical, medicinal, and health-related products. Specifically, C. asiatica acts as a natural source of antimicrobial agents, providing an alternative solution to overcome antibiotic resistance, which has become a concern due to the widespread usage of antibiotics [12]. The interest in C. asiatica has resulted in patents for topical skincare formulations [24], an oral rinse [25], a lower limb ulcer treatment [26], a memory enhancer in Alzheimer’s disease [27], and a topical hair composition [28]. More than 100 health care formulations based on C. asiatica are sold on the market [29], and an asiaticoside and madecassoside content of at least 2% is required for the product to be used in herbal industries. In India, the price for 500–1000 MT of C. asiatica is 0.4–0.5 U.S. dollars/kg [30]. The demand for C. asiatica in Indonesia has reached 25 tons/year, but the supply is only four tons/year [31]. Since 90% of the supply of C. asiatica is exploited from wild plants [32], it is crucial to ensure that harvesting is sustainable to meet the demand. Extraction is also essential to obtain the maximum amount of the desired compounds using the minimum amount of C. asiatica. The typical stages of separating bioactive compounds from plant materials are sample preparation, extraction, and purification. It is crucial to select the appropriate extraction method since more than 60% of the total time is used for the sample preparation stage [33]. Moreover, the right choice of extraction method can improve the extract’s quality and prevent loss of the target compounds. Since extraction is the crucial stage in obtaining the desired compounds from C. asiatica, here we review the methods used for the extraction of C. asiatica. This review covers the utilization of both simple technologies and advanced extraction techniques that have been reported to get the best yield from the herb, and the in vitro and in vivo antimicrobial activity of the extracts from the selected extraction methods. Finally, the safety and toxicology of the C. asiatica extract are further discussed for a better understanding. 2. Techniques for Extracting Bioactive Compounds from Centella asiatica Extraction is the separation of medicinally active compounds from plant parts using selective solvents through chosen procedures, leaving behind the insoluble compounds. The extracts after solvent removal are obtained in liquid form or as a dry powder consisting of a mixture of compounds. In general, the extraction of plant material is done by washing out the analyte from the matrix into the solvent and diffusion through the cell wall [34]. Over the five years from 2016 to 2020, many papers on the extraction of C. asiatica have been published in Scopus, mainly from India (305), China (191), Malaysia (153), Indonesia (80), and Thailand (66). Figure 1 shows that the number of publications on C. asiatica extraction has increased during the period, thus proving the interest in this plant. Several methods for preparing C. asiatica extracts have been reported, for example, the one developed by Duval [35], who used at least 30 wt.% of an alcoholic solvent in the extraction to obtain a mixture of madecassoside and terminoloside in a refined extract of C. asiatica. Kim et al. [36] patented a method to obtain a water-soluble extract of asiaticoside and madecassoside from C. asiatica; Loiseau et al. [37] reported a method to obtain an extract consisting of a mixture of madecassoside, asiaticoside, and terminoloside that was more than 75 wt.% relative to the extract total weight, and an extract consisting of a mixture of madecassoside and terminoloside that was more than 95 wt.% pure relative to the total weight of the mixture. extract consisting of a mixture of madecassoside, asiaticoside, and terminoloside that was more than 75 wt.% relative to the extract total weight, and an extract consisting of a mixof 24 ture of madecassoside and terminoloside that was more than 95 wt.% pure relative to3the total weight of the mixture. Antibiotics 2021, 10, 457 Figure1.1.Trend Trendofofpublications publicationson onCentella Centellaasiatica asiaticaextraction extractionfrom from2016–2020 2016–2020ininScopus. Scopus. Figure Usually, extraction of C. asiatica is done for its phenolic and flavonoid compounds. Usually, extraction of C. asiatica is done for its phenolic and flavonoid compounds. A A wide range of technologies has been used until now to yield a high-quality extract of wide range of technologies has been used until now to yield a high-quality extract of C. asiatica at moderate cost and with a shorter extraction time. Selection of the extraction C. asiatica at moderate cost and with a shorter extraction time. Selection of the extraction technique depends on the economic feasibility and suitability of the process for the target technique depends on the economic feasibility and suitability of the process for the target compounds. Since the amount of bioactive compounds in C. asiatica are relatively small, compounds. Since the amount of bioactive compounds in C. asiatica are relatively small, the extraction method must be chosen carefully to obtain the desired compound from the the extraction method must be chosen carefully to obtain the desired compound from the herb. The types of technique utilized to extract bioactive compounds/essential oil from herb. The types of technique utilized to extract bioactive compounds/essential oil from C. asiatica are shown in Table 1. C. asiatica are shown in Table 1. Table 1. Methods for the extraction of compounds in C. asiatica. Extraction Method Extraction Time Sample Type Solvent Type Sample to Solvent Ratio (g/mL) Additional Information Compounds Extracted References 6h Dried 90% methanol 10:100 66 ◦ C Phenolics, saponins [38] 24 h Fresh Dried Ethanol 5:25 0.5:25 Room temperature Saponins [39] 5h Dried Water, ethanol, methanol 3:100 Solvent boiling point Asiaticoside, asiatic acid [40] Madecassoside, asiaticoside, asiatic acid, madecassic acid [6] 120 min Dried 95% ethanol 20:100 60 ◦ C 30–90 min Dried Ethanol–water 1:20 30–60 ◦ C Polyphenols, carotenoids [16] [41] 72 h Dried Distilled water 100:1000 - Gluconic acid, ferulic acid, kaempferol, chlorogenic acid, asiatic acid 24 h Fresh/dried 80% ethanol–water 1:20 Room temperature Phenolics, triterpene saponins [42] - Dried Methanol, petroleum ether, chloroform - 60–80 ◦ C Triterpenoids, saponins, tannins, flavonoids [7] Room temperature Valine, triparanol, butamben, neuraminic acid, aesculin, esculetin, famciclovir, isocitretin, rhoifoline, gentiopicrin, urocortisone, pelargonic acid, gabapentin, ivermectin, sarmentoside, khivorin [43] Maceration 24 h Powder Methanol 2:100 Antibiotics 2021, 10, 457 4 of 24 Table 1. Cont. Extraction Method Distillation Extraction Time Sample Type Solvent Type Sample to Solvent Ratio (g/mL) Additional Information Compounds Extracted 75 min Dried Xylene 0.4:100 - Essential oil, α-caryophyllene, germacrene D [44] 3h Fresh Water - - Essential oil [45] 4h Dried Water - - Essential oil [32] - Phenolics, flavonoids, ascorbic acid [46] [47] 12–24 h Soxhlet Ultrasoundassisted extraction (UAE) Microwaveassisted extraction (MAE) Dried Methanol - References 8h Dried Ethanol 500:125 - Saponins, terpenoids, alkaloids, and phenols but no steroids, flavonoids, tannins, proteins, carbohydrates, or glycosides 1h Dried Methanol– water (9:1) 2:50 - Asiatic acid, asiaticoside, madecassoside [48] 8h Dried Methanol 1:100 - Asiaticoside, madecassic acid, madecassoside, asiatic acid [49] 6h Dried Methanol - 60 ◦ C Alkaloids, phenols, tannins, flavonoids, terpenoids, and saponins [10] - Dried Ethanol–water (1:1) 1:10 45 ± 2 ◦ C Total polyphenols, flavonoids, β-carotene, tannins, and vitamin C [50] 1h Dried Methanol– water (9:1) 2:50 - Asiatic acid, asiaticoside, madecassoside [48] 20 min Dried Water 0.6:50 125 W Asiatic acid [51] Alkaloids, flavonoids, saponins, terpenoids [52] 5h Dried Ethyl acetate–water (99:1) 6:150 Frequency: 40 kHz; temperature: 70 ◦ C; power: 216 W 3 × 10 min Dried Methanol– water (9:1) 1:10 - Asiatic acid, asiaticoside, madecassoside, madecassic acid [53] 20 min Dried Methanol– water (9:1) 1:25 - Asiatic acid, asiaticoside, madecassoside [48] 6 min Dried Ethanol 1:25 Microwave power: 50%; 40% Phenolics, triterpenoids Flavonoids [54] 5 min with 2 min pauses Dried Methanol– water (9:1) 10:100 Microwave power: 100% Phenolics, saponins [38] 20 min Fresh Dried Ethanol 10:50 1:50 Atmospheric Triterpene saponins, TPC [39] 20, 40, 60 kPa Triterpene saponins [39] Vacuum microwaveassisted extraction (VMAE) 20 min Solvent-free microwave extraction (SFME) 15 min Fresh - - Microwave power: 300 W Asiaticoside [55] Enzymatic pretreatment microwave extraction (EPME) 110 s Dried 3% cellulase solution 3:108 Enzymolysis 30 min, 45 ◦ C Asiaticoside [56] Subcritical water extraction 5h Dried Deionized water - 250 ◦ C, 40 MPa Asiatic acid, asiaticoside [40] 10:50 Fresh Dried Ethanol 1:50 2.1. Maceration Maceration is a technique in which plant materials are soaked in a solvent at a specific temperature and time [57]. The extract from this method is concentrated using a rotary evaporator to obtain the final solvent-free crude extract. This process softens the plant cells and eventually releases the bioactive compounds from the cells. The solvent used for maceration is based on the study objectives. Maceration is a simple extraction method and Antibiotics 2021, 10, 457 Maceration is a technique in which plant materials are soaked in a solvent at a specific temperature and time [57]. The extract from this method is concentrated using a rotary evaporator to obtain the final solvent-free crude extract. This process softens the plant cells and eventually releases the bioactive compounds from the cells. The solvent used for 5 of 24 maceration is based on the study objectives. Maceration is a simple extraction method and has broad applicability. However, it has a long extraction time, high temperature, high mass transfer resistance, low extraction efficiency, and requires a large volume of solvent [57,58]. has broad applicability. However, it has a long extraction time, high temperature, high mass In the resistance, case of C. asiatica, organicefficiency, solvents like or a mixture of alcotransfer low extraction andethanol, requiresmethanol, a large volume of solvent [57,58]. hol andInwater are typically used. Aqueous extracts obtained by this technique usually the case of C. asiatica, organic solvents like ethanol, methanol, or a mixture of show antioxidant andare cytotoxic activity [6,59]. Maceration has also been used in cosmetics alcohol and water typically used. Aqueous extracts obtained by this technique usually manufacturing, in which propylene glycol[6,59]. and water are used solvents, and the leaves show antioxidant and cytotoxic activity Maceration hasasalso been used in cosmetics and stalks of C. asiatica are extracted a few days [60]. a study by Monton al.leaves [6] manufacturing, in which propylenefor glycol and water areInused as solvents, andet the using maceration, the highest amounts of madecassoside, asiaticoside, madecassic acid, and stalks of C. asiatica are extracted for a few days [60]. In a study by Monton et al. [6] and asiatic acid (0.855%, 0.174%, amounts 0.053%, and 0.025%, respectively) were extracted at opusing maceration, the highest of madecassoside, asiaticoside, madecassic acid, timal of (0.855%, 60 °C and 120 min extraction (Figure 2). In a study cold at andconditions asiatic acid 0.174%, 0.053%, and time 0.025%, respectively) were using extracted maceration, Pittella et al. obtained and flavonoid constituents C. asiatoptimal conditions of [59] 60 ◦ C and 120phenolic min extraction time (Figure 2). Infrom a study using ica.cold Maceration of C.Pittella asiaticaetisal. also able to extract various types of compounds such from as maceration, [59] obtained phenolic and flavonoid constituents triterpenoids, flavonoids, of phenolics, alkaloids, tannins, andtypes carotenoids, priC. asiatica. Maceration C. asiaticasaponins, is also able to extract various of compounds marily on the solvents used and period of extraction [7,42,43]. such based as triterpenoids, flavonoids, phenolics, saponins, alkaloids, tannins, and carotenoids, primarily based on the solvents used and period of extraction [7,42,43]. Figure 2. Triterpenoids found in Centella asiatica extract using the maceration method of Monton et al. [6]. Figure 2. Triterpenoids found in Centella asiatica extract using the maceration method of Monton et al. 2.2. [6]. Distillation Distillation is the separation of components at a particular boiling point and condensation. There are two types of distillation used in extraction: steam distillation that is performed by passing dry steam through the plant material, and water distillation in which elevated pressure is used with plants whose essential oil is difficult to extract at a higher temperature [61]. Steam distillation using distilled water and vinegar has been used to extract dry and fresh leaves of C. asiatica. Steam distillation is an efficient technique for obtaining the best quality of oil, and by employing fresh leaves over dry leaves in extraction, many constituents can be detected [44]. The essential oil of C. asiatica obtained by Florczak [44] from steam distillation yielded more sesquiterpenoid hydrocarbons; 43 constituents were identified, representing 98.60% of the composition of the oil. On the other hand, water distillation is an excellent method for extracting caryophyllene and monoterpenoid hydrocarbons from C. asiatica; 54 constituents were identified, representing 98.29% of the Antibiotics 2021, 10, 457 fresh leaves of C. asiatica. Steam distillation is an efficient technique for obtaining the best quality of oil, and by employing fresh leaves over dry leaves in extraction, many constituents can be detected [44]. The essential oil of C. asiatica obtained by Florczak [44] from steam distillation yielded more sesquiterpenoid hydrocarbons; 43 constituents were identified, representing 98.60% of the composition of the oil. On the other hand, water distil6 of 24 lation is an excellent method for extracting caryophyllene and monoterpenoid hydrocarbons from C. asiatica; 54 constituents were identified, representing 98.29% of the total composition. However, water distillation utilizes a tremendous amount of water besides contotal composition. utilizes amount of water suming a lot of energyHowever, and time water [62]. Adistillation study by Orhan et aal.tremendous [63] successfully extracted consuming a lot of energy and [62]. A study Orhan etcompound al. [63] successfully 47 besides components representing 88.9% of thetime essential oil. The by dominant was αextracted 47 components representing 88.9% of the essential oil. The dominant compound copaene (22.0%), followed by alloaromadendrene (7.6%), β-caryophyllene (7.1%), α-huwas α-copaene (22.0%), followed by alloaromadendrene (7.6%), β-caryophyllene (7.1%), mulene (6.7%), and β-cubebene (5.9%). Paudel et al. [32] discovered that distillation of 85 α-humulene (6.7%), and β-cubebene (5.9%). Paudel et al. [32] discovered that distillation g of dry C. asiatica leaves for 4 h extracted a 0.05% yield of essential oil composed of 33 of 85 g of dry asiatica leaves 4 hinextracted a 0.05% yield of essential oiland composed compounds. TheC.essential oil wasfor rich sesquiterpene hydrocarbons (74.1%) oxyof 33 compounds. The essential was rich abundant in sesquiterpene hydrocarbons (74.1%) and genated sesquiterpenoids (13.0%),oil the most compounds being β-farnesene oxygenated sesquiterpenoids (13.0%), the most abundant compounds(8.8%) being(Figure β-farnesene (26.5%), α-humulene (20.9%), β-caryophyllene (13.3%), and falcarinone 3). (26.5%), α-humulene (20.9%), β-caryophyllene (13.3%), and falcarinone (8.8%) (Figure 3). Figure 3. Compounds obtained from of Paudel Paudelet etal. al.[32]. Figure 3. Compounds obtained fromCentella Centellaasiatica asiaticausing usingthe thedistillation distillation method method of [32]. 2.3. Soxhlet Extraction Soxhlet extraction is a technique used to obtain semi-volatile and non-volatile compounds from C. asiatica [64]. In this method, the herb is placed in a thimble or porous bag in the Soxhlet rig. The extracting solvent in a round-bottomed flask is boiled at the desired temperature and its vapors are condensed in a condenser. The cooled vapor drips onto the sample of herbs and extracts by contact. When the liquid in the thimble rises to the overflow level, a siphon aspirates the solution into the round-bottomed flask. This cycle is continued for several hours until an adequate phytochemical is acquired. The solvent mixture is then concentrated by a rotary evaporator. Since the sample is frequently exposed to the solvent, and the temperature of the extraction is higher than room temperature, more analytes can be extracted from the sample. Additionally, no filtration is required. The downsides of this method are that it involves a long extraction time, and a large amount of solvent is consumed, which is expensive to dispose of and can cause environmental pollution [64]. In terms of acquiring a volatile compound, the Soxhlet method has been found to be terrible for extraction [44]. The long extraction time and high temperature increase the probability of thermolabile substances being degraded [34]. It has also become Antibiotics 2021, 10, 457 7 of 24 an unattractive method for analyzing a high number of samples due to its long extraction time, and samples can only be extracted one at a time for each apparatus. Soxhlet extraction has also been used to obtain the crude extract of C. asiatica, which will be later screened for its antimicrobial activity. Several organic solvents such as hexane, chloroform, methanol [17], petroleum ether, acetone [65], and water [11] have been used in the extraction for this purpose. Byakodi et al. [10] discovered that the methanolic extract of C. asiatica from the Soxhlet method contained phenols, tannins, flavonoids, terpenoids, saponin, and alkaloids. Another study by Thamarai Selvi et al. [47] found that 500 g of powdered C. asiatica subjected to Soxhlet extraction for 8 h using an ethanol to solid ratio of 1:4 resulted in extracts containing saponins, terpenoids, alkaloids, and phenols but no steroids, flavonoids, tannins, proteins, carbohydrates, or glycosides. In a preliminary phytochemical screening of the C. asiatica extract, Jayaprakash and Nagarajan [65] discovered the existence of alkaloids, saponins, flavonoids, phenols, steroids, tannins, glycosides, triterpenoids, and terpenoids. The extract also contained 1–8% saponins, 0.1% volatile oils, triterpenic acids (e.g., terminolic acid, brahmic acid, centellic acid, madasiatic acid), and glycosides (e.g., madasiaticoside, brahminoside, centelloside) [60]. Rahman et al. [50] used 100% ethanol, 50% ethanol, and water as solvents for Soxhlet extraction to obtain total polyphenols, flavonoids, β-carotene, tannins, and vitamin C from C. asiatica. The study showed that the 50% ethanol extract of C. asiatica contained a significantly higher Antibiotics 2021, 10, x FOR PEER REVIEW 9 of 26 amount of polyphenols and flavonoids while 100% ethanol extracted the highest amount of β-carotene and tannins. On the other hand, the water extract of C. asiatica contained more vitamin C than the 50 and 100% ethanol extracts (Figure 4). Figure 50% ethanol, ethanol, and and water waterin Figure4.4.Compounds Compoundsextracted extracted from from Centella Centella asiatica asiatica using 100% ethanol, 50% in the Soxhlet extraction method Rahman [50]. the Soxhlet extraction method byby Rahman et et al.al. [50]. 2.4.Ultrasound-Assisted Ultrasound-AssistedExtraction Extraction 2.4. Ultrasound-assistedextraction extraction(UAE) (UAE)isisan an extraction extraction method method in in which which ultrasonic ultrasonic Ultrasound-assisted waves produce acoustic cavitation in the solvent and cause the disruption of cells. This waves produce acoustic cavitation in the solvent and cause the disruption of cells. This disruption promotes the release of bioactive compounds and enhances the contact surface area between solid and liquid phases [66–68]. Ultrasound causes surface exfoliation, abrasion, and particle disintegration, which increase the mass transfer from the cell cytoplasm to the surrounding solvent [51]. However, frequencies of more than 20 kHz may affect the Antibiotics 2021, 10, 457 8 of 24 disruption promotes the release of bioactive compounds and enhances the contact surface area between solid and liquid phases [66–68]. Ultrasound causes surface exfoliation, abrasion, and particle disintegration, which increase the mass transfer from the cell cytoplasm to the surrounding solvent [51]. However, frequencies of more than 20 kHz may affect the bioactive compounds through the formation of free radicals [18]. Although this method has a short extraction time and low solvent usage, a minimal sample size is needed to achieve better extraction efficiency [51]. This method is the best approach to recover a high yield of phenolic and flavonoid compounds with the highest antioxidant activity such as betacyanin and anthocyanin, and lipids and protein [69]. It is also highly efficient and causes less destruction of the bioactive compounds since elevated temperature is not used [70]. Thus, it is applicable for extracting thermolabile and unstable compounds [34]. Sellathoroe et al. [52] found that UAE gives a higher yield than both Soxhlet and maceration methods for an extract consisting of 0.34% terpenoids, 0.47% saponins, 0.03% alkaloids, and 0.11% flavonoids, as shown in Figure 5 [52]. A study by Nithyanandam et al. [71] demonstrated that UAE is the best extraction method for C. asiatica for the significant recovery of antioxidant compounds. Furthermore, the extract contained 79% 1,1-diphenyl2-picrylhydrazyl (DPPH) scavenging activity, a total phenolic content (TPC) of 1350 mg Antibiotics 2021, 10, x FOR PEER REVIEW 10 of 26 GAE/100 g DW, and total flavonoid content (TFC) of 599 mg CE/100 g DW, all of which were higher than those obtained using maceration, Soxhlet, and hot-water extraction. This method also extracted 9.24 ± 0.88% of bioactive compounds, which increased to 11.80 ± 0.58% using focused high ultrasound, of which 30.21% was polyphenols [72]. The diffusion and solubility of the solvent were increased by using a 120 kHz ultrasonic wave diffusion and solubility of the solvent were increased by using a 120 kHz ultrasonic wave energy or higher, thus causing more polyphenols to be extracted [72]. energy or higher, thus causing more polyphenols to be extracted [72]. Figure 5.5. Bioactive usingthe theultrasonic ultrasonicextraction extraction method of Sellathoroe Figure Bioactivecompounds compoundsextracted extracted using method of Sellathoroe et al.et[52]. al. [52]. Methanol extract of C. asiatica leaf has more constituents (alkaloids, terpenoids, glyMethanol extract of C. asiatica leaf has constituents terpenoids, glycosides, steroids, tannins, flavonoids, andmore reducing sugars) (alkaloids, than the acetone extract (no cosides, steroids, tannins, flavonoids, and reducing sugars)UAE thanusing the acetone extract (no terpenoids) and chloroform extract (no tannins). However, those three solvents terpenoids) chloroform (nostudy tannins). However, UAE using three solis unable toand extract saponinsextract [73]. The by Nithyanandam et al. [71]those also showed that vents is unable extract [73]. binary The study by Nithyanandam et al. [71] a better extractto yield was saponins obtained with solvent (40:60 ethanol–water) thanalso with showed that a better extract yield was obtained with binary solvent (40:60 ethanol–water) than with pure solvent (100% water and 100% ethanol). Increased yield of TPC and TFC, and DPPH antioxidant activity was proportional to an increase in temperature, but above 45 °C, DPPH scavenging activity decreased drastically while TPC and TFC yield was maintained [71]. The high recovery of phenolic and flavonoid compounds using the UAE Antibiotics 2021, 10, 457 9 of 24 pure solvent (100% water and 100% ethanol). Increased yield of TPC and TFC, and DPPH antioxidant activity was proportional to an increase in temperature, but above 45 ◦ C, DPPH scavenging activity decreased drastically while TPC and TFC yield was maintained [71]. The high recovery of phenolic and flavonoid compounds using the UAE technique causes a high percentage of DPPH scavenging activity, which is an important assay to test antioxidant properties. The disadvantage of this technique is the long exposure of the sample to ultrasound activity, which may degrade the yield of asiaticoside and asiatic acid [51]. 2.5. Microwave Extraction Microwave extraction is a technique that utilizes fast heating of aqueous samples. The key concept of the method is that the solvent absorbs the microwave energy, which is then transferred in the form of heat to the sample [48]. The energy is transferred into the solvent by the twin mechanism of ionic conduction and dipole rotation [69]. Usually, a solvent such as methanol or water mixture is used for polar compounds and hexane is used for non-polar compounds. Extraction of active compounds from the sample to solvent is influenced predominantly by the temperature and the nature of the solvent. Contrary to classical heating, microwaves heat the whole sample concurrently. Microwaveassisted extraction (MAE) is a process that operates under higher temperature and in an oxygen-rich environment. It is an excellent technique for reducing the time and solvent consumed to extract target compounds compared to conventional methods [48,56]. This method has a higher extraction rate and produces better products but at a lower cost. Many studies have already proven that it is a feasible option to conventional techniques for many kinds of samples. Shen et al. [48] found for extraction of C. asiatica by MAE that using 90% methanol for 20 min was the optimal condition to extract the highest yield. In another study by Desai et al. [38], the extracts yielded by the MAE method had a 26% increase of TPC and 8% increase of saponin content compared to the traditional solvent extraction. However, the extracts did not exhibit anti-inflammatory activity. A study by Sen et al. [54] using MAE found that it was a much better method for obtaining TPC and triterpenoids than Soxhlet: the extraction time was only 6 min compared to 36 h by Soxhlet, but also the microwave produced 200 times less carbon load. Figure 6 shows the difference in compound concentrations obtained using MAE and Soxhlet extraction. Increasing the microwave power from 170 to 425 W also increased the TPC from 20 to 50%, due to the rapid transfer of electromagnetic energy in the microwave, thus simultaneously heating the plant sample. This situation causes internal stress inside the plant and eventually bursts the cell wall, releasing its contents [54]. Regardless of the benefits, the interaction between the sample and oxygen during MAE operated at a very high temperature may cause a destructive effect on the desired bioactive compounds, especially if they are oxygen- and heat-sensitive [39,74]. Hence, vacuum microwave-assisted extraction (VMAE) has been developed as an option to acquire that type of bioactive compounds [39]. It is a method in which the boiling temperature of an extraction solvent is lowered, consequently reducing the extraction temperature. The pressure is also reduced so that less oxygen is present for the unfavorable process of oxidation. Fresh and dried leaves of C. asiatica were treated with VMAE to extract triterpene saponins and phenolics at various pressures. It was found that the fresh leaves of C. asiatica contained the maximum triterpene saponins when VMAE was performed at 60 kPa, while the maximum TPC was obtained when extraction was carried out in atmospheric conditions. On the other hand, the dried leaves contained the highest triterpene saponins and TPC when they were extracted at atmospheric pressure: the temperature was higher, which caused greater diffusivity of bioactive compounds and resulted in a more elevated amount of triterpene saponins and TPC. Otherwise, at 20 kPa, the temperature of extraction was reduced (~40 ◦ C), thus minimizing diffusion of the bioactive compounds [39]. Antibiotics 2021, 10, 457 produced 200 times less carbon load. Figure 6 shows the difference in compound concentrations obtained using MAE and Soxhlet extraction. Increasing the microwave power from 170 to 425 W also increased the TPC from 20 to 50%, due to the rapid transfer of electromagnetic energy in the microwave, thus simultaneously heating the plant sample. of 24 This situation causes internal stress inside the plant and eventually bursts the cell10wall, releasing its contents [54]. Figure6.6.Difference Differenceofofcompound compoundconcentrations concentrationsextracted extractedfrom fromCentella Centellaasiatica asiaticausing usingmicrowave microwave Figure and Soxhlet extraction methods by Sen et al. [54]. and Soxhlet extraction methods by Sen et al. [54]. MAE has also beenbenefits, improved using solvent-free extraction (SFME). The Regardless of the thebyinteraction betweenmicrowave the sample and oxygen during method utilized a vacuum and stirrer during extraction and operated at 300 W for 15 min MAE operated at a very high temperature may cause a destructive effect on the desired to extractcompounds, 158 µg/mL especially of asiaticoside from C. asiatica. Solventless extraction thevacuse bioactive if they are oxygenand heat-sensitive [39,74].saves Hence, of organic solvents, and the extraction is performed in a shorter time due to direct heat uum microwave-assisted extraction (VMAE) has been developed as an option to acquire radiation to the samples, which results in rapid extraction [55]. that type of bioactive compounds [39]. It is a method in which the boiling temperature of A study by Wang et al. [56] on enzymatic pretreatment and microwave extraction an extraction solvent is lowered, consequently reducing the extraction temperature. The (EPME), combining enzymolysis and microwave extraction, showed that it has high expressure is also reduced so that less oxygen is present for the unfavorable process of oxitraction efficiency, takes less time, and is an environmentally friendly way of extracting dation. Fresh and dried leaves of C. asiatica were treated with VMAE to extract triterpene asiaticoside from C. asiatica. However, the downside of this technique is that it is expensive saponins and phenolics at various pressures. It was found that the fresh leaves of C. because enzymes are used, and the complexity of industrializing EPME would restrict further application. 2.6. Subcritical Water Extraction Subcritical water extraction is a method that modifies the physical properties of water under high pressure and increases the temperature above its boiling point (up to 374 ◦ C) to maintain the water in its liquid state, thus improving it as an extraction solvent. Subcritical water extraction is an efficient, harmless, and eco-friendly method for extracting polar compounds from samples, which is an excellent substitute for conventional organic solvent extraction methods. Furthermore, this method requires less extraction time and solvent, but acquires a higher quality of extracts [40]. There are not many studies on the compounds, in vitro analysis, or in vivo analysis of C. asiatica obtained by subcritical water extraction. A study by Kim et al. [40] on the extraction of C. asiatica with the subcritical water extraction method demonstrated a rise in the concentration of asiatic acid from 0 to 7.0 mg/g, and in that of asiaticoside from 1.1 to 8.4 mg/g, when the temperature was increased from 100 to 250 ◦ C (Figure 7). The enhanced solubility and enhanced transport properties of subcritical water at a higher temperature may be responsible for the higher extraction yields of asiatic acid and asiaticoside. On the other hand, the outcome was not heavily reliant on the pressure of the extraction process because there was little increment in the yield of asiaticoside (from 4.6 to 8.1 mg/g) or asiatic acid (from 2.4 to 3.4 mg/g) on increasing the pressure from 10 to 40 MPa. Regarding the Antibiotics 2021, 10, 457 11 of 24 Antibiotics 2021, 10, x FOR PEER REVIEW 13 of 26 effect of pressure on the extraction, there were not many changes toward the polarity of the subcritical water. Hence, subcritical water pressure lacks influence on the extraction yield. Figure7.7.Compounds Compounds obtained from Centella asiatica using the subcritical water extraction method Figure obtained from Centella asiatica using the subcritical water extraction method ◦ C. Kimetetal.al.[40] [40]atat100 100and and250 250°C. byby Kim 3. Antimicrobial Activity 3. Antimicrobial Activity Plant-based antimicrobials are often studied and used in medicine as they have fewer Plant-based antimicrobials are often studied and used in medicine as they have fewer side effects compared to synthetic antimicrobials. C. asiatica is one of the plants that have side effects compared to synthetic antimicrobials. C. asiatica is one of the plants that have antimicrobial activity against many types of bacteria and fungi since the triterpenoids antimicrobial activity against many types of bacteria and fungi since the triterpenoids in in C. asiatica can be regarded as phytoanticipins due to their selective cytotoxicity and C. asiatica can be regarded as phytoanticipins due to their selective cytotoxicity and proprotective role in preventing infections from a pathogen [75,76]. Flavonoids and tannins tective role in preventing infections from a pathogen [75,76]. Flavonoids and tannins are are known for their antimicrobial activity [65]. An antimicrobial susceptibility test is known for their antimicrobial activity [65]. antimicrobial susceptibility test of is essential essential to determine the efficiency of aAn plant extract against the growth microbes. toBesides determine the efficiency of a plant extract against the growth of microbes. Besides tartargeting the desired compounds in C. asiatica, the extracts obtained from selected geting the desired compounds in C. asiatica, the extracts obtained from selected extraction extraction methods were further analyzed for their antimicrobial properties. Many studies methods were further analyzed their antimicrobial properties. Many studies have have been done on the activityfor of C. asiatica extracts against pathogens. Still, they arebeen hard done on the activity of C. asiatica extracts against pathogens. Still, they are hard to compare to compare depending on the type of extraction method, solvents, strains of microbes, and depending on the of extraction method, solvents, strains of microbes, and antimicroantimicrobial testtype methods used [77]. bial test methods used [77]. 3.1. In Vitro Studies 3.1. In The Vitrodisc Studies diffusion method is widely used as it is suitable for preliminary testing for The disc diffusion method is widely usedextracts as it is suitable preliminary of testing for screening the antimicrobial activity of plant [78]. Thefor susceptibility microbes screening the extract antimicrobial activity of extractsmicrobial [78]. The suspension susceptibility of microbes to to the plant is determined byplant inoculating onto the medium the plant extract is determined inoculating microbial ontoa the medium agar surface and swabbing itbyevenly all over the agarsuspension surface with cotton swab.agar The filter paper discs are dipped with incubated 24–48 swab. h. TheThe inhibition surface and swabbing it evenly all plant over extract the agarand surface with for a cotton filter zonesdiscs are observed, diameter measured, indicated byh. a The clearinhibition area around the paper are dippedand withtheir plant extract and incubated for 24–48 zones filter paper disc. A clear zone of inhibition reveals by theamicrobes’ to an are observed, and their diameter measured, indicated clear areavulnerability around the filter extract, while the absence of such a zone shows the microbes’ resistancetotoan theextract, extract. paper disc. A clear zone of inhibition reveals the microbes’ vulnerability This method is a low-cost straightforward technique to determine theextract. efficiency while the absence of such aand zone shows the microbes’ resistance to the Thisof a plant extract in inhibiting the growth of microbes [79]. For the agar well diffusion method is a low-cost and straightforward technique to determine the efficiency of a plant method, the agar isthe diffused the plant[79]. extract it well is solidified, each plate extract in inhibiting growthwith of microbes For before the agar diffusionand method, the agar is diffused with the plant extract before it is solidified, and each plate is inoculated with microbial culture and spread evenly with a sterile bent glass rod. The agar medium Antibiotics 2021, 10, 457 12 of 24 is inoculated with microbial culture and spread evenly with a sterile bent glass rod. The agar medium is cut with a sterile cork borer, and different concentrations of plant extract solution are loaded by micropipette into the agar well. Plates are incubated for 24–48 h before the zones of inhibition are observed. There is also the determination of the minimum inhibitory concentration (MIC) by the microdilution method using 96-well microtitration plates, the hole plate diffusion method, the two-fold method in a microtiter plate, the Versa Max Tunable microplate reader, and particular MIC methods such as the tetrazolium microplate method, liquid dilution method, and serial dilution method [80]. In terms of MIC, concentrations of extracts are varied in the preferred antimicrobial test to determine the lowest concentration of extract that effectively prevents microbial growth [81]. A study by Zheng et al. [82] using an extract of C. asiatica obtained by the maceration method found that at 2 mg/mL, it inhibited the growth of Helicobacter pylori by the agar well diffusion method; the MIC against the strains tested ranged from 0.125 to 8 mg/mL. Methanol extract obtained by the maceration method showed a broad spectrum of antimicrobial activity using the agar diffusion method against both Gram-positive and -negative bacteria, with the zones of inhibition ranging from 9 to 29 mm and the MIC ranging from 1.25 to >10 mg/mL [76]. However, while maceration of C. asiatica using methanol produced an extract that was unable to inhibit Escherichia coli [83,84], positive inhibition was achieved when using ethanol as the solvent [85,86]. Vadlapudi et al. [17] used the agar well diffusion method to determine the antimicrobial activity of C. asiatica from inhibition zones and MIC values. Methanol extract of C. asiatica from Soxhlet extraction was able to inhibit Aspergillus niger (14–19 mm), Fusarium oxysporum (13–14 mm), Xanthomonas campestris (10–13 mm), Lactobacillus acidophilus (11–13 mm), Pseudomonas marginalis (10–25 mm), Pseudomonas syringae (18–22 mm), Staphylococcus salivarius (13–177 mm), and Staphylococcus aureus (9–12 mm), but exhibited no activity toward Penicillium expansum, Pseudomonas aeruginosa, or Staphylococcus mutans. The MIC found were 0–155 mg/mL. In another study, aqueous extract from maceration exhibited better antimicrobial activity toward Bacillus cereus, E. coli, P. aeruginosa, S. aureus, and S. mutans by the well diffusion method compared to the ethanol extract [87]. The MIC of the aqueous extract was 25 mg/mL for B. cereus (3.00 ± 0.00 mm), E. coli (6.00 ± 0.00 mm), and S. mutans (3.00 ± 0.00 mm). On the other hand, the MIC of the aqueous extract was 50 mg/mL for P. aeruginosa (6.33 ± 0.58 mm) and S. aureus (10.0 ± 0.00 mm). Purkait et al. [88] studied the efficacy of aqueous, methanol, and chloroform extracts of C. asiatica against the fish pathogenic bacteria Aeromonas hydrophila and Edwardsiella tarda using agar disc diffusion, agar overlay well diffusion, and broth dilution assays. All types of C. asiatica extract failed to inhibit A. hydrophila. On the other hand, E. tarda was inhibited by chloroform extract (11.25 ± 0.35 mm) using the agar disc diffusion assay. In the agar overlay well diffusion assay, 50 µL of chloroform and methanol extracts of C. asiatica inhibited E. tarda (30.50 ± 6.40 and 7.50 ± 0.70 mm, respectively) (Figure 8). The crude chloroform extract of C. asiatica using agar overlay well-diffusion assay produced the largest zone of inhibition (30.50 ± 6.40 mm), which was comparable to those of the chloramphenicol (40.75 ± 1.76 mm). In the broth dilution assay, increasing the concentration of crude chloroform C. asiatica extract resulted in increased inhibition of E. tarda growth (Figure 9). It was concluded that the chloroform extract of C. asiatica is the most effective to control E. tarda infection, especially in aquaculture. Antibiotics2021, 2021, 10,457 x FOR PEER REVIEW Antibiotics Antibiotics 2021, 10,10, x FOR PEER REVIEW 15 of 24 26 13 of 15 of 26 Figure8. 8. Inhibition Inhibitionof ofEdwardsiella Edwardsiellatarda tarda by (a)agar agardisc discdiffusion diffusion and agar overlay well diffuFigure and (b)(b) agar overlay well diffusion Figure 8. Inhibition of Edwardsiella tarda byby(a)(a)agar disc diffusion and (b) agar overlay well diffusion assays. A: aqueous control; B: aqueous test; C: chloroform control; D: chloroform test; E: assays. A:A: aqueous control; B: B: aqueous test; C:C: chloroform control; D:D: chloroform test; E:E: methanol sion assays. aqueous control; aqueous test; chloroform control; chloroform test; methanol control; F: methanol test; C 30: chloramphenicol, 30 µg/disc; 10 µL/disc and 50 µL/well methanol F: methanol test;chloramphenicol, C 30: chloramphenicol, 30 µg/disc; 10 µL/disc and 50 µL/well control;control; F: methanol test; C 30: 30 µg/disc; 10 µL/disc and 50 µL/well used for used for agar disc diffusion and agar overlay well diffusion assays, respectively. Figure and capused fordisc agar disc diffusion andoverlay agar overlay well diffusion assays, respectively. Figure and capagar diffusion and agar well diffusion assays, respectively. Figure and caption tion reused from Purkait et al. [88]. Used under the Creative Commons License (http://crea-reused tion reused frometPurkait al. [88]. Used the Creative Commons License (http://creafrom Purkait al. [88].etUsed under theunder Creative tivecommons.org/licenses/by/4.0/ (accessed on 26Commons February License 2021)). (http://creativecommons.org/ tivecommons.org/licenses/by/4.0/ (accessed 26 February 2021)). licenses/by/4.0/ (accessed on 26 Februaryon2021)). Figure 9. Growth inhibition of Edwardsiella tarda by crude Centella asiatica chloroform extract in the Figure 9. 9. Growth inhibition of of Edwardsiella tarda byby crude Centella asiatica chloroform extract in the Figure Growth inhibition Edwardsiella tarda crude Centella asiatica extract in the broth dilution assay. OD: optical density. Figure and caption reused fromchloroform Purkait et al. [88]. Used broth dilution assay. OD: optical density. Figure and caption reused from Purkait et al. [88]. Used broth assay.Commons OD: optical density. Figure and caption reused from Purkait et(accessed al. [88]. Used underdilution the Creative License (http://creativecommons.org/licenses/by/4.0/ on 26 under the Creative Commons License (http://creativecommons.org/licenses/by/4.0/ (accessed on 26 under the 2021)). Creative Commons License (http://creativecommons.org/licenses/by/4.0/ (accessed on February February 2021)). 26 February 2021)). Methanol extract of C. asiatica from Soxhlet extraction was found to inhibit various Methanol extract of of C.C. asiatica from Soxhlet extraction was found totoinhibit various Methanol asiatica from Soxhlet extraction was inhibit types of fungi extract (e.g., Aspergillus niger, Alternaria alternata, and F.found oxysporum) and various bacteria types of fungi (e.g., Aspergillus niger, Alternaria alternata, and F. oxysporum) and bacteria types of syringae, fungi (e.g., Aspergillus niger, Alternaria alternata, and F. oxysporum) and bacteria (e.g., P. S. aureus, and Bacillus subtilis) [17,65,89], but petroleum ether extract did (e.g., P.P.syringae, S.S. aureus, and Bacillus subtilis) [17,65,89], but petroleum ether extract did (e.g., syringae, aureus, and Bacillus subtilis) [17,65,89], petroleum ether extract not show any antimicrobial activity against the microbesbut tested [65]. Byakodi et al. did [10] not show any antimicrobial activity against the microbes microbestested tested[65]. [65].Byakodi Byakodi al. [10] not any antimicrobial activity against the et et al.antimicro[10] also alsoshow found that a methanolic extract obtained by the Soxhlet method exhibited also found that a methanolic extract obtained by the Soxhletmethod methodexhibited exhibitedantimicrobial antimicrofound that a methanolic extract obtained by the Soxhlet bial activity against Gram-positive and -negative strains of bacteria. Ethanolic extract obbial activity against Gram-positive -negative strains of bacteria. Ethanolic extract obactivity against Gram-positive andand -negative strains of bacteria. Ethanolic extract obtained tained by the Soxhlet method was found to exhibit better antimicrobial activity toward E. tained by the Soxhlet method was found to exhibit better antimicrobial activity toward by Soxhlet method wasβ-lactamase-producing found to exhibit better antimicrobial toward E. E. coli colithe(extended-spectrum (ESBL) and activity carbapenem-resistant coli (extended-spectrum β-lactamase-producing (ESBL) and carbapenem-resistant (extended-spectrum β-lactamase-producing (ESBL) and carbapenem-resistant strains), strains), Klebsiella pneumoniae (carbapenem-resistant strains), and P. aeruginosa (carstrains), Klebsiella pneumoniae (carbapenem-resistant strains), and (carbapenem-resistant P. aeruginosa (carKlebsiella pneumoniae (carbapenem-resistant strains), and P. aeruginosa bapenem-resistant strains) in contrast to the petroleum ether extract, which was only effibapenem-resistant strains) inpetroleum contrast toether the petroleum etherwas extract, which was effistrains) contrast to the extract,However, which only efficient inonly inhibiting cient in in inhibiting K. pneumoniae (ESBL strains). both extracts were unable to cient in inhibiting K. pneumoniae (ESBL strains). However, both extracts were unable to K. pneumoniae (ESBL strains). However, both extracts were unable to inhibit the growth inhibit the growth of Acinetobacter baumannii [90]. ESBL-producing Enterobacteriaceae inhibit the growthbaumannii of Acinetobacter baumannii [90]. Enterobacteriaceae ESBL-producing Enterobacteriaceae of Acinetobacter [90]. ESBL-producing cause infections in cause infections in the urinary tract, and drug-resistant infections were found to cause cause infections in and the urinary tract, and drug-resistant infections the urinary tract, drug-resistant infections were found to causewere 9000found deaths,toofcause which 9000 deaths, of which 600 were due to carbapenem-resistant K. pneumoniae and E. coli 9000 deaths, of which 600 were due to carbapenem-resistant K. pneumoniae and E. coli 600 were due to carbapenem-resistant K. pneumoniae and E. coli [90,91]. Therefore, new Antibiotics 2021, 10, 457 14 of 24 antibacterial agents are needed to counter the pathogen, and the extract of C. asiatica could be one of them. Mostly, extraction of C. asiatica to determine the antimicrobial activity of the extract is done by a conventional method such as maceration or Soxhlet. Still, several studies have used methods such as UAE and MAE: extracts of C. asiatica obtained using these methods also showed positive inhibition of the growth of studied bacteria. UAE of C. asiatica produced an extract that could inhibit P. aeruginosa (6.5 mm zone of inhibition) and E. coli (8.5 mm) [52]. Similarly, Sellathoroe et al. [52] found that the extract from UAE was the best for inhibiting the growth of Gram-negative bacteria. At a concentration of 100 µg/mL, methanol extract of C. asiatica leaves obtained by UAE showed the largest zone of inhibition against E. coli (30 mm), followed by B. cereus (29 mm), P. aeruginosa, and S. aureus (both 28 mm) [73]. Asiaticoside and asiatic acid are prominent bioactive compounds in C. asiatica leaves; these compounds display efficacy against Gram-negative bacteria such as S. aureus and E. coli [92]. In general, the methanolic extract of C. asiatica shows a better inhibitory effect than acetone, chloroform, and water extracts [65,73,77]. The extract is also rich in compounds like terpenoids, saponins, phenols, flavonoids, and tannins that consequently contribute to better antimicrobial activity [10]. Antimicrobial activity tests also indicate that C. asiatica is a good source of an antimicrobial agent against a wide range of microbes. Details of the effectiveness of C. asiatica extracts obtained by several different extraction methods toward various types of microbes are shown in Table 2. Table 2. Antimicrobial activity of C. asiatica extracts obtained by various methods. Extraction Method Solvent Antimicrobial Method Microbes Effect References Maceration Methanol, water Open hole diffusion, 2-fold dilution method Bacillus subtilis Escherichia coli Aeromonas hydrophila Citrobacter freundii + − − − [83] Maceration Ethanol Agar diffusion Bacillus cereus Listeria monocytogenes + + [93] + + + + − − − + [85] Maceration Ethanol Disc diffusion Escherichia coli Bacillus subtilis Vibrio cholerae Shigella sonnei Bacillus cereus Shigella dysenteriae Staphylococcus aureus Salmonella paratyphi Maceration Ethanol Disc diffusion Staphylococcus aureus + [92] Disc diffusion, agar well diffusion Broth dilution, agar well diffusion, disc diffusion Aeromonas hydrophila Edwardsiella tarda Aeromonas hydrophila Edwardsiella tarda − − + + [88] − − − − [94] Water, methanol Maceration Chloroform Maceration Water Agar diffusion, Disc diffusion Salmonella enterica Shigella flexneri Escherichia coli Enterobacter cloacae Maceration Ethanol Agar well diffusion Helicobacter pylori + [82] Micro broth dilution Mycobacterium sp. Staphylococcus aureus Bacillus subtilis Aspergillus niger Candida albicans Escherichia coli + + + + + − [84] Maceration Methanol Antibiotics 2021, 10, 457 15 of 24 Table 2. Cont. Extraction Method Solvent Antimicrobial Method Microbes Effect References Micro broth dilution Bacillus cereus Serratia sp. Rhodotorula mucilaginosa Aspergillus flavus Penicillium citrinum Bacillus cereus Serratia sp. Rhodotorula mucilaginosa Aspergillus flavus Penicillium citrinum Bacillus cereus Serratia sp. Rhodotorula mucilaginosa Aspergillus flavus Penicillium citrinum + − − − − + + + + + + − + + + [77] + + + + + [81] [42] Acetone Maceration Methanol Ethanol Maceration Dichloromethane:methanol Disc diffusion, micro broth dilution Escherichia coli Salmonella typhi Bacillus subtilis Staphylococcus aureus Shigella sonnei Maceration Ethanol aqueous Disc diffusion, agar dilution Staphylococcus aureus + Pseudomonas aeruginosa Staphylococcus aureus Streptococcus agalactiae Bacillus cereus Enterococcus hirae Enterococcus faecalis (clinical isolate) Enterococcus gallinarum Escherichia coli + + + + + Streptococcus pyogenes Pseudomonas aeruginosa Escherichia coli Staphylococcus aureus Staphylococcus albus Streptococcus pneumoniae Candida albicans Microsporum boulardii Aspergillus niger Aspergillus flavus Streptococcus pneumoniae Streptococcus pyogenes Pseudomonas aeruginosa Escherichia coli Staphylococcus aureus Staphylococcus albus Escherichia coli Staphylococcus aureus Staphylococcus albus Pseudomonas aeruginosa Streptococcus pyogenesis Streptococcus pneumoniae − − + + + + − − + + − − + + + + + + + + + + Maceration Methanol, acetone, ethyl acetate Aqueous Agar diffusion, microplate dilution assay Disc diffusion Maceration Aqueous Open hole diffusion Chloroform Disc diffusion [76] + + + [86] Antibiotics 2021, 10, 457 16 of 24 Table 2. Cont. Extraction Method Solvent Antimicrobial Method Microbes Effect References Soxhlet Ethanol, methanol Disc diffusion Aspergillus niger Bacillus subtilis + + [89] Disc diffusion Escherichia coli Klebsiella pneumoniae Staphylococcus aureus Streptococcus pyogenes + + + + [11] Agar well diffusion Escherichia coli Staphylococcus aureus Bacillus megaterium Vibrio parahaemolyticus Vibrio mimicus Shigella boydii Bacillus cereus Bacillus subtilis Shigella dysenteriae Salmonella typhi Salmonella Paratyphi Pseudomonas aeruginosa Escherichia coli Sarcina lutea Staphylococcus aureus + + + + + + + + + + + + + + + [95] Disc diffusion Methicillin-resistant Staphylococcus aureus (MRSA) Staphylococcus aureus Klebsiella pneumoniae Pseudomonas aeruginosa Escherichia coli + − − − Micrococcus luteus Staphylococcus aureus Bacillus subtilis Bacillus cereus Escherichia coli Pseudomonas aeruginosa Zymomonas mobilis Aspergillus niger Aspergillus sydouri Trichoderma reesei + + + + + + + + + + [10] Aspergillus niger Penicillium expansum Fusarium oxysporum Xanthomonas campestris Lactobacillus acidophilus Pseudomonas marginalis Pseudomonas syringae Pseudomonas aeruginosa Streptococcus mutans Streptococcus salivarius Staphylococcus aureus + − + + + + + − − + + [17] Soxhlet Soxhlet Soxhlet Water Aqueous Methanol Agar well diffusion Soxhlet Methanol Disc diffusion Soxhlet Methanol Agar well diffusion + [96] Antibiotics 2021, 10, 457 17 of 24 Table 2. Cont. Extraction Method Solvent Antimicrobial Method Microbes Effect References Disc diffusion Proteus mirabilis Streptococcus faecalis Streptococcus pyogenes Escherichia coli Fusarium oxysporum Alternaria alternata Curvularia lunata Staphylococcus aureus Bacillus subtilis Bacillus thuringiensis Enterococcus faecalis Serratia marcescens Pseudomonas aeruginosa Proteus vulgaris Proteus mirabilis Klebsiella pneumoniae Escherichia coli + + + + + + + − − − − − − − − − − [65] Disc diffusion ESBL strains Escherichia coli + Klebsiella pneumoniae − Carbapenem-resistant strains Acinetobacter baumannii − Klebsiella pneumoniae + Pseudomonas aeruginosa + ESBL strains Escherichia coli − Klebsiella pneumoniae + Carbapenem-resistant strains Acinetobacter baumannii − Klebsiella pneumoniae − Pseudomonas aeruginosa − [90] + + + + [73] Methanol Soxhlet Petroleum ether Ethanol Soxhlet Petroleum ether UAE Methanol, acetone, chloroform, water Agar well diffusion Bacillus cereus Escherichia coli Staphylococcus aureus Pseudomonas aeruginosa UAE Methanol Disc diffusion Microbes in fish surimi + [97] MAE Ethanol Disc diffusion Streptococcus mutans Streptococcus mitis Streptococcus pyogenes + + + [98] 3.2. In Vivo Studies Although many studies have been done on the antimicrobial properties of C. asiatica in vitro, they are not a complete test to summarize the antimicrobial activity and analogize the way the extract will act in vivo. There are other circumstances such as first-pass metabolism, microbial defense, drug resistance, and conditions of the patient’s pathology that will also influence the effectiveness of the test [79]. Few antimicrobial studies are done in vivo due to their complexity and expense; not only does the activity against the microbes need to be assessed, but there is also a concern regarding possible allergic reactions and mammalian cell toxicity [80]. Most of the studies of C. asiatica in animal models have focused on wound healing since the plant is famous for its antioxidant and anti-inflammatory activity [21,41,99,100]. Antibiotics 2021, 10, 457 18 of 24 Antibiotics 2021, 10, x FOR PEER REVIEW 20 of 26 Figure 10 shows the effectiveness in vivo of C. asiatica extracts obtained using maceration in reducing H. pylori gastric mucosal colonization in a C57BL/6 mouse model. The optimum concentration was 50 mg/kg after oral administration once daily for three weeks [82]. Figure Effectsof ofCentella Centellaasiatica asiaticaleaf leafextract extract (CAE) against Helicobacter pylori colonization Figure 10. Effects (CAE) against Helicobacter pylori colonization in in C57BL/6 mice.CAE CAEwas was administered orally at 50 mg/kg, for three weeks. C57BL/6 mice. administered orally at 50 andand 250250 mg/kg, onceonce dailydaily for three weeks. The results are expressed as mean ± SEM = 10). < 0.05 with thewith control. The results are expressed as mean ± (n SEM (n =p10). pcompared < 0.05 compared the Figure control.and Figure caption reused fromfrom Zheng et al. [82]. Creative Commons License License (http://creaand caption reused Zheng et al.Used [82]. under Used the under the Creative Commons (http:// tivecommons.org/licenses/by/4.0/ (accessed on 26 February 2021)). creativecommons.org/licenses/by/4.0/ (accessed on 26 February 2021)). There an in in vivo vivo study study of of C. C. asiatica asiatica as as an an anti-acne anti-acnegel, gel,ininwhich whichthe theantianThere was was also also an tibacterial effect on Propionibacterium acne was observed and measured through sebum bacterial effect on Propionibacterium acne was observed and measured through sebum sesecretion. Clinicalassessment assessmentusing usingaaskin skinanalysis analysistool toolon on 12 12 volunteers volunteers showed showed decreased cretion. Clinical decreased symptoms of inflammation, the number of papules, nodules, and pustules, symptoms of inflammation, the number of papules, nodules, and pustules, and and shifts shifts in in sebum levels [101]. sebum levels [101]. The The aqueous aqueous extract extract of of C. C. asiatica asiatica was was used used to to control control the the fish fishdisease disease columnaris columnaris in in Nile tilapia caused by the bacterium Flavobacterium columnare. Fish mortality was decreased Nile tilapia caused by the bacterium Flavobacterium columnare. Fish mortality was dedepending on the doses (0,used 20, 40, and40, 60and mg/L) and atand 100atmg/L of extract, no creased depending on theused doses (0, 20, 60 mg/L) 100 mg/L of extract, mortality or adverse effects were found in the infected fish [102]. no mortality or adverse effects were found in the infected fish [102]. 4. Safety and Toxicology 4. Safety and Toxicology Toxicity is the level of adverse health effects on living organisms from the interaction Toxicity is the level of adverse health effects on living organisms from the interaction between living cells and selective toxicants. A toxicity test is crucial to ensure the safety and between and selective toxicants. A toxicity testshrimp is crucial to ensure the(BSLA) safety efficacy ofliving plantcells extracts. This can be done using the brine lethality assay and efficacy plant extracts. beextracts done using the brine lethalitysalina) assay using differentofconcentrations of This crudecan plant [103,104]. Brine shrimp shrimp (Artemia (BSLA) using different concentrations of crude plant extracts [103,104]. Brine shrimp (Arhave been used in over 90% of studies using Artemia as a test organism since it is simple, temia salina) have been used in over 90% of studies using Artemia as a test organism since inexpensive, rapid, convenient, and requires a small amount of test material. LD50 is it is concentration simple, inexpensive, rapid, convenient, and requires a small amount of test with material. the required to obtain the death of 50% of the test population the LD50 is the concentration required to obtain the death of 50% of the test population with BSLA. An extract is considered toxic if the LD50 is less than 1000 µg/mL, weakly toxic if the BSLA. extractand is considered if the is LD50 is than less than µg/mL, 500 to 1000An µg/mL, non-toxic iftoxic the LD50 more 1000 1000 µg/mL [105].weakly toxic if 500Itto 1000 µg/mL, and non-toxic if the LD50 is more 1000 µg/mL [105]. was found that C. asiatica has cytotoxic activity fromthan 500 µg/mL to above 1000 µg/mL, It was found that C. asiatica has cytotoxic activity from 500 µg/mL toofabove 1000 suggesting that it is weakly or insignificantly cytotoxic [103]. Ethanolic extract C. asiatica µg/mL, suggesting that it is weakly or insignificantly cytotoxic [103]. Ethanolic extract of was discovered to have an LD50 of more than 1000 µg/mL at three different concentrations C. asiatica was discovered to have an LD50 of more than 1000 µg/mL at three different of C. asiatica (100, 500, and 1000 µg/mL) against 28.7 µg/mL potassium dichromate [106]. concentrations of C. asiatica (100, 500, and 1000 µg/mL) against 28.7 µg/mL potassium diAll concentrations posed an insignificant toxicity level after 24 h of exposure. The mortality chromate [106]. All concentrations posed an insignificant toxicity level after 24 h of exposure. The mortality rate of brine shrimp was 3.33% at a concentration of 10 µg/mL, 20.00% at 100 µg/mL, and 40.00% at 1000 µg/mL of C. asiatica ethanolic extract, after 24 h of Antibiotics 2021, 10, 457 19 of 24 rate of brine shrimp was 3.33% at a concentration of 10 µg/mL, 20.00% at 100 µg/mL, and 40.00% at 1000 µg/mL of C. asiatica ethanolic extract, after 24 h of exposure [107]. Calculation of LD50 was 1926 µg/mL against the standard drug, etoposide. The maximum toxic concentration was 60,822 µg/mL, and the minimum limit of toxic concentration was 606 µg/mL, showing a lower cytotoxicity level of C. asiatica [103]. Selvi et al. [108] reported an LD50 for aqueous and chloroform extracts of C. asiatica of 840 and 765 µg/mL, respectively, also showing low toxicity of C. asiatica. The type of solvent used to obtain extracts influenced the LD50 results because different solvents have different extraction potential for toxicity screening. Furthermore, some solvents are better than others for extracting bioactive compounds from plants that might be toxic [103]. Preparation of stock solutions and dilution factors of sample extracts can also affect the concentrations of sample solutions, thus directly influencing the toxicity results for sample extracts during biological screening [105]. The efficacy, performance, and safety of C. asiatica have been witnessed and widely applied in traditional Indian, Asian, and Chinese medicines, herbs or food and beverages, and pharmaceutical products [109]. A report by the Cosmetic Review Ingredient [60] expert panel confirmed the safety of various C. asiatica extracts in the current practices of use, based on the reported research including limited data on in vitro human cell cultures and oral administration in human studies. According to the World Health Organization (WHO), the recommended dosage for oral intake of C. asiatica is 1.00 to 2.00 g per day for scar surface or wound healing. In terms of tea or juice preparation using the dried plant, the dosage is 0.33–0.68 g per meal [1,42]. To date, no toxic effects from C. asiatica intake have been reported by the WHO [42]. C. asiatica extracts and asiatic acid were orally administered in an experimental hamster and rabbit model. No toxic effect was observed after intake of 1.0 mg/kg of asiatic acid [59] or 1 mg/kg of asiaticoside [110]. In fact, administering asiaticoside at 1.00 g/kg of the patient’s body weight has been demonstrated to be non-toxic in the oral application of C. asiatica extract [1]. Acute oral administration of 1 g/kg body weight of an ethanolic 50% extract and alcoholic extracts of C. asiatica also showed no toxicity at doses of 350 mg/kg when given to rats [111]. The European Medicine Agency has reported clinical studies on the effects of C. asiatica on chronic venous insufficiency (CVI), periodontitis, psoriasis, ulcer cicatrization, burn recovery, anxiety, and atherosclerosis. Clinical studies showed that C. asiatica improves microcirculation and leg volume associated with decreased edema and symptoms. The safety profile of C. asiatica extracts appears satisfactory and tolerable as it is emerging from clinical studies in patients affected by CVI and from its use in products on the market. Clinical trials have reported no drug-related serious adverse events. The recommended doses for non-toxic nature with no or infrequent adverse side effects are 60–180 mg daily. However, at a higher dose, there may be occasional burning pain or skin allergy following injection or topical application. In addition, gastric complaints and nausea have occasionally been reported following oral administration of C. asiatica extract [32,110]. In general, there are no reasons for concerns relating to safety, and the tolerability of oral C. asiatica preparations was good in all studies. No adverse events from pharmacovigilance data are known [111]. 5. Conclusions Several methods have been used to extract compounds from C. asiatica. The efficiency of extraction is based on the extraction method, extraction solvent, and extraction time. The desired compounds extracted also influence the choice of extraction method aside from cost and availability. Among the applications, the antimicrobial action of C. asiatica has been widely studied, mostly in extracts obtained conventionally. To date, very few microbial species have been tested using C. asiatica extracts obtained by modern extraction techniques. Thus, more studies are necessary for these extracts to determine their effect on microorganisms. The modern extraction techniques also seem to be more promising for obtaining antimicrobial compounds in terms of cost, time, and better efficacy toward certain microbes compared to conventional techniques. In particular, solventless extraction Antibiotics 2021, 10, 457 20 of 24 hinders the possible retention of the chemical solvent in the extract. Therefore, the extracts obtained from these modern techniques are worthy as antimicrobial agents. Both in vitro and in vivo studies have shown that C. asiatica possesses antimicrobial activity, although there have been few in vivo studies due to their complexity. Nevertheless, the extracts have the potential to be used in the medicinal, cosmeceutical, and food sectors. Author Contributions: Concept, F.N.I.; Writing–review and editing, M.M.N. and F.N.I.; Project administration, M.M.N. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Fundamental Research Grant Scheme (203/PJKIMIA/6071379). Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Data sharing not applicable. Acknowledgments: The authors would like to thank the Ministry of Higher Education Malaysia for their support through the Fundamental Research Grant Scheme (203/PJKIMIA/6071379). Conflicts of Interest: The authors declare no conflict of interest. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Ekor, M. The growing use of herbal medicines: Issues relating to adverse reactions and challenges in monitoring safety. Front. Neurol. 2014, 4, 1–10. [CrossRef] Hussin, F.; Eshkoor, S.A.; Rahmat, A.; Othman, F.; Akim, A. The Centella asiatica juice effects on DNA damage, apoptosis and gene expression in hepatocellular carcinoma (HCC). BMC Complement. Altern. Med. 2014, 32, 1–7. [CrossRef] Sabaragamuwa, R.; Perera, C.O.; Fedrizzi, B. Centella asiatica (Gotu kola) as a neuroprotectant and its potential role in healthy ageing. Trends Food Sci. Technol. 2018, 79, 88–97. [CrossRef] Mohd Heikal, M.Y.; Siti Mariam, H.; Mohd Ilham, A.; Mee Fong, C.; Aminuddin, B.S.; Ruszymah, B.H.I. Anti-proliferative activities of Centella asiatica extracts on human respiratory epithelial cells in vitro. J. Med. Plant Res. 2014, 8, 864–869. [CrossRef] Alfarra, H.Y.; Omar, M.N. Centella asiatica: From folk remedy to the medicinal biotechnology—A state revision. Int. J. Biosci. 2013, 3, 49–67. Monton, C.; Settharaksa, S.; Luprasong, C.; Songsak, T. An optimization approach of dynamic maceration of Centella asiatica to obtain the highest content of four centelloids by response surface methodology. Braz. J. Pharmacogn. 2019, 29, 254–261. [CrossRef] Saha, S.; Guria, T.; Singha, T.; Maity, T.K. Evaluation of analgesic and anti-inflammatory activity of chloroform and methanol extracts of Centella asiatica Linn. ISRN Pharmacol. 2013, 2013, 1–6. [CrossRef] [PubMed] Nurlaily, A.; Noor Baitee, A.; Musalmah, M. Comparative antioxidant and anti-inflammatory activity of different extracts of Centella asiatica (L.) urban and its active compounds, asiaticoside and madecassoside. Med. Health 2012, 7, 62–72. Abdulla, M.A.; Al-Bayaty, F.H.; Younis, L.T.; Abu Hassan, M.I. Anti-ulcer activity of Centella asiatica leaf extract against ethanolinduced gastric mucosal injury in rats. J. Med. Plants Res. 2010, 4, 1253–1259. Byakodi, M.K.; Bagewadi, Z.K.; Muddapur, U.M. Phytoconstituents profiling and evaluation of antimicrobial and antioxidant attributes of methanolic extract of Centella asiatica. Res. J. Pharm. Biol. Chem. Sci. 2018, 9, 493–500. Sarumathi, A.; Anbu, S.; Saravanan, N. Antibacterial activity of Centella asiatica (Linn.,) leaves. Int. J. Chem. Pharm. Sci. 2013, 1, 393–395. Harun, N.H.; Septama, A.W.; Wan Ahmad, W.A.N.; Suppian, R. The potential of Centella asiatica (Linn.) urban as an anti-microbial and immunomodulator agent: A review. Nat. Prod. Sci. 2019, 25, 92–102. [CrossRef] Soumyanath, A.; Zhong, Y.P.; Henson, E.; Wadsworth, T.; Bishop, J.; Gold, B.G.; Quinn, J.F. Centella asiatica extract improves behavioral deficits in a mouse model of Alzheimer’s disease: Investigation of a possible mechanism of action. Int. J. Alzheimer’s Dis. 2012, 381974. [CrossRef] Seevaratnam, V.; Banumathi, P.; Premalatha, M.R.; Sundaram, S.P.; Arumugam, T. Functional properties of Centella asiatica (L.): A review. Int. J. Pharm. Pharm. Sci. 2012, 4, 8–14. Long, H.S.; Stander, M.A.; Van Wyk, B.E. Notes on the occurrence and significance of triterpenoids (asiaticoside and related compounds) and caffeoylquinic acids in Centella species. S. Afr. J. Bot. 2012, 82, 53–59. [CrossRef] Gunathilake, K.D.P.P.; Ranaweera, K.K.D.S.; Rupasinghe, H.P.V. Response surface optimization for recovery of polyphenols and carotenoids from leaves of Centella asiatica using an ethanol-based solvent system. Food Sci. Nutr. 2019, 7, 528–536. [CrossRef] [PubMed] Vadlapudi, V.; Behara, M.; Kaladhar, D.S.V.G.K.; Suresh Kumar, S.V.N.; Seshagiri, B.; John Paul, M. Antimicrobial profile of crude extracts Calotropis procera and Centella asiatica against some important pathogens. Indian J. Sci. Technol. 2012, 5, 3132–3136. [CrossRef] Antibiotics 2021, 10, 457 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 21 of 24 Azwanida, N.N. A review on the extraction methods use in medicinal plants, principle, strength and limitation. Med. Aromat. Plants 2015, 4, 1–6. [CrossRef] Rengganaten, V. In Vitro Screening of Cytotoxic Effect and Antioxidant Activity of Pereskia bleo and Centella asiatica Crude Extracts. Bachelor’s Thesis, Universiti Tunku Abdul Rahman, Perak, Malaysia, 2013. Bylka, W.; Znajdek-awiżeń, P.; Studzińska-sroka, E.; Brzezińska, M. Centella asiatica in cosmetology. Post˛epy Dermatol. Alergol. 2013, 63, 46–49. [CrossRef] [PubMed] Azis, H.A.; Taher, M.; Ahmed, A.S.; Sulaiman, W.M.A.W.A.W.; Susanti, D.; Chowdhury, S.R.; Zakaria, Z.A. In vitro and in vivo wound healing studies of methanolic fraction of Centella asiatica extract. S. Afr. J. Bot. 2017, 108, 163–174. [CrossRef] Lu, L.; Ying, K.; Wei, S.; Fang, Y.; Liu, Y.; Lin, H.; Ma, L.; Mao, Y. Asiaticoside induction for cell-cycle progression, proliferation and collagen synthesis in human dermal fibroblasts. Basic Sci. Rev. 2004, 43, 801–807. [CrossRef] [PubMed] Transparency Market Research. Centella Asiatica Market—Global Industry Analysis, Size, Share, Trends, Growth, and Forecasts, 2020 to 2030. Available online: https://www.transparencymarketresearch.com/centella-asiatica-market.html (accessed on 25 February 2021). Sene, G.; Loiseau, A.; Lepetit, J.-C. Use of Compounds from Centella asiatica. WO 2007054211, 18 October 2017. Loewy, Z.G.; Levine, W.Z.; Saffer, A.J. Oral Rinse. U.S. Patent 807,592,4B2, 13 December 2011. Montanez-Soto, F.L. Pharmaceutical Composition Based on Centella asiatica (Hydrocotyle asiatica L.) for the Treatment of Lower Limb Ulcers. U.S. Patent 10,232,004, 17 March 2016. Palpu, P.; Rao, C.V.; Kishore, K.; Gupta, Y.K.; Kartik, R.; Govindrajan, R. Herbal Formulation as Memory Enhancer in Alzheimer Condition. U.S. Patent 7,429,397, 29 June 2006. Kodgule, M.M.; Kumar, Y. Topical Hair Composition. WO 2020234794A1, 26 November 2020. Prasad, A.; Mathur, A.K.; Mathur, A. Advances and emerging research trends for modulation of centelloside biosynthesis in Centella asiatica (L.) Urban—A review. Ind. Crops Prod. 2019, 141, 111768. [CrossRef] Ved, D.K.; Goraya, G.S. Demand and Supply of Medicinal Plants in India; National Medicinal Plants Board, New Delhi & FRLHT: Bangalore, India, 2007. Vinolina, N.S. Centelloside content of cultivated pegagan (Centella asiatica) with application of phosporus fertilizer. J. Phys. Conf. Ser. 2018, 1116, 052072. [CrossRef] Paudel, P.; Satyal, P.; Dosoky, N.S.; Setzer, W.N. Chemical composition and biological activity of Centella asiatica essential oil from Nepal. Am. J. Essent. Oils Nat. Prod. 2017, 5, 5–8. Yahya, N.A.; Attan, N.; Wahab, R.A. An overview of cosmeceutically relevant plant extracts and strategies for extraction of plant-based bioactive compounds. Food Bioprod. Process. 2018, 112, 69–85. [CrossRef] Zhang, Q.W.; Lin, L.G.; Ye, W.C. Techniques for extraction and isolation of natural products: A comprehensive review. Chin. Med. 2018, 13, 1–26. [CrossRef] [PubMed] Duval, C. Method for Preparing Extract of Centella asiatica. U.S. Patent 948,755,2B2, 4 July 2013. Kim, K.; Lee, S.-Y.; Seo, S.-K.; Hwang, B.-R.; Park, J.-K. Water-Soluble Extract of Asiaticoside and Madecassoside from Centella asiatica and Isolating Method Thereof. U.S. Patent 641,734,9B1, 9 July 2002. Loiseau, A.; Sene, G.; Theron, E. Method for Preparing Centella asiatica Extract rich in Madecassoside and in Terminoloside. U.S. Patent 848,690,0B2, 16 July 2013. Desai, J.V.; Thomas, R.; Kamat, S.D.; Kamat, D.V. Microwave assisted extraction of saponins from Centella asiatica and its in vitro anti-inflammatory study. Asian J. Biochem. Pharm. Res. 2011, 1, 330–334. Hiranvarachat, B.; Devahastin, S.; Soponronnarit, S. Comparative evaluation of atmospheric and vacuum microwave-assisted extraction of bioactive compounds from fresh and dried Centella asiatica L. leaves. Int. J. Food Sci. Technol. 2015, 50, 750–757. [CrossRef] Kim, W.; Kim, J.; Veriansyah, B.; Kim, J.; Lee, Y.; Oh, S.; Tjandrawinata, R.R. Extraction of bioactive components from Centella asiatica using subcritical water. J. Supercrit. Fluids 2009, 48, 211–216. [CrossRef] Kumari, S.; Deori, M.; Elancheran, R.; Kotoky, J.; Devi, R. In vitro and in vivo antioxidant, anti-hyperlipidemic properties and chemical characterization of Centella asiatica (L.) extract. Front. Pharmacol. 2016, 7, 1–12. [CrossRef] Niamnuy, C.; Charoenchaitrakool, M.; Mayachiew, P.; Devahastin, S. Bioactive compounds and bioactivities of Centella asiatica (L.) urban prepared by different drying methods and conditions. Dry. Technol. Int. J. 2013, 31, 2007–2015. [CrossRef] Agme-ghodke, V.; Agme, R.N.; Sagar, A.D. Analysis of bioactive compounds in leaves extract of Centella asiatica by using HRLC-MS & IR techniques. J. Chem. Pharm. Res. 2016, 8, 122–125. Florczak, M.C. Extraction and Characterisation of the Essential Oil from Centella asiatica (Pennywort) Growing in South Africa. Master’s Thesis, University Witwatersrand, Johannesbg, South Africa, 2014. Joshi, D.R.; Adhikari, N. An overview on common organic solvents and their toxicity. J. Pharm. Res. Int. 2019, 28, 1–18. [CrossRef] Govarthanan, M.; Rajinikanth, R.; Kamala-kannan, S.; Selvankumar, T. A comparative study on bioactive constituents between wild and in vitro propagated Centella asiatica. J. Genet. Eng. Biotechnol. 2015, 13, 25–29. [CrossRef] Thamarai Selvi, P.; Senthil Kumar, M.; Rajesh, R.; Kathiravan, T. Antidepressant activity of ethanolic extract of leaves of Centella asiatica. Linn by in vivo methods. Asian J. Res. Pharm. Sci. 2012, 2, 76–79. Shen, Y.; Liu, A.; Ye, M.; Wang, L.; Chen, J.; Wang, X.; Han, C. Analysis of biologically active constituents in Centella asiatica by microwave-assisted extraction combined with LC–MS. Chromatographia 2009, 70, 431–438. [CrossRef] Antibiotics 2021, 10, 457 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 22 of 24 Rafamantanana, M.H.; Rozet, E.; Raoelison, G.E.; Cheuk, K.; Ratsimamanga, S.U.; Hubert, P.; Quetin-Leclercq, J. An improved HPLC-UV method for the simultaneous quantification of triterpenic glycosides and aglycones in leaves of Centella asiatica (L.) Urb (APIACEAE). J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 2009, 877, 2396–2402. [CrossRef] Rahman, M.; Hossain, S.; Rahaman, A.; Fatima, N.; Nahar, T.; Uddin, B. Antioxidant activity of Centella asiatica (Linn.) urban: Impact of extraction solvent polarity. J. Pharmacogn. Phytochem. 2013, 1, 27–32. Borhan, M.Z.; Ahmad, R.; Rusop, M.; Abdullah, S. Green extraction: Enhanced extraction yield of asiatic acid from Centella asiatica (L.) nanopowders. J. Appl. Chem. 2013, 2013, 1–7. [CrossRef] Sellathoroe, S.; Marimuthu, S.; Ramays, T.R. Comparison of different extraction methods to study the antimicrobial activity of Centella asiatica leaf extracts. Int. J. Adv. Res. 2019, 7, 344–347. [CrossRef] Randriamampionona, D.; Diallo, B.; Rakotoniriana, F.; Rabemanantsoa, C.; Cheuk, K.; Corbisier, A.M.; Mahillon, J.; Ratsimamanga, S.; El Jaziri, M. Comparative analysis of active constituents in Centella asiatica samples from Madagascar: Application for ex situ conservation and clonal propagation. Fitoterapia 2007, 78, 482–489. [CrossRef] [PubMed] Sen, K.K.; Chouhan, K.B.S.; Tandey, R.; Mehta, R.; Mandal, V. Impact of microwaves on the extraction yield of phenolics, flavonoids, and triterpenoids from Centella leaves: An approach toward digiti botanical extraction. Pharmacogn. Mag. 2019, 15, S267–S273. [CrossRef] Idris, F.N.; Nadzir, M.M.; Abd Shukor, S.R. Optimization of solvent-free microwave extraction of Centella asiatica using Taguchi method. J. Environ. Chem. Eng. 2020, 8, 103766. [CrossRef] Wang, C.; Han, W.; Fan, L.; Wang, C. Enzymatic pretreatment and microwave extraction of asiaticoside from Centella asiatica. J. Biomed. Sci. Eng. 2009, 2, 526–531. [CrossRef] Albuquerque, B.R.; Prieto, M.A.; Barreiro, M.F.; Rodrigues, A.E.; Curran, T.P.; Barros, L.; Ferreira, I.C.F.R. Catechin-based extract optimization obtained from Arbutus unedo L. fruits using maceration/microwave/ultrasound extraction techniques. Ind. Crop. Prod. 2017, 95, 404–415. [CrossRef] Safdar, M.N.; Kausar, T.; Jabbar, S.; Mumtaz, A.; Ahad, K.; Saddozai, A.A. Extraction and quantification of polyphenols from kinnow (Citrus reticulate L.) peel using ultrasound and maceration techniques. J. Food Drug Anal. 2017, 25, 488–500. [CrossRef] Pittella, F.; Dutra, R.C.; Junior, D.D.; Lopes, M.T.P.; Barbosa, N.R. Antioxidant and cytotoxic activities of Centella asiatica (L). Int. J. Mol. Sci. 2009, 10, 3713–3721. [CrossRef] Cosmetic Ingredient Review. Safety Assessment of Centella asiatica-Derived Ingredients as Used in Cosmetics. 2015. Available online: https://www.cir-safety.org/sites/default/files/centel062015FR.pdf (accessed on 2 February 2021). Zhao, C.; He, X.; Li, C.; Yang, L.; Fu, Y.; Wang, K.; Zhang, Y.; Ni, Y. A microwave-assisted simultaneous distillation and extraction method for the separation of polysaccharides and essential oil from the leaves of Taxus chinensis Var. mairei. Appl. Sci. 2016, 6, 19. [CrossRef] Lucchesi, M.E.; Chemat, F.; Smadja, J. Solvent-free microwave extraction of essential oil from aromatic herbs: Comparison with conventional hydro-distillation. J. Chromatogr. A 2004, 1043, 323–327. [CrossRef] Orhan, I.E.; Atasu, E.; Senol, F.S.; Ozturk, N.; Demirci, B.; Das, K.; Sekeroglu, N. Comparative studies on Turkish and Indian Centella asiatica (L.) Urban (gotu kola) samples for their enzyme inhibitory and antioxidant effects and phytochemical characterization. Ind. Crops Prod. 2013, 47, 316–322. [CrossRef] Zhuoyan, H. Microwave Assisted Extraction: The Effects, Mechanisms and Applications on Selected Plant Materials. Ph.D. Thesis, The Hong Kong Polytechnic University, Hong Kong, China, 2011. Jayaprakash, S.B.; Nagarajan, N. Studies on the bioactive compounds and antimicrobial activities of medicinal plant Centella asiatica (Linn). J. Med. Plants Stud. 2016, 4, 181–185. Chew, K.K.; Khoo, M.Z.; Ng, S.Y.; Thoo, Y.Y.; Aida, W.M.W.; Ho, C.W. Effect of ethanol concentration, extraction time and extraction temperature on the recovery of phenolic compounds and antioxidant capacity of Centella asiatica extracts. Int. Food Res. J. 2011, 18, 1427–1435. [CrossRef] Bimakr, M.; Rahman, R.A.; Saleena Taip, F.; Adzahan, N.M.; Islam Sarker, Z.; Ganjloo, A. Ultrasound-assisted extraction of valuable compounds from winter melon (Benincasa hispida) seeds. Int. Food Res. J. 2013, 20, 331–338. Jemain, S.F.P.; Jamal, P.; Raus, A.R.; Amid, A.; Jaswir, I. Effects of process conditions on the ultrasonic extraction of phenolics scavenger from Curcuma caesia rhizome. Int. Food Res. J. 2017, 24, 422–427. Danlami, J.M.; Arsad, A.; Ahmad Zaini, M.A.; Sulaiman, H. A comparative study of various oil extraction techniques from plants. Rev. Chem. Eng. 2014, 30, 605–626. [CrossRef] Mutalib, L.Y. Comparison between conventional and modern methods for extraction of Rosmarinus officinalis leaves. Zanco J. Med. Sci. 2015, 19, 1029–1034. [CrossRef] Nithyanandam, R.; Shapheri, M.R.; Nasir, M.H. Antioxidant potential of Malaysian Herb Centella asiatica. Int. Conf. Environ. Chem. Biol. 2014, 78, 56–60. [CrossRef] Seo, S.B.; Kim, Y.M. Improving cosmetic activity by optimizing Centella asiatica extraction process. Nat. Prod. Commun. 2019, 14, 1–4. [CrossRef] Arumugam, T.; Ayyanar, M.; Justin, Y.; Pillai, K. Phytochemical screening and antibacterial activity of leaf and callus extracts of Centella asiatica. Bangladesh J. Pharmacol. 2011, 6, 55–60. [CrossRef] Chan, C.H.; Yusoff, R.; Ngoh, G.C.; Kung, F.W.L. Microwave-assisted extractions of active ingredients from plants. J. Chromatogr. A 2011, 1218, 6213–6225. [CrossRef] [PubMed] Antibiotics 2021, 10, 457 23 of 24 James, J.T.; Dubery, I.A. Pentacyclic triterpenoids from the medicinal herb, Centella asiatica (L.) Urban. Molecules 2009, 14, 3922–3941. [CrossRef] [PubMed] 76. Soyingbe, O.S.; Mongalo, N.I.; Makhafola, T.J. In vitro antibacterial and cytotoxic activity of leaf extracts of Centella asiatica (L.) Ur, Warburgia salutaris (Bertol. F.) Chiov and Curtisia dentata (Burm. F.) C.A.Sm-medicinal plants used in South Africa. BMC Complement. Altern. Med. 2018, 3, 1–10. 77. Dhiman, R.; Aggarwal, N.; Aneja, K.R.; Kaur, M. In vitro antimicrobial activity of spices and medicinal herbs against selected microbes associated with juices. Int. J. Microbiol. 2016, 1–9. [CrossRef] 78. Kalemba, D.; Kunicka, A. Antibacterial and antifungal properties of essential oils. Curr. Med. Chem. 2005, 10, 813–829. [CrossRef] 79. Arifullah, M.; Vikram, P.; Chiruvella, K.K.; Shaik, M.M.; Abdullah Ripain, I.H. A review on Malaysian plants used for screening of antimicrobial activity. Annu. Res. Rev. Biol. 2014, 4, 2088–2132. [CrossRef] 80. Meena, A.K.; Khan, S.; Rao, M.M.; Reddy, R.K.; Padhi, M.M. Antifungal metabolites from plants used in Ayurvedic system of medicine in India. In Antifungal Metabolites from Plants; Razzaghi-Abyaneh, M., Rai, M., Eds.; Springer: Berlin/Heidelberg, Germany, 2013; pp. 303–331, ISBN 9783642380761. 81. Sieberi, B.M.; Omwenga, G.I.; Wambua, R.K.; Samoei, J.C.; Ngugi, M.P. Screening of the dichloromethane: Methanolic extract of Centella asiatica for antibacterial activities against Salmonella typhi, Escherichia coli, Shigella sonnei, Bacillus subtilis, and Staphylococcus aureus. Sci. World J. 2020, 2020, 1–8. [CrossRef] [PubMed] 82. Zheng, H.M.; Choi, M.J.; Kim, J.M.; Lee, K.W.; Park, Y.H.; Lee, D.H. In vitro and in vivo anti-Helicobacter pylori activities of Centella asiatica leaf extract. Prev. Nutr. Food Sci. 2016, 21, 197–201. [CrossRef] 83. Jamal, P.; Karim, I.A.; Abdullah, E.; Raus, R.A.; Zuhanis, Y. Phytochemical screening for antibacterial activity of potential Malaysian medicinal plants. Afr. J. Biotechnol. 2011, 10, 18795–18799. [CrossRef] 84. Gautam, R.; Saklani, A.; Jachak, S.M. Indian medicinal plants as a source of antimycobacterial agents. J. Ethnopharmacol. 2007, 110, 200–234. [CrossRef] 85. Ferdous, N.; Rahman, M.; Alamgir, A.N.M. Investigation on phytochemical, cytotoxic and antimicrobial properties of ethanolic extracts of Centella asiatica (L.) Urban. J. Med. Plants Stud. 2017, 5, 187–188. 86. Nasution, M.Y.; Restuati, M.; Pulungan, A.S.S.; Pratiwi, N. Antimicrobial activities of Centella asiatica leaf and root extracts on selected pathogenic micro-organisms. J. Med. Sci. 2018, 18, 198–2018. [CrossRef] 87. Edgar, S.A. Analysis of Phenolics from Centella asiatica and Vernonia Amygdalina and Their Roles as Antibacterial and Antioxidant Compounds. Master’s Thesis, University Malaya, Kuala Lumpur, Malaysia, 2014. 88. Purkait, S.; Jawahar Abraham, T.; Karmakar, S.; Dey, B.; Roy, A. Inhibition of fish pathogenic Aeromonas hydrophila and Edwardsiella tarda by Centella asiatica in-vitro. J. Aquac. Res. Dev. 2018, 9, 3–6. [CrossRef] 89. Idris, F.N.; Nadzir, M.M. Antimicrobial activity of Centella asiatica on Aspergillus niger and Bacillus subtilis. Chem. Eng. Trans. 2017, 56, 1381–1386. [CrossRef] 90. Suresh, M.; Alfonisan, M.; Alturaiki, W.; Al Aboody, M.S.; Alfaiz, F.A.; Premanathan, M.; Vijayakumar, R.; Umamagheswari, K.; Ghamdi, S.A.; Alsagaby, S.A. Investigations of bioactivity of Acalypha indica (L.), Centella asiatica (L.) and Croton bonplandianus (Baill) against multidrug resistant bacteria and cancer cells. J. Herb. Med. 2020, 100359. [CrossRef] 91. Centers for Disease Control and Prevention. Antibiotic Resistance Threats in the United States, Atlanta, USA. Available online: https://www.cdc.gov/drugresistance/pdf/ar-threats-2013-508.pdf (accessed on 16 February 2021). 92. Taemchuay, D.; Rukkwamsuk, T.; Sakpuram, T.; Ruangwises, N. Antibacterial activity of crude extracts of Centella asiatica against Staphylococcus aureus in bovine mastitis. Kasetsart Vet. 2009, 19, 119–128. 93. Pitinidhipat, N.; Yasurin, P. Antibacterial activity of Chrysanthemum indicum, Centella asiatica and Andrographis paniculata against Bacillus cereus and Listeria monocytogenes under osmotic stress. Assumpt. Univ. J. Technol. 2012, 15, 239–245. 94. Al-talib, H.; Mohamad Ali, N.D.; Suhaimi, M.H.; Rosli, S.S.N.; Othman, N.H.; Mansor, N.A.S.; Sulaiman Shah, A.K.; Ariffin, N.S.; Al-khateeb, A. Antimicrobial effect of Malaysian vegetables against enteric bacteria. Asian Pac. J. Trop. Biomed. 2016, 6, 211–215. [CrossRef] 95. Kalita, D.; Saikia, J. Ethonomedicinal, antibacterial and antifungal potentiality of Centella asiatica, Nerium indicum and Cuscuta reflexa—Widely used in Tiwa tribe of Morigaon district of Assam, India. Int. J. Phytomed. 2012, 4, 380–385. 96. Zaidan, M.R.S.; Noor Rain, A.; Badrul, A.R.; Adlin, A.; Norazah, A.; Zakiah, I. In vitro screening of five local medicinal plants for antibacterial activity using disc diffusion method. Trop. Biomed. 2005, 22, 165–170. 97. Siti Nasiroh, I.; Zaleha, I.; Norazian, N. Antimicrobial effect of Centella asiatica extract on fish surimi. Politek. Kolej Komuniti J. Eng. Technol. 2016, 1, 101–107. 98. Puttarak, P.; Panichayupakaranant, P. Short Communication: A new method for preparing pentacyclic triterpene rich Centella asiatica extracts. Nat. Prod. Res. 2013, 27, 684–686. [CrossRef] 99. Bylka, W.; Znajdek-awi, P.; Studzińska-Sroka, E.; Dańczak-Pazdrowska, A.; Brzezińska, M. Centella asiatica in dermatology: An overview. Phyther. Res. 2014, 1–8. [CrossRef] 100. Arora, R.; Kumar, R.; Agarwal, A.; Reeta, K.H.; Gupta, Y.K. Comparison of three different extracts of Centella asiatica for antiamnesic, antioxidant and anticholinergic activities: In vitro and in vivo study. Biomed. Pharmacother. 2018, 105, 1344–1352. [CrossRef] 101. Hastuti, N.S.; Taurhesia, S.; Wibowo, A.E. Aktivitas secara in vitro dan in vivo kombinasi ekstrak daun kelor (Moringa oleifera lam.) dan pegagan (Centella asiatica (L.). Urb.) sebagai gel anti jerawat. Intisari Sains Medis. 2019, 10, 629–636. [CrossRef] 75. Antibiotics 2021, 10, 457 24 of 24 102. Rattanachaikunsopon, P.; Phumkhachorn, P. Use of Asiatic pennywort Centella asiatica aqueous extract as a bath treatment to control columnaris in Nile tilapia. J. Aquat. Anim. Health 2010, 22, 14–20. [CrossRef] [PubMed] 103. Ramli, S.; Xian, W.J.; Azira, N.; Mutalib, A. A Review: Antibacterial activities, antioxidant properties and toxicity profile of Centella asiatica. Educ. JSMT 2020, 7, 39–47. 104. Syahmi, A.R.M.; Vijayarathna, S.; Sasidharan, S.; Latha, Y.L.; Kwan, Y.P.; Lau, Y.L.; Shin, L.N.; Chen, Y. Acute oral toxicity and brine shrimp lethality of Elaeis guineensis jacq., (oil palm leaf) methanol extract. Molecules 2010, 15, 8111–8121. [CrossRef] [PubMed] 105. Hamidi, M.R.; Jovanova, B.; Panovska, T.K. Toxicological evaluation of the plant products using brine shrimp (Artemia salina L.) model. Maced. Pharm. Bull. 2014, 60, 9–18. [CrossRef] 106. Padmaja, R.; Arun, P.C.; Prashanth, D.; Deepak, M.; Amit, A.; Anjana, M. Brine shrimp lethality bioassay of selected Indian medicinal plants. Fitoterapia 2002, 73, 508–510. [CrossRef] 107. Sultan, R.A.; Mahmood, S.B.Z.; Azhar, I.; Ahmed, S.W.; Mahmood, Z.A. Biological activities assessment of Centella asiatica (Linn.). J. Herbsspices Med. Plants 2014, 20, 319–327. [CrossRef] 108. Selvi, C.K. Brine shrimp lethality assay of some medicinal plants using Artemia franciscana and Artemia salina. Int. J. Sci. Res. 2018, 7, 1901–1905. [CrossRef] 109. Hashim, P. MiniReview: Centella asiatica in food and beverage applications and its potential antioxidant and neuroprotective effect. Int. Food Res. J. 2011, 18, 1215–1222. 110. Gohil, K.J.; Patel, J.A.; Gajjar, A.K. Pharmacological review on Centella asiatica: A potential herbal cure-all. Indian J. Pharm. Sci. 2010, 72, 546–556. [CrossRef] [PubMed] 111. European Medicines Agency. Assessment Report on Centella asiatica (L.) Urban, Herba. Available online: https://www.ema. europa.eu/en/documents/herbal-report/final-assessment-report-centella-asiatica-l-urban-herba-first-version_en.pdf (accessed on 22 February 2021).
https://openalex.org/W3151998923	https://hal.inrae.fr/hal-03313734/document	English	null	Shared Odds of Borrelia and Rabies Virus Exposure in Serbia	Pathogens	2,021	cc-by	9,389	Shared Odds of Borrelia and Rabies Virus Exposure in Serbia Pavle Banović, Adrian Alberto Díaz-Sánchez, Dragana Mijatović, Dragana Vujin, Zsolt Horváth, Nenad Vranješ, Zorana Budakov-Obradović, Nevenka Bujandrić, Jasmina Grujić, Abdul Ghafar, et al. To cite this version: Pavle Banović, Adrian Alberto Díaz-Sánchez, Dragana Mijatović, Dragana Vujin, Zsolt Horváth, et al.. Shared Odds of Borrelia and Rabies Virus Exposure in Serbia. Pathogens, 2021, 10 (4), pp.1-11. 10.3390/pathogens10040399. hal-03313734 Distributed under a Creative Commons Attribution 4.0 International License j g j ( ) 8 Blood Transfusion Institute Vojvodina, 21000 Novi Sad, Serbia 9 Department of Veterinary Biosciences, Melbourne Veterinary School, Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Werribee, VIC 3030, Australia; Sciences, The University of Melbourne, Werribee, VIC 3030, Australia; aghafar@student.unimelb.edu.au (A.G.); jabbara@unimelb.edu.au (A.J.) y aghafar@student.unimelb.edu.au (A.G.); jabbara@unimelb.edu.au (A.J.) 10 School of Environmental Sciences, University of Guelph, Guelph, ON N1G 2W1, Canada; dasieloa@uoguelph.ca 11 Center for Nuclear Energy in Agriculture, University of São Paulo, Piracicaba, São Paulo 13400-970, Brazil 12 Anses, INRAE, Ecole Nationale Vétérinaire d’Alfort, UMR BIPAR, Laboratoire de Santé Animale, F-94700 Maisons-Alfort, France * Correspondence: pavle.banovic.mf@gmail.com (P.B.); alejandro.cabezas@vet-alfort.fr (A.C.-C.) * Correspondence: pavle.banovic.mf@gmail.com (P.B.); alejandro.cabezas@vet-alfort.fr (A.C.-C.) † Equal contribution * Correspondence: pavle.banovic.mf@gmail.com (P.B.); alejandro.cabezas@vet-alfort.fr (A.C.-C.) † Equal contribution. Abstract: Lyme borreliosis (LB) is the most common tick-borne disease in Serbia and other European countries. Rabies is a fatal zoonosis distributed worldwide and is caused by the rabies virus. Professionals at risk of rabies—including veterinarians, hunters, communal service workers, and forestry workers—overlap with some professions at a higher risk of exposure to tick bites and tick- borne pathogen infections. We hypothesized that individuals identiﬁed by the public health system as at risk of rabies virus infection, and consequently vaccinated against rabies virus, also share a higher likelihood of Borrelia exposure. To test our hypothesis, a case-control study was carried out during 2019 in Serbia to determine the seroprevalence of anti-Borrelia antibodies in two case groups (individuals at risk and vaccinated against rabies virus) and a control group (individuals without risk of rabies). Individuals vaccinated against rabies following either “pre-exposure protocol” (PrEP, n = 58) or “post-exposure protocol” (PEP, n = 42) were considered as rabies risk groups and healthy blood donors (n = 30) as the control group. The results showed higher Borrelia seroprevalence in PrEP (17.2%; 10/58) and PEP (19.0%; 8/42) groups compared with the control group (6.67%; 2/30). Furthermore, odds ratio (OR) analysis showed that risk of rabies (in either the PrEP (OR = 2.91) or PEP (OR = 3.29) groups) is associated with increased odds of being seropositive to Borrelia. However, the difference in Borrelia seroprevalence between groups was not statistically signiﬁcant (Chi-square (χ2) test p > 0.05). HAL Id: hal-03313734 https://hal.inrae.fr/hal-03313734v1 Submitted on 4 Aug 2021 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License pathogens pathogens pathogens Article Article Shared Odds of Borrelia and Rabies Virus Exposure in Serbia https://doi.org/10.3390/ pathogens10040399 Academic Editor: Ángeles Sonia Olmeda Received: 26 February 2021 Accepted: 25 March 2021 Published: 28 March 2021 Citation: Banovi´c, P.; Díaz-Sánchez, A.A.; Mijatovi´c, D.; Vujin, D.; Horváth, Z.; Vranješ, N.; Budakov-Obradovi´c, Z.; Bujandri´c, N.; Gruji´c, J.; Ghafar, A.; et al. Shared Odds of Borrelia and Rabies Virus Exposure in Serbia. Pathogens 2021, 10, 399. https://doi.org/10.3390/ pathogens10040399 Academic Editor: Ángeles Sonia Olmeda Received: 26 February 2021 Accepted: 25 March 2021 Published: 28 March 2021 Article Shared Odds of Borrelia and Rabies Virus Exposure in Serbia Pavle Banovi´c 1,2,,† , Adrian Alberto Díaz-Sánchez 3,† , Dragana Mijatovi´c 1, Dragana Vujin 4, Zsolt Horváth 5, Nenad Vranješ 6, Zorana Budakov-Obradovi´c 7,8, Nevenka Bujandri´c 7,8 , Jasmina Gruji´c 7,8 , Abdul Ghafar 9 , Abdul Jabbar 9 , Verica Simin 5, Dasiel Obregón 10,11 and Alejandro Cabezas-Cruz 12, 1 Ambulance for Lyme Borreliosis and Other Tick-Borne Diseases, Department of Prevention of Rabies and Other Infectious Diseases, Pasteur Institute Novi Sad, 21000 Novi Sad, Serbia; Draganav77@gmail.com 1 Ambulance for Lyme Borreliosis and Other Tick-Borne Diseases, Department of Prevention of Rabies and Other Infectious Diseases Pasteur Institute Novi Sad 21000 Novi Sad Serbia; Draganav77@gmail com 1 Ambulance for Lyme Borreliosis and Other Tick-Borne Diseases, Department of Prevention of Rabies and Other Infectious Diseases, Pasteur Institute Novi Sad, 21000 Novi Sad, Serbia; Draganav77@gmail.com 2 Department of Microbiology with Parasitology and Immunology, Faculty of Medicine in Novi Sad, University of Novi Sad, 21000 Novi Sad, Serbia g g 2 Department of Microbiology with Parasitology and Immunology, Faculty of Medicine in Novi Sad, University of Novi Sad, 21000 Novi Sad, Serbia 3 Department of Biology, University of Saskatchewan, 112 Science Place, Saskatoon, SK S7N 5E2, Canada; adiasanz88@gmail.com 3 Department of Biology, University of Saskatchewan, 112 Science Place, Saskatoon, SK S7N 5E2, Canada; adiasanz88@gmail.com 4 National Reference Laboratory for Rabies, Department of Microbiology, Pasteur Institute Novi Sad, 21000 N i S d S bi f t @ il 4 National Reference Laboratory for Rabies, Department of Microbiology, Pasteur Institute Novi Sad, 21000 Novi Sad, Serbia; favn.paster@gmail.com 5 Agricultural School, Maršala Tita 167, 24300 Baˇcka Topola, Serbia; cibi@stcable.net (Z.H.); Luketic.s@mts.rs (V.S.) 6 Department for Research & Monitoring of Rabies & Other Zoonoses, Pasteur Institute Novi Sad, 21000 Novi Sad, Serbia; nenad.vranjes@gmail.com 6 Department for Research & Monitoring of Rabies & Other Zoonoses, Pasteur Institute Novi Sad, 21000 Novi Sad Serbia; nenad vranjes@gmail com 7 Faculty of Medicine in Novi Sad, University of Novi Sad, 21000 Novi Sad, Serbia; zorana.budakov-obradovic@mf.uns.ac.rs (Z.B.-O.); nevenka.bujandric@mf.uns.ac.rs (N.B.); jasmina.grujic@mf.uns.ac.rs (J.G.) Citation: Banovi´c, P.; Díaz-Sánchez, A.A.; Mijatovi´c, D.; Vujin, D.; Horváth, Z.; Vranješ, N.; Budakov-Obradovi´c, Z.; Bujandri´c, N.; Gruji´c, J.; Ghafar, A.; et al. Shared Odds of Borrelia and Rabies Virus Exposure in Serbia. Pathogens 2021, 10, 399. 1. Introduction Ecological perturbations due to land use by humans, growing human population, globalization, and climate change are associated with an increased incidence and prevalence of vector-borne diseases. Higher average temperatures associated with climate change are suitable for most arthropod vectors, which may explain the territorial expansion and increased abundance of different vectors. Ticks are the main vectors for animal and human diseases in Europe, and their geographic ranges have expanded in recent years [1–3]. Ixodes ricinus (Genus: Ixodes, Family: Ixodidae), the most important tick vector of human diseases in Europe, is a generalist tick that can feed on several animal species, including humans— who are considered accidental hosts [4,5]. Ixodes ricinus is mostly found in deciduous forests where small mammals and deer serve as its main hosts. However, this tick species can also be found in swamps and meadows during periods of high rainfalls [6,7]. Diseases transmitted by I. ricinus include tick-borne encephalitis, human granulocytic anaplasmosis and other rickettsial diseases, babesiosis and Lyme borreliosis (LB) [8,9]. Borrelia infection in Europe is caused by the members of the Borrelia burgdorferi sensu lato (s.l.) complex, including Borrelia burgdorferi sensu stricto (s.s.), Borrelia afzelii, and Borrelia garinii, and the estimated incidence of LB in the region ranges from 85,000 to more than 200,000 cases per year [10,11]. Although less frequent, LB can also be caused by Borrelia spielmanii and some pathogenic strains of Borrelia lusitaniae, Borrelia valaisiana, and Borrelia bissettii [11,12]. The spirochaete are transmitted enzootically between ticks and their hosts. In Serbia, as in other countries of Europe, I. ricinus is responsible for the transmission of several tick-borne diseases (TBDs) including LB, tick-borne encephalitis, and rickettsioses [13–16]. Among these TBDs, LB is the most common in Serbia [13,17], where Borrelia DNA have been detected in I. ricinus collected from several regions [18,19], and about 50% of the tested ticks were found positive for at least one member of B. burgdorferi s.l. complex [18]. Due to the absence of a national guideline for diagnostics and treatment of TBDs, LB is possibly an underestimated cause of disease in Serbia. Rabies is a lethal zoonosis with worldwide distribution and is caused by rabies virus (Genus: Lyssavirus, Family: Rhabdoviridae), which affects the central nervous system of mammals and leads to signs of encephalomyelitis with almost 100% fatality rate [20]. The virus normally circulates within its natural hosts (i.e., species of the orders Carnivora and Chiroptera). The shared odds of LB and rabies exposure found in this study suggest that, in countries where both diseases occur, the common citizen can be at risk of both diseases when in a risky habitat. These ﬁndings are important to guide physicians in targeting high-risk groups, and diagnose LB, and to guide decision-makers in targeting control and prevention measures for both infections in risk areas. Received: 26 February 2021 Accepted: 25 March 2021 Published: 28 March 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). https://www.mdpi.com/journal/pathogens Pathogens 2021, 10, 399. https://doi.org/10.3390/pathogens10040399 2 of 11 Pathogens 2021, 10, 399 Keywords: rabies; Borrelia; seroprevalence; exposure; relative odds 1. Introduction Hosts, such as humans, who are not reservoirs for rabies virus, get infected as a consequence of spillover events when an infected natural host (e.g., fox, dog, bat) transmit the virus via licking, scratching, or biting [21]. The number of human deaths globally due to dog-associated rabies is estimated to be 59,000 annually, and most of the victims are children under 15 years [21]. In Serbia, dog-associated rabies was a public health problem until 1980s. Since then, most rabies cases in Serbia have been associated with foxes, and thus this animal has become the main reservoir species for rabies virus in the country [22]. In humans, without preventive vaccination or earlier post-exposure vaccination/treatment, the disease is fatal. Accordingly, “pre-exposure protocol” (PrEP) immunization against rabies is recommended for professionals (such as veterinarians, hunters, communal service workers, and forestry workers) in close contact with wild and stray animals, and “post-exposure protocol” (PEP) immunization is recommended for individuals who were in contact with a rabid or suspected-rabid animal, independent of profession-associated risks [23]. Annually, about 700 people receive anti-rabies prophylaxis in Serbia, from which 50% are indications of PEP. Rabies is a notiﬁable disease in Serbia, and except for a case of fox-transmitted rabies detected in 2018 (unpublished data), the effective vaccination strategy has controlled the occurrence of human rabies cases in Serbia [24]. Due to their professional activities and/or lifestyle, risk groups included in the PrEP and PEP vaccination protocols could also be at high risk of exposure to tick bites and LB. 3 of 11 3 of 11 Pathogens 2021, 10, 399 Considering the above evidence, we hypothesized that individuals identiﬁed by the public health system as at risk of rabies virus infection also share a high likelihood of being bitten by I. ricinus during ﬁeld activities and, in consequence, to be more exposed to Borrelia infection than individuals with no regular access to risk areas. To test our hypothesis (i.e., the odds of Borrelia seropositivity are higher among individuals at risk of rabies virus exposure than among individuals at no risk of rabies virus exposure), we designed a case-control study to compare the seroprevalence of anti-Borrelia antibodies in individuals immunized against rabies (two case groups) and healthy blood donors with no profession-associated accesses to risk areas (control group) in Serbia and Bosnia and Herzegovina. 2.1. Ethical Statement This study was approved by the Novi Sad Ethical Committee at the Faculty of Medicine, the University of Novi Sad (approval number 01-39/266/1), and was conducted according to the Declaration of Helsinki and The Patient Rights Law of the Republic of Serbia. Written informed consents were obtained from blood donors to allow the use of their blood samples for this study. 2.2. Study Design and Sampling Strategy An observational, case-control study was conducted during the year 2019 in Serbia and Bosnia and Herzegovina to compare the seroprevalence of anti-Borrelia IgG among individ- uals at risk, or not, of rabies virus exposure. Accordingly, two groups of individuals were considered at risk of rabies virus exposure. The ﬁrst group included individuals who live, or spend time, in rural areas or forests and after being bitten by rabid or rabies-suspected animals, were vaccinated against rabies and treated with anti-rabies sera (hereafter referred to as the “post-exposure protocol” group, PEP). The second group included professionals (i.e., veterinarians, community service workers, foresters) who had received preventive rabies immunization (hereafter referred as the “pre-exposure protocol” group, PrEP) due to profession-associated increased risk of rabies virus exposure. Sera from clinically healthy blood donors without an identiﬁed profession-associated risk of rabies virus exposure were selected as the control group. Anti-Borrelia antibodies were tested in all sera samples by indirect immunoﬂuorescent test (IFAT). The association between the seroprevalence of Borrelia spp. and rabies virus exposure was explored in potential risk groups, including age (children and teenagers: 0–19; adults: 20–59; and seniors: >60 years), gender (male and female), residence (North, South, and Central Serbia and Bosnia and Herzegovina (Sarajevo)) and the anti-rabies immunization protocol (PEP and PrEP). Socio-demographic data were obtained through medical documentation available at the Pasteur Institute Novi Sad (PEP and PrEP groups) and Blood Transfusion Institute of Vojvodina (control group). 1. Introduction Such valuable information would be highly useful for the planning of prevention and control measures, particularly for personnel at high risk of LB. 2.3. Selection of Samples in PEP and PrEP Groups For this study, sera samples of PEP (n = 42, Supplementary Materials Table S1) and PrEP (n = 58, Supplementary Materials Table S2) individuals were randomly selected from the sera bank at the National Reference Laboratory for Rabies (NRLR), Pasteur Institute Novi Sad, Serbia. The NRLR is a national-level institution that archives sera samples from individuals immunized against rabies at local “anti-rabies stations” scattered across the territory of Serbia. All individuals in the PEP and PrEP groups were immunized against rabies with the vaccine Verorab® (SanoﬁPasteur, Paris, France). In the case of PEP, along with the ﬁrst vaccine dose, human rabies immune globulin (HRIG) (Blood Transfusion Institute of Serbia, Belgrade) was administered to each patient as passive immunization. Accordingly, all the sera samples in the PEP and PrEP groups contained neutralizing anti-rabies antibodies in titer 0.5 ≤IU detected with the golden standard assay i.e., rapid Pathogens 2021, 10, 399 4 of 11 ﬂuorescent foci inhibition test (RFFIT). Sera were stored at −80 ◦C for at least one year after the detection of rabies-neutralizing antibodies titer. ﬂuorescent foci inhibition test (RFFIT). Sera were stored at −80 ◦C for at least one year after the detection of rabies-neutralizing antibodies titer. 2.4. Selection of Healthy Blood Donor Samples Healthy blood donor sera samples (n = 30, Supplementary Materials Table S3) used as the control group were acquired from the Blood Transfusion Institute of Vojvodina, Serbia. None of the individuals in this group had professions associated with risk of exposure to rabies virus (Supplementary Materials Table S3). As part of the routine screening of blood donors by the Blood Transfusion Institute of Vojvodina, control samples were analyzed for and tested negative for the presence of antibodies against Treponema pallidum and antigen, as well as for antibodies and the DNA of the human immunodeﬁciency and hepatitis C viruses. 2.5. Borrelia Culture and IFAT Antigen Preparation Borrelia spirochetes were grown in complete BSK-H medium (Sigma-Aldrich, Saint Louis, MO, USA, Cat. B8291) at 33 ◦C [25]. The culture (2 mL) was pelleted down by centrifugation at 1200 × g for 15 min and washed twice in sterile phosphate buffered saline (PBS; pH 7.4). The density of Borrelia culture was adjusted to approximately 3 × 107 organ- isms/mL. We used the whole spirochetes as antigen and 0.2 mL of the suspension of Borrelia afzelii in PBS was added to microscopic slide coated with poly-L-lysine (Sigma-Aldrich, Saint Louis, MO, USA, Cat. P0425) and left to dry at room temperature. Antigenic ﬁelds were then encircled with hydrophobic PAP pen (Abcam, Cambridge, United Kingdom, Cat. ab2601). The B. afzelii strain, part of a Borrelia spp. collection, was kindly donated by the Group of Medical Entomology, Institute for Medical Research, University of Belgrade. 3. Results Association between Regional Distribution of Borrelia spp. Seroprevalence and Rabies Virus Exposure Risk When the municipalities of residency of all individuals included in the study were considered, we found a different distribution of Borrelia spp. seroprevalence across the Serbian territory. Anti-Borrelia IgG were detected in sera from North (17/99, 17.17%; CI, 11.31–27.09), Central (1/11, 9.09%; CI, 0.00–29.26), and South (1/14, 7.14%; CI, 0.00–22.57) Serbia, as well as the Belgrade area (1/2, 50%; CI, 35.22–64.78), while anti-Borrelia IgG were not detected in sera samples from Sarajevo in Bosnia and Herzegovina (0/4, 0.0%). Excluding the results from the Belgrade area (due to small sample size n = 2), Borrelia spp. seroprevalence was higher in Northern Serbia. The local administrative unit level Borrelia spp. seroprevalence in PEP (χ2 = 2.24, p = 0.13) and PrEP (χ2 = 1.87, p = 0.17) case groups was not statistically signiﬁcant compared with the healthy donors control group. In addition, no signiﬁcant difference (χ2 = 0.05, p = 0.81) was found in the Borrelia spp. seroprevalence of the case groups. Although the differences did not reach statisti- cal signiﬁcance, odds ratio (OR) analysis showed that Borrelia spp. seroprevalence was positively correlated (OR > 1) with risk of rabies virus exposure in the PEP (OR = 3.29; CI, 0.65–16.78; p = 0.13) and PrEP (OR = 2.91; CI, 0.59–14.27; p = 0.17) case groups compared with the healthy donors control group. In contrast, comparison of Borrelia spp. seropreva- lence in PEP and PrEP case groups revealed an OR very close to 1 (OR = 1.12; CI, 0.40–3.15; p = 0.13), suggesting that the outcome (i.e., Borrelia spp. seroprevalence) is independent of the vaccination protocol (i.e., PEP and PrEP). 3.2. Association between Regional Distribution of Borrelia spp. Seroprevalence and Rabies Virus Exposure Risk 3.2. Association between Regional Distribution of Borrelia spp. Seroprevalence and Rabies Virus Exposure Risk When the municipalities of residency of all individuals included in the study were considered, we found a different distribution of Borrelia spp. seroprevalence across the Serbian territory. Anti-Borrelia IgG were detected in sera from North (17/99, 17.17%; CI, 11.31–27.09), Central (1/11, 9.09%; CI, 0.00–29.26), and South (1/14, 7.14%; CI, 0.00–22.57) Serbia, as well as the Belgrade area (1/2, 50%; CI, 35.22–64.78), while anti-Borrelia IgG were not detected in sera samples from Sarajevo in Bosnia and Herzegovina (0/4, 0.0%). 2.6. Detection of Anti-Borrelia Antibodies by IFAT Anti-Borrelia IgG were detected by an in-house IFAT using B. afzelii antigens on microscopic slides. Previous to the analysis of the samples of the study, the in-house IFAT assay was validated using Euroimmun commercial IFAT assay (FI 2131-1 G, Euroimmun, Lubeck, Germany) as a reference and according to the guidelines and methodology for validation of diagnostic tests described in the Manual of Diagnostic Tests and Vaccines for Terrestrial Animals of the World Organization for Animal Health (OIE) [26]. Optimal antigen concentration as well as sera and antibody dilutions were deﬁned using titration assays, where PBS (pH 7.4) was used for dilution of examined sera. Test sera were diluted to 1:100 and incubated with antigen-coated slides for 45 min in a humidiﬁed chamber at 37 ◦C. Following two PBS washes, goat anti-human IgG labelled with the green ﬂuorescent dye CF™488A (Sigma-Aldrich, Saint Louis, MO, USA, Cat. SAB4600041) was added as a secondary antibody at 1:20 working dilution (antibody was diluted in PBS and Evans blue was added in ﬁnal mixture to achieve counterstain concentration of 0.5%) and incubated for 45 min in humidiﬁed chamber at 37 ◦C. The slide was then washed twice with PBS (pH 7.2, Thermo Scientiﬁc, Waltham, MA, USA). Presence of B. afzelii antigens on the slides was conﬁrmed by adding to each slide a positive control (i.e., B. afzelii antigen recognition by polyclonal anti-Borrelia antibodies diluted 1:20 and labeled with ﬂuorescein isothiocyanate (FITC) (Invitrogen, Carlsbad, CA, USA, Cat. PA1-73005). The absence of non-speciﬁc reactions of the secondary antibody was conﬁrmed by adding a negative control (i.e., B. afzelii antigen directly exposed to secondary antibodies without primary antibodies) on each slide. Fluorescence reactions were visualized on a Leica DM 3000 microscope (Leica, Wetzlar, Germany) with an excitation wavelength of 515–560 nm. Results were considered positive when an intense ﬂuorescence reaction was detected at 1:100 or higher sera dilution. In case of weak ﬂuorescence at 1:100 dilution, the sample was declared as borderline and additionally tested using the second-tier recomBlot test (Mikrogen Diagnostik, Neuried, Germany, Cat. 4272), based on recombinant Borrelia spp. antigens. Inconclusive results in the second-tier test were considered negative. 5 of 11 Pathogens 2021, 10, 399 2.7. Data Analysis 2.7. Data Analysis The association between the seroprevalence of anti-Borrelia IgG and the exposure to rabies virus between groups of cases (PEP and PrEP) and control (healthy donors), and further between demographic groups (i.e., gender, age, residence, and immunization protocol), was explored on a series of contingency tables (i.e., observed and expected events for each group), assessed using two-tailed Chi-square (χ2) test; Yates’s correction was used to prevent overestimation of statistical signiﬁcance because of the small sample size (n) in the study. The test was implemented in the statistical software package IBM SPSS Statistics 25 (IBM, Armonk, NY, USA). The association between Borrelia spp. seropositivity and rabies virus risk exposure in the case groups (i.e., PEP and PrEP) and control group (i.e., healthy donors) were calculated by the odds ratio (OR) (CI 95%). Statistical signiﬁcance was considered for p-values < 0.05. Data analysis was performed using GraphPad Prism v.8.0.1 (GraphPad Software Inc., La Jolla, CA, USA). 3. Results 3.1. Association between Borrelia Seroprevalence and Risk of Rabies Exposure Anti-Borrelia IgG were detected in 20 of the 130 analyzed samples (15.38%; CI, 9.10–21.66) In a second-tier test, three samples (2 PEP and 1 PrEP) were borderline and considered as negative. The highest Borrelia spp. seroprevalence (18/100, 18%) was found among individuals at risk of rabies virus exposure (i.e., PEP and PrEP case groups). Speciﬁcally, anti-Borrelia IgG were detected in 19.0% (8/42) and 17.2% (10/58) samples of the PEP and PrEP groups, respectively. In contrast, only 6.67% (2/30) samples were seropositive for Borrelia spp. in the healthy donors group (i.e., control group), composed by individuals without risk of rabies virus exposure (Table 1). Table 1. Borrelia seroprevalence in individuals at risk of rabies and healthy blood donors. Groups Rabies Virus Exposure Risk Anti-Borrelia IgG Detection Borrelia spp. Seroprevalence (%) CI (95%) Negative Positive Healthy donors No 28 2 6.67 1.16–23.51 “Post-exposure protocol” (PEP) Yes 34 8 19.05 9.14–34.63 “Pre-exposure protocol” (PrEP) Yes 48 10 17.24 9.00–29.88 Table 1. Borrelia seroprevalence in individuals at risk of rabies and healthy blood donors. Table 1. Borrelia seroprevalence in individuals at risk of rabies and healthy blood donors. Groups Rabies Virus Exposure Risk Anti-Borrelia IgG Detection Borrelia spp. Seroprevalence (%) CI (95%) Negative Positive Healthy donors No 28 2 6.67 1.16–23.51 “Post-exposure protocol” (PEP) Yes 34 8 19.05 9.14–34.63 “Pre-exposure protocol” (PrEP) Yes 48 10 17.24 9.00–29.88 Borrelia spp. seroprevalence in PEP (χ2 = 2.24, p = 0.13) and PrEP (χ2 = 1.87, p = 0.17) case groups was not statistically signiﬁcant compared with the healthy donors control group. In addition, no signiﬁcant difference (χ2 = 0.05, p = 0.81) was found in the Borrelia spp. seroprevalence of the case groups. Although the differences did not reach statisti- cal signiﬁcance, odds ratio (OR) analysis showed that Borrelia spp. seroprevalence was positively correlated (OR > 1) with risk of rabies virus exposure in the PEP (OR = 3.29; CI, 0.65–16.78; p = 0.13) and PrEP (OR = 2.91; CI, 0.59–14.27; p = 0.17) case groups compared with the healthy donors control group. In contrast, comparison of Borrelia spp. seropreva- lence in PEP and PrEP case groups revealed an OR very close to 1 (OR = 1.12; CI, 0.40–3.15; p = 0.13), suggesting that the outcome (i.e., Borrelia spp. seroprevalence) is independent of the vaccination protocol (i.e., PEP and PrEP). 3.2. 3. Results Excluding the results from the Belgrade area (due to small sample size n = 2), Borrelia spp. seroprevalence was higher in Northern Serbia. The local administrative unit level 6 of 11 G were xclud- Pathogens 2021, 10, 399 analysis showed that most seropositive patients in Northern Serbia were concentrated in the municipalities Šid, Baˇcki Petrovac, Srbobran, Temerin, Plandište, and In ¯dija (Figure 1). p g y showed that most seropositive patients in Northern Serbia were concentrated in the mu- nicipalities Šid, Bački Petrovac, Srbobran, Temerin, Plandište, and Inđija (Figure 1). Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality name (green is for seropositive and orange for seronegative). Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality ( i f i i d f i ) Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality name (green is for seropositive and orange for seronegative). Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. 3. Results Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality name (green is for seropositive and orange for seronegative). Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality name (green is for seropositive and orange for seronegative). Figure 1. Distribution of seropositive and seronegative samples in study subjects from Serbia (state and district borders marked with blue line) and Bosnia and Herzegovina (state and region borders marked with black line). Municipalities with positive samples are shaded in green, while locations of seronegative samples are crossed with orange lines. Ratio of seropositive and seronegative samples in individuals from one municipality is shown in pie chart near each municipality name (green is for seropositive and orange for seronegative). When considering Borrelia spp. seropositive cases in both PEP and PrEP case groups, no significant differences (χ² = 1.54, p = 0.46) were found in seroprevalence between the North (15/69, 21.73%), Central (1/12, 8.33%) and South (2/15, 13.33%) regions of Serbia. As all healthy donor samples were from Northern Serbia, we compared the Borrelia spp. se- roprevalence among the two case and control groups in Northern Serbia. The difference When considering Borrelia spp. seropositive cases in both PEP and PrEP case groups, no signiﬁcant differences (χ2 = 1.54, p = 0.46) were found in seroprevalence between the North (15/69, 21.73%), Central (1/12, 8.33%) and South (2/15, 13.33%) regions of Serbia. As all healthy donor samples were from Northern Serbia, we compared the Borrelia spp. seroprevalence among the two case and control groups in Northern Serbia. The difference in Borrelia spp. seroprevalence between PEP (6/26, 23.08%; χ2 = 3.1, p = 0.08) and PrEP (9/43, 20.93%; χ2 = 2.8, p = 0.09) case groups in the North region was not statistically signiﬁcant compared with that of the healthy donors control group. However, the OR analysis revealed greater odds of association of Borrelia spp. 3. Results There was no significant difference (χ² = 0.23, p = 0.62) in the Borrelia spp. seroprev- alence of PEP (3/22, 13.64%) and PrEP (10/45, 18.18%) case groups. No significant differ- ences (Chi-square test p = 1) in Borrelia spp. seroprevalence were detected in seniors when PEP (3/14, 21.42%) and PrEP (0/3) case groups were compared between them. In the healthy donors control group, we found anti-Borrelia IgG only in two individuals from the adult age group (2/27; 7.41%), while no seropositive individuals were found in seniors (0/1) or children and teenagers (0/2) age groups. In the adults, Borrelia spp. seroprevalence was associated with rabies virus exposure risk in the PEP (OR = 2.0; CI, 0.37–12.5; χ² = 0.51, p = 0.47) and PrEP (OR = 2.8; CI, 0.65–13; χ² = 1.7, p = 0.19) case groups, compared with the healthy donors control group (Figure 2a). In addition, compared with the healthy donors control group of the same age groups, rabies virus exposure risk was not found to be sig- nificantly associated with Borrelia spp. seroprevalence in children and teenagers (Chi- square test p = 1) or seniors (Chi-square test p = 1), regardless of immunization protocol. Figure 2. Seroprevalence of anti-Borrelia IgG among individuals of the proposed case-control study groups. The case groups PEP (“post-exposure protocol” group) and PrEP (“pre-exposure protocol” group) at risk of rabies virus exposure are compared with a control group (unexposed healthy blood donors). (a) Borrelia spp. seroprevalence and its association with rabies virus exposure were compared between individuals belonging to the age groups Children (children and teen- agers), Adults, and Seniors, as well as overall (Total). (b) Comparison of Borrelia spp. seroprevalence and association with risk of rabies virus exposure among male and female individuals from the different case-control study groups. OR: odds ratio, ∞ (inﬁnite upper limit) when the prevalence in the control group was 0. Asterisks denote statistically signiﬁcant differences (* p < 0.05, ns: non-significant). Figure 2. Seroprevalence of anti-Borrelia IgG among individuals of the proposed case-control study groups. The case groups PEP (“post-exposure protocol” group) and PrEP (“pre-exposure protocol” group) at risk of rabies virus exposure are compared with a control group (unexposed healthy blood donors). (a) Borrelia spp. seroprevalence and its association with rabies virus exposure were compared between individuals belonging to the age groups Children (children and teenagers), Adults, and Seniors, as well as overall (Total). 3. Results (b) Comparison of Borrelia spp. seroprevalence and association with risk of rabies virus exposure among male and female individuals from the different case-control study groups. OR: odds ratio, ∞ (inﬁnite upper limit) when the prevalence in the control group was 0. Asterisks denote statistically signiﬁcant differences (* p < 0.05, ns: non-signiﬁcant). Figure 2. Seroprevalence of anti-Borrelia IgG among individuals of the proposed case-control study groups. The case groups PEP (“post-exposure protocol” group) and PrEP (“pre-exposure protocol” group) at risk of rabies virus exposure are compared with a control group (unexposed healthy blood donors). (a) Borrelia spp. seroprevalence and its association with rabies virus exposure were compared between individuals belonging to the age groups Children (children and teen- agers), Adults, and Seniors, as well as overall (Total). (b) Comparison of Borrelia spp. seroprevalence and association with risk of rabies virus exposure among male and female individuals from the different case-control study groups. OR: odds ratio, ∞ (inﬁnite upper limit) when the prevalence in the control group was 0. Asterisks denote statistically signiﬁcant differences (* p < 0.05, ns: non-significant). Figure 2. Seroprevalence of anti-Borrelia IgG among individuals of the proposed case-control study groups. The case groups PEP (“post-exposure protocol” group) and PrEP (“pre-exposure protocol” group) at risk of rabies virus exposure are compared with a control group (unexposed healthy blood donors). (a) Borrelia spp. seroprevalence and its association with rabies virus exposure were compared between individuals belonging to the age groups Children (children and teenagers), Adults, and Seniors, as well as overall (Total). (b) Comparison of Borrelia spp. seroprevalence and association with risk of rabies virus exposure among male and female individuals from the different case-control study groups. OR: odds ratio, ∞ (inﬁnite upper limit) when the prevalence in the control group was 0. Asterisks denote statistically signiﬁcant differences (* p < 0.05, ns: non-signiﬁcant). The analysis of gender distribution in enrolled individuals from case groups showed that 15% (6/40; CI, 3.43–26.57) and 20% (12/60; CI, 9.58–30.42) of seropositive samples to anti-Borrelia IgG antibodies were women and men, respectively. No signiﬁcant difference was found in the Borrelia spp. seroprevalence between women and men (χ2 = 0.29, p = 0. 58). For the healthy donor group, all seropositive samples were from women (2/13, 15.38%; CI, 0.00–38.07). The comparison of the relative odds showed that male individuals of the PEP group had more chance of being seropositive to Borrelia spp. 3. Results seroprevalence with the rabies virus exposure in PEP (OR = 4.2; CI, 0.88–22.00, p = 0.08) and PrEP (OR = 3.7; CI, 0.82–18.00, p = 0.09) compared with the healthy donors control group. 3.3. Association between Borrelia spp. Seroprevalence and Rabies Virus Exposure Risk According to Age and Gender Individuals at risk of rabies virus exposure were unequally distributed in the age groups, and the highest seroprevalence was detected in children and teenagers (2/6, 33.33%; CI, 5.90–75.80), followed by seniors (3/17, 17.64%; CI, 4.67–44.19), and adults (13/77, 16.88%; CI, 9.64–27.50). Due to the fact that PrEP requires the existence of a profession- 7 of 11 g rs (2/6, d lt Pathogens 2021, 10, 399 associated risk of rabies virus exposure, no children or teenagers could undergo this type of vaccination protocol; therefore, comparison of Borrelia spp. seroprevalence between PEP and PrEP case groups in this age category was not possible in the present study. There was no signiﬁcant difference (χ2 = 0.23, p = 0.62) in the Borrelia spp. seroprevalence of PEP (3/22, 13.64%) and PrEP (10/45, 18.18%) case groups. No signiﬁcant differences (Chi-square test p = 1) in Borrelia spp. seroprevalence were detected in seniors when PEP (3/14, 21.42%) and PrEP (0/3) case groups were compared between them. In the healthy donors control group, we found anti-Borrelia IgG only in two individuals from the adult age group (2/27; 7.41%), while no seropositive individuals were found in seniors (0/1) or children and teenagers (0/2) age groups. In the adults, Borrelia spp. seroprevalence was associated with rabies virus exposure risk in the PEP (OR = 2.0; CI, 0.37–12.5; χ2 = 0.51, p = 0.47) and PrEP (OR = 2.8; CI, 0.65–13; χ2 = 1.7, p = 0.19) case groups, compared with the healthy donors control group (Figure 2a). In addition, compared with the healthy donors control group of the same age groups, rabies virus exposure risk was not found to be signiﬁcantly associated with Borrelia spp. seroprevalence in children and teenagers (Chi-square test p = 1) or seniors (Chi-square test p = 1), regardless of immunization protocol. sion-associated risk of rabies virus exposure, no children or teenagers could undergo this type of vaccination protocol; therefore, comparison of Borrelia spp. seroprevalence be- tween PEP and PrEP case groups in this age category was not possible in the present study. 4. Discussion Lyme borreliosis (LB) is the most common TBD in Europe and is the cause of major concern for both public and veterinary health due to its considerable impact on animal health and human life quality. To the best of our knowledge, this is the ﬁrst study to consider that individuals at risk of rabies virus exposure also have high odds of Borrelia infection. The present study revealed an 18% of Borrelia spp. seroprevalence in individuals immunized against rabies, which is higher than any value previously reported by other authors in the healthy population from Serbia [14,27]. In addition, the OR analysis showed a strong association between Borrelia spp. seroprevalence with rabies virus exposure risk in the PEP and PrEP case groups, which could be due to activities shared by individuals at risk of rabies and LB, including extensive contact with animals and exposure to tick bites, as reported by most of the study individuals. The lack of statistically signiﬁcant differences in the comparisons of Borrelia spp. seroprevalence between groups may be associated with one of the main limitations of our study (i.e., the small sample size of case and control groups). This ﬁnding is in agreement with previous studies conducted in Europe that have described high exposure risk to B. burgdorferi s.l. complex in many occupational groups, including forestry workers, farmers, veterinarians, military recruits, and outdoor workers [28–30]. It is worth mentioning that anti-Borrelia antibodies are not necessarily associated with clinical symptoms of LB and that it is currently unknown for how long these antibodies last in the bloodstream [31]. Therefore, the use of serological tests alone does not sufﬁce to distinguish past from newly acquired infections [32]. The identiﬁcation of new risk groups related to TBDs is crucial for the improvement and implementation of surveillance programs aimed at the prevention and control of this group of diseases. In the present study, subjects immunized against rabies showed higher Borrelia spp. seroprevalence and increased likelihood of exposure to ticks and tick-borne pathogens, compared with all previously tested groups from Serbia [17,33]. When the immunization protocol was considered, it became evident that individuals immunized against rabies via PrEP (17.2%) showed similar seroprevalence to soldiers working in the Belgrade area (17.1%) [17]. 3. Results compared with the healthy donors (OR = ∞, CI: 1.3–∞; χ2 = 4.9, p = 0.03) (Figure 2b). No signiﬁcant differences were found between the PrEP and healthy donor groups (OR = ∞, CI, 0.7–∞; χ2 = 3.4, p = 0.07). No signiﬁcant association was found between Borrelia spp. seropositivity and the exposure to rabies virus risk in females from PEP (0.87, CI, 0.16-5.5; χ2 = 0.02, p = 0. 89) and PrEP (1.1, CI, 0.2–7; χ2 = 0.01, p = 0.92) case groups (Figure 2b). 8 of 11 Pathogens 2021, 10, 399 4. Discussion This ﬁnding is explained by the fact that those professionals are more fre- quently working outdoors in natural environments, where they can be at risk of tick bites, as well as in close contact with animals, including military dogs, hunting dogs, and ani- mals at the veterinary examination centers [17]. On the other hand, subjects immunized via PEP (19.0%) showed similar seroprevalence as Belgrade park maintenance workers (23.5%) and Slovenian forest workers (23.8%) [33,34]. The reason for such similarity in seroprevalence could be the similar lifestyle of these cohorts since edges of forests and places with transitional vegetation have been identiﬁed as high-risk areas for humans to ticks exposure [35]. If we consider that the majority of contact with rabid or suspected rabid animals in Serbia could happen in rural areas or wilderness, there is the possibility that PEP patients spend a considerable amount of time in rural environment activities (e.g., recreation, animal husbandry, agriculture, beekeeping, mushroom picking, among others) and possibly live in rural areas where garden maintenance is a common hobby and can share a similar risk of tick bite as those who are involved in park maintenance activity in cities or who work in a forest environment. Nevertheless, it is necessary to conduct further research to clarify which activities related to high risk of tick bite exposure are present in patients that are immunized against rabies via PEP. Unfortunately, there are no data about seroprevalence in the general population from Central and South Serbia, which makes it impossible to compare with subjects immunized against rabies from these regions. In addition, seroprevalence found in this study group from Northern Serbia was similar to seroprevalence in professionals at increased risk of LB in Slovenia (21.73% vs. 23.8%) [34]. Previous studies conducted in high-risk profession- als for LB from Belgrade reported that seroprevalence of anti-Borrelia IgG antibodies in forest workers, park maintenance workers, and soldiers was 11.76%, 23.5%, and 17.14%, respectively [8,21]. In the present study, only two persons immunized against rabies from Belgrade were included, and one of them was seropositive to Borrelia spp. Although this Pathogens 2021, 10, 399 9 of 11 9 of 11 is an interesting result, the small sample size makes it impossible to make a meaningful comparison with previous seroprevalence studies in Belgrade. 4. Discussion Observed seroprevalence in a population immunized against rabies suggests possible overlap of Borrelia exposure risk that was previously linked only to speciﬁc professions. Since the majority of seropositive persons are living in northern parts of Serbia, it should be further investigated whether the shared odds of Borrelia and rabies virus exposure are present only to speciﬁc regions of the country. In regions where shared odds of Borrelia and rabies virus exposure are conﬁrmed, a series of preventive measures related to TBDs can be implemented simultaneously with rabies prophylaxis, including educational and screening programs, as well as recommenda- tion of preventive immunizations against ticks or most common TBDs for which vaccines are available. In persons who have been infected with Borrelia spp. and have achieved the serocon- version, IgG remain detectable for longer periods and have greater speciﬁcity for binding to epitopes compared with IgM. For this reason, we only tested exposure to Borrelia spp. by screening anti-Borrelia IgG. It is considered that Borrelia seroprevalence in a region is dependent upon LB endemicity. Except for few studies, there are no published data on Borrelia seroprevalence in the human population from other countries bordering Serbia, including Bulgaria, Bosnia and Herzegovina, Montenegro, North Macedonia, and Albania. Countries neighboring Serbia on the north and west with LB endemic regions are Hun- gary, Slovenia, and Croatia [27,34,36–38]. In endemic regions from Slovenia and Croatia, the reported seroprevalence values of anti-Borrelia IgG in healthy persons were between 9.7% and 44% [27,34,36]. On the other hand, a study conducted in a population from a non-endemic region in Croatia reported lower seroprevalence values of Borrelia spp. (2.7–4.7%) [27]. Similarly, Hristea et al. (2001) reported an overall 4.3% seroprevalence of anti-Borrelia IgG antibodies in a healthy population of voluntary blood donors from Romania, which is bordering Serbia on the east [38]. In Serbia, the presence of anti-Borrelia IgG has been previously determined in a healthy population from South Baˇcka District (3.22%) in Northern Serbia and in Belgrade city (2.9–8.57%) [17,33,39]. The variation in Borrelia spp. seroprevalence rates in Belgrade could be likely attributed to several fac- tors, including the demography of the sampled population (i.e., the type and number of samples analyzed), the extent of tick infestations, and the sensitivity of diagnostic assays employed [40]. Other studies reported 9.7% and 4.3% Borrelia spp. seroprevalence in the Slovenian general population [19] and in blood donors in Romania, respectively [23]. 5. Conclusions Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Faculty of Medicine, the University of Novi Sad (protocol code 01-39/266/1 and approved 16 October 2020). Informed Consent Statement: Informed consent was obtained from subjects involved in the study. Informed Consent Statement: Informed consent was obtained from subjects involved in the study. Acknowledgments: The authors would like to thank Snežana Tomanovi´c (the University of Belgrade, Institute for Medical Research, Group of Medical Entomology) for providing B. afzelli culture. UMR BIPAR is supported by the French Government’s Investissement d’Avenir program, Laboratoire d’Excellence “Integrative Biology of Emerging Infectious Diseases” (grant no. ANR-10-LABX-62- IBEID). Acknowledgments: The authors would like to thank Snežana Tomanovi´c (the University of Belgrade, Institute for Medical Research, Group of Medical Entomology) for providing B. afzelli culture. UMR BIPAR is supported by the French Government’s Investissement d’Avenir program, Laboratoire d’Excellence “Integrative Biology of Emerging Infectious Diseases” (grant no. ANR-10-LABX-62- IBEID). Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Madison-Antenucci, S.; Kramer, L.D.; Gebhardt, L.L.; Kauffman, E. Emerging Tick-Borne Diseases. Clin. Microbiol. Rev. 2020, 33, e00083-18. [CrossRef] [PubMed] 2. Mysterud, A.; Jore, S.; Østerås, O.; Viljugrein, H. Emergence of Tick-Borne Diseases at Northern Latitudes in Europe: A Comparative Approach. Sci. Rep. 2017, 7, 16316. [CrossRef] [PubMed] 3. Medlock, J.M.; Hansford, K.M.; Bormane, A.; Derdakova, M.; Estrada-Peña, A.; George, J.-C.; Golovljova, I.; Jaenson, T.G.T.; Jensen, J.-K.; Jensen, P.M.; et al. Driving Forces for Changes in Geographical Distribution of Ixodes Ricinus Ticks in Europe. Parasites Vectors 2013, 6, 1. [CrossRef] [PubMed] 4. Ehrmann, S.; Liira, J.; Gärtner, S.; Hansen, K.; Brunet, J.; Cousins, S.A.O.; Deconchat, M.; Decocq, G.; De Frenne, P.; De Smedt, P.; et al. Environmental Drivers of Ixodes Ricinus Abundance in Forest Fragments of Rural European Landscapes. BMC Ecol. 2017, 17, 31. [CrossRef] [PubMed] 5. Ruiz-Fons, F.; Fernández-de-Mera, I.G.; Acevedo, P.; Gortázar, C.; de la Fuente, J. Factors Driving the Abundance of Ixodes Ricinus Ticks and the Prevalence of Zoonotic I. Ricinus-Borne Pathogens in Natural Foci. Appl. Environ. Microbiol. 2012, 78, 2669–2676. [CrossRef] 6. Gray, J.S. The Ecology of Ticks Transmitting Lyme Borreliosis. Exp. Appl. Acarol. 1998, 22, 249–258. 7. Gryczy´nska, A.; Zgódka, A.; Płoski, R.; Siemiatkowski, M. Borrelia Burgdorferi Sensu Lato Infection in Passerine Birds from the Mazurian Lake Region (Northeastern Poland). Avian Pathol. 2004, 33, 69–75. [CrossRef] 7. Gryczy´nska, A.; Zgódka, A.; Płoski, R.; Siemiatkowski, M. Borrelia Burgdorferi Sensu Lato Infectio Mazurian Lake Region (Northeastern Poland). Avian Pathol. 2004, 33, 69–75. [CrossRef] g 8. Azagi, T.; Hoornstra, D.; Kremer, K.; Hovius, J.W.R.; Sprong, H. Evaluation of Disease Causality of Rare Ixodes Ricinus-Borne Infections in Europe. Pathogens 2020, 9, 150. [CrossRef] p g 9. Rizzoli, A.; Silaghi, C.; Obiegala, A.; Rudolf, I.; Hubálek, Z.; Földvári, G.; Plantard, O.; Vayssier-Taussat, M.; Bonnet, S.; Spitalská, E.; et al. Ixodes Ricinus and Its Transmitted Pathogens in Urban and Peri-Urban Areas in Europe: New Hazards and Relevance for Public Health. Front. Public Health 2014, 2, 251. [CrossRef] 10. Sykes, R.A.; Makiello, P. An Estimate of Lyme Borreliosis Incidence i 10. Sykes, R.A.; Makiello, P. An Estimate of Lyme Borreliosis Incidence in Western Europe†. J. Public Health 2 10. Sykes, R.A.; Makiello, P. An Estimate of Lyme Borreliosis Incidence in Western Europe†. J. Public Health 2017, 39, 74–81. [CrossRef] 11. Rudenko, N.; Golovchenko, M.; Grubhoffer, L.; Oliver, J.H. 5. Conclusions The results of the present study strongly suggest that individuals at risk of rabies virus exposure also have a high likelihood of exposure to tick bites and Borrelia infection in Serbia. Although the small sample size and unequal distribution of individuals in both case and control groups are a main limitation of our study, the obtained results suggest that Borrelia spp. infection could be a newly recognized occupational hazard in Serbia for professionals working in rabies-risky areas. Individuals following either the PrEP or PEP prophylaxis protocols are good cohorts when considering shared odds of rabies virus and Borrelia exposure. The higher distribution of Borrelia seroprevalence in several municipalities from Northern Serbia suggests that risk of rabies virus exposure is highly associated with Borrelia infection in Northern Serbia. The data obtained in this study indicate that further research is needed to increase the sample size and extend the observations of this study to other TBDs and regions of Serbia. These ﬁndings should be considered by physicians and policy-makers to guide risk assessment and public health policies for TBDs at the population level. Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/pathogens10040399/s1, Table S1: Demographic data of individuals vaccinated against rabies following the “post-exposure protocol” (PEP), Table S2: Demographic data of individuals vaccinated against rabies following the “pre-exposure protocol” (PrEP), Table S3: Demographic data of healthy blood donors included in the study. 10 of 11 Pathogens 2021, 10, 399 10 of 11 Author Contributions: Conceptualization, P.B. and A.C.-C.; methodology, P.B., D.O.; validation, D.M., V.S., N.V. and D.V.; formal analysis, P.B., A.A.D.-S., D.O.; investigation, D.M., V.S., Z.H., P.B., Z.B.-O.; resources, P.B., N.V., D.M., D.V., J.G., Z.B.-O.; data curation, P.B., N.B., Z.H.; writing—original draft preparation, P.B., A.A.D.-S., D.O., A.C.-C.; writing—review and editing, P.B., A.C.-C., A.A.D.-S., D.O., A.J., A.G.; visualization, D.O.; supervision, P.B., A.C.-C.; All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Faculty of Medicine, the University of Novi Sad (protocol code 01-39/266/1 and approved 16 October 2020). References WHO Expert Consultation on Rabies: Third Report; World Health Organizati 2018; ISBN 978-92-4-121021-8. nisation. WHO Expert Consultation on Rabies: Third Report; World Health Organization: Geneva, Switzerland -121021-8. 22. Picot, V.; Rasuli, A.; Abella-Rider, A.; Saadatian-Elahi, M.; Aikimbayev, A.; Barkia, A.; Benmaiz, S.; Bouslama, Z.; De Balogh, K.; Dehove, A.; et al. The Middle East and Eastern Europe Rabies Expert Bureau (MEEREB) Third Meeting: Lyon-France (7–8 April 2015). J. Infect. Public Health 2017, 10, 695–701. [CrossRef] [PubMed] 23. Pattanaik, A.; Mani, R.S. WHO’s New Rabies Recommendations: Implications for High Incidence Cou Dis. 2019, 32, 401–406. [CrossRef] 24. Taylor, E.; Del Rio Vilas, V.; Scott, T.; Coetzer, A.; Prada, J.M.; Alireza, G.; Alqadi, N.A.; Berry, A.; Bazzal, B.; Barkia, A.; et al. Rabies in the Middle East, Eastern Europe, Central Asia, North Africa: Building. J. Infect. Public Health 2021, in press. [CrossRef] 25. Barbour, A.G. Isolation and Cultivation of Lyme Disease Spirochetes. Yale J. Biol. Med. 1984, 57, 521–525. 24. Taylor, E.; Del Rio Vilas, V.; Scott, T.; Coetzer, A.; Prada, J.M.; Alireza, G.; Alqadi, N.A.; Berry, A.; Bazzal, B.; Barkia, A.; et al. Rabies in the Middle East, Eastern Europe, Central Asia, North Africa: Building. J. Infect. Public Health 2021, in press. [CrossRef] p g f p 25. Barbour, A.G. Isolation and Cultivation of Lyme Disease Spirochetes. Yale J. Biol. Med. 1984, 57, 521–525. 26. OIE. Manual of Diagnostic Tests and Vaccines for Terrestrial Animals (Mammals, Birds and Bees), 6th ed.; OIE Animal Health: Paris, France, 2008; ISBN 978-92-9044-718-4. 27. Poljak, I.; Troselj-Vuki´c, B.; Mileti´c, B.; Morovi´c, M.; Ruzi´c-Sablji´c, E.; Vucemilovi´c, A.; Materljan, E. Low Sero-Prevalence of Lyme Borreliosis in the Forested Mountainous Area of Gorski Kotar, Croatia. Croat. Med. J. 2000, 41, 433–436. [PubMed] 28. De Keukeleire, M.; Vanwambeke, S.O.; Somassè, E.; Kabamba, B.; Luyasu, V.; Robert, A. Scouts, Forests, and Ticks: Impact of Landscapes on Human-Tick Contacts. Ticks Tick-Borne Dis. 2015, 6, 636–644. [CrossRef] [PubMed] 29. De Keukeleire, M.; Robert, A.; Kabamba, B.; Dion, E.; Luyasu, V.; Vanwambeke, S.O. Individual and Environmental Factors Associated with the Seroprevalence of Borrelia Burgdorferi in Belgian Farmers and Veterinarians. Infect. Ecol. Epidemiol. 2016, 6, 32793. [CrossRef] [PubMed] 30. Tomao, P.; Ciceroni, L.; D’Ovidio, M.C.; De Rosa, M.; Vonesch, N.; Iavicoli, S.; Signorini, S.; Ciarrocchi, S.; Ciufolini, M.G.; Fiorentini, C.; et al. Prevalence and Incidence of Antibodies to Borrelia Burgdorferi and to Tick-Borne Encephalitis Virus in Agricultural and Forestry Workers from Tuscany, Italy. Eur. References Updates on Borrelia Burgdorferi Sensu Lato Complex with Respect to Public Health Ticks Tick-Borne Dis 2011 2 123–128 [CrossRef] 10. Sykes, R.A.; Makiello, P. An Estimate of Lyme Borreliosis Incidence in Western Europe†. J. Public Health 2017, 39, 74–81. [CrossRef] 11 Rudenko N ; Golovchenko M ; Grubhoffer L ; Oliver J H Updates on Borrelia Burgdorferi Sensu Lato Complex with Respect to 11. Rudenko, N.; Golovchenko, M.; Grubhoffer, L.; Oliver, J.H. Updates on Borrelia Burgdorferi Sensu Lato Public Health. Ticks Tick-Borne Dis. 2011, 2, 123–128. [CrossRef] 12. Banovi´c, P.; ˇCapo, I.; Ogorelica, D.; Vranješ, N.; Simin, V.; Laloševi´c, D. Mysterious Path of Borrelia Spielmanii: Spreading without Morphological Alteration of Collagen Type I and IV. Future Microbiol. 2019, 14, 1469–1475. [CrossRef] 13. Steere, A.C.; Strle, F.; Wormser, G.P.; Hu, L.T.; Branda, J.A.; Hovius, J.W.R.; Li, X.; Mead, P.S. Lyme Borreliosis. Nat. Rev. Dis. Primers 2016, 2, 16090. [CrossRef] [PubMed] 14. Potkonjak, A.; Savi´c, S.; Ruzi´c-Sablji´c, E.; Vraˇcar, V.; Lako, B.; Juriši´c, A.; Petrovi´c, A.; Rajkovi´c, D. Molecular Characterization of Borrelia Strains Isolated from Ticks in Vojvodina. Parasites Vectors 2014, 7, P18. [CrossRef] 15. Savi´c, S.; Vidi´c, B.; Lazi´c, S.; Lako, B.; Potkonjak, A.; Lepšanovi´c, Z. Borrelia Burgdorferi in Ticks and Dogs in the Province of Vojvodina, Serbia. Parasite 2010, 17, 357–361. [CrossRef] [PubMed] 16. Potkonjak, A.; Petrovi´c, T.; Ristanovi´c, E.; Lali´c, I.; Vraˇcar, V.; Savi´c, S.; Turkulov, V.; ˇCanak, G.; Miloševi´c, V.; Vidanovi´c, D.; et al. Molecular Detection and Serological Evidence of Tick-Borne Encephalitis Virus in Serbia. Vector-Borne Zoonotic Dis. 2017, 17, 813–820. [CrossRef] 11 of 11 Pathogens 2021, 10, 399 17. Jovanovic, D.; Atanasievska, S.; Protic-Djokic, V.; Rakic, U.; Lukac-Radoncic, E.; Ristanovic, E. Seroprevalence of Borrelia Burgdorferi in Occupationally Exposed Persons in the Belgrade Area, Serbia. Braz. J. Microbiol. Publ. Braz. Soc. Microbiol. 2015, 46, 807–814. [CrossRef] 18. Tomanovi´c, S.; Radulovi´c, Ž.; Masuzawa, T.; Milutinovi´c, M. Coexistence of Emerging Bacterial Path in Serbia. Parasite 2010, 17, 211–217. [CrossRef] [PubMed] 19. Banovi´c, P.; Díaz-Sánchez, A.A.; Galon, C.; Simin, V.; Mijatovi´c, D.; Obregón, D.; Moutailler, S.; Cabezas-Cruz, A. Humans Infested with Ixodes Ricinus Are Exposed to a Diverse Array of Tick-Borne Pathogens in Serbia. Ticks Tick-Borne Dis. 2020, 12, 101609. [CrossRef] 20. Walker, P.J.; Blasdell, K.R.; Calisher, C.H.; Dietzgen, R.G.; Kondo, H.; Kurath, G.; Longdon, B.; Stone, D.M.; Tesh, R.B.; Tordo, N.; et al. ICTV Virus Taxonomy Proﬁle: Rhabdoviridae. J. Gen. Virol. 2018, 99, 447–448. [CrossRef] 21. World Health Organisation. References J. Clin. Microbiol. Infect. Dis. Off. Publ. Eur. Soc. Clin. Microbiol. 2005, 24, 457–463. [CrossRef] 31. Kalish, R.A.; McHugh, G.; Granquist, J.; Shea, B.; Ruthazer, R.; Steere, A.C. Persistence of Immunoglobulin M or Immunoglobulin G Antibody Responses to Borrelia Burgdorferi 10-20 Years after Active Lyme Disease. Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am. 2001, 33, 780–785. [CrossRef] 32. Borchers, A.T.; Keen, C.L.; Huntley, A.C.; Gershwin, M.E. Lyme Disease: A Rigorous Review of Diagnostic Criteria and Treatment. J. Autoimmun. 2015, 57, 82–115. [CrossRef] 33. Krstic, M.; Stajkovic, N. Risk for infection by Lyme disease cause in green surfaces maintenance workers in Belgrade. Vojnosanit. Pregl. 2007, 64, 313–318. [CrossRef] [PubMed] 34. Rojko, T.; Ruži´c-Sablji´c, E.; Strle, F.; Lotriˇc-Furlan, S. Prevalence and Incidence of Lyme Borreliosis among Slovene Forestry Workers during the Period of Tick Activity. Wien. Klin. Wochenschr. 2005, 117, 219–225. [CrossRef] [PubMed] 35. Lambin, E.F.; Tran, A.; Vanwambeke, S.O.; Linard, C.; Soti, V. Pathogenic Landscapes: Interactions betwe Vectors, and Their Animal Hosts. Int. J. Health Geogr. 2010, 9, 54. [CrossRef] Vectors, and Their Animal Hosts. Int. J. Health Geogr. 2010, 9, 54. [CrossRef] 36. Burek, V.; Miši´c-mayerus, L.; Maretic, T. Antibodies to Borrelia Burgdorferi in Various Population Groups in Croatia. Scand. J. Infect. Dis. 1992, 24, 683–684. [CrossRef] 36. Burek, V.; Miši´c-mayerus, L.; Maretic, T. Antibodies to Borrelia Burgdorferi in Various Population Gr Infect. Dis. 1992, 24, 683–684. [CrossRef] 37. Zöldi, V.; Juhász, A.; Nagy, C.; Papp, Z.; Egyed, L. Tick-Borne Encephalitis and Lyme Disease in Hungary: The Epidemiological Situation Between 1998 and 2008. Vector-Borne Zoonotic Dis. 2013, 13, 256–265. [CrossRef] 38. Hristea, A.; Hristescu, S.; Ciufecu, C.; Vasile, A. Seroprevalence of Borrelia Burgdorferi in Romania. Eur. J. Epidemiol. 2001, 17, 891–896. [CrossRef] 39. Hrnjakovi´c Cvjetkovi´c, I.; Petrovi´c, T.; Petri´c, D.; Cvjetkovi´c, D.; Kovaˇcevi´c, G.; Radovanov, J.; Jovanovi´c Galovi´c, A.; Pat, A.; Nikoli´c, N.; Stefan Miki´c, S.; et al. Seroprevalence of Mosquito-Born and Tick-Born Microorganisms in Human Population of South Baˇcka District. Arch. Veter Med. 2016, 9, 23–30. [CrossRef] South Baˇcka District. Arch. Veter Med. 2016, 9, 23–30 40. Lakos, A.; Igari, Z.; Solymosi, N. Recent Lesson from a Clinical and Seroepidemiological Survey: Low Positive Predictive Value of Borrelia Burgdorferi Antibody Testing in a High Risk Population. Adv. Med. Sci. 2012, 57, 356–363. [CrossRef] [PubMed]
https://openalex.org/W2325018865	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0152134&type=printable	English	null	The Hemagglutinin of Bat-Associated Influenza Viruses Is Activated by TMPRSS2 for pH-Dependent Entry into Bat but Not Human Cells	PloS one	2,016	cc-by	11,521	RESEARCH ARTICLE OPEN ACCESS Citation: Hoffmann M, Krüger N, Zmora P, Wrensch F, Herrler G, Pöhlmann S (2016) The Hemagglutinin of Bat-Associated Influenza Viruses Is Activated by TMPRSS2 for pH-Dependent Entry into Bat but Not Human Cells. PLoS ONE 11(3): e0152134. doi:10.1371/journal.pone.0152134 Editor: Michael CW Chan, Centre of Influenza Research, The University of Hong Kong, HONG KONG Editor: Michael CW Chan, Centre of Influenza Research, The University of Hong Kong, HONG KONG Received: December 1, 2015 Accepted: March 9, 2016 Published: March 30, 2016 Copyright: © 2016 Hoffmann et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Received: December 1, 2015 Accepted: March 9, 2016 Published: March 30, 2016 Copyright: © 2016 Hoffmann et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: © 2016 Hoffmann et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Markus Hoffmann1, Nadine Krüger2, Pawel Zmora1, Florian Wrensch1, Georg Herrler2, Stefan Pöhlmann1 1 Infection Biology Unit, German Primate Center, Göttingen, Germany, 2 Institute of Virology, University of Veterinary Medicine Hannover, Hannover, Germany 1 Infection Biology Unit, German Primate Center, Göttingen, Germany, 2 Institute of Virology, University of Veterinary Medicine Hannover, Hannover, Germany * spoehlmann@dpz.eu (SP); mhoffmann@dpz.eu (MH) * spoehlmann@dpz.eu (SP); mhoffmann@dpz.eu (MH) Abstract New World bats have recently been discovered to harbor influenza A virus (FLUAV)-related viruses, termed bat-associated influenza A-like viruses (batFLUAV). The internal proteins of batFLUAV are functional in mammalian cells. In contrast, no biological functionality could be demonstrated for the surface proteins, hemagglutinin (HA)-like (HAL) and neuramini- dase (NA)-like (NAL), and these proteins need to be replaced by their human counterparts to allow spread of batFLUAV in human cells. Here, we employed rhabdoviral vectors to study the role of HAL and NAL in viral entry. Vectors pseudotyped with batFLUAV-HAL and -NAL were able to enter bat cells but not cells from other mammalian species. Host cell entry was mediated by HAL and was dependent on prior proteolytic activation of HAL and endosomal low pH. In contrast, sialic acids were dispensable for HAL-driven entry. Finally, the type II transmembrane serine protease TMPRSS2 was able to activate HAL for cell entry indicating that batFLUAV can utilize human proteases for HAL activation. Collectively, these results identify viral and cellular factors governing host cell entry driven by batFLUAV surface proteins. They suggest that the absence of a functional receptor precludes entry of batFLUAV into human cells while other prerequisites for entry, HAL activation and proton- ation, are met in target cells of human origin. The Hemagglutinin of Bat-Associated Influenza Viruses Is Activated by TMPRSS2 for pH-Dependent Entry into Bat but Not Human Cells Markus Hoffmann1, Nadine Krüger2, Pawel Zmora1, Florian Wrensch1, Georg Herrler2, Stefan Pöhlmann1 Markus Hoffmann1, Nadine Krüger2, Pawel Zmora1, Florian Wrensch1, Georg Herrler2, Stefan Pöhlmann1 Host Cell Entry by Bat-Associated Influenza Viruses Waterfowl has been shown to constitute the natural reservoir of FLUAV [4, 5] from which viruses with pandemic potential can be directly transmitted to humans or can emerge upon reassortment of avian and human FLUAV [2, 4, 6]. Influenza pandemics might have dramatic consequences, as highlighted by the 30–50 million deaths attributed to the influenza pandemic of the years 1918/1919 (Spanish influenza [7, 8]). Competing Interests: The authors have read the journal's policy and the authors of this manuscript have the following competing interests: Stefan Pöhlmann currently serves as an academic editor for PLoS One. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. The viral surface proteins hemagglutinin (HA) and neuraminidase (NA) facilitate FLUAV entry and release from target cells, respectively. HA facilitates viral attachment to cells by bind- ing to sialic acids on cell surface proteins or lipids [9–11] and, upon proteolytic activation by a host cell protease and exposure to endosomal low pH, mediates fusion of the viral membrane with the endosomal membrane [12–14]. In contrast, NA promotes release of progeny particles from infected cells by removing sialic acids from cell surface factors. Based on sequence and antigenic properties, sixteen HA (H1-16) and nine NA (N1-9) subtypes have been identified, and viruses representing all HA and NA subtypes are circulating in waterfowl [4, 15]. However, FLUAV-related viruses were recently discovered in New World bats [16, 17], provisionally termed bat-associated influenza A-like viruses (batFLUAV), and were shown to harbor HA- and NA-like proteins (termed HAL and NAL), which constitute new subtypes, H17/H18 (HL17/HL18) and N10/N11 (NL10/NL11), respectively. The question whether these viruses have the potential to infect and spread in humans is the focus of current research efforts. Attempts to isolate batFLUAV were unsuccessful [16, 17] but, employing reverse genetics, it was demonstrated that the viral replication machinery and interferon antagonists are func- tional in mammalian cells [16, 18–21]. In contrast, the HAL and NAL proteins of batFLUAV were incompatible with viral spread in the cell culture systems examined so far [18, 19] for at present unknown reasons. Biochemical and structural studies imply that batFLUAV-HAL, unlike FLUAV-HA, does not engage sialic acids (SA) for host cell entry [17, 22–24], and that batFLUAV-NAL, unlike FLUAV-NA, neither shows neuraminidase activity nor possesses an active site that would allow interaction with sialic acids [17, 25–27]. However, it is currently unclear whether HAL and NAL can facilitate viral entry into certain target cells and it is unknown which determinants control the entry process. Here, we utilize rhabdoviral vectors to analyze host cell entry driven by batFLUAV-HAL and -NAL. We show that HAL facilitates entry into certain bat but not human cell lines and that entry is independent of sialic acids. In contrast, HAL-driven entry was dependent on prior proteolytic activation of HAL and endosomal acidification. Moreover, we provide evidence that HAL can utilize the cellular protease TMPRSS2 for its activation, suggesting that bat- FLUAV entry into human cells is mainly restricted at the stage of receptor engagement while proteolytic activation and triggering of HAL are not limiting the entry process. Introduction Funding: This work was supported by the Deutsche Forschungsgemeinschaft (DFG) grant PO 716/6-1, Gӧttingen Graduate School for Neurosciences, Biophysics, and Molecular Biosciences, Deutsche Forschungsgemeinschaft (DFG) grants GSC 226/1 and GSC 226/2 (www.dfg.de). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Influenza A viruses (FLUAV) are enveloped, negative stranded RNA viruses that pose a major threat to public health [1]. The ability of FLUAV to constantly adapt to immune pressure allows these viruses to continuously circulate in the human population, resulting in annual influenza epidemics (seasonal influenza [2, 3]). Infants, children and the elderly are at particu- lar risk of developing severe disease upon infection with seasonal FLUAV and it has been esti- mated that world-wide 250,000 to 500,000 people die each year of seasonal influenza [1]. 1 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 NE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Competing Interests: The authors have read the journal's policy and the authors of this manuscript have the following competing interests: Stefan Pöhlmann currently serves as an academic editor for PLoS One. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Plasmids Shuttle vectors harboring codon-optimized (for expression in human cells) open reading frames coding for the published amino acid (aa) sequences of the HAL of the two batFLUAV, A/little yellow-shouldered bat/Guatemala/153/2009 (H17/N10) (GenBank: CY103876.1, HAL17) and A/flat-faced bat/Peru/033/2010 (H18/N11) (GenBank: CY125945.1, HAL18), were purchased from a commercial service (Eurofins MWG Operon) and cloned into the pCG1 expression vector, that was kindly provided by R. Cattaneo, via BamHI and XbaI restric- tion sites. The pCAGGS-based expression plasmid for NAL of batFLUAV A/little yellow- shouldered bat/Guatemala/153/2009 (H17/N10) (GenBank: CY103878.1, NAL10) was pro- vided by M. Schwemmle. HAL equipped with a C-terminal FLAG epitope (DYKDDDDK, HAL17-FLAG and HAL18-FLAG) were constructed by PCR and controlled for sequence integrity by automated sequence analysis. In addition, we used pCAGGS-based expression plasmids for the HA and NA of A/WSN/33 (H1N1) and A/Singapore/1/57 (H2N2) (GenBank: AY209895.1) [32, 33]. Furthermore, we employed pCAGGS-based expression plasmids for the HA of A/South Carolina/1/18 (H1N1) (GenBank: AF117241.1) and the NA of A/Brevig Mis- sion/1/1918 (H1N1) (GenBank: AF250356.2) that were generated from previously used con- structs [32]. The expression plasmid for the glycoprotein (G) of vesicular stomatitis virus (VSV, Indiana strain, VSV-G; GenBank: AJ318514.1) was generated by inserting the VSV-G ORF into the pCG1 expression vector and has been used in previous studies [28, 31, 34, 35]. Furthermore, expression plasmids for Nipah virus fusion protein (F, GenBank: AF212302) and glycoprotein (G, synthetic, GenBank: AF212302; derived from NiV/MY/HU/1999/CDC) were used [36]. For experiments analyzing proteolytic activation of HAL17 and HAL18 by human type-II transmembrane serine proteases (TTSPs), we employed expression plasmids for TMPRSS2, TMPRSS11E (DESC-1) and TMPRSS13 (MSPL), which have been described previ- ously [32, 33]. Materials and Methods Cell culture The following cell lines were used as targets for transduction and expression experiments and were maintained in Dulbecco's modified Eagle's medium (PAA Laboratories), supplemented with 10% fetal bovine serum (Biochrom) and antibiotics (penicillin/streptomycin, PAA Labo- ratories): HEK-293T, Huh7, Vero, MDCK, BHK-21, as well as chiropteran cell lines from five different bat species, RoNi/7, HypNi/1.1, EidNi/41, EpoNi/22.1 and CpKd (Table 1). All non- bat-derived cell lines were obtained from collaborators. The fruit bat cell lines (RoNi/7, HypNi/ 1.1, EidNi/41, EpoNi/22.1) were a kind gift of C. Drosten and M. A. Müller and have been described previously [28–31]. The CpKd cell line was described elsewhere [28]. All cell lines were grown in a humidified atmosphere at 37°C and 5% CO2. For passaging and seeding, cells 2 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses Table 1. Cell lines used to study batFLUAV tropism. Name Species Organ HEK-293T Human (Homo sapiens) Kidney Huh7 Human (Homo sapiens) Liver Vero African green monkey (Chlorocebus aethiops) Kidney MDCK Dog (Canis lupus familaris) Kidney RoNi/7 Egyptian fruit bat (Rousettus aegyptiacus) Kidney HypNi/1.1 Hammer-headed fruit bat (Hypsignathus monstrosus) Kidney EidNi/41 Straw-colored fruit bat (Eidolon helvum) Kidney EpoNi/22.1 Buettikofer's epauletted fruit bat (Epomops buettikoferi) Kidney CpKd Seba’s short-tailed bat (Carollia perspicillata) Kidney doi:10.1371/journal.pone.0152134.t001 Table 1. Cell lines used to study batFLUAV tropism. were detached by either resuspension in fresh culture medium (HEK-293T cells) or by the use of trypsin/EDTA (PAA Laboratories). Treatment of cell lines with neuraminidase and ammonium chloride To investigate the roles of sialic acids and endosomal acidification in batFLUAV entry, we used recombinantly-produced, bacterial sialidase (Clostridium welchii, Sigma-Aldrich) and ammo- nium chloride (Sigma-Aldrich). For treatment, the cell culture supernatant was removed and the cells were washed with phosphate buffered saline (PBS) before culture medium containing water (negative control), ammonium chloride (50 mM) or different concentrations of bacterial sialidase (1.5, 15 or 150 mU) was added. After 2 h of incubation at 37°C and 5% CO2, the supernatant was removed, the cells washed and then inoculated with pseudotypes, as described below. Production of rhabdoviral pseudotypes We employed a replication-deficient VSV vector for pseudotyping that contains two separate open reading frames, coding for enhanced green fluorescent protein (EGFP) and firefly lucifer- ase (fLuc), instead of the genetic information for VSV-G, VSVΔG-fLuc [28, 31, 35]. Propaga- tion of VSVΔG-fLuc was performed in a previously described VSV-G-expressing, transgenic cell line [37]. Generation of VSV pseudotypes (VSVpp) was performed as follows: HEK-293T 3 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses cells were transfected by calcium-phosphate precipitation with expression plasmids encoding viral surface proteins, VSV-G (positive control), NiV-F/G, FLUAV-HA and/or NA and bat- FLUAV-HAL and/or NAL, or empty plasmid (pCAGGS) as negative control. In order to inves- tigate the potential of human TTSPs to proteolytically activate batFLUAV-HAL for host cell entry, we additionally cotransfected the cells with expression plasmids for TMPRSS2, DESC-1 or MSPL. At 16 h post transfection, the cells were inoculated with VSVΔG-fLuc at a multiplic- ity of infection of 3 for 1 h at 37°C and 5% CO2. Subsequently, the cells were washed and incu- bated with an anti-VSV serum to neutralize residual input virus. Finally, the cells received fresh culture medium and were further incubated for 16–20 h, before the VSVpp-containing supernatants were collected, clarified from cell debris by centrifugation and aliquoted. Aliquots were stored at 4°C for a maximum of 7 days. For proteolytic activation of HA/HAL by trypsin, pseudotypes were incubated with bovine trypsin (Sigma-Aldrich; final concentration: 50 μg/ml) for 20 min at 37°C. Subsequently, tryp- sin was inactivated by addition of soybean trypsin inhibitor (Sigma-Aldrich; final concentra- tion: 50 μg/ml). Transduction of cell lines with rhabdoviral pseudotypes and quantification of fLuc activity All transduction experiments were performed in 96-well plates using quadruplicate samples. At 24 h post seeding, the cell culture medium was removed and the cells were washed with PBS. The cells were either directly inoculated with VSVpp or treated as specified above and then inoculated. VSVpp inoculation was performed for 1 h at 37°C and 5% CO2. Afterwards, the inoculum was removed and the cells were again washed and incubated with fresh culture medium for 16–18 h at 37°C and 5% CO2. For the quantification of the fLuc activity as an indi- cator of transduction efficiency, the cell culture supernatant was removed and the cells were washed with PBS. Next, 50 μl of 1x Luciferase Cell Culture Lysis Reagent (Promega) in PBS was added to each well and incubated for 30 min at room temperature, before the cell lysate was transferred to a white, opaque-walled 96-well plate (Thermo Scientific). The measurement of the fLuc activity was carried out in a microplate reader, Plate CHAMELEON (Hidex), using the MicroWin2000 software (version 4.44, Mikrotek Laborsysteme GmbH) and fLuc substrates from the Luciferase Assay System (Promega) or Beetle-Juice (PJK) kits. Transduction effi- ciency, represented by fLuc activity, was either displayed in counts per second (cps) or as nor- malized values. Immunofluorescence analysis of HAL expression To assess expression of HAL proteins, we transfected BHK-21 cells that were grown on cover- slips with expression plasmids for HAL17-FLAG or HAL18-FLAG using the Lipofecta- mine2000 reagent (ThermoFisher Scientific) according to manufacturers’ protocol. Cells 4 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses transfected with an empty expression vector served as negative control. At 24 h post transfec- tion, cells were fixed with 4% paraformaldehyde/PBS, permeabilized by incubation with 0.2 M Triton X-100/PBS (10 min at room temperature) and subsequently incubated with anti-FLAG (mouse, 1:1,000, Sigma-Aldrich) and Cy3-labeled anti-mouse (1:750, Sigma-Aldrich) antibod- ies. After each antibody incubation, the cells were washed three times with PBS and finally incubated with DAPI (Roth, 5 min/37°C) to stain the nuclei before the coverslips were mounted on glass slides using Mowiol (AppliChem) supplemented with anti-bleaching reagent (DABCO, Roth). Representative pictures were taken at a 10x magnification using a Nikon Eclipse Ti fluorescence microscope and the NIS Elements AR software (Nikon). Preparation of cell lysates and rhabdoviral pseudotypes for Western blot analysis BatFLUAV-HAL cleavage was investigated by cotransfection of HEK-293T cells, which were grown in 6-well plates, with expression plasmids for batFLUAV-HAL and different TTSPs (TMPRSS2, DESC-1 or MSPL) or by incubation of HAL-expressing cells with trypsin (1, 5, 10, 50 μg/ml; 20 min/37°C) directly before cell lysates were produced. Cells cotransfected with empty plasmid and not subjected to trypsin treatment served as negative control. The 1918-HA served as positive control, since cleavage by TTSPs and trypsin has been previously shown [33]. At 24 h post transfection, the cells were washed with PBS, resuspended in 100 μl 2x SDS-containing lysis buffer (50 mM Tris [pH 6.8], 10% glycerol, 2% SDS, 5% β-mercap- toethanol, 0.1% bromophenol blue, 1 mM EDTA) and boiled for 20 min at 96°C. To assess incorporation of HAL into VSVpp, 1 ml of the respective pseudotypes was loaded onto a 20% sucrose cushion in PBS and centrifuged at 17,000x g for 2 h at 4°C. After discarding the super- natant, the pelleted pseudotypes were mixed with 30 μl 2x SDS-containing lysis buffer and boiled for 20 min at 96°C. Finally, all samples were subjected to immunoblot analysis. For this, anti-FLAG (mouse, 1:1,000, Sigma-Aldrich), anti-FLUAV (goat, 1:1,000, Millipore), anti- VSV-M (mouse, 1:1,000, Kerafast), anti-VSV-G (I1, mouse hybridoma supernatant from CRL- 2700, ATCC, 1:200) and anti-ß-actin (mouse, 1:1,000, Sigma-Aldrich) served as primary anti- bodies. Peroxidase-coupled anti-mouse (1:10,000, Dianova) and anti-goat (1:5,000, Dianova) antibodies served as secondary antibodies. Signal detection was carried out in a ChemoCam imager together with the ChemoStar professional software (both Intas) using a self-made chemiluminescence substrate (recipe available upon request). Statistical analysis In order to assess statistical significance, two-tailed student’s t-tests were performed. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Both HAL17 and HAL18 are comparably expressed in mammalian cells and incorporated into rhabdoviral pseudotypes The results were confirmed in an independent experiment. doi:10.1371/journal.pone.0152134.g001 Using antibodies specific for the FLAG epitope, VSV-G and VSV matrix protein (VSV-M), we found that VSV-G, as expected, as well as both HAL17 and HAL18 proteins were incorporated into particles, with incorporation of HAL17 being more efficient than that of HAL18. (Fig 1B). Thus, both HAL17 and HAL18 were robustly expressed and incorporated into VSVpp, allow- ing their functional characterization. Using antibodies specific for the FLAG epitope, VSV-G and VSV matrix protein (VSV-M), we found that VSV-G, as expected, as well as both HAL17 and HAL18 proteins were incorporated into particles, with incorporation of HAL17 being more efficient than that of HAL18. (Fig 1B). Thus, both HAL17 and HAL18 were robustly expressed and incorporated into VSVpp, allow- ing their functional characterization. Both HAL17 and HAL18 are comparably expressed in mammalian cells and incorporated into rhabdoviral pseudotypes Both HAL17 and HAL18 are comparably expressed in mammalian cells and incorporated into rhabdoviral pseudotypes To test whether HAL17 and HAL18 are comparably expressed and incorporated into rhabdo- viral pseudotypes, both proteins were equipped with a C-terminal FLAG epitope (HAL17-F- LAG, HAL18-FLAG), since no batFLUAV-HAL-specific antibody was available. Upon transfection of BHK-21 cells similar numbers of HAL-expressing cells were detected by fluores- cence microscopy (Fig 1A) and the intensity of the fluorescence signals emitted by the cells was comparable, indicating robust expression of both batFLUAV-HAL proteins. In order to assess HAL incorporation into rhabdoviral pseudotypes, we pelleted pseudotype preparations through a sucrose cushion and subjected the samples to SDS-PAGE and immunoblotting. 5 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses Fig 1. HAL of batFLUAV are robustly expressed and incorporated into rhabdoviral particles. (A) BHK-21 cells were transfected with the indicated HAL proteins harboring a C-terminal FLAG antigenic tag and protein expression was analyzed by immunofluorescence microscopy of permeabilized cells (magnification, 10x). Cells transfected with an empty expression vector served as negative control. Nuclei were visualized by DAPI staining. Similar results were obtained in a separate experiment. (B) For analysis of HAL incorporation into rhabdoviral particles, vesicular stomatitis virus (VSV)-based pseudotypes were pelleted though a 20% sucrose cushion and analyzed by SDS-PAGE and Western blotting with antibodies against the FLAG tag (α-FLAG), VSV glycoprotein (α-VSV-G) and matrix protein (α-VSV-M). The numbers on the left side of the blots indicate the molecular weight in kilo Daltons (kDa). The results were confirmed in an independent experiment. doi:10 1371/journal pone 0152134 g001 Fig 1. HAL of batFLUAV are robustly expressed and incorporated into rhabdoviral particles. (A) BHK-21 cells were transfected with the indicated HAL proteins harboring a C-terminal FLAG antigenic tag and protein expression was analyzed by immunofluorescence microscopy of permeabilized cells (magnification, 10x). Cells transfected with an empty expression vector served as negative control. Nuclei were visualized by DAPI staining. Similar results were obtained in a separate experiment. (B) For analysis of HAL incorporation into rhabdoviral particles, vesicular stomatitis virus (VSV)-based pseudotypes were pelleted though a 20% sucrose cushion and analyzed by SDS-PAGE and Western blotting with antibodies against the FLAG tag (α-FLAG), VSV glycoprotein (α-VSV-G) and matrix protein (α-VSV-M). The numbers on the left side of the blots indicate the molecular weight in kilo Daltons (kDa). PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 HAL of batFLUAV mediates cellular entry into bat cell lines and entry depends on proteolytic activation It has been previously reported that HAL and NAL of batFLUAV are not compatible with viral spread in the cell lines tested so far [18, 19], suggesting that these proteins mediate entry into a restricted set of target cells or are even inactive. In order to gain insights into the functional activity of batFLUAV-HAL and NAL, we employed rhabdoviral vectors pseudotyped with these proteins. We inoculated cell lines from different host species, including those standardly used for FLUAV research, as well as cell lines derived from different bat species (Table 1), as they are the natural reservoir for batFLUAV. We chose bat cell lines known to be susceptible to infection by viruses of different families [28, 29, 31] and previously used to functionally charac- terize surface glycoproteins of bat-borne viruses [28, 36, 38]. It is well established that FLUAV-HA depends on proteolytic cleavage by host cell proteases to transit into an active form [12] and it has been previously reported that batFLUAV can be activated by exogenous trypsin [39–41]. Therefore, we assessed whether trypsin treatment of the pseudotypes impacts transduction efficiency. As controls, we included the surface glycoprotein(s) of well-character- ized FLUAV strains, A/WSN/33 (H1N1) (WSN-HA, WSN-NA), A/South Carolina/1/18 6 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses (H1N1) (1918-HA), A/Brevig Mission/1/1918 (H1N1) (1918-NA), and A/Singapore/1/57 (H2N2) (H2N2-HA, H2N2-NA), as well as the glycoproteins of Nipah virus (NiV-G, NiV-F) and vesicular stomatitis virus (VSV-G). (H1N1) (1918-HA), A/Brevig Mission/1/1918 (H1N1) (1918-NA), and A/Singapore/1/57 (H2N2) (H2N2-HA, H2N2-NA), as well as the glycoproteins of Nipah virus (NiV-G, NiV-F) and vesicular stomatitis virus (VSV-G). We found that none of the human, simian and canine cell lines was susceptible to entry driven by batFLUAV surface proteins (Fig 2A). In contrast, pseudotypes bearing VSV-G, NiV-F/G or WSN-HA/NA could readily enter these cells, whereas pseudotypes that harbored 1918- or H2N2-HA/NA required activation by exogenous trypsin for efficient transduction (Fig 2A). These results are in agreement with expectations, since activation of WSN-HA is known to be independent of trypsin [42–44], although viral infectivity can be enhanced by trypsin treatment. When we focused on bat-derived cell lines, VSV-G, NiV-F/G and WSN-HA/NA again per- mitted pseudotype entry without prior trypsin treatment, while pseudotypes harboring 1918-HA/NA or H2N2-HA/NA were only able to transduce some of the bat cell lines after incubation with trypsin (Fig 2B). HAL of batFLUAV mediates cellular entry into bat cell lines and entry depends on proteolytic activation CpKd cells remained refractory to entry mediated by 1918-HA/NA or H2N2-HA/NA but also showed the lowest susceptibility to all other tested pseudotypes. Notably, three bat cell lines (EidNi/41, HypNi/1.1 and EpoNi/22.1) were suscepti- ble to entry of pseudotypes bearing HAL and NAL of batFLUAV (Fig 2B), demonstrating that surface glycoproteins of batFLUAV can mediate cellular entry. Entry into the three bat cell lines was robust (1–3 log units above the threshold) and required prior treatment of pseudo- types with trypsin, which presumably resulted in the proteolytic activation of HAL. Further- more, pseudotypes bearing HAL17/NAL10 or HAL18/NAL10 were both able to enter HypNi/ 1.1 and EpoNi/22.1 cells while EidNi/41 cells could only be transduced by pseudotypes bearing HAL18/NAL10 (Fig 2B), suggesting that batFLUAV of the HL17NL10 and HL18NL11 subtype might exhibit subtle differences in entry efficiency or cell tropism. Finally, HAL-proteins with a C-terminal FLAG tag facilitated host cell entry, although with somewhat reduced efficiency as compared to their untagged counterparts, indicating that the proteins used to study HAL expression and virion incorporation (Fig 1) were functional (data not shown). Taken together, we showed that batFLUAV surface proteins can mediate entry into certain bat cell lines. For further studies on the entry process, we focused on EpoNi/22.1 cells since they showed the highest susceptibility to entry driven by batFLUAV surface proteins. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses Fig 2. Surface glycoproteins of batFLUAV enable pseudoty stomatitis virus-based pseudotypes (VSVpp) harboring no or the (black bars), before they were inoculated onto mammalian cell li (RoNi/7, EidNi/41, HypNi/1.1, EpoNi/22.1, CpKd) representing f measured by quantification of the activity of the VSVpp-encoded representative experiment carried out with quadruplicate sample separate pseudotype preparations. Error bars indicate standard p < 0.05). doi:10.1371/journal.pone.0152134.g002 Fig 2. Surface glycoproteins of batFLUAV enable pseudotype entry into different bat but not human, simian and canine cell line stomatitis virus-based pseudotypes (VSVpp) harboring no or the indicated viral glycoproteins were either left untreated (white bars) or tre (black bars), before they were inoculated onto mammalian cell lines of human (HEK-293T, Huh7), simian (Vero) and canine (MDCK) orig (RoNi/7, EidNi/41, HypNi/1.1, EpoNi/22.1, CpKd) representing five different bat species (B). At 18–20 h post inoculation, the transductio measured by quantification of the activity of the VSVpp-encoded luciferase (given as counts per second, cps, on a logarithmic scale). Th representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in four independent experimen separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statis p < 0.05). doi:10.1371/journal.pone.0152134.g002 entry into different bat but not human, simian and canine cell ndicated viral glycoproteins were either left untreated (white bars) o s of human (HEK-293T, Huh7), simian (Vero) and canine (MDCK) different bat species (B). At 18–20 h post inoculation, the transdu uciferase (given as counts per second, cps, on a logarithmic scale) is shown. Similar results were obtained in four independent exper eviations. A two-tailed, unpaired student’s t-test was used to test s g 2. Surface glycoproteins of batFLUAV enable pseudotype entry into different bat but not human, simian and canine cell lines. Vesicular omatitis virus-based pseudotypes (VSVpp) harboring no or the indicated viral glycoproteins were either left untreated (white bars) or treated with trypsin ack bars), before they were inoculated onto mammalian cell lines of human (HEK-293T, Huh7), simian (Vero) and canine (MDCK) origin (A) or bat cell lines oNi/7, EidNi/41, HypNi/1.1, EpoNi/22.1, CpKd) representing five different bat species (B). At 18–20 h post inoculation, the transduction efficiency was easured by quantification of the activity of the VSVpp-encoded luciferase (given as counts per second, cps, on a logarithmic scale). The result of a single presentative experiment carried out with quadruplicate samples is shown. Expression of batFLUAV-NAL in pseudotype producing cells does not impact particle infectivity The NA proteins of human FLUAV facilitate release of progeny particles from infected cells by removing sialic acids from the cell surface. To study the impact of the batFLUAV-NAL on transduction efficiency, we produced pseudotypes bearing batFLUAV-HAL, -NAL or both proteins. For comparison, we generated pseudotypes harboring WSN-HA, WSN-NA or both proteins. These pseudotypes were then used for inoculation of MDCK (inoculated with pseu- dotypes bearing WSN proteins) and EpoNi/22.1 (inoculated with pseudotypes bearing WSN or batFLUAV proteins) cells. Pseudotypes harboring only WSN-NA were not able to transduce target cells (Fig 3) while pseudotypes bearing either WSN-HA alone or in combination with WSN-NA transduced both MDCK and EpoNi/22.1, as expected. Transduction efficiency was ~500–1,500-fold higher when WSN-NA was expressed in cells used for pseudotype produc- tion, in keeping with the findings that the presence of NA is required for efficient release of HA-bearing vectors and infectious FLUAV [32, 45, 46]. Pseudotypes harboring NAL were not infectious while pseudotypes bearing HAL robustly transduced EpoNi/22.1 cells, indicating that batFLUAV-HAL, like WSN-HA, is sufficient to mediate host cell entry. However, unlike WSN-NA, the expression of NAL in pseudotype producer cells did not increase transduction 7 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 The combined data from three independent experiments (quadruplicate samples) with separate pseudotype preparations are shown. Transduction efficiencies were normalized against pseudotypes harboring only HA or HAL (set as 1) and are given as x-fold changes on a logarithmic scale. Error bars indicate standard error of the mean. A two-tailed, paired student’s t-test was used to test statistical significance (* = p < 0.05). Fig 3. NAL of batFLUAV has no impact on the transduction efficiency of vectors bearing HAL. Vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated surface proteins of FLUAV or batFLUAV were treated with trypsin before inoculation of MDCK and EpoNi/22.1 cells. At 18–20 h post inoculation, the transduction efficiency was measured by quantification of the activity of the VSVpp-encoded luciferase. The combined data from three independent experiments (quadruplicate samples) with separate pseudotype preparations are shown. Transduction efficiencies were normalized against pseudotypes harboring only HA or HAL (set as 1) and are given as x-fold changes on a logarithmic scale. Error bars indicate standard error of the mean. A two-tailed, paired student’s t-test was used to test statistical significance (* = p < 0.05). doi:10.1371/journal.pone.0152134.g003 efficiency of HAL-harboring pseudotypes (Fig 3), suggesting that NAL is not required for release and/or infectivity of HAL containing particles, at least in the experimental system chosen. BatFLUAV-HAL does not use sialic acids for host cell entry FLUAV employ alpha-2,3- (avian viruses) and alpha-2,6-linked (human viruses) sialic acids as receptors for host cell entry [47–51]. In order to assess the potential role of sialic acids in HAL- driven entry, we pre-treated the cells with escalating doses of bacterial neuraminidase before transduction. Neuraminidase treatment of EpoNi/22.1 cells reduced transduction mediated by the FLUAV-HA proteins, as expected (Fig 4). In contrast, sialidase treatment had no effect on pseudotype entry mediated by batFLUAV-HAL, NiV-F/G or VSV-G. Moreover, pre-treatment of EpoNi/22.1 cells at the highest dose (150 mU) even enhanced transduction driven by HAL and VSV-G (Fig 4). These results indicate that HAL does not use sialic acids for host cell entry and suggest that removal of sialic acids might even increase batFLUAV infectivity, potentially by increasing accessibility of a cellular receptor. Similar results were obtained in four independent experiments carried out with parate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance (* = < 0.05). Fig 2. Surface glycoproteins of batFLUAV enable pseudotype entry into different bat but not human, simian and canine cell lines. Vesicular stomatitis virus-based pseudotypes (VSVpp) harboring no or the indicated viral glycoproteins were either left untreated (white bars) or treated with trypsin (black bars), before they were inoculated onto mammalian cell lines of human (HEK-293T, Huh7), simian (Vero) and canine (MDCK) origin (A) or bat cell lines (RoNi/7, EidNi/41, HypNi/1.1, EpoNi/22.1, CpKd) representing five different bat species (B). At 18–20 h post inoculation, the transduction efficiency was measured by quantification of the activity of the VSVpp-encoded luciferase (given as counts per second, cps, on a logarithmic scale). The result of a single representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in four independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance (* = p < 0 05) doi:10.1371/journal.pone.0152134.g002 8 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses Fig 3. NAL of batFLUAV has no impact on the transduction efficiency of vectors bearing HAL. Vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated surface proteins of FLUAV or batFLUAV were treated with trypsin before inoculation of MDCK and EpoNi/22.1 cells. At 18–20 h post inoculation, the transduction efficiency was measured by quantification of the activity of the VSVpp-encoded luciferase. The combined data from three independent experiments (quadruplicate samples) with separate pseudotype preparations are shown. Transduction efficiencies were normalized against pseudotypes harboring only HA or HAL (set as 1) and are given as x-fold changes on a logarithmic scale. Error bars indicate standard error of the mean. A two-tailed, paired student’s t-test was used to test statistical significance (* = p < 0.05). doi:10 1371/journal pone 0152134 g003 Fig 3. NAL of batFLUAV has no impact on the transduction efficiency of vectors bearing HAL. Vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated surface proteins of FLUAV or batFLUAV were treated with trypsin before inoculation of MDCK and EpoNi/22.1 cells. At 18–20 h post inoculation, the transduction efficiency was measured by quantification of the activity of the VSVpp-encoded luciferase. Fig 4. HAL of batFLUAV does not require sialic acids for host cell entry. EpoNi/22.1 cells were incubated for 1.5 h in the absence (black bars) or presence (white and grey bars) of increasing concentrations of exogenous sialidase and were subsequently inoculated with trypsin-treated vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated viral glycoproteins. At 1 h post inoculation, the inoculum was removed, the cells were washed and further incubated for 18–20 h with fresh medium before transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase. Transduction of sialidase-treated cells is shown relative to that measured for mock-treated cells (on a linear scale), which was set at 100%. The result of a single representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in two independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance of differences measured for sialidase- versus mock-treated samples (* = p < 0.05). Fig 4. HAL of batFLUAV does not require sialic acids for host cell entry. EpoNi/22.1 cells were incubated for 1.5 h in the absence (black bars) or presence (white and grey bars) of increasing concentrations of exogenous sialidase and were subsequently inoculated with trypsin-treated vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated viral glycoproteins. At 1 h post inoculation, the inoculum was removed, the cells were washed and further incubated for 18–20 h with fresh medium before transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase. Transduction of sialidase-treated cells is shown relative to that measured for mock-treated cells (on a linear scale), which was set at 100%. The result of a single representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in two independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance of differences measured for sialidase- versus mock-treated samples (* = p < 0.05). doi:10.1371/journal.pone.0152134.g004 NiV-F and -G was unaffected, again in keeping with published data [52, 53]. Finally, HAL- driven entry was markedly reduced by ammonium chloride, demonstrating that the membrane fusion activity of batFLUAV-HAL is triggered by acidification (Fig 5). Host Cell Entry by Bat-Associated Influenza Viruses Fig 4. HAL of batFLUAV does not require sialic acids for host cell entry. EpoNi/22.1 cells were incubated for 1.5 h in the absence (black bars) or presence (white and grey bars) of increasing concentrations of exogenous sialidase and were subsequently inoculated with trypsin-treated vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated viral glycoproteins. At 1 h post inoculation, the inoculum was removed, the cells were washed and further incubated for 18–20 h with fresh medium before transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase. Transduction of sialidase-treated cells is shown relative to that measured for mock-treated cells (on a linear scale), which was set at 100%. The result of a single representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in two independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance of differences measured for sialidase- versus mock-treated samples (* = p < 0.05). d i 10 1371/j l 0152134 004 HAL-driven entry of batFLUAV relies on endosomal acidification Endosomal low pH triggers FLUAV-HA for membrane fusion. Therefore, we investigated whether increasing the endosomal pH in EpoNi/22.1 cells by ammonium chloride treatment impacts HAL-driven entry. As expected, ammonium chloride treatment led to a decrease in transduction efficiency mediated by pseudotypes bearing the HA-proteins of FLUAV of the H1N1 and H2N2 subtype and VSV-G (Fig 5). In contrast, pseudotype entry orchestrated by PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 9 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Fig 5. Entry driven by batFLUAV-HAL relies on an acidic pH. EpoNi/22.1 cells incubated for 3 h in the absence (black bars) or presence (white bars) of ammonium chloride (50 mM) were subsequently inoculated with trypsin-treated vesicular stomatitis virus-based pseudotypes (VSVpp) harboring the indicated viral glycoproteins. At 1 h post inoculation, the inoculum was removed and the cells were further incubated for 18– 20 h in the presence or absence of ammonium chloride before transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase. For each of the different pseudotypes, transduction efficiency (given as percentage on a linear scale) was normalized against the respective control (water). The result of a single representative experiment carried out with quadruplicate samples is shown. Similar results were obtained in two independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance (* = p < 0.05). doi:10.1371/journal.pone.0152134.g005 pseudotypes bearing 1918-HA and -NA were included in this experiment. The pseudotypes were treated with trypsin to activate HA/HAL or were mock-treated before addition to EpoNi/22.1 cells. Pseudotypes bearing 1918-HA and -NA and produced in the presence of TMPRSS2, DESC-1 and MSPL or treated with trypsin robustly transduced target cells (Fig 6B). In contrast, infectivity of FLUAV-HA pseudotypes produced in the absence of TTSPs or not treated with trypsin was in the background range (Fig 6B). Similarly, batFLUAV-HAL- bearing pseudotypes were activated by trypsin or TTSPs, including TMPRSS2 (Fig 6B). How- ever, differences in the activation of HAL17 and HAL18 were observed and correlated with the efficiency of HAL protein cleavage, as determined above (Fig 6A). Thus, expression of TMPRSS2 but not DESC-1 and MSPL conferred robust infectivity to HAL17-bearing pseu- dotypes while all proteases were able to efficiently activate HAL18. Moreover, transfection of escalating amounts of TMPRSS2-encoding plasmids increased infectivity of HAL17-bearing pseudotypes in a concentration-dependent manner. In contrast, transfection of even the low- est amount of TMPRSS2 plasmid was sufficient to confer maximal infectivity of HAL18- bearing pseudotypes, confirming that the efficiency of TMPRSS2-mediated activation of HAL is subtype specific (Fig 6C). In sum, proteolytic activation of batFLUAV-HAL is critical for HAL-driven cell entry and proteases able to activate HA can also activate HAL. pseudotypes bearing 1918-HA and -NA were included in this experiment. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 TMPRSS2 activates batFLUAV-HAL FLUAV-HA is synthesized as an inactive precursor and requires activation by host cell prote- ases to be responsive to low pH, the trigger for HA-driven membrane fusion [12–14]. Members of the TTSP family activate FLUAV-HA in cell culture [33, 54–58] and TMPRSS2 was previ- ously shown to be essential for FLUAV-HA activation and viral spread in mice [59]. Therefore, we asked whether TTSPs able to activate FLUAV-HA can also activate batFLUAV-HAL. For this, we first investigated batFLUAV-HAL cleavage by TMPRSS2, DESC-1 and MSPL, and compared it to cleavage by trypsin. Cleavage of the 1918-HA served as positive control. We found that 1918-HA was efficiently processed by all proteases tested, as expected. Moreover, we could show that coexpression of TMPRSS2, DESC-1 and MSPL, and trypsin treatment resulted in cleavage of the HAL precursor (HAL0) determined by the appearance of bands cor- responding to the HAL2 subunit (Fig 6A). While HAL18 was comparably cleaved by all tested TTSPs, HAL17 cleavage by TMPRSS2 was more pronounced than proteolysis by DESC-1 and MSPL (Fig 6A). Moreover, HAL18 was generally more sensitive to cleavage by TTSPs than HAL17 (Fig 6A). In order to assess whether batFLUAV-HAL cleavage by TTSPs also leads to HAL activation for host cell entry, we produced pseudotypes harboring batFLUAV-HAL (HAL17 or HAL18) in the presence of TMPRSS2, DESC-1 and MSPL. As a control, 10 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses Host Cell Entry by Bat-Associated Influenza Viruses Fig 6. Human proteases that activate FLUAV-HA for cell entry also activate batFLUAV-HAL. (A) HEK-293T cells were transfected with plasmids encoding HA or HAL proteins and either trypsin treated or cotransfected with plasmids encoding type II transmembrane serine proteases. Transfection of empty vector served as negative control. Cleavage of HA/HAL proteins was analyzed by SDS-PAGE and Western blotting, employing antibodies against FLUAV-HA (α-FLUAV) and the FLAG epitope (α-FLAG). Detection of ß-actin served as loading control. Signals corresponding to uncleaved precursor proteins are marked by black circles, while products of proteolytic cleavage are indicated by white circles. The results were confirmed in a separate experiment. To assess proteolytic activation of HA/HAL proteins, vesicular stomatitis virus-based pseudotypes (VSVpp) were produced in cells transfected to express the indicated type II transmembrane serine proteases (B) or different amounts of TMPRSS2 (C). Pseudotypes were either directly used for transduction of EpoNi/22.1 cells (black bars) or previously treated with trypsin (white bars). At 24 h post inoculation, transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase in cell lysates. For normalization, transduction by HA- or HAL-bearing pseudotypes that were produced in the absence of type II transmembrane serine protease expression (empty vector) and not treated with trypsin was set as 1. The result of a single representative experiment carried out with quadruplicate samples is presented. Similar results were obtained in three independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance (* = p < 0.05). doi:10 1371/journal pone 0152134 g006 Fig 6. Human proteases that activate FLUAV-HA for cell entry also activate batFLUAV-HAL. (A) HEK-293T cells were transfected with plasmids Fig 6. Human proteases that activate FLUAV-HA for cell entry also activate batFLUAV-HAL. (A) HEK-293T cells were transfected with plasmids encoding HA or HAL proteins and either trypsin treated or cotransfected with plasmids encoding type II transmembrane serine proteases. Transfection of empty vector served as negative control. Cleavage of HA/HAL proteins was analyzed by SDS-PAGE and Western blotting, employing antibodies against FLUAV-HA (α-FLUAV) and the FLAG epitope (α-FLAG). Detection of ß-actin served as loading control. Signals corresponding to uncleaved precursor proteins are marked by black circles, while products of proteolytic cleavage are indicated by white circles. The results were confirmed in a separate experiment. To assess proteolytic activation of HA/HAL proteins, vesicular stomatitis virus-based pseudotypes (VSVpp) were produced in cells transfected to express the indicated type II transmembrane serine proteases (B) or different amounts of TMPRSS2 (C). Pseudotypes were either directly used for transduction of EpoNi/22.1 cells (black bars) or previously treated with trypsin (white bars). At 24 h post inoculation, transduction efficiency was measured by quantification of the activity of VSVpp-encoded luciferase in cell lysates. For normalization, transduction by HA- or HAL-bearing pseudotypes that were produced in the absence of type II transmembrane serine protease expression (empty vector) and not treated with trypsin was set as 1. The result of a single representative experiment carried out with quadruplicate samples is presented. Similar results were obtained in three independent experiments carried out with separate pseudotype preparations. Error bars indicate standard deviations. A two-tailed, unpaired student’s t-test was used to test statistical significance (* = p < 0.05). doi:10.1371/journal.pone.0152134.g006 The pseudotypes were treated with trypsin to activate HA/HAL or were mock-treated before addition to EpoNi/22.1 cells. Pseudotypes bearing 1918-HA and -NA and produced in the presence of TMPRSS2, DESC-1 and MSPL or treated with trypsin robustly transduced target cells (Fig 6B). In contrast, infectivity of FLUAV-HA pseudotypes produced in the absence of TTSPs or not treated with trypsin was in the background range (Fig 6B). Similarly, batFLUAV-HAL- bearing pseudotypes were activated by trypsin or TTSPs, including TMPRSS2 (Fig 6B). How- ever, differences in the activation of HAL17 and HAL18 were observed and correlated with the efficiency of HAL protein cleavage, as determined above (Fig 6A). Thus, expression of TMPRSS2 but not DESC-1 and MSPL conferred robust infectivity to HAL17-bearing pseu- dotypes while all proteases were able to efficiently activate HAL18. Moreover, transfection of escalating amounts of TMPRSS2-encoding plasmids increased infectivity of HAL17-bearing pseudotypes in a concentration-dependent manner. In contrast, transfection of even the low- est amount of TMPRSS2 plasmid was sufficient to confer maximal infectivity of HAL18- bearing pseudotypes, confirming that the efficiency of TMPRSS2-mediated activation of HAL is subtype specific (Fig 6C). In sum, proteolytic activation of batFLUAV-HAL is critical for HAL-driven cell entry and proteases able to activate HA can also activate HAL. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 11 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 doi:10.1371/journal.pone.0152134.g006 Discussion The identification of two new FLUAV, subtypes H17N10 (HL17NL10) and H18N11 (HL18NL11), in new-world bats [16, 17] suggests that bats could serve as a natural reservoir of FLUAV [16, 17]. Unraveling the zoonotic potential of batFLUAV is of great importance since FLUAV are major human pathogens, responsible for influenza epidemics and pandemics. While the batFLUAV replication machinery appears to be functional in different mammalian (including human) cells [16, 18–21, 60], reassortment with FLUAV and FLUBV is unlikely [18, 19]. Regarding batFLUAV tropism of the viral surface proteins, HAL and NAL, only 12 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 12 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses limited information is available, which indicate that batFLUAV do not engage with canonical FLUAV receptors [17, 22–24]. However, until very recently no proof of functional activity of batFLUAV surface proteins was available [17, 22–27]. Here, we employed a vector system to analyze batFLUAV-HAL and -NAL. We show that HAL mediates entry into certain bat cell lines and that entry does not depend on the presence of sialic acids on the cell surface. More- over, we demonstrate that NAL is not required for production of infectious HAL-bearing parti- cles, at least under the conditions examined. Finally, our studies revealed that trypsin and TTSPs activate HAL for host cell entry. We used a VSV-based vector system to study cellular entry of HAL and NAL-bearing parti- cles. VSV pseudotypes allow convenient analysis of entry driven by diverse glycoproteins [28, 31, 61], although one should keep in mind that due to differences in particle shape and effi- ciency of glycoprotein incorporation pseudotypes might not adequately mirror all aspects of cellular entry of authentic viruses [62, 63]. We found that cell lines frequently used for FLUAV propagation were not susceptible to transduction by HAL and NAL bearing particles, which is in agreement with the finding that replacement of batFLUAV-HAL and -NAL by their FLUAV counterparts is required for spread of batFLUAV in the cell lines studied so far [18, 19]. In con- trast, inoculation of bat cell lines originating from five different species of micro- and mega- chiropteran bats revealed that three cell lines, EidNi/41, HypNi/1.1, EpoNi/22.1, were susceptible to entry mediated by batFLUAV surface proteins. Discussion EpoNi/22.1 cells showed the highest susceptibility and were thus used for further studies, while EidNi/41 cells were found to be robustly susceptible only to transduction by pseudotypes harboring the HAL18. Collectively, these findings suggest that batFLUAV surface proteins can mediate entry into certain bat cells and that entry efficiency might differ between batFLUAV subtypes. Our finding that certain bat cell lines are susceptible to pseudotypes harboring HAL of batFLUAV are in line with observations very recently documented by Maruyama et al. who found that out of a diverse panel of bat cell lines tested, cells from Miniopterus fuliginosus, Miniopterus schreibersii and Pteropus giganteus were susceptible to pH-dependent, HAL-driven entry [64]. A cell line derived from Eidolon helvum spleen cells was found to be non-susceptible in contrast to our findings with a kidney cell line established from the same species, suggesting that receptor expression might differ between organs. Somewhat more surprising, Maruyama and colleagues also observed HAL-driven entry into MDCK cells [64], which was not detected in the present study, and these discrepant results might be attributed to use of MDCK cells from different sources or to differences in the method used to quantify pseudotype entry. Finally, it is note- worthy that cell lines from bats inhabiting different geographical locations (Africa, Asia, Europe) were found to be susceptible to HAL-driven entry, suggesting that entry is not a bottle- neck for spread of batFLUAV between bat species. The finding that batFLUAV surface proteins can facilitate entry into bat-derived target cells allowed us to investigate which viral and cellular components contribute to the entry process. We first focused on NAL. The expression of this protein, unlike expression of NA, in pseudo- type-producing cells did not increase the titers of vectors harboring the corresponding bat- FLUAV-HAL protein. However, this finding does not exclude that NAL, like the NA of FLUAV, acts as a receptor-destroying enzyme. Thus, HEK-293T cells used for pseudotype preparation were not susceptible to HAL-driven entry, most likely because they do not express the appropriate receptor. It thus remains to be analyzed whether batFLUAV-NAL increases release of HAL-bearing vectors and authentic batFLUAV from susceptible bat cell lines. These endeavors might be challenging since transfection of bat cell lines by calcium-phosphate pre- cipitation and liposome-based reagents was inefficient (not shown). PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Discussion The FLUAV-HA is sufficient to mediate viral binding and entry into target cells and our find- ings indicate that the same applies to batFLUAV-HAL. In contrast to FLUAV-HA, however, HAL 13 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses does not depend on the presence of sialic acids for entry. Thus, treatment of EpoNi/22.1 cells with sialidase did not decrease HAL-mediated pseudotype entry. These findings are in keeping with the work by Maruyama et al. [64] and with structural data indicating that HAL does possess a dis- torted putative sialic acid binding site [17, 24]. Contrarily, high amounts of sialidase increased entry efficiency, potentially by increasing access to the elusive receptor. In addition, removal of sialic acids might increase electrostatic interactions of batFLUAV-HAL with cell surface factors, since sialic acids are negatively charged. Although HAL-driven entry was independent of sialic acids, it did require endosomal acidification (in accordance with Maruyama et al. [64]), which is known to trigger the membrane fusion activity of HA. Most likely, protonation also triggers bat- FLUAV-HAL for membrane fusion. However, it cannot be disregarded that the inhibitory effect of ammonium chloride was due to blockade of pH-dependent endosomal cysteine proteases, which activate the surface proteins of several coronaviruses and ebolaviruses [65–68]. Cleavage-activation of FLUAV-HA by host cell proteases is essential for FLUAV infectivity. Several TTSPs can cleave and activate HA in cell culture and recent studies demonstrated that TMPRSS2 is essential for FLUAV spread in mice [32, 33, 58, 59]. Moreover, polymorphisms in TMPRSS2 were shown to impact severity of influenza in humans [69]. Treatment of bat- FLUAV-HAL-expressing cells with trypsin led to proteolytic cleavage of HAL and exposure of HAL-bearing pseudotypes to trypsin was required for efficient transduction of target cells, indi- cating that proteolytic processing is also required for HAL function. Moreover, coexpression of batFLUAV-HAL with TMPRSS2, DESC-1 or MSPL resulted in proteolytic cleavage of HAL and rendered the particles infectious in the absence of trypsin treatment, suggesting that bat- FLUAV-HAL can utilize human proteases for their activation. Finally, titration of the amounts of TMPRSS2 had differential effects on the proteolytic cleavage of HAL17 and HAL18 and on infectivity of pseudotypes bearing these proteins, hinting towards subtle differences in the effi- ciency of TMPRSS2 use by these subtypes. Whether bat TMPRSS2 is also able to cleave and activate batFLUAV-HAL remains to be investigated. Acknowledgments We would like to thank C. Drosten, M. A. Müller and M. Schwemmle for cell lines and expres- sion plasmids. Furthermore, we thank E. Berger and I. Nehlmeier for excellent technical support. Discussion Collectively, our results are most compatible with a scenario in which human cells allow for batFLUAV-HAL activation and triggering but fail to express a receptor, which can be employed by HAL for host cell entry. These results, jointly with the documented observation that the batFLUAV replication machinery is functional in human cells [16, 18, 19] suggest that HAL adaptation to a human receptor might be the major hurdle batFLUAV need to overcome to spread in humans. It will thus be highly interesting to identify the nature of this receptor and its interface with batFLUAV-HAL. Of note, during the preparation of this manuscript, Maruyama and colleagues published a manuscript reporting batFLUAV-HAL-driven entry into bat cell lines different from those used in the present study (Maruyama et al., 2015, doi: 10.1016/j.virol.2015.11.002.). Both stud- ies show that HAL-driven entry requires prior proteolytic HAL-activation by trypsin and endo- somal acidification but is independent of sialic acids. The present work extends these findings by demonstrating that HAL can utilize the human enzyme TMPRSS2 for its activation. References 1. WHO. Influenza (Seasonal) Fact sheet No. 211: World Health Organisation; 2014 [updated March 2014; cited 2015 26.11.2015]. Available from: http://www.who.int/mediacentre/factsheets/fs211/en/. 2. Cox NJ, Subbarao K. Global epidemiology of influenza: past and present. Annual review of medicine. 2000; 51:407–21. doi: 10.1146/annurev.med.51.1.407 PMID: 10774473. 3. Parrish CR, Kawaoka Y. The origins of new pandemic viruses: the acquisition of new host ranges by canine parvovirus and influenza A viruses. Annual review of microbiology. 2005; 59:553–86. doi: 10. 1146/annurev.micro.59.030804.121059 PMID: 16153179. 4. Webster RG, Bean WJ, Gorman OT, Chambers TM, Kawaoka Y. Evolution and ecology of influenza A viruses. Microbiological reviews. 1992; 56(1):152–79. PMID: 1579108; PubMed Central PMCID: PMC372859. 5. Liu D, Shi W, Shi Y, Wang D, Xiao H, Li W, et al. Origin and diversity of novel avian influenza A H7N9 viruses causing human infection: phylogenetic, structural, and coalescent analyses. Lancet. 2013; 381 (9881):1926–32. doi: 10.1016/S0140-6736(13)60938-1 PMID: 23643111. 6. Taubenberger JK, Morens DM. Pandemic influenza—including a risk assessment of H5N1. Revue scientifique et technique. 2009; 28(1):187–202. PMID: 19618626; PubMed Central PMCID: PMC2720801. 7. Palese P. Influenza: old and new threats. Nature medicine. 2004; 10(12 Suppl):S82–7. doi: 10.1038/ nm1141 PMID: 15577936. 8. Taubenberger JK, Morens DM. 1918 Influenza: the mother of all pandemics. Emerging infectious dis- eases. 2006; 12(1):15–22. doi: 10.3201/eid1201.050979 PMID: 16494711; PubMed Central PMCID: PMC3291398. 9. Carroll SM, Higa HH, Paulson JC. Different cell-surface receptor determinants of antigenically similar influenza virus hemagglutinins. The Journal of biological chemistry. 1981; 256(16):8357–63. PMID: 6167577. 10. Suzuki Y, Matsunaga M, Nagao Y, Taki T, Hirabayashi Y, Matsumoto M. Ganglioside GM1b as an influ- enza virus receptor. Vaccine. 1985; 3(3 Suppl):201–3. PMID: 4060848. 11. Suzuki Y, Kato H, Naeve CW, Webster RG. Single-amino-acid substitution in an antigenic site of influ- enza virus hemagglutinin can alter the specificity of binding to cell membrane-associated gangliosides. Journal of virology. 1989; 63(10):4298–302. PMID: 2476569; PubMed Central PMCID: PMC251045. 12. Bertram S, Glowacka I, Steffen I, Kuhl A, Pohlmann S. Novel insights into proteolytic cleavage of influ- enza virus hemagglutinin. Reviews in medical virology. 2010; 20(5):298–310. doi: 10.1002/rmv.657 PMID: 20629046. 13. Bouvier NM, Palese P. The biology of influenza viruses. Vaccine. 2008; 26 Suppl 4:D49–53. PMID: 19230160; PubMed Central PMCID: PMC3074182. 14. Palese P, Shaw ML. Orthomyxoviridae. In: Knipe DM, Howley PM, editors. Fields Virology. 2. 5th ed: Lippincott Williams & Wilkins; 2007. p. 1647–89. 15. Fouchier RA, Munster V, Wallensten A, Bestebroer TM, Herfst S, Smith D, et al. Author Contributions Conceived and designed the experiments: MH NK GH SP. Performed the experiments: MH NK PZ FW. Analyzed the data: MH NK PZ FW GH SP. Wrote the paper: MH SP. 14 / 18 PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 Host Cell Entry by Bat-Associated Influenza Viruses PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 References Characterization of a novel influenza A virus hemagglutinin subtype (H16) obtained from black-headed gulls. Journal of virol- ogy. 2005; 79(5):2814–22. doi: 10.1128/JVI.79.5.2814–2822.2005 PMID: 15709000; PubMed Central PMCID: PMC548452. 16. Tong S, Li Y, Rivailler P, Conrardy C, Castillo DA, Chen LM, et al. A distinct lineage of influenza A virus from bats. Proceedings of the National Academy of Sciences of the United States of America. 2012; 109(11):4269–74. doi: 10.1073/pnas.1116200109 PMID: 22371588; PubMed Central PMCID: PMC3306675. 17. Tong S, Zhu X, Li Y, Shi M, Zhang J, Bourgeois M, et al. New world bats harbor diverse influenza A viruses. PLoS pathogens. 2013; 9(10):e1003657. doi: 10.1371/journal.ppat.1003657 PMID: 24130481; PubMed Central PMCID: PMC3794996. 18. Juozapaitis M, Aguiar Moreira E, Mena I, Giese S, Riegger D, Pohlmann A, et al. An infectious bat- derived chimeric influenza virus harbouring the entry machinery of an influenza A virus. Nature commu- nications. 2014; 5:4448. doi: 10.1038/ncomms5448 PMID: 25055345. 19. Zhou B, Ma J, Liu Q, Bawa B, Wang W, Shabman RS, et al. Characterization of uncultivable bat influ- enza virus using a replicative synthetic virus. PLoS pathogens. 2014; 10(10):e1004420. doi: 10.1371/ journal.ppat.1004420 PMID: 25275541; PubMed Central PMCID: PMC4183581. 20. Poole DS, Yu S, Cai Y, Dinis JM, Muller MA, Jordan I, et al. Influenza A virus polymerase is a site for adaptive changes during experimental evolution in bat cells. Journal of virology. 2014; 88(21):12572– 85. doi: 10.1128/JVI.01857-14 PMID: 25142579; PubMed Central PMCID: PMC4248895. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 15 / 18 Host Cell Entry by Bat-Associated Influenza Viruses 21. Turkington HL, Juozapaitis M, Kerry PS, Aydillo T, Ayllon J, Garcia-Sastre A, et al. Novel Bat Influenza Virus NS1 Proteins Bind Double-Stranded RNA and Antagonize Host Innate Immunity. Journal of virol- ogy. 2015; 89(20):10696–701. doi: 10.1128/JVI.01430-15 PMID: 26246567; PubMed Central PMCID: PMC4580192. 22. Sun X, Shi Y, Lu X, He J, Gao F, Yan J, et al. Bat-derived influenza hemagglutinin H17 does not bind canonical avian or human receptors and most likely uses a unique entry mechanism. Cell reports. 2013; 3(3):769–78. doi: 10.1016/j.celrep.2013.01.025 PMID: 23434510. 23. Wu Y, Wu Y, Tefsen B, Shi Y, Gao GF. Bat-derived influenza-like viruses H17N10 and H18N11. Trends in microbiology. 2014; 22(4):183–91. doi: 10.1016/j.tim.2014.01.010 PMID: 24582528. 24. Zhu X, Yu W, McBride R, Li Y, Chen LM, Donis RO, et al. Hemagglutinin homologue from H17N10 bat influenza virus exhibits divergent receptor-binding and pH-dependent fusion activities. References Proceedings of the National Academy of Sciences of the United States of America. 2013; 110(4):1458–63. doi: 10. 1073/pnas.1218509110 PMID: 23297216; PubMed Central PMCID: PMC3557073. 25. Garcia-Sastre A. The neuraminidase of bat influenza viruses is not a neuraminidase. Proceedings of the National Academy of Sciences of the United States of America. 2012; 109(46):18635–6. doi: 10. 1073/pnas.1215857109 PMID: 23100536; PubMed Central PMCID: PMC3503194. 26. Li Q, Sun X, Li Z, Liu Y, Vavricka CJ, Qi J, et al. Structural and functional characterization of neuramini- dase-like molecule N10 derived from bat influenza A virus. Proceedings of the National Academy of Sciences of the United States of America. 2012; 109(46):18897–902. doi: 10.1073/pnas.1211037109 PMID: 23012237; PubMed Central PMCID: PMC3503196. 27. Zhu X, Yang H, Guo Z, Yu W, Carney PJ, Li Y, et al. Crystal structures of two subtype N10 neuramini- dase-like proteins from bat influenza A viruses reveal a diverged putative active site. Proceedings of the National Academy of Sciences of the United States of America. 2012; 109(46):18903–8. doi: 10. 1073/pnas.1212579109 PMID: 23012478; PubMed Central PMCID: PMC3503178. 28. Hoffmann M, Muller MA, Drexler JF, Glende J, Erdt M, Gutzkow T, et al. Differential sensitivity of bat cells to infection by enveloped RNA viruses: coronaviruses, paramyxoviruses, filoviruses, and influenza viruses. PloS one. 2013; 8(8):e72942. doi: 10.1371/journal.pone.0072942 PMID: 24023659; PubMed Central PMCID: PMC3758312. 29. Muller MA, Raj VS, Muth D, Meyer B, Kallies S, Smits SL, et al. Human coronavirus EMC does not require the SARS-coronavirus receptor and maintains broad replicative capability in mammalian cell lines. mBio. 2012; 3(6). doi: 10.1128/mBio.00515-12 PMID: 23232719; PubMed Central PMCID: PMC3520110. 30. Biesold SE, Ritz D, Gloza-Rausch F, Wollny R, Drexler JF, Corman VM, et al. Type I interferon reaction to viral infection in interferon-competent, immortalized cell lines from the African fruit bat Eidolon hel- vum. PloS one. 2011; 6(11):e28131. doi: 10.1371/journal.pone.0028131 PMID: 22140523; PubMed Central PMCID: PMC3227611. 31. Kuhl A, Hoffmann M, Muller MA, Munster VJ, Gnirss K, Kiene M, et al. Comparative analysis of Ebola virus glycoprotein interactions with human and bat cells. The Journal of infectious diseases. 2011; 204 Suppl 3:S840–9. doi: 10.1093/infdis/jir306 PMID: 21987760; PubMed Central PMCID: PMC3189982. 32. Chaipan C, Kobasa D, Bertram S, Glowacka I, Steffen I, Tsegaye TS, et al. Proteolytic activation of the 1918 influenza virus hemagglutinin. Journal of virology. 2009; 83(7):3200–11. doi: 10.1128/JVI.02205- 08 PMID: 19158246; PubMed Central PMCID: PMC2655587. 33. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 References Plasminogen-binding activity of neuraminidase determines the pathogenicity of influenza A virus. Journal of virology. 2001; 75(19):9297–301. doi: 10.1128/JVI.75. 19.9297–9301.2001 PMID: 11533192; PubMed Central PMCID: PMC114497. 44. Lazarowitz SG, Goldberg AR, Choppin PW. Proteolytic cleavage by plasmin of the HA polypeptide of influenza virus: host cell activation of serum plasminogen. Virology. 1973; 56(1):172–80. PMID: 4795670. 45. Air GM, Laver WG. The neuraminidase of influenza virus. Proteins. 1989; 6(4):341–56. doi: 10.1002/ prot.340060402 PMID: 2482974. 46. Roberts NA. Anti-influenza drugs and neuraminidase inhibitors. Progress in drug research Fortschritte der Arzneimittelforschung Progres des recherches pharmaceutiques. 2001; 56:195–237. PMID: 11417114. 47. Rogers GN, Paulson JC, Daniels RS, Skehel JJ, Wilson IA, Wiley DC. Single amino acid substitutions in influenza haemagglutinin change receptor binding specificity. Nature. 1983; 304(5921):76–8. PMID: 6191220. 48. Nobusawa E, Aoyama T, Kato H, Suzuki Y, Tateno Y, Nakajima K. Comparison of complete amino acid sequences and receptor-binding properties among 13 serotypes of hemagglutinins of influenza A viruses. Virology. 1991; 182(2):475–85. PMID: 2024485. 49. Matrosovich M, Tuzikov A, Bovin N, Gambaryan A, Klimov A, Castrucci MR, et al. Early alterations of the receptor-binding properties of H1, H2, and H3 avian influenza virus hemagglutinins after their intro- duction into mammals. Journal of virology. 2000; 74(18):8502–12. PMID: 10954551; PubMed Central PMCID: PMC116362. 50. Gambaryan AS, Tuzikov AB, Piskarev VE, Yamnikova SS, Lvov DK, Robertson JS, et al. Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylgly- copolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6'-sialyl(N-acetyllactosamine). Virology. 1997; 232(2):345–50. doi: 10.1006/ viro.1997.8572 PMID: 9191848. 51. Baigent SJ, McCauley JW. Influenza type A in humans, mammals and birds: determinants of virus viru- lence, host-range and interspecies transmission. BioEssays: news and reviews in molecular, cellular and developmental biology. 2003; 25(7):657–71. doi: 10.1002/bies.10303 PMID: 12815721. 52. Diederich S, Moll M, Klenk HD, Maisner A. The nipah virus fusion protein is cleaved within the endoso- mal compartment. The Journal of biological chemistry. 2005; 280(33):29899–903. doi: 10.1074/jbc. M504598200 PMID: 15961384. 53. Diederich S, Sauerhering L, Weis M, Altmeppen H, Schaschke N, Reinheckel T, et al. Activation of the Nipah virus fusion protein in MDCK cells is mediated by cathepsin B within the endosome-recycling compartment. Journal of virology. 2012; 86(7):3736–45. doi: 10.1128/JVI.06628-11 PMID: 22278224; PubMed Central PMCID: PMC3302499. 54. Okumura Y, Takahashi E, Yano M, Ohuchi M, Daidoji T, Nakaya T, et al. References Zmora P, Blazejewska P, Moldenhauer AS, Welsch K, Nehlmeier I, Wu Q, et al. DESC1 and MSPL acti- vate influenza A viruses and emerging coronaviruses for host cell entry. Journal of virology. 2014; 88 (20):12087–97. doi: 10.1128/JVI.01427-14 PMID: 25122802; PubMed Central PMCID: PMC4178745. 34. Zimmer G, Locher S, Berger Rentsch M, Halbherr SJ. Pseudotyping of vesicular stomatitis virus with the envelope glycoproteins of highly pathogenic avian influenza viruses. The Journal of general virol- ogy. 2014; 95(Pt 8):1634–9. doi: 10.1099/vir.0.065201–0 PMID: 24814925. 35. Berger Rentsch M, Zimmer G. A vesicular stomatitis virus replicon-based bioassay for the rapid and sensitive determination of multi-species type I interferon. PloS one. 2011; 6(10):e25858. doi: 10.1371/ journal.pone.0025858 PMID: 21998709; PubMed Central PMCID: PMC3187809. 36. Kruger N, Hoffmann M, Weis M, Drexler JF, Muller MA, Winter C, et al. Surface glycoproteins of an Afri- can henipavirus induce syncytium formation in a cell line derived from an African fruit bat, Hypsignathus monstrosus. Journal of virology. 2013; 87(24):13889–91. doi: 10.1128/JVI.02458-13 PMID: 24067951; PubMed Central PMCID: PMC3838219. 37. Hanika A, Larisch B, Steinmann E, Schwegmann-Wessels C, Herrler G, Zimmer G. Use of influenza C virus glycoprotein HEF for generation of vesicular stomatitis virus pseudotypes. The Journal of general virology. 2005; 86(Pt 5):1455–65. doi: 10.1099/vir.0.80788–0 PMID: 15831958. 38. Kruger N, Hoffmann M, Drexler JF, Muller MA, Corman VM, Sauder C, et al. Functional properties and genetic relatedness of the fusion and hemagglutinin-neuraminidase proteins of a mumps virus-like bat PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 16 / 18 Host Cell Entry by Bat-Associated Influenza Viruses virus. Journal of virology. 2015; 89(8):4539–48. doi: 10.1128/JVI.03693-14 PMID: 25741010; PubMed Central PMCID: PMC4442385. 39. Klenk HD, Rott R, Orlich M, Blodorn J. Activation of influenza A viruses by trypsin treatment. Virology. 1975; 68(2):426–39. PMID: 173078. 40. Klenk HD, Rott R, Orlich M. Further studies on the activation of influenza virus by proteolytic cleavage of the haemagglutinin. The Journal of general virology. 1977; 36(1):151–61. doi: 10.1099/0022-1317- 36-1-151 PMID: 142124. 41. Lazarowitz SG, Choppin PW. Enhancement of the infectivity of influenza A and B viruses by proteolytic cleavage of the hemagglutinin polypeptide. Virology. 1975; 68(2):440–54. PMID: 128196. 42. Goto H, Kawaoka Y. A novel mechanism for the acquisition of virulence by a human influenza A virus. Proceedings of the National Academy of Sciences of the United States of America. 1998; 95 (17):10224–8. PMID: 9707628; PubMed Central PMCID: PMC21489. 43. Goto H, Wells K, Takada A, Kawaoka Y. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 References Novel type II transmembrane serine proteases, MSPL and TMPRSS13, Proteolytically activate membrane fusion activity of the hem- agglutinin of highly pathogenic avian influenza viruses and induce their multicycle replication. Journal of virology. 2010; 84(10):5089–96. doi: 10.1128/JVI.02605-09 PMID: 20219906; PubMed Central PMCID: PMC2863848. 55. Hamilton BS, Gludish DW, Whittaker GR. Cleavage activation of the human-adapted influenza virus subtypes by matriptase reveals both subtype and strain specificities. Journal of virology. 2012; 86 (19):10579–86. doi: 10.1128/JVI.00306-12 PMID: 22811538; PubMed Central PMCID: PMC3457293. 56. Baron J, Tarnow C, Mayoli-Nussle D, Schilling E, Meyer D, Hammami M, et al. Matriptase, HAT, and TMPRSS2 activate the hemagglutinin of H9N2 influenza A viruses. Journal of virology. 2013; 87 (3):1811–20. doi: 10.1128/JVI.02320-12 PMID: 23192872; PubMed Central PMCID: PMC3554176. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 17 / 18 Host Cell Entry by Bat-Associated Influenza Viruses 57. Garten W, Braden C, Arendt A, Peitsch C, Baron J, Lu Y, et al. Influenza virus activating host proteases: Identification, localization and inhibitors as potential therapeutics. European journal of cell biology. 2015; 94(7–9):375–83. doi: 10.1016/j.ejcb.2015.05.013 PMID: 26095298. 58. Bottcher E, Matrosovich T, Beyerle M, Klenk HD, Garten W, Matrosovich M. Proteolytic activation of influenza viruses by serine proteases TMPRSS2 and HAT from human airway epithelium. Journal of virology. 2006; 80(19):9896–8. doi: 10.1128/JVI.01118-06 PMID: 16973594; PubMed Central PMCID: PMC1617224. 59. Hatesuer B, Bertram S, Mehnert N, Bahgat MM, Nelson PS, Pohlmann S, et al. Tmprss2 is essential for influenza H1N1 virus pathogenesis in mice. PLoS pathogens. 2013; 9(12):e1003774. doi: 10.1371/ journal.ppat.1003774 PMID: 24348248; PubMed Central PMCID: PMC3857797. 60. Tefsen B, Lu G, Zhu Y, Haywood J, Zhao L, Deng T, et al. The N-terminal domain of PA from bat- derived influenza-like virus H17N10 has endonuclease activity. Journal of virology. 2014; 88(4):1935– 41. doi: 10.1128/JVI.03270-13 PMID: 24284327; PubMed Central PMCID: PMC3911528. 61. Hofmann H, Li X, Zhang X, Liu W, Kuhl A, Kaup F, et al. Severe fever with thrombocytopenia virus gly- coproteins are targeted by neutralizing antibodies and can use DC-SIGN as a receptor for pH-depen- dent entry into human and animal cell lines. Journal of virology. 2013; 87(8):4384–94. doi: 10.1128/JVI. 02628-12 PMID: 23388721; PubMed Central PMCID: PMC3624395. 62. Jung C, Le Doux JM. Lentiviruses inefficiently incorporate human parainfluenza type 3 envelope pro- teins. Biotechnology and bioengineering. 2008; 99(4):1016–27. doi: 10.1002/bit.21622 PMID: 17705232. 63. Steffen I, Simmons G. Pseudotyping Viral Vectors With Emerging Virus Envelope Proteins. Current gene therapy. 2016; 16(1):47–55. PMID: 26785737. PLOS ONE \| DOI:10.1371/journal.pone.0152134 March 30, 2016 References 64. Maruyama J, Nao N, Miyamoto H, Maeda K, Ogawa H, Yoshida R, et al. Characterization of the glyco- proteins of bat-derived influenza viruses. Virology. 2016; 488:43–50. doi: 10.1016/j.virol.2015.11.002 PMID: 26605499. 65. Chandran K, Sullivan NJ, Felbor U, Whelan SP, Cunningham JM. Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection. Science. 2005; 308(5728):1643–5. doi: 10.1126/science. 1110656 PMID: 15831716. 66. Simmons G, Gosalia DN, Rennekamp AJ, Reeves JD, Diamond SL, Bates P. Inhibitors of cathepsin L prevent severe acute respiratory syndrome coronavirus entry. Proceedings of the National Academy of Sciences of the United States of America. 2005; 102(33):11876–81. doi: 10.1073/pnas.0505577102 PMID: 16081529; PubMed Central PMCID: PMC1188015. 67. Kawase M, Shirato K, Matsuyama S, Taguchi F. Protease-mediated entry via the endosome of human coronavirus 229E. Journal of virology. 2009; 83(2):712–21. doi: 10.1128/JVI.01933-08 PMID: 18971274; PubMed Central PMCID: PMC2612384. 68. Qiu Z, Hingley ST, Simmons G, Yu C, Das Sarma J, Bates P, et al. Endosomal proteolysis by cathep- sins is necessary for murine coronavirus mouse hepatitis virus type 2 spike-mediated entry. Journal of virology. 2006; 80(12):5768–76. doi: 10.1128/JVI.00442-06 PMID: 16731916; PubMed Central PMCID: PMC1472567. 69. Cheng Z, Zhou J, To KK, Chu H, Li C, Wang D, et al. Identification of TMPRSS2 as a Susceptibility Gene for Severe 2009 Pandemic A(H1N1) Influenza and A(H7N9) Influenza. The Journal of infectious diseases. 2015; 212(8):1214–21. doi: 10.1093/infdis/jiv246 PMID: 25904605. 18 / 18
https://openalex.org/W2726537687	https://www.redalyc.org/pdf/6617/661770079003.pdf	Russian	null	Constructions of squaring the circle, doubling the cube and angle trisection	Vojnotehnički glasnik	2,017	cc-by	11,174	Vojnotehnicki glasnik/Military Technical Courier ISSN: 0042-8469 vojnotehnicki.glasnik@mod.gov.rs University of Defence Serbia Vojnotehnicki glasnik/Military Technical Courier ISSN: 0042-8469 vojnotehnicki.glasnik@mod.gov.rs University of Defence Serbia Available in: https://www.redalyc.org/articulo.oa?id=661770079003 How to cite Complete issue More information about this article Journal's homepage in redalyc.org Scientific Information System Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Non-profit academic project, developed under the open access initiative How to cite Complete issue More information about this article Journal's homepage in redalyc.org Scientific Information System Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Non-profit academic project, developed under the open access initiative muš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 CONSTRUCTIONS OF SQUARING THE CIRCLE, DOUBLING THE CUBE AND ANGLE TRISECTION Veselin M. Rmuš Vocational Secondary School, Berane, Montenegro, e-mail: veselinrmus12@gmail.com, ORCID iD: http://orcid.org/0000-0001-6104-7281 http://dx.doi.org/10.5937/vojtehg65-13404 Veselin M. Rmuš Vocational Secondary School, Berane, Montenegro, e-mail: veselinrmus12@gmail.com, ORCID iD: http://orcid.org/0000-0001-6104-7281 FIELD: Mathematics ARTICLE TYPE: Original Scientific Paper ARTICLE LANGUAGE: English Abstract: The constructions of three classical Greek problems (squaring the circle, doubling the cube and angle trisection) using only a ruler and a compass are considered unsolvable. The aim of this article is to explain the original methods of construction of the above-mentioned problems, which is something new in geometry. For the construction of squaring the circle and doubling the cube the Thales' theorem of proportional lengths has been used, whereas the angle trisection relies on a rotation of the unit circle in the Cartesian coordinate system and the axioms of angle measurement. The constructions are not related to the precise drawing figures in practice, but the intention is to find a theoretical solution, by using a ruler and a compass, under the assumption that the above- mentioned instruments are perfectly precise. Keywords: construction, squaring the circle, doubling the cube, angle trisection, coordinate system, unit circle, rotation, proportion. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Is Squaring the circle is related to constructing a square with the same area as a given circle. Doubling the cube is the problem of determining the length of the sides of a cube whose volume is double that of a given cube. Angle trisecection concerns the construction of an angle equal to one third of a given arbitrary angle. The above-mentioned problems are allowed to be constructed using only a straightedge and a compass, i.e. using elementary Euclidean construction. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COUR Through an original method based on pure geometry, the three problems have been solved. The work methodology is based on the problem-solving process, i.e. constructive task-solving, consisting of four parts: analysis (description of the construction), construction, proof and discussion. A reader should use a straightedge (ruler), a compass and a sheet of paper to follow the procedure for solving the problems. The construction of squaring the circle In everyday speech we may hear the expression “squaring the circle” used as a metaphor for trying to solve something impossible. The origin of the phrase is not familiar to many of those who resort to its usage in conversation, but it is widely known among mathematicians that it refers to a problem proposed by ancient Greeks, related to constructing a square with the same area as a given circle by using only a compass and a straightedge (ruler). g g ( ) When the length X = 2 is constructed, one can notice that the construction of a side of a square, which meets the requirements of the problem, is similar to the construction of the length X = 2 . Introduction Three problems were proposed in the time of the ancient Greeks, between 600 and 450 BC. Even though the problems of squaring the circle, doubling the cube and angle trisection date back to Thales’s times, it is not known who proposed them. Many Greeks from that period until 500 AD attempted to solve the problems using only Euclidean constructions, but without success. However, they did find a series of solutions using tools other than a straight edge and a compass which made a significant contribution to mathematics at the time. No progress on the unsolved problems was made until 19th century when abstract algebra was developed and concluded that the three Greek 617 TEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 problems cannot be solved. The arguments put forward to prove the unsolvability of squaring the circle, doubling the cube and angle trisection were the impossibility of constructing the square root of , the cube root of 2 and the angle trisection of 60º, respectively (Courant, Robbins, 1973, pp.108-113). These individual cases prejudiced mathematicians against the unsolvability of the three Greek problems. Description of the length construction X = 2 We will consider a line p to contain an arbitrary length AB (Fig. 1). 618 muš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Figure 1 – Division of the length AB in the ratio 1:2 Рис. 1 – Длина отрезка АВ, в отношении 1:2 Слика 1 – Подјела дужи AB у односу 1:2 The point C divides the given length in the ratio of integers 2:1, i.e. AC:CB = 2:1, in the following way: We construct an arbitrary ray Aq and by a compass determine the points M and N so that the length AM=2, and the length MN=1. Then we construct the length NB. The line s passing through the point M is parallel to the length NB. We denote the intersection of the lines s and p by C. Then we construct the line l passing through the point C parallel to the ray Aq and denote its intersection with the length NB by L. Let us prove that the length AB is divided by the point C in the ratio 2:1. In Fig. 1, the triangles ACM and CBL are similar because the corresponding angles at the vertices A and C, i.e. M and L are as equal as angles with the parallel arms in the same direction. Тhe following proportion is true: AC : AM = CB : CL (1) Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Figure 1 – Division of the length AB in the ratio 1:2 Рис. 1 – Длина отрезка АВ, в отношении 1:2 Слика 1 – Подјела дужи AB у односу 1:2 Figure 1 – Division of the length AB in the ratio 1:2 Рис. 1 – Длина отрезка АВ, в отношении 1:2 Слика 1 – Подјела дужи AB у односу 1:2 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Figure 1 – Division of the length AB in the ratio 1:2 Рис. 1 – Длина отрезка АВ, в отношении 1:2 Слика 1 – Подјела дужи AB у односу 1:2 The point C divides the given length in the ratio of integers 2:1, i.e. AC:CB = 2:1, in the following way: We construct an arbitrary ray Aq and by a compass determine the points M and N so that the length AM=2, and the length MN=1. Then we construct the length NB. Description of the length construction X = 2 However, in geometry, as we have shown, 2 is the length, because there are no approximate lengths. Therefore, the value of 2 corresponds to the real number between 1 and 2, i.e. the relation is the following: 1< 2 <2. The relation can be proven in a classical, well-known way described below. The Cartesian coordinate system is given (Fig. 2). In the first quadrant, we construct the square OABC whose side OA equals 1. Figure 2 – Value of the length X between real numbers 1 and 2 Рис. 2 – Значение длины X между действительными числами 1 и 2 Слика 2 – Bриједност дужи X између природних вриједности 1 и 2 OTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 The right-angled triangles ACD and DCB are similar, because the angles at the vertices A and D are as equal as angles with the perpendicular arms. The vertex angle D of the triangle ADB is right-angled because it is peripheral, whereas the straight angle BOA is 180° as the central angle, which is two times as great as the peripheral angle. g g The following proportion is true: (2) VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL C Q.E.D. (Quod erat demonstrandum). In the 19th and 20th century, many mathematicians were trying to prove the unsolvability of squaring the circle using an algebraic method relying on the fact that 2 cannot be written as a fraction; that is why it is considered an approximate number. However, in geometry, as we have shown, 2 is the length, because there are no approximate lengths. Therefore, the value of 2 corresponds to the real number between 1 and 2, i.e. the relation is the following: 1< 2 <2. The relation can be proven in a classical, well-known way described below. The Cartesian coordinate system is given (Fig. 2). In the first quadrant, we construct the square OABC whose side OA equals 1. Figure 2 – Value of the length X between real numbers 1 and 2 Р 2 З д X д д й 1 2 620 VOJ Figure 2 – Value of the length X between real numbers 1 and 2 Рис. Description of the length construction X = 2 The line s passing through the point M is parallel to the length NB. We denote the intersection of the lines s and p by C. Then we construct the line l passing through the point C parallel to the ray Aq and denote its intersection with the length NB by L. Let us prove that the length AB is divided by the point C in the ratio 2:1. In Fig. 1, the triangles ACM and CBL are similar because the corresponding angles at the vertices A and C, i.e. M and L are as equal as angles with the parallel arms in the same direction. g p Тhe following proportion is true: g p Тhe following proportion is true: (1) g p p AC : AM = CB : CL (1) by replacing: АМ = 2 and МN = CL = 1 in we obtain: AC : 2 = CB : 1  AC : CB = 2 : 1, Q.E.D. (Quod erat demonstrandum). Q.E.D. (Quod erat demonstrandum). Q.E.D. (Quod erat demonstrandum). Further, let us construct a semicircle on the length AB (Fig. 1). With a compass and a straightedge, we construct the line n perpendicular to the length AB through the point C and denote its intersection with the semicircle by the point D. We construct the lengths AD and BD. Let us prove that the length CD equals the real number X= 2 . 619 620 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 The right-angled triangles ACD and DCB are similar, because the angles at the vertices A and D are as equal as angles with the perpendicular arms. The vertex angle D of the triangle ADB is right-angled because it is peripheral, whereas the straight angle BOA is 180° as the central angle, which is two times as great as the peripheral angle. The following proportion is true: AC : CD = CD : CB CD2 = AC · CB (2) if we replace AC = 2 and CB = 1 in (2) we obtain CD2 = 2 · 1 CD = 2 , Q.E.D. (Quod erat demonstrandum). In the 19th and 20th century, many mathematicians were trying to prove the unsolvability of squaring the circle using an algebraic method relying on the fact that 2 cannot be written as a fraction; that is why it is considered an approximate number. Description of the length construction X = 2 2 – Значение длины X между действительными числами 1 и 2 Слика 2 – Bриједност дужи X између природних вриједности 1 и 2 Figure 2 – Value of the length X between real numbers 1 and 2 Рис. 2 – Значение длины X между действительными числами 1 и 2 Слика 2 – Bриједност дужи X између природних вриједности 1 и 2 Figure 2 – Value of the length X between real numbers 1 and 2 Рис. 2 – Значение длины X между действительными числами 1 и 2 Слика 2 – Bриједност дужи X између природних вриједности 1 и 2 620 ctions of squaring the circle, doubling the cube and angle trisection, pp.617-640 We denote the diagonal OB by X. According to the Pythagorean Theorem, the equation of the right-angled triangle OAB is the following: Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 X2 = 12 +12  X2 = 2  X = 2 If we rotate the length OB around the point O as a centre of rotation by a negative angle α = - 45º, the point B will be mapped onto the point B1 which is situated between the points A and D on the axis Ox. (Fig. 2) The length OB1 corresponds to the real number and it is bigger than the length OA and smaller than the length OD. Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection “On the basis of Cantor’s axiom which states that there is a one-to- one correspondence between real numbers and points on a line, every point on the real number line corresponds to a real number” (Dolićanin, 1984, p.62). It can be concluded that the real point B1 is situated between integers 1 and 2. 1 Instead of a circle, in Figure 3 a semicircle is constructed for the sake of clarity. 1 Instead of a circle, in Figure 3 a semicircle is constructed for the sake of clarity. Squaring the circle using only a straightedge and a compass is possible Description of the construction: 1 Description of the construction: 1 A given circle1 with the central point O and the radius r are denoted by k(O, r). The length AB is the diameter of an arbitrary circle k (Fig. 3). As shown by the previous method, when constructing the length X= 2 , we divide the diameter AB by the point C in the ratio of integers 11000000 and 3005681, i.e. AC : CB = 11000000 : 3005681, in the following way: g y On the arbitrary ray Aq, we determine the point M by “transferring” 11000000 arbitrary unit lengths. Then we determine the point N so that the length MN equals 3005681 arbitrary unit lengths. p g q y g Then we construct the length NB. Through the point M, we draw a line s parallel to the length NB. The intersection of the line s and the length AB is denoted by C. Through the point C we construct the line l so that it is parallel to the ray Aq and its intersection with the length NB we denote by the point L (Fig. 3). The length AB is divided in the above-mentioned ratio by the point C. 621 RY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Proof: Figure 3 – Proportion of the lengths AC and CB in the ratio 11000000 : 3005681 Рис. 3 – Пропорциональность длины отрезка АС и СВ в отношении 11000000 : 3005681 С 3 П д AC CB д 11000000 3005681 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Proof: VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL CO VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICA Figure 3 – Proportion of the lengths AC and CB in the ratio 11000000 : 3005681 Figure 3 Proportion of the lengths AC and CB in the ratio 11000000 : 3005681 Рис. 3 – Пропорциональность длины отрезка АС и СВ в отношении 11000000 : 3005681 Слика 3 – Пропорционалност дужи AC и CB у односу 11000000 : 3005681 g p g Рис. 3 Drawing the function graph P(t) is left to the reader. Squaring the circle using only a straightedge and a compass is possible 3 – Пропорциональность длины отрезка АС и СВ в отношении 11000000 : 3005681 Слика 3 – Пропорционалност дужи AC и CB у односу 11000000 : 3005681 The triangles AMC and CLB are similar, so we can form the proportion: p p AC : AM = CB : CL (3) Based on relation (3), we replace: AM = 11 · 106 and MN = CL = 3005681 = 3.005681 · 106 It follows that AC : CB = 11 · 106 : 3.005681 · 106 After having it shortened with 106, we get: AC : CB = 11 : 3.005681 = t (4) Based on relation (4) AC : 11 = t  AC =11t and CB : 3.005681 = t  CB = 3.005681t, (5) where t is a non-negative real number, i.e. t > 0 and t € R. (3) (4) (5) AC : 11 = t  AC =11t and CB : 3.005681 = t  CB = 3.005681t, (5) where t is a non-negative real number, i.e. t > 0 and t € R. Let us construct a line n through the point C to be perpendicular to the diameter AB, and denote its (one) intersection with the periphery of the circle by D. Then we draw the lengths AD and BD. AD represents the side of the square whose area is equal to the area of the given circle. Then we construct the square ADGH (Fig. 3). q ( g ) Discussion: The problem of squaring the circle always has two solutions, because the line n with the circle k(O,r), apart from the point D, has one more intersection point D1 and thus the lengths AD and AD1 are 622 equal. We can construct one more square of the same area as the given circle. Thus the problem of squaring the circle has been proven solvable. equal. We can construct one more square of the same area as the given circle. Thus the problem of squaring the circle has been proven solvable. 2 Drawing the function graph r(t) is left to the reader. 3 Proof of squaring the circle by calculation By calculation, we shall now prove that the area of the given circle k(O, r) equals the area of the square ADGH (Fig. 3). Radius r(t) = (11t + 3.005681t) : 2 (6) Radius r(t) = (11t + 3.005681t) : 2 (6) Radius r(t) = (11t + 3.005681t) : 2 r(t) = 7.0028405 · t (6) ( ) ( r(t) = 7.0028405 · t ( ) ( ) r(t) = 7.0028405 · t (6) ( ) ( r(t) = 7.0028405 · t r(t) is a linear function whose graph (the part of line) belongs to the first quadrant and is defined for every t > 0.2 q y By using the formula to calculate the area of circle P = r2, r = 7.0028405 ·t, By using the formula to calculate the area of circle P = r2, r = 7.0028405 ·t, we obtain the equation: we obtain the equation: we obtain the equation: we obtain the equation: P(t) = (7.0028405t)2 P(t) = 49.039775t2 · P(t) = 154.062t2 P0 = 154.06 t2 (7) P(t) = (7.0028405t)2 P(t) = 49.039775t2 · P(t) = 154.062t2 P0 = 154.06 t2 (7) (7) P(t) is a square function whose graph (part of the parabola) belongs to the first quadrant, and is defined for every t > 0.3 q , y Now we shall calculate the area of the square ADGH. 2 q The area of the square ADGH equals AD2 (Fig. 3). If we apply the Pythagorean Theorem on the right-angled triangle ACD in Fig. 3, we obtain the relation: AD2= AC2 + CD2 AD2= AC2 + CD2 (8 AD2= AC2 + CD2 (8) AC = 11t  AC2 = 121t2 (9) AC = 11t  AC2 = 121t2 (9) AC = 11t  AC2 = 121t2 (9) AC = 11t  AC2 = 121t2 Based on the similarity of the triangles ACD and DCB in Fig. Proof of squaring the circle by calculation 3, the proportion is true: AC : CD = CD : CB  AC2 =AC · CB if AC = 11t  AC = 121t2 (10) AC : CD = CD : CB  AC2 =AC · CB if AC = 11t  AC = 121t2 (10) AC : CD = CD : CB  AC2 =AC · CB if AC = 11t  AC = 121t2 (10) By replacing AC = 11t and CB = 3.005681t in (9), we obtain CD2 = 11t · 3.005681t CD2 = 33.062t2 (11) (10) By replacing AC = 11t and CB = 3.005681t in (9), we obtain CD2 = 11t · 3.005681t CD2 = 33.062t2 (11) (11) 623 OTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 If we replace relation (10) and (11) with relation (8), we get VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 AD2  the P0 = solv edg a co is p leng Рис AD2 = 121t2 + 33,062t2  P□ = 154.06t2 (12) (12) By comparing relations (7) and (12), one may notice that the area of the circle is equal to the area of the square, i.e. By comparing relations (7) and (12), one may notice that the area of the circle is equal to the area of the square, i.e. P0 = P□ Thus the Greek problem of squaring the circle has been proven solvable by calculation. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL CO Constuction of doubling the cube (hexahedron) The problem of doubling the cube relates to the construction of the edge of a second cube whose volume is double that of the first, using only a compass and a straightedge. Doubling of the cube using only a straightedge and a compass is possible p Description of the construction to determine the edges of the cube: Based on the similarity of the triangles, the following proportion is true: AB : AM = AC : AN (13) by replacing AB = а and AC = x, AM = 107 and AN = 1.2599211 · 107 q g p Based on the similarity of the triangles, the following proportion is true: AB : AM = AC : AN (13) AB : AM = AC : AN (13) by replacing AB = а and AC = x, AM = 107 and AN = 1.2599211 · 107 in (13) we obtain: by replacing AB = а and AC = x, AM = 107 and AN = 1.2599211 · 107 in (13) we obtain: (14) a : 107 = x : 1.2599211 · 107. a : 107 = x : 1.2599211 · 107. After having it shortened with 107 in relation (14), we get After having it shortened with 107 in relation (14), we get a : 1 = x : 1.2599211 it follows that x = 1.1599211 · a. (15) The cubed equation (15) x3 = (1.2599211· a)3 x3 = 1.25992113 · a3 x3 = 2 · a3 (16) If a3 = V1 and x3 = V2 (17) V2 = 2 · V1 (15) (16) (17) Q.E.D. (Quod erat demonstrandum) p Description of the construction to determine the edges of the cube: On the arbitrary line p, we determine the points A and B so that the length AB is equal to the edge of the given cube, i.e. AB = a in Fig. 4. Figure 4 – Proportion of the lengths AB and AC in the ratio 10000000 : 12599211 Рис. 4 – Пропорциональность длины отрезка АС и СВ в отношении 10000000 : 12599211 Слика 4 – Пропорционалност дужи AB и AC у односу 10000000 : 12599211 Figure 4 – Proportion of the lengths AB and AC in the ratio 10000000 : 12599211 Рис. 4 – Пропорциональность длины отрезка АС и СВ в отношении 10000000 : 12599211 Слика 4 – Пропорционалност дужи AB и AC у односу 10000000 : 12599211 Figure 4 – Proportion of the lengths AB and AC in the ratio 10000000 : 12599211 Рис. 4 – Пропорциональность длины отрезка АС и СВ в отношении 10000000 : 12599211 Слика 4 – Пропорционалност дужи AB и AC у односу 10000000 : 12599211 624 V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 We will construct an arbitrary ray Aq and determine the points M and N on the ray so that the length AM = 107 arbitrary unit lengths, and the length AN=12599211 arbitrary unit lengths, i.e. 7 g AM : AN = 107 : 12599211. Let us construct the length BM. Then we construct a line s through the point N parallel to the length BM. The intersection of the line s and the line p is denoted by C. We will prove that the length AC is the edge of the cube, whose volume is double that of the given cube. g Let AC equals x. The triangles ABM and ACN are similar because the angle at the vertex A is common, and the angles at the vertices B and C are as equal as angles with parallel arms in the same direction. are as equal as angles with parallel arms in the same direction. Q.E.D. (Quod erat demonstrandum) 625 Figure 5 – Oblique projection of a cube Рис. 5 – Наклонная проекция куба Слика 5 – Коса пројекција коцке VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Figure 5 – Oblique projection of a cube Рис. 5 – Наклонная проекция куба Слика 5 – Коса пројекција коцке VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Is VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL C Figure 5 – Oblique projection of a cube Рис. 5 – Наклонная проекция куба Слика 5 – Коса пројекција коцке For the oblique picture to be more transparent (clearer), we will take the angle α = 30º, and q = 1 : 2 (α = 30º, angle –yOx). The ratio q = 1 : 2 represents the ratio of the length on the Ox-axis and O-z-axes, i.e. if an arbitrary length on the Ox-axis equals 1, then the length on the Oz-axis is twice longer (α is called a reduction angle, q is a reduction ratio). ) On the negative part of the z-axis we determine a point L so that the length OL is equal to the edge of the cube in its true size (the length AB = a in Fig. 4). Through the point L we construct the line n1 perpendicular to Ox and we denote their intersection by B (Fig. 5). p p y ( g ) The point D is determined on the ray Oy so that the length OD is equal to the edge of the given cube in Fig. 4 in its true size. The point A coincides with the vertex of the trihedral. Then, we determine the point C by constructing a parallelogram ABCD (it is the oblique picture of the lower base of the given cube). Through the points B, C and D we construct lines parallel to the z-axis. On the z-axis and the parallel lines we determine the points A1B1C1D1 so that AA1 = BB1 = CC1 = DD1 = a. Let us construct other lengths where ABCDA1B1C1D1 presents an oblique picture of the given cube. In a similar way, we construct a cube so that its volume is double that of the given cube, i.e. whose edge is x (it is the length AC in Fig. 4). On the negative part of the z-axis in Fig. Q.E.D. (Quod erat demonstrandum) 5 we determine the point R so that the length OR equals the edge of the cube x = AC in Fig. 4. Through the point R we construct the line n2 perpendicular to the axis Ox and denote the intersection with the Ox-axis by N. The point M coincides with the trihedral vertex. The point Q is determined on the Oy axis so that MQ = x (in its true 626 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 size). Let us construct a parallelogram MNPQ (it is the oblique picture of the lower base of the new cube). Through the points N, P, Q in Fig. 5 we construct lines parallel to the z-axis. size). Let us construct a parallelogram MNPQ (it is the oblique picture of the lower base of the new cube). Through the points N, P, Q in Fig. 5 we construct lines parallel to the z-axis. p On the z-axis and all the parallel lines we construct the length x in its true size by a compass and we determine the points M1, N1, P1 and Q1 so that MM1=NN1=PP1=QQ1=x. (Fig. 4). The parallelogram M1N1P1Q1 represents the oblique picture of the upper base of the new cube, whereas MNPQM1N1P1Q1 is the oblique picture of the new cube. In this way, we have constructed the cube MNPQM1N1P1Q1 whose volume is double that of the given cube ABCDA1B1C1D1. On the z-axis and all the parallel lines we construct the length x in its true size by a compass and we determine the points M1, N1, P1 and Q1 so that MM1=NN1=PP1=QQ1=x. (Fig. 4). The parallelogram M1N1P1Q1 represents the oblique picture of the upper base of the new cube, whereas MNPQM1N1P1Q1 is the oblique picture of the new cube. In this way, we have constructed the cube MNPQM1N1P1Q1 whose y volume is double that of the given cube ABCDA1B1C1D1. Discussion: The above method has proven the solvability of doubling the cube using only a straightedge and a compass. The problem always has a solution, i.e. every cube can be doubled. The construction of angle trisection 627 OTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 (10) The central angle (in a circle) is twice the size of the periphery angle which lies over the same arc of the circle. (11) Given the circle with the centre at the origin of Cartesian coordinate system. Each chord constructed to be parallel to the coordinate axis cuts equal circular arcs on the given circle. q g (12) In mathematics, a unit circle is a circle with a radius of one. (12) In mathematics, a unit circle is a circle with a radius of one. (13) An angle bisector is a ray that divides an angle into two equal angles. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017 g (14) Each length can be divided into any (arbitrary) number of equal parts. The construction of angle trisection Angle trisection is related to dividing an arbitrary angle into three equal parts in a constructive way using only a straightedge and a compass. Dividing an angle into 3 equal parts does not seem to be a particular problem. For instance, it is easy to construct one-third of the angles of 45º, 67º 30’, 90º, 135º, 180º, 202º 30’, 270º, 360º, etc. using a straightedge and a compass. However, the general problem arises when an arbitrary angle should be divided into three equal parts. q p In order to solve this problem, we present some wellknown geometry properties (axioms, theorems, definitions) which will be used here. p p ( ) (1) Axiom on the measurement of angles: The degree measure of an angle equals the sum of degree measures of the angle divided by an arbitrary ray which passes through its arms. (1) Axiom on the measurement of angles: The degree measure of an angle equals the sum of degree measures of the angle divided by an arbitrary ray which passes through its arms. y y g (2) By convention, the rotation of an angle arm counter-clockwise is called positive rotation, whereas negative rotation goes clockwise. Positive rotation is denoted by R(O, α), negative by R(O, -α), where the point O is the centre of rotation and the oriented angle α is the angle of rotation. g g (3) The base angles of an isosceles triangle are always equal. (3) The base angles of an isosceles triangle are always equal. (4) The exterior angle of a triangle equals the sum of two interior opposite non-adjacent angles. (4) The exterior angle of a triangle equals the sum of two interior opposite non-adjacent angles. pp j g (5) Vertically opposite angles are equal. (6) Corresponding angles are equal in measure if and only if two parallel lines are cut by a transversal. y (7) Alternate angles are equal in measure if and only if two parallel lines are cut by a transversal. y (8) A circle is a centrally symmetric figure. (9) A circle is an axially symmetric figure (8) A circle is a centrally symmetric figure (8) A circle is a centrally symmetric figure. ( ) y y g (9) A circle is an axially symmetric figure. 9) A circle is an axially symmetric figure. (9) A circle is an axially symmetric figure. 4 The division of the length FE has been constructed separately for the sake of clarity in Figure 8. Angle trisection using a straightedge and a compass is possible VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL C Proof: Proof: Given an acute angle with the vertex at point O and the arms p and q (Fig. 6). Figure 6 – Model of the angle pOq Рис. 6 – Модель угла pOq Слика 6 – Модел угла pOq Figure 6 – Model of the angle pOq Рис. 6 – Модель угла pOq Слика 6 – Модел угла pOq Figure 6 – Model of the angle pOq Рис. 6 – Модель угла pOq Слика 6 – Модел угла pOq gure 6 – Model of the angle pOq Рис. 6 – Модель угла pOq Слика 6 – Модел угла pOq Let us construct the Cartesian coordinate system xOy so that the positive part of the x-axis corresponds to the arm p of the given angle (marked as x ≡ p). (Fig. 7) Figure 7 – Unit circle with with the centre at the origin of coordinate system Рис. 7 – Единичная окружность с центром в начале системы координат Слика 7 – Јединична кружница са центром у координатном почетку Figure 7 – Unit circle with with the centre at the origin of coordinate system Рис. 7 – Единичная окружность с центром в начале системы координат Слика 7 – Јединична кружница са центром у координатном почетку 628 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Then we construct the circle k with the centre O and a radius that equals one (in Figure 7 marked as k(O, 1)). q ( g ( )) On the basis of property (12), this will hereinafter be referred to as the unit circle. The intersections of the unit circle k(O,1) with the x-axis and the y-axis are denoted by A and B, and C and D, respectively. y p y The intersection of the unit circle k(O,1) and the arm q is denoted by E (Fig. 7). Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trise The problem arises when we want to divide the angle BOE into three equal parts in a constructive way using only a straightedge and a compass. The first step is to divide the angle β = 22º 30’ into three equal parts. On the unit circle, we construct the angle β by a 45-degree bisector (Fig. 8). Angle trisection using a straightedge and a compass is possible 629 Rmuš, V., Constructions of squaring the circle, doubling Figure 8 – Angle trisection β = 22º 30’ Рис. 8 – Трисекция угла β = 22º 30’ Слика 8 – Трисекција угла β = 22º 30’ Description of the construction: Through the point E (Fig. 8) we draw a perpendicular n to the x-axis and its intersection with the x-axis is denoted by F. Then, we divide the length FE into three equal parts and denote the points by G and H (Fig. 9)4 4 The division of the length FE has been constructed separately for the sake of clarity in Figure 8. Figure 8 – Angle trisection β = 22º 30’ Рис. 8 – Трисекция угла β = 22º 30’ Слика 8 – Трисекција угла β = 22º 30’ Figure 8 – Angle trisection β = 22º 30’ Рис. 8 – Трисекция угла β = 22º 30’ Слика 8 – Трисекција угла β = 22º 30’ Figure 8 – Angle trisection β = 22º 30’ Рис. 8 – Трисекция угла β = 22º 30’ Слика 8 – Трисекција угла β = 22º 30’ Figure 8 – Angle trisection β = 22º 30’ Рис. 8 – Трисекция угла β = 22º 30’ Слика 8 – Трисекција угла β = 22º 30’ Description of the construction: p Through the point E (Fig. 8) we draw a perpendicular n to the x-axis and its intersection with the x-axis is denoted by F. Then, we divide the length FE into three equal parts and denote the points by G and H (Fig. 9)4 629 Figure 9 – Division of the length FE for angle trisection of β = 22º 30’ Рис. 9 – Длина деления отрезка FE по трисекции угла β = 22º 30’ Слика 9 – Подјела дужи FE за трисекцију угла β = 22º 30’ OTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Figure 9 – Division of the length FE for angle trisection of β = 22º 30’ Рис. 9 – Длина деления отрезка FE по трисекции угла β = 22º 30’ Слика 9 – Подјела дужи FE за трисекцију угла β = 22º 30’ VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COUR Figure 9 – Division of the length FE for angle trisection of β = 22º 30’ Рис. Angle trisection using a straightedge and a compass is possible 9 – Длина деления отрезка FE по трисекции угла β = 22º 30’ Слика 9 – Подјела дужи FE за трисекцију угла β = 22º 30’ From the x-axis, the set of points on the length FE is F-G-H-E (Ostojić, 1980, p.164). Let us construct a line s passing through the point G parallel to the x-axis and denote its intersections with the unit circle k in the first quadrant by the point L and in the second quadrant by the point M. (Fig.8) According to (11), the circular arcs AM and BL are equal on the unit circle k(O, 1). Given BOE = β, and BOL = α, we will prove that BOL= 3 1 BOE, ie. 3   . Based on (9), the point L is the symmetric point of M with respect to the y-axis. Let the point M be the symmetric point of N with respect to the origin of the Coordinate system based on (8) and the point L is the symmetric point of N with respect to the x-axis based on property (9). Further, if the point L is the symmetric point of R with respect to the origin, then the point M is the symmetric point of R with respect to the x-axis (Fig. 8). y ( g ) Based on the properties of the circle as an axially symmetric and centrally symmetric figure, it follows that the circular arcs are equal, i.e. BL = BN = AM = AR. Therefore, on the unit circle, the central angles which lie over the equal circular arcs are equal. Based on (5), it follows that BON = AOM = α and BOL = AOR = α. Applying property (10), the central angle NOL is twice the size of the periphery angle NML which lies over the arc of the circle NL, i.e. NML = (2α) : 2 = α. Since the triangle OLM is an isosceles triangle (the base is the chord ML), then according to (3) OLM = OML = α. 630 muš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Let the symmetric point E with respect to the y-axis be Q (Fig. 8). The circular arcs AQ and BE are equal in accordance with property (11). It was pointed out that circular arcs AM and BL are equal. LTE = ТЕМ + ТМЕ LTE = ТЕМ + ТМЕ If we replace the angles with Greek letters ( LTE = β, TEM = 2α TME = α), we obtain If we replace the angles with Greek letters ( LTE = β, TEM = 2α, TME = α), we obtain β = 2α + α, β = 3α, α = 3  Q.E.D β = 2α + α, β = 3α, α = 3  Q.E.D The angle trisection of 22º 30’ is α = 7º 30’, and thus the angle trisection has been proven. Angle trisection using a straightedge and a compass is possible Accordingly, the chords EQ and LM are parallel (marked as EQ \|\| LM). Based on (7) MEQ = LME as alternate angles are equal and the line determined by M and E is called transversal and it follows that MEQ = LME = α. Let us construct the length LQ. The chords ME and QL intersect at the point S. OMSL quadrilateral is a rhombus. The diagonal ML of the rhombus divides the angle OME into two equal parts. It follows that ОМЕ = 2α. OME is an isosceles triangle and the angle OEM = 2α. g g The angle NOE is central, and NME peripheral over the same arc NE. Based on (10) it follows that: α + β = 2 · 2α  α + β = 4α  β = 3α  α = 3  α + β = 2 · 2α  α + β = 4α  β = 3α  α = 3  Alternative proof: p The intersection of the arm q and the line s is determined by the point T. According to (3), the triangle OEM is an isosceles triangle (the chord ME is the base of the triangle), so that ОМЕ = 2α and ОЕМ = ТЕМ = 2α. The angle LME = ТМЕ =α (because the points L and T belong to the line s). ) The angle BOE = LTE = β, as they are corresponding angles in accordance with property (6). Further, based on (4), the external angle of the triangle TME is equal to the sum of the two internal non-adjacent angles, i.e. Angle trisection of β < 22º 30’ f the acute angle is less than 45º, then two angle trisections are erformed – of 22º 30’ and of less than 22º 30’, because the given ngle was divided into two parts by the ray. The angle of 22º 30’ is onstructed using the angle bisector of 45º. In accordance with roperty 1, thirds of the circular arcs (which have been described in Figs. 8, 9, 10 and 11) are summed up and as a whole they epresent the third of the angle. f the acute angle is less than 67º 30’, then the angle is divided into wo angles: one angle of 45º and the other of less than 22º 30’ by he ray. The angle trisection of 45º is performed first. The angle of 5º is obtained by constructing the angle bisector of 30º. The angle f less than 22º30’ is divided (as described in Figs. 10 and 11). The hird of the circular arc of the angle of 45º and the third of the ircular arc less than 22º 30’ are added up and as a whole epresent the third of the given angle. f the acute angle is less than 90º, then the angle is divided into one ngle of 67º 30’ and the other angle less than 22º 30’ by the ray. Trisections of both angles are performed separately. The angle risection of 67º 30’ is simple as it is the angle of 22º 30’ and it is btained by the angle bisector of 45º. The third of the angle of less han 22º 30’has been described in Figs. 10 and 11. The sum of hirds of circular arcs, as a whole, is transferred to the circular arc of he given angle (three times) and thus the division of the unit circle nto three equal parts has been performed, i.e. the trisection of an rbitrary acute angle. trary angle trisection he acute angle trisection has been described. iven the obtuse angle β`= 90º + β (the arm p coincides with the part of the x-axis, and the arm q belongs to the second quadrant), Then, º + 3 . a) The trisection of an angle less than 22º 30’ has been described in Figs. 10 and 11. Angle trisection of β < 22º 30’ b) If the acute angle is less than 45º, then two angle trisections are performed – of 22º 30’ and of less than 22º 30’, because the given angle was divided into two parts by the ray. The angle of 22º 30’ is constructed using the angle bisector of 45º. In accordance with property 1, thirds of the circular arcs (which have been described in Figs. 8, 9, 10 and 11) are summed up and as a whole they represent the third of the angle. p g c) If the acute angle is less than 67º 30’, then the angle is divided into two angles: one angle of 45º and the other of less than 22º 30’ by the ray. The angle trisection of 45º is performed first. The angle of 15º is obtained by constructing the angle bisector of 30º. The angle of less than 22º30’ is divided (as described in Figs. 10 and 11). The third of the circular arc of the angle of 45º and the third of the circular arc less than 22º 30’ are added up and as a whole represent the third of the given angle. p g g d) If the acute angle is less than 90º, then the angle is divided into one angle of 67º 30’ and the other angle less than 22º 30’ by the ray. Trisections of both angles are performed separately. The angle trisection of 67º 30’ is simple as it is the angle of 22º 30’ and it is obtained by the angle bisector of 45º. The third of the angle of less than 22º 30’has been described in Figs. 10 and 11. The sum of thirds of circular arcs, as a whole, is transferred to the circular arc of the given angle (three times) and thus the division of the unit circle into three equal parts has been performed, i.e. the trisection of an arbitrary acute angle. Angle trisection of β < 22º 30’ Proof: Given the acute angle β less than 22º 30’ in Fig.10. 631 Description of the construction: ( ) f f VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Description of the construction: The procedure, denoting (marking) and proof of the angle trisection less than 22º 30’ are completely the same as the procedure, denoting (marking) and proof we explained when we constructed the angle trisection of β = 22º 30’ (shown in Figs. 8 and 9). The construction of β < 22º 30’ is shown in Figs. 10 and 11. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Iss VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 201 Figure 10 – Аngle trisection β < 22º 30’ Рис. 10 – Трисекция угла β < 22º 30’ Слика 10 –Трисекција угла β < 22º 30’ Figure 11 – Division of the length FE for the angle trisection of β < 22º 30’ Рис. 11 – Длина деления отрезка FE по трисекции угла β < 22º 30’ Слика 11 – Подјела дужи FE за трисекцију угла β < 22º 30’ Accordingly if the acute angle pOq is given and if it belongs to th Figure 10 – Аngle trisection β < 22º 30’ Рис. 10 – Трисекция угла β < 22º 30’ Слика 10 –Трисекција угла β < 22º 30’ Figure 10 – Аngle trisection β < 22º 30’ Рис. 10 – Трисекция угла β < 22º 30’ Слика 10 –Трисекција угла β < 22º 30’ Figure 11 – Division of the length FE for the angle trisection of β < 22º 30’ Рис. 11 – Длина деления отрезка FE по трисекции угла β < 22º 30’ Слика 11 – Подјела дужи FE за трисекцију угла β < 22º 30’ Accordingly, if the acute angle pOq is given and if it belongs to the first quadrant, then we compare it to the angle of 22º 30’. Every acute angle may be: less than 22º 30’, less than 45º, less than 67º 30’ and less than 90º. 632 We shall explain the angle trisection for all the abovementioned cases: 633 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 shall explain the angle trisection for all the abovementioned The trisection of an angle less than 22º 30’ has been described in Figs. 10 and 11. Arbitrary angle trisection 1. The acute angle trisection has been described. 1. The acute angle trisection has been described. g 2. Given the obtuse angle β`= 90º + β (the arm p coincides with the positive part of the x-axis, and the arm q belongs to the second quadrant), Fig. 12. Then, g 2. Given the obtuse angle β`= 90º + β (the arm p coincides with the positive part of the x-axis, and the arm q belongs to the second quadrant), Fig. 12. Then, 3  = 30º + 3 . 2. Given the obtuse angle β`= 90º + β (the arm p coincides with the positive part of the x-axis, and the arm q belongs to the second quadrant), Fig. 12. Then, 3  = 30º + 3 . 633 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 Figure 12 – Arm q of β’ angle belongs to the second quadrant Рис. 12 – Луч q угла β’ принадлежит второму квадранту Слика 12 – Крак q угла β’ припада другом квадранту VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL CO Figure 12 – Arm q of β’ angle belongs to the second quadrant Рис. 12 – Луч q угла β’ принадлежит второму квадранту Слика 12 – Крак q угла β’ припада другом квадранту Applying property (2) we perform a rotation of β’ for the angle (-90º), or R(O, -90º). Then the arm p ≡ coincides with the negative part of the y-axis and the arm q belongs to the first quadrant. A rotation of the points A, B, C and D has also been performed. (Fig. 13) Figure 13 – Рotation of β’ for the angle (-90º) Рис. 13 – Поворот угла β’ на (-90º) Слика 13 – Ротација угла β’ за (-90º) Figure 13 – Рotation of β’ for the angle (-90º) Рис. 13 – Поворот угла β’ на (-90º) Слика 13 – Ротација угла β’ за (-90º) 634 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 After the rotation of β’, the angle β becomes acute and its trisection explained in the abovementioned cases for the acute angle can be applied (see a, b, c, d in 3.2.). The third of circular arc β is added to the circular arc of 30º on the unit circle (Fig.14). Arbitrary angle trisection Figure 14 – Angle trisection of β’ when the arm q belongs to the second quadrant Рис. 14 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит второму квадранту Слика 14 – Трисекција угла β’ када крак q угла β’ припада другом квадранту Figure 14 – Angle trisection of β’ when the arm q belongs to the second quadrant Рис. 14 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит второму квадранту Слика 14 – Трисекција угла β’ када крак q угла β’ припада другом квадранту Figure 14 – Angle trisection of β’ when the arm q belongs to the second quadrant Рис. 14 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит второму квадранту Слика 14 – Трисекција угла β’ када крак q угла β’ припада другом квадранту Figure 14 – Angle trisection of β’ when the arm q belongs to the second quadrant Рис. 14 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит второму квадранту Слика 14 – Трисекција угла β’ када крак q угла β’ припада другом квадранту The angle BOE is divided into three equal parts by the dashed lines a and b, and it follows BON = NOQ = QOE = 3 . The angle BOT equals 30º. (Fig. 14) q ( g ) The circular arc TN in Fig. 14 equals the circular arc CL in Fig. 13. The sum of the circular arcs BT and TN is equal to the circular arc BN. It follows that BON = 3 1 BOE. (The construction has been performed on the unit circle in Fig. 14). 3. If the angle pOq is less than 270º, i.e. if the arm q belongs to the third quadrant, then we write down β` =180º + β (Fig. 15). The rotation of the angle β’ for the angle (-180º) is performed, i.e. R(O,-180º). Then the p is congruent to the negative part of the x-axis, while the q belongs to the first quadrant. The angle β as the acute angle belongs to the first quadrant. (Fig. 16) The angle φ is straight. (Fig. 15) Then, Then, 635 , 3  = 60º + 3 . 3  = 60º + 3 . 635 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. Arbitrary angle trisection 65, Issue 3 It follows that the circular arc of 60º is added to the third of circular arc of the acute angle β. VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issu Figure 15 – Arm q of the β’ angle belongs to the third quadrant Рис. 15 – Луч q угла β’ принадлежит третьему квадранту Слика 15 – Крак q угла β’ припада трећем квадранту Figure 15 – Arm q of the β’ angle belongs to the third quadrant Рис. 15 – Луч q угла β’ принадлежит третьему квадранту Слика 15 – Крак q угла β’ припада трећем квадранту Figure 16 – Рotation of β’ for the angle (-180º) Рис. 16 – Поворот угла β’ на (-180º) Слика 16 – Ротација угла β’ за (-180º) Figure 16 – Рotation of β’ for the angle (-180º) Рис. 16 – Поворот угла β’ на (-180º) Слика 16 – Ротација угла β’ за (-180º) 636 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 637 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-640 hen the arm q belongs to the third quadrant луч q угла β’ принадлежит третьему квадранту крак q угла β’ припада трећем квадранту o three equal parts by the dashed lines a TN in Fig. 17 equals the circular arc AL BT and the circular arc TN equals the or BON = 3 1 BOE in Fig. 17 where less than 360º, i.e. if the arm q belongs β (Fig. 18), then we perform a rotation of 70º, marked as R(O,-270º). The arm p the y-axis and the arm q belongs to the angle belongs to the fourth quadrant надлежит четвертому квадранту ’ припада четвртом квадранту Figure 17 – Angle trisection of β’ when the arm q belongs to the third quadrant Рис. 17 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит третьему квадранту Слика 17 – Трисекција угла β’ када крак q угла β’ припада трећем квадранту Figure 17 – Angle trisection of β’ when the arm q belongs to the third quadrant Рис. 17 – Трисекция угла β’ в случае, когда луч q угла β’ принадлежит третьему квадранту Слика 17 – Трисекција угла β’ када крак q угла β’ припада трећем квадранту The angle BOE is divided into three equal parts by the dashed lines a and b (in Fig. Arbitrary angle trisection 17). The circular arc TN in Fig. 17 equals the circular arc AL in Fig.16. g The sum of the circular arc BT and the circular arc TN equals the circular arc BN (on the unit circle), or BON = 3 1 BOE in Fig. 17 where the angle BOT equals 60º. 4. Finally, if the angle pOq is less than 360º, i.e. if the arm q belongs to the fourth quadrant, β`= 270º + β (Fig. 18), then we perform a rotation of the angle β’ for the angle of -270º, marked as R(O,-270º). The arm p coincides with the positive part of the y-axis and the arm q belongs to the first quadrant (in Fig. 19). 637 Figure 18 – Arm q of the β’ angle belongs to the fourth quadrant Рис. 18 – Луч q угла β’ принадлежит четвертому квадранту Слика 18 – Крак q угла β’ припада четвртом квадранту Figure 18 – Arm q of the β’ angle belongs to the fourth quadrant Рис. 18 – Луч q угла β’ принадлежит четвертому квадранту Слика 18 – Крак q угла β’ припада четвртом квадранту 637 Figure 19 – Рotation of β’ for the angle (-270º) Рис. 19 – Поворот угла β’на (-270º) Слика 19 – Ротација угла β’ за (-270º)   Figure 19 – Рotation of β’ for the angle (-270º) Рис. 19 – Поворот угла β’на (-270º) Слика 19 – Ротација угла β’ за (-270º) OTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 (wh con an cor Р С an eq VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL CO Figure 19 – Рotation of β’ for the angle (-270º) Рис. 19 – Поворот угла β’на (-270º) Слика 19 – Ротација угла β’ за (-270º) Figure 19 – Рotation of β’ for the angle (-270º) Рис. Arbitrary angle trisection 19 – Поворот угла β’на (-270º) Слика 19 – Ротација угла β’ за (-270º) Since in this case 3  = 90º + 3  the trisection of the acute angle β (which is situated in the first quadrant after the rotation of the angle β’) is constructed in the same way as that described of the trisection of acute angles, then the third of the circular arc β is added to the circular arc corresponding to the angle of 90º which belongs to the unit circle (Fig. 20). (which is situated in the first quadrant after the rotation of the angle β’) is constructed in the same way as that described of the trisection of acute angles, then the third of the circular arc β is added to the circular arc corresponding to the angle of 90º which belongs to the unit circle (Fig. 20). Figure 17 – Angle trisection of β’ when the arm q belongs to the fourth quadrant Рис. 17– Трисекция угла β’ в случае, когда луч q угла β’ принадлежит четвертому квадранту Слика 17 – Трисекција угла β’ када крак q угла β’ припада четвртом квадранту Figure 17 – Angle trisection of β’ when the arm q belongs to the fourth quadrant Рис. 17– Трисекция угла β’ в случае, когда луч q угла β’ принадлежит четвертому квадранту Слика 17 – Трисекција угла β’ када крак q угла β’ припада четвртом квадранту Figure 17 – Angle trisection of β’ when the arm q belongs to the fourth quadrant Рис. 17– Трисекция угла β’ в случае, когда луч q угла β’ принадлежит четвертому квадранту Слика 17 – Трисекција угла β’ када крак q угла β’ припада четвртом квадранту The angle BOE is divided into three equal parts by the dashed lines a and b. (Fig. 20) The angle BOC equals 90º. The circular arc CN in Fig. 20 equals the circular arc DL in Fig. 19. It follows that BON = 3 1 BOE. It follows that BON = 3 1 BOE. It follows that BON = 3 1 BOE. 638 ctions of squaring the circle, doubling the cube and angle trisection, pp.617-640 Discussion: Applying the described method, the angle trisection can be performed for every angle. The diameter of the unit circle has been taken arbitrarily and cannot be changed while constructing without changing the lengths of circle arcs. Arbitrary angle trisection Depending on the size of the angle, thirds of a circle arc less than 22º 30’ on the unit circle can be added to a circle arc of 7º 30’, 15º, 22º 30’, 30º, 60º and 90º. 639 Rmuš, V., Constructions of squaring the circle, doubling the cube and angle trisection, pp.617-6 cussion: Applying the described method, the angle trisection can rmed for every angle. The diameter of the unit circle has been rbitrarily and cannot be changed while constructing without g the lengths of circle arcs. Depending on the size of the angle, a circle arc less than 22º 30’ on the unit circle can be added to a c of 7º 30’, 15º, 22º 30’, 30º, 60º and 90º. ferences rant, R., Robbins, H., 1973. What is Mathematics? An Elementary h to Ideas and Methods. Oxford University Press, London. ćanin, Ć., 1984. Маthematics. Institute for Textbooks and Educational , Priština. ojić, О., 1980. Theoretical Basis of Initial Mathematics Teaching, k for Math Teachers, Republic Institute for the Improvement of n, Titograd. KOНСТРУКЦИИ КВАДРАТУРА КРУГА, УДВОЕНИЕ КУБА И ТРИСЕКЦИЯ УГЛА Веселин М. Рмуш Средняя профессиональная школа, г. Беране, Черногория ОБЛАСТЬ: математика ВИД СТАТЬИ: оригинальная научная статья ЯЗЫК СТАТЬИ: английский Резюме: Построение трех классических античных задач (квадратура круга, удвоение куба и трисекция угла), с помощью линейки и циркуля до сих пор считается неразрешимым. В данной статье представлены оригинальные методы построения вышеупомянутых задач, что представляет собой инновационный подход в геометрии. Для построения квадратуры круга и удвоения куба была использована теорема Фалеса о пропорциональных отрезках, а трисекция угла основанная на повороте единичного круга в декартовой системе координат и аксиомах измерения угла. Данные построения не претендуют на создание точных чертежей на практике, так как главная наша цель заключается в нахождении теоретического решения, основанного на построении данных фигур с помощью линейки и циркуля, с учетом предположения, что вышеупомянутые инструменты являются совершенно точными. Ключевые слова: построение, квадратура круга, удвоение куба, трисекция угла, система координат, единичный круг, вращение, пропорция. KOНСТРУКЦИИ КВАДРАТУРА КРУГА, KOНСТРУКЦИИ КВАДРАТУРА КРУГА, Rmuš, V., Constructions of squaring the circle, d Ц Д УДВОЕНИЕ КУБА И ТРИСЕКЦИЯ УГЛА References Courant, R., Robbins, H., 1973. What is Mathematics? An Elementary Approach to Ideas and Methods. Oxford University Press, London. Courant, R., Robbins, H., 1973. What is Mathematics? An Elementary Approach to Ideas and Methods. Oxford University Press, London. y Dolićanin, Ć., 1984. Маthematics. Institute for Textbooks and Educational Materials, Priština. y Dolićanin, Ć., 1984. Маthematics. Institute for Textbooks and Educational Materials, Priština. Ostojić, О., 1980. Theoretical Basis of Initial Mathematics Teaching, Handbook for Math Teachers, Republic Institute for the Improvement of Education, Titograd. Ostojić, О., 1980. Theoretical Basis of Initial Mathematics Teaching, Handbook for Math Teachers, Republic Institute for the Improvement of Education, Titograd. Ц УДВОЕНИЕ КУБА И ТРИСЕКЦИЯ УГЛА Веселин М. Рмуш Средняя профессиональная школа, г. Беране, Черногория Веселин М. Рмуш Средняя профессиональная школа, г. Беране, Черногория ОБЛАСТЬ: математика ВИД СТАТЬИ: оригинальная научная статья ЯЗЫК СТАТЬИ: английский Резюме: Построение трех классических античных задач (квадратура круга, удвоение куба и трисекция угла), с помощью линейки и циркуля до сих пор считается неразрешимым. В данной статье представлены оригинальные методы построения вышеупомянутых задач, что представляет собой инновационный подход в геометрии. Для построения квадратуры круга и удвоения куба была использована теорема Фалеса о пропорциональных отрезках, а трисекция угла основанная на повороте единичного круга в декартовой системе координат и аксиомах измерения угла. Данные построения не претендуют на создание точных чертежей на практике, так как главная наша цель заключается в нахождении теоретического решения, основанного на построении данных фигур с помощью линейки и циркуля, с учетом предположения, что вышеупомянутые инструменты являются совершенно точными. Ключевые слова: построение, квадратура круга, удвоение куба, трисекция угла, система координат, единичный круг, вращение, пропорция. 639 TEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER, 2017, Vol. 65, Issue 3 KOНСТРУКЦИЈЕ КВАДРАТУРЕ КРУГА, УДВАЈАЊА КОЦКЕ И ТРИСЕКЦИЈЕ УГЛА Веселин М. Рмуш Средња стручна школа, Беране, Црна Гора ОБЛАСТ: математика ВРСТА ЧЛАНКА: оригинални научни чланак ЈЕЗИК ЧЛАНКА: енглески ОБЛАСТ: математика ВРСТА ЧЛАНКА: оригинални научни чланак ЈЕЗИК ЧЛАНКА: енглески ЈЕЗИК ЧЛАНКА: енглески VOJNOTEHNIČKI GLASNIK / MILITARY TECHNICAL COURIER Сажетак: Kонструкције три класична грчка проблема (квадратура круга, удавајање коцке и трисекција угла), уз употребу само лењира и шестара, до данас се сматрају нерешивим. Циљ овог чланка јесте да се оригиналним методама објасне конструкције поменутих проблема, што представља новину у геометрији. За конструкцију квадратуре круга и удвајање коцке коришћена је Талесова теорема о пропорционалним дужима, а за трисекцију угла ротација јединичне кружнице у правоуглом координатном систему и аксиоме о мјерењу угла. Констукције се не односе на прецизно цртање фигура у пракси, већ је намјера да се употребом лењира и шестара нађе теоријско рјешење под претпоставком да су поменути инструменти савршено прецизни. VOJNOTEHNIČKI GLASNIK / MILITARY TECH Кључне ријечи: конструкција, квадратура круга, удвајање коцке, трисекција угла, координатни систем, јединична кружница, ротација, пропорција. 640 VOJNOTEHNIČ Paper received on / Дата получения работы / Датум пријема чланка: 08.03.2017. Manuscript corrections submitted on / Дата получения исправленной версии работы / Датум достављања исправки рукописа: 24.03.2017. Paper accepted for publishing on / Дата окончательного согласования работы / Датум коначног прихватања чланка за објављивање: 26.03.2017. © 2017 The Author. Published by Vojnotehnički glasnik / Military Technical Courier (www.vtg.mod.gov.rs, втг.мо.упр.срб). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/rs/). © 2017 Автор. Опубликовано в «Военно-технический вестник / Vojnotehnički glasnik / Military Technical Courier» (www.vtg.mod.gov.rs, втг.мо.упр.срб). Данная статья в открытом доступе и распространяется в соответствии с лицензией «Creative Commons» (http://creativecommons.org/licenses/by/3.0/rs/). © 2017 Аутор. Објавио Војнотехнички гласник / Vojnotehnički glasnik / Military Technical Courier (www.vtg.mod.gov.rs, втг.мо.упр.срб). Ово је чланак отвореног приступа и дистрибуира се у складу са Creative Commons licencom (http://creativecommons.org/licenses/by/3.0/rs/). Paper received on / Дата получения работы / Датум пријема чланка: 08.03.2017. Manuscript corrections submitted on / Дата получения исправленной версии работы / Датум достављања исправки рукописа: 24.03.2017. у р ру Paper accepted for publishing on / Дата окончательного согласования работы / Датум коначног прихватања чланка за објављивање: 26.03.2017. © 2017 The Author. Published by Vojnotehnički glasnik / Military Technical Courier (www.vtg.mod.gov.rs, втг.мо.упр.срб). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/rs/). © 2017 Автор. Опубликовано в «Военно-технический вестник / Vojnotehnički glasnik / Military Technical Courier» (www.vtg.mod.gov.rs, втг.мо.упр.срб). Данная статья в открытом доступе и распространяется в соответствии с лицензией «Creative Commons» (http://creativecommons.org/licenses/by/3.0/rs/). © 2017 Аутор. Објавио Војнотехнички гласник / Vojnotehnički glasnik / Military Technical Courier (www.vtg.mod.gov.rs, втг.мо.упр.срб). Сажетак: Ово је чланак отвореног приступа и дистрибуира се у складу са Creative Commons licencom (http://creativecommons.org/licenses/by/3.0/rs/). 640
https://openalex.org/W2443273310	https://dr.ntu.edu.sg/bitstream/10356/89294/1/Epidemic%20resurgence%20of%20dengue%20fever%20in%20Singapore%20in%202013-2014%20a%20virological%20and%20entomological%20perspective.pdf	English	null	Epidemic resurgence of dengue fever in Singapore in 2013-2014: A virological and entomological perspective	BMC infectious diseases	2,016	cc-by	10,858	Epidemic resurgence of dengue fever in Singapore in 2013‑2014 : a virological and entomological perspective Hapuarachchi, Hapuarachchige Chanditha; Koo, Carmen; Rajarethinam, Jayanthi; Chong, Chee‑Seng; Lin, Cui; Yap, Grace; Liu, Lilac; Lai, Yee‑Ling; Ooi, Peng Lim; Cutter, Jeffery; Ng, Lee‑Ching 2016 2016 Hapuarachchi, H. C., Koo, C., Rajarethinam, J., Chong, C.‑S., Lin, C., Yap, G., . . . Ng, L.‑C. (2016). Epidemic resurgence of dengue fever in Singapore in 2013‑2014: A virological and entomological perspective. BMC Infectious Diseases, 16, 300‑. doi:10.1186/s12879‑016‑1606‑z https://hdl.handle.net/10356/89294 https://doi.org/10.1186/s12879‑016‑1606‑z © 2016 Hapuarachchi et al. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Downloaded on 24 Oct 2024 12:44:55 SGT Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 DOI 10.1186/s12879-016-1606-z Open Access * Correspondence: ng_lee_ching@nea.gov.sg 1Environmental Health Institute, National Environment Agency, 11, Biopolis Way, #06-05-08, Singapore 138667, Singapore 4School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore Full list of author information is available at the end of the article Epidemic resurgence of dengue fever in Singapore in 2013-2014: A virological and entomological perspective Hapuarachchige Chanditha Hapuarachchi1, Carmen Koo1, Jayanthi Rajarethinam1, Chee-Seng Chong1, Cui Lin3, Grace Yap1, Lilac Liu1, Yee-Ling Lai1, Peng Lim Ooi2, Jeffery Cutter2 and Lee-Ching Ng1,4* Abstract Background: Dengue resurged in Singapore during 2013-14, causing an outbreak with unprecedented number of cases in the country. In the present study, we summarise the epidemiological, virological and entomological findings gathered through the dengue surveillance programme and highlight the drivers of the epidemic. We also describe how the surveillance system facilitated the preparedness to moderate epidemic transmission of dengue in the country. Methods: The case surveillance was based on a mandatory notification system that requires all medical practitioners to report clinically-suspected and laboratory-confirmed cases within 24 hours. The circulating Dengue virus (DENV) populations were monitored through an island wide virus surveillance programme aimed at determining the serotypes and genotypes of circulating virus strains. Entomological surveillance included adult Aedes surveillance as well as premise checks for larval breeding. Results: A switch in the dominant serotype from DENV-2 to DENV-1 in March 2013 signalled a potential spike in cases, and the alert was corroborated by an increase in average Aedes house index. The alert triggered preparedness and early response to moderate the impending outbreak. The two-year outbreak led to 22,170 cases in 2013 and 18,338 in 2014, corresponding to an incidence rate of 410.6 and 335.0 per 100,000 population, respectively. DENV-1 was the dominant serotype in 2013 (61.7 %, n = 5,071) and 2014 (79.2 %, n = 5,226), contributed largely by a newly-introduced DENV-1 genotype III strain. The percentage of houses with Ae. aegypti breeding increased significantly (p < 0.001) from 2012 (annual average of 0.07 %) to 2013 (annual average of 0.14 %), followed by a drop in 2014 (annual average of 0. 10 %). Aedes breeding data further showed a wide spread distribution of Ae. aegypti in the country that corresponded with the dengue case distribution pattern in 2013 and 2014. The adult Aedes data from 34 gravitrap sentinel sites revealed that approximately 1/3 of the monitored sites remained at high risk of DENV transmission in 2013. Conclusions: The culmination of the latest epidemic is likely to be due to a number of demographic, social, virological, entomological, immunological, climatic and ecological factors that contribute to DENV transmission. A multi-pronged approach backed by the epidemiological, virological and entomological understanding paved way to moderate the case burden through an integrated vector management approach. Keywords: Dengue, Aedes aegypti, House index, Genotype, Epidemic, Surveillance, Virology, Entomology, Control © 2016 Hapuarachchi et al. © 2016 Hapuarachchi et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Background low Aedes house index (below 1 %) [14–16]. It has been postulated that the resurgence is due to multiple factors that facilitate DENV transmission: an increase in hu- man population density, low herd immunity resulting from long periods of low transmission [17–20], im- proved local transportation that facilitates virus dissem- ination, increased travel-related virus importations and the geo-expansion of Ae. aegypti in parallel to urban developments. It could also be partly due to improved diagnostic and notification rates over the years. These changing epidemiological settings signaled a need for a strategic revision of the integrated Aedes mosquito con- trol programme. Based on programme reviews in the last decade, an enhanced approach for dengue control organised around three key approaches; inter-epidemic surveillance and control, risk based prevention and intervention as well as coordinated intersectoral co- operation, has been introduced. The current integrated surveillance framework is supported by four main pil- lars; 1). Improved operational response through en- hanced case surveillance by using rapid diagnostics; 2). Early warning of outbreaks based on virus surveillance; 3). Understanding the distribution of vectors and their density fluctuations through entomological surveillance; 4). Understanding the relationship between environ- mental parameters and outbreak risk. Currently, all four serotypes of DENV circulate in Singapore [21] where one serotype is typically dominated for a particular period until it is replaced by another serotype. Such “serotype switch” events have preceded previous epi- demics [15]. A switch from DENV-2 to DENV-1 in early 2004 was associated with an outbreak during 2004-2005. In early 2007, DENV-2 gained dominance from DENV-1 and drove the 2007-08 epidemic of over 8,000 cases [15]. Dengue incidence stabilized subse- quently around an annual figure of 5,000 cases until the end of 2012. Singapore experienced yet another unpre- cedented increase in dengue cases in 2013 and 2014, recording 22,170 and 18,338 cases, respectively [22]. g Dengue fever is currently the most prevalent mosquito- borne viral disease, especially in the Americas and Asia, caused by Dengue virus (DENV) which is transmitted to humans primarily by Aedes aegypti and Ae. albopictus mosquitoes [1, 2]. DENV is a positive sense single stranded enveloped RNA virus of the genus Flavivirus [3]. Its genome is approximately 11.8 kb in size, and en- codes for a single polypeptide flanked by highly struc- tured 5’ and 3’ untranslated regions. Background DENV exists in four phylogenetically and antigenically distinct serotypes (DENV1-4) and exposure to a particular serotype elicits type-specific life-long immunity [4]. The virus is further subdivided into genotypes based on envelope (E) gene and envelope/non-structural protein 1 (E/NS1) gene junction regions [5, 6]. These genotypes show a charac- teristic geographical distribution [7], implying a com- petitive advantage for individual genotypes that differ in their ability to spread and cause disease in a particular region. DENV infection causes a clinical syndrome that is pri- marily benign, but can seldom be fatal due to increased vascular permeability, leading to systemic shock and multi-organ failure [8]. It is estimated to affect approxi- mately 50–200 million individuals each year. More than 125 countries distributed across Americas, Southeast Asia and the Western Pacific regions are endemic to DENV, where almost 50 % of the world population lives at risk [1, 9]. The magnitude of dengue incidence has increased by 30-fold in the past five decades, mainly contributed by rapid urbanization, overcrowding, in- creased global travel and expansion of vector popula- tions. Except for the tetravalent vaccine (Dengvaxia®) approved recently in a few countries, there is neither an effective antiviral agent nor a licensed vaccine against DENV in many endemic settings where vector control remains as the sole strategy for epidemic control and prevention of dengue. Dengue fever replaced malaria as the most important mosquito-borne disease in Singapore in the mid-1960s. As a result, a nationwide integrated Aedes mosquito control programme was introduced at that time to sup- press the vector population and thereby DENV trans- mission in the country. The programme primarily focused on vector source reduction through surveillance, enforcement, community engagement, careful urban plan- ning and operational research. Consequently, the Aedes house index was successfully brought down from about 50 % in 1960s to less than 5 % by the late 1970s. This re- sulted in a corresponding decline in dengue incidence [10–13]. However, dengue fever started to resurge in the country in 1980s, in a typical 5–6 year epidemic cycle. The magnitude of epidemics escalated within each cycle during the last decade, despite maintaining a consistently In the present study, we present an account of the re- surgence of dengue fever in Singapore during 2013-14, the worst known dengue fever encounter in the country. Background We describe how the surveillance system led to an early warning of the outbreak, and how it facilitated prepared- ness within the country. We also present the virological and entomological findings gathered through the surveil- lance programme. Abstract Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Page 2 of 13 Page 2 of 13 Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Virological findings DENV serotypes were confirmed in 8,216 samples in 2013 and 6,598 samples in 2014, which corresponded to 37.1 and 36.0 % coverage of all dengue cases reported in respective years. These viruses also included 45 and 12 specimens of Ae. aegypti and Ae. albopictus collected in 2013 and 2014, respectively. In 2013, the overall serotype composition was dominated by DENV-1 (61.7 %, n = 5,071), followed by DENV-2 (24.8 %, n = 2,034), DENV-3 (11.5 %, n = 941) and DENV-4 (2.0 %, n = 170). The trend continued in 2014; DENV-1 (79.2 %, n = 5,226), DENV-2 (18.1 %, n = 1,194), DENV-3 (2.5 %, n = 165) and DENV-4 (0.2 %, n = 13). As shown in Fig. 2, the serotype composition was more diverse in 2013, espe- cially during the first half of the year, than in 2014, sug- gesting that a mixed viral population circulated during the early epidemic establishment period. Early warning of an impending outbreak Early warning of an impending outbreak In March 2013, the proportion of DENV-1 rose from 20–30 % to more than 50 %, replacing DENV-2 as the predominant serotype. The switch in predominant sero- type signaled a potential spike in cases in the approaching Page 3 of 13 Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 per 100,000 population, n = 14,210). The diagnostic rate at Environmental Health Institute (EHI) diagnostics surged from 27 % (n = 86) of the total number of samples tested in January 2013 to 41 % (n = 139) in March, 12 weeks ahead of the first peak (Fig. 1). The rate dropped gradually and maintained between 20–30 % from July 2013 to reach the lowest (15 %) in February 2014. An upward momen- tum occurred subsequently to reach the peak diagnostic rate (46 %) in April followed by another peak (49 %) in July 2014. The second peak coincided with the highest number of weekly cases in 2014. The relationship between the EHI diagnostic rate and total number of reported cases from 2012 to 2014 showed a positive correlation (Spearman’s correlation =0.73, P < 0.05). “dengue season” during May-August. The alert was fur- ther corroborated by the earlier than usual escalation of weekly cases at the beginning of 2013, and an increase in average Aedes house index from 0.17 (range: 0.12– 0.27) in 2012 to 0.19 (range: 0.11–0.25) in 2013. In epi- demiological week 9 (March 2013), a newly-developed statistical model [23] provided a forecast of about 800 cases per week at the peak of the epidemic in June. The temporal model utilizes data obtained through case and vector surveillance as well as weather parameters [23]. Based on the multiple warning signs, the National En- vironment Agency (NEA) enhanced its vector control in the community and initiated preparation to respond to the outbreak. All stakeholders were also warned through the Inter-Agency Dengue Task Force [14] of the impending epidemic risk. The community was galvanized through a campaign in April 2013. Epidemiological findings The epidemic took off in January 2013 (week 1–5) and weekly cases showed a rapid surge in April (week 14– 18) to reach the first peak in June 2013 (week 23–26) (Fig. 1). There were 134 cases in the first week of January 2013, which escalated to reach 842 cases in 25th week. The trend declined and stabilized between September 2013 (week 36–39) and January 2014 (week 1–5). The second peak was observed in July 2014, reaching 891 cases in 27th week, the highest weekly number of dengue cases ever recorded in Singapore. The epidemic started to subside in November 2014 (week 45–48), recording the lowest number of weekly cases in 47th week (n = 149). The total number of laboratory-confirmed cases re- ported in 2013 (n = 22,170) and 2014 (n = 18,338) corre- sponded to an incidence rate of 410.6 and 335.0 per 100,000 population, respectively. The incidence rate in 2013 was the highest reported so far in Singapore and exceeded the previous highest recorded in 2005 (333.1 A total of 1,270 complete E gene sequences was gener- ated in 2013 (Table 1). The corresponding number for 2014 was 1,531. The sequence collection included 38 mosquito-derived virus sequences in 2013 and another six sequences detected among mosquitoes trapped in 2014. The overall genotype coverage was 4.5 % and 8.4 % Fig. 1 Epidemiological curve for weekly dengue cases in 2013 and 2014. All cases are laboratory confirmed Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 4 of 13 Fig. 2 Weekly distribution of DENV serotypes in Singapore: 2013-2014. The serotype proportions were calculated based on 14,814 patient sera successfully typed in 2013 (n = 8,216) and 2014 (n = 6,598) Phylogenetic analysis revealed that DENV-1 genotype III was the dominant genotype among all four serotypes sequenced in 2013 (n = 666, 52.4 %) and 2014 (n = 1,010, 65.9 %). This “epidemic strain” was first detected in a sporadic case in November 2012, at a time when DENV-2 cosmopolitan genotype was leading the trans- mission. In January 2013, the epidemic strain of DENV- 1 genotype III was detected in a major dengue cluster in the Southeastern part of Singapore, indicating the es- tablishment of indigenous transmission of this strain. It took the lead from DENV-2 cosmopolitan genotype in March 2013 and continued to expand throughout the country. The epidemic strains showed a rapid accumu- lation rate (Fig. Epidemiological findings 3) and formed a lineage distinct from those reported earlier locally [21], with 99 % bootstrap support (Fig. 4a). This genetic distinction of the epi- demic lineage suggested that its appearance in the country was due to an introduction. Its closest relatives were DENV-1 genotype III strains reported from India, Bangladesh and China during 2008-2011 periods (Fig. 4a), indicating the circulation of genetically- related virus strains in Asia preceding the epidemic and thereby the likelihood of its introduction from a re- gional country. of all reported cases in 2013 and 2014, respectively. The majority (60.6 %, n = 770) of virus strains sequenced in 2013 were DENV-1, followed by DENV-2 (26.1 %, n = 332) (Table 1). Similarly, DENV-1 and DENV-2 consti- tuted 75 % (n = 1150) and 21.8 % (n = 334) of genotyped samples in 2014, respectively. The proportion of DENV-3 sequences generated in 2013 (n = 133, 10.5 %) was 3.6 times higher than that in 2014 (n = 43, 2.9 %). of all reported cases in 2013 and 2014, respectively. The majority (60.6 %, n = 770) of virus strains sequenced in 2013 were DENV-1, followed by DENV-2 (26.1 %, n = 332) (Table 1). Similarly, DENV-1 and DENV-2 consti- tuted 75 % (n = 1150) and 21.8 % (n = 334) of genotyped samples in 2014, respectively. The proportion of DENV-3 sequences generated in 2013 (n = 133, 10.5 %) was 3.6 times higher than that in 2014 (n = 43, 2.9 %). Epidemiological findings Table 1 Composition of DENV genotypes circulated during the epidemic in 2013 and 2014 Serotype Genotypes/clades 2013 2014 DENV-1 Genotype I 97 (7.6 %) 136 (8.9 %) Genotype II 04 (0.3 %) 0 (0 %) Genotype IIIa 669 (52.7 %) 1,014 (66.2 %) DENV-2 Asian I 0 (0 %) 01 (0.1 %) Cosmopolitan 02 (0.2 %) 12 (0.8 %) Cosmopolitan Clade I 28 (2.2 %) 11 (0.7 %) Cosmopolitan Clade 1a 151 (11.9 %) 13 (0.8 %) Cosmopolitan Clade 1b 33 (2.6 %) 269 (17.6 %) Cosmopolitan Clade III 74 (5.8 %) 0 (0 %) Cosmopolitan Clade V 12 (0.9 %) 0 (0 %) Cosmopolitan Indian 32 (2.5 %) 28 (1.8 %) DENV-3 Genotype I 02 (0.2 %) 26 (1.7 %) Genotype II 03 (0.2 %) 01 (0.1 %) Genotype III 128 (10.1 %) 16 (1 %) DENV-4 Genotype I 01 (0.1 %) 01 (0.1 %) Genotype II 34 (2.7 %) 03 (0.2 %) Total 1,270 1,531 aincludes epidemic strains in 2013 (n = 666) and 2014 (n = 1,010) The proportion of each genotype among all genotyped cases for the respective year is shown within brackets. The phylogeny of genotypes and clades of DENV serotypes is given in Fig. 4a and b Table 1 Composition of DENV genotypes circulated during the epidemic in 2013 and 2014 In addition, the early establishment phase of the epi- demic saw an uprising of two new strains: DENV-2 cosmopolitan genotype clade I (named as clade Ia) and DENV-3 genotype III (Fig. 4a and b). Both strains formed genetically distinct clusters in respective phylog- enies (Fig. 4a and b) and were likely to be newly intro- duced. Clade Ia shared ancestry with DENV-2 strains reported earlier from Indonesia during 2010-2012 pe- riods, whereas DENV-3 genotype III strains were related to those reported from India during 2009-2010 periods. The first evidence of DENV-2 cosmopolitan clade Ia in Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 5 of 13 Fig. 3 Cumulative dynamics of cases due to the most common DENV strains detected during the epidemic and their case contributory pattern in 2013 and 2014. The weekly national serotype data and weekly EHI genotype data from 2007 to 2014 were used to estimate the historical genotype proportions. Epidemiological findings In order to obtain smooth estimates of genotype proportions over time, a Bayesian approach was used assuming multinomial distribution of serotypes and genotypes, and an auto-correlated prior distribution for logarithm transformed proportions. Bayesian estimates of the weekly genotype proportions were sampled from the posterior distribution, which were used together with weekly national case count to calculate the weekly cases attributed to each genotype as well the cumulative case count. The analysis was done using R software version 3.1.1 [33]. Only the genotypes dominant during the epidemic years have been plotted in the graph Fig. 3 Cumulative dynamics of cases due to the most common DENV strains detected during the epidemic and their case contributory pattern in 2013 and 2014. The weekly national serotype data and weekly EHI genotype data from 2007 to 2014 were used to estimate the historical genotype proportions. In order to obtain smooth estimates of genotype proportions over time, a Bayesian approach was used assuming multinomial distribution of serotypes and genotypes, and an auto-correlated prior distribution for logarithm transformed proportions. Bayesian estimates of the weekly genotype proportions were sampled from the posterior distribution, which were used together with weekly national case count to calculate the weekly cases attributed to each genotype as well the cumulative case count. The analysis was done using R software version 3.1.1 [33]. Only the genotypes dominant during the epidemic years have been plotted in the graph Fig. 3 Cumulative dynamics of cases due to the most common DENV strains detected during the epidemic and their case contributory pattern in 2013 and 2014. The weekly national serotype data and weekly EHI genotype data from 2007 to 2014 were used to estimate the historical genotype proportions. In order to obtain smooth estimates of genotype proportions over time, a Bayesian approach was used assuming multinomial distribution of serotypes and genotypes, and an auto-correlated prior distribution for logarithm transformed proportions. Bayesian estimates of the weekly genotype proportions were sampled from the posterior distribution, which were used together with weekly national case count to calculate the weekly cases attributed to each genotype as well the cumulative case count. The analysis was done using R software version 3.1.1 [33]. Epidemiological findings Only the genotypes dominant during the epidemic years have been plotted in the graph When the mosquito population size is small, the domin- ant strains are likely to saturate the infectious mosquito pool relatively fast, leaving little opportunity for “minor” strains to settle down. On the other hand, when the mosquito population expands, especially during peak pe- riods of transmission, the chances for more virus strains to establish transmission also increase. As the control measures are targeted to suppress the Aedes population, their numbers fluctuate and virus populations are ex- pected to go through bottlenecks during periods of low Aedes density. Importantly, the vector density fluctua- tions are not universal across the country as areas with higher number of disease clusters, especially major clus- ters, require more effort and time to bring down the mosquito density than in areas with localized transmis- sion. Consequently, the effects of vector control efforts are more likely to be profound on the “minor” strains that show localized transmission than dominant strains. These findings suggested that favourable conditions for virus transmission prevailed in 2013, but sustained disease control measures may have facilitated the elimination of many DENV strains in 2014. It is noteworthy that immune pressure and evolutionary forces may also have affected the survival of virus strains and may have contributed to fluctuations in virus population independent of vector density. Singapore was in June 2012 (23rd week). Prior to the set- ting out of the epidemic, clade Ia strains maintained a low profile, but showed a rapid surge in case accumula- tion from January 2013 and started to subside at the end of the year. There were only 30 clade Ia strains in 2012, whereas 151 sequences were generated in 2013. On the other hand, DENV-3 genotype III strains emerged in January 2013, at the beginning of the epidemic. Its circu- lation lasted until 42nd week of 2013. These observations suggested that prevailing favourable conditions for virus transmission during the early phase of the epidemic allowed multiple newly-introduced virus strains to pro- liferate and spread. When the mosquito population size is small, the domin- ant strains are likely to saturate the infectious mosquito pool relatively fast, leaving little opportunity for “minor” strains to settle down. On the other hand, when the mosquito population expands, especially during peak pe- riods of transmission, the chances for more virus strains to establish transmission also increase. Epidemiological findings BMC Infectious Diseases (2016) 16:300 Page 7 of 13 (See figure on previous page.) Fig. 4 E-gene based phylogeny of DENV illustrating different types of virus strains circulated during 2013-2014. Phylogenetic analysis was performed in MEGA6 program [32] using the maximum-likelihood method based on the general time reversible model with gamma distribution and invariant sites. The robustness of the original tree was tested with 1000 bootstrap replications. a. DENV-1 and DENV-2 strains circulated in Singapore in both years have been highlighted in red and blue respectively. b. DENV-3 and DENV-4 strains circulated in Singapore in both years have been highlighted in green and purple respectively. The major groups summarized in Table 1 are shown in triangular cartoons. Each taxon is named with sample ID/NCBI accession number, reported year, country and genotype information. Numbers on branches are bootstrap support values. GI, GII, and GIII = genotypes I, II and III; cosmo = cosmopolitan genotype percentage of Gravitraps that caught at least one adult Ae. aeygpti in a particular location, was 9.7 % (±0.3 % SE) in 2013. The index declined to 6.2 % (±0.1 % SE) in 2014. Furthermore, the index indicated that 29.4 % of became gradually extinct in early 2014. Interestingly, clade Ib was the dominant strain circulated in Johor and Melaka, the southern states of Malaysia, in 2013 [22]. Johor is the closest Malaysian state to Singapore. The detection of identical clade Ib strains in Singapore indi- cated the possibility of virus exchange between the two countries [22]. Clade Ib’s dominance in Malaysian states also suggested its outbreak potential. In fact, clade Ib has been the dominant DENV-2 strain in Singapore since May 2014. Fig. 5 Temporal variation of House Index in different types of residential premises in 2013 and 2014. a. Housing Development Board (HDB) apartments, b. Private apartments/condominiums and c. Landed properties. House Index for each residential premise type is defined as the number of houses detected with Aedes mosquito breeding per 100 houses inspected. The analysis is based on the outcome of routine inspections carried out by approximately 800 ground officers from the Environmental Public Health Operations, NEA. Aedes immatures were morphologically identified to the species level at EHI. E-week = epidemiological week Epidemiological findings As the control measures are targeted to suppress the Aedes population, their numbers fluctuate and virus populations are ex- pected to go through bottlenecks during periods of low Aedes density. Importantly, the vector density fluctua- tions are not universal across the country as areas with higher number of disease clusters, especially major clus- ters, require more effort and time to bring down the mosquito density than in areas with localized transmis- sion. Consequently, the effects of vector control efforts are more likely to be profound on the “minor” strains that show localized transmission than dominant strains. These findings suggested that favourable conditions for virus transmission prevailed in 2013, but sustained disease control measures may have facilitated the elimination of many DENV strains in 2014. It is noteworthy that immune pressure and evolutionary forces may also have affected the survival of virus strains and may have contributed to fluctuations in virus population independent of vector density. As summarised in Table 1, besides the major strains described above, multiple virus strains belonging to all four DENV serotypes circulated during the epidemic period. The phylogeny representative of all these strains has been illustrated under respective serotypes (Fig. 4a and b). The temporal fluctuation of proportions of each strain among all genotyped cases demonstrated a highly heterogeneous DENV population at any given time, though a few strains were dominant. Those dominant strains were widely distributed across the country, and contributed to 70–80 % of cases during 2013-14. On the other hand, the “minor” strains were less well established and their transmission was generally localized. One of the notable observations was that a relatively higher number of genetically distinct strains, including “minor” strains, circulated in substantial proportions in 2013 than in 2014 (Table 1). Besides the availability of a sus- ceptible human pool, the sustainability of vector-borne virus transmission is affected by the vector density. Nevertheless, disappearance of major strains provided the opportunity for less-common strains to dominate. One such example is the replacement of DENV-2 cosmopolitan clade Ia by clade Ib strains in April 2014. Clade Ib has been circulating since February 2013, but was unable to dominate over clade Ia until the latter Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 6 of 13 Fig. 4 (See legend on next page.) Fig 4 (See legend on next page ) Fig. 4 (See legend on next page.) Hapuarachchi et al. Entomological findings The house index for Aedes mosquitoes in three main types of residential premises, namely government- sponsored Housing Development Board (HDB) flats, private apartments and condominiums as well as landed properties, increased gradually from the end of September 2012 (Fig. 5). The increase was most obvi- ous in HDB and landed properties categories. Among the Aedes immatures collected, the proportion of Ae. aegypti mosquito breeding, the primary vector of DENV in Singapore, in residential premises was 59.5 and 54.6 % in 2013 and 2014, respectively. The percentage of houses with Ae. aegypti mosquito breeding (named as Ae. aegypti house index) was signifi- cantly higher (p < 0.001) in 2013 (annual average of 0.14) than in 2012 (annual average of 0.07). The index reached its peak at the beginning of the inclination of cases in Feb- ruary 2013 (Fig. 6). The trend continued in 2014 though at a lower magnitude (annual average of 0.10) than in 2013 (Fig. 6). Moreover, the distribution of Ae. aegypti ex- panded dramatically during the decade from 2003 to 2013 in parallel to infrastructure development in the country (Fig. 7a). The new areas generally represented newly- developed urban residential sites that are mainly distrib- uted in south-central and western parts of Singapore (Fig. 7a). As expected, the dengue case distribution pattern in 2013 and 2014 was in line with the geographical spread of Ae. aegypti in the country (Fig. 7b). Fig. 5 Temporal variation of House Index in different types of residential premises in 2013 and 2014. a. Housing Development Board (HDB) apartments, b. Private apartments/condominiums and c. Landed properties. House Index for each residential premise type is defined as the number of houses detected with Aedes mosquito breeding per 100 houses inspected. The analysis is based on the outcome of routine inspections carried out by approximately 800 ground officers from the Environmental Public Health Operations, NEA. Aedes immatures were morphologically identified to the species level at EHI. E-week = epidemiological week In the third quarter of 2013, an adult Aedes sentinel surveillance system was established with about 3,000 Gravitraps in 34 locations throughout the island, in addition to Gravitraps that were deployed in ad hoc cluster management [24]. The surveillance showed that the average Gravitrap aegypti index, which expresses the Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 8 of 13 Fig. 6 Weekly trends of cases and Ae. aegypti house index during 2012-2014. Entomological findings The Ae. aegypti house index is expressed as the percentage of houses with Ae. aegypti mosquito breeding. The comparison shows that the fluctuation of Ae. aegypti house index generally preceded that of cases Fig. 6 Weekly trends of cases and Ae. aegypti house index during 2012-2014. The Ae. aegypti house index is expressed as the percentage of houses with Ae. aegypti mosquito breeding. The comparison shows that the fluctuation of Ae. aegypti house index generally preceded that of cases Fig. 7 Spatial distribution of Ae. aegypti and dengue cases in 2013 and 2014. a. Spatial expansion of Ae. aegypti breeding in 2013 as compared to 2003. The breeding sites were identified based on island wide Aedes breeding data collected on a daily basis as part of vector control operations. The map was generated using the ArcGIS 10.1 ArcMap software (ESRI, CA, USA). b. Spatial distribution of dengue cases in 2013 and 2014. The spatial distribution of dengue cases in Singapore was generated using the kernel density tool in the spatial analyst toolbox of ArcGIS 10.1 ArcMap software (ESRI, CA, USA) based on a search radius of 400 m. Case density values were classified into four classes of < 25th (2 cases/km2), 25th-50th (16-25 cases/km2), 51st-75th (56-61 cases/km2) and more than 75th (217 cases/km2) quantiles, using the quantile classification method and were displayed in tones of pink as shown in the legend Fig. 7 Spatial distribution of Ae. aegypti and dengue cases in 2013 and 2014. a. Spatial expansion of Ae. aegypti breeding in 2013 as compared to 2003. The breeding sites were identified based on island wide Aedes breeding data collected on a daily basis as part of vector control operations. The map was generated using the ArcGIS 10.1 ArcMap software (ESRI, CA, USA). b. Spatial distribution of dengue cases in 2013 and 2014. The spatial distribution of dengue cases in Singapore was generated using the kernel density tool in the spatial analyst toolbox of ArcGIS 10.1 ArcMap software (ESRI, CA, USA) based on a search radius of 400 m. Case density values were classified into four classes of < 25th (2 cases/km2), 25th-50th (16-25 cases/km2), 51st-75th (56-61 cases/km2) and more than 75th (217 cases/km2) quantiles, using the quantile classification method and were displayed in tones of pink as shown in the legend Fig. 7 Spatial distribution of Ae. aegypti and dengue cases in 2013 and 2014. a. Drivers of the epidemic Case burden of the dengue epidemic in 2013 and 2014 was unprecedented, recording 1.7 times more cases than those reported during the last DENV-1 epidemic during 2004-05. The culmination of the latest epidemic is likely to be due to a number of demographic, social, viro- logical, entomological, immunological, climatic and eco- logical factors that contribute to DENV transmission. The most obvious driving force of the epidemic was emergence of a new strain of DENV-1 genotype III. In average, this epidemic strain contributed to 68.5 % of cases among all genotyped cases (n = 2,803) during the 2013-14 periods. Serotype data that provided even a higher coverage of reported cases (an average of 36.6 % in both years) confirmed the dominance of DENV-1 during the same period (in average, 69.5 % of serotyped cases). The peak of the epidemic during May-June 2013 was preceded by a change in the dominant serotype from DENV-2 to DENV-1, approximately four months after the first evidence of epidemic strains in November 2012. A similar event has occurred in Singapore during a dengue epidemic during 2007-08, which anchored the phenomenon of “serotype switch” as a warning sign of epidemic risk in the country [15]. The entomological surveillance data indicated an in- crease in both larval and adult Ae. aegypti population during the epidemic period. The average percentage of houses with Ae. aegypti mosquito breeding doubled in 2013 as compared to 2012. Moreover, the data sug- gested Ae. aegypti breeding in a high proportion of resi- dential premises in the latest epidemic years. Aedes aegypti adult surveillance with Gravitraps also demon- strated high vector density in many locations prior to and during the epidemic. These observations supported the notion that prevailing vector and host factors were favourable for virus transmission at the time of intro- duction of a new virus lineage and explained its rapid establishment and spread across the island. The ground conditions were conducive to sustain virus transmission on a longer term causing an unprecedented epidemic. Besides highly-potential virus strains, the epidemic transmission of DENV requires the connectivity be- tween two important contributors: a susceptible human population and an optimal vector density. The latter is partially affected by climatic and ecological factors. Once exposed to a particular serotype of DENV, an in- dividual elicits type-specific life-long immunity, but re- mains immunologically susceptible to a heterotypic DENV infection [4]. Entomological findings Spatial expansion of Ae. aegypti breeding in 2013 as compared to 2003. The breeding sites were identified based on island wide Aedes breeding data collected on a daily basis as part of vector control operations. The map was generated using the ArcGIS 10.1 ArcMap software (ESRI, CA, USA). b. Spatial distribution of dengue cases in 2013 and 2014. The spatial distribution of dengue cases in Singapore was generated using the kernel density tool in the spatial analyst toolbox of ArcGIS 10.1 ArcMap software (ESRI, CA, USA) based on a search radius of 400 m. Case density values were classified into four classes of < 25th (2 cases/km2), 25th-50th (16-25 cases/km2), 51st-75th (56-61 cases/km2) and more than 75th (217 cases/km2) quantiles, using the quantile classification method and were displayed in tones of pink as shown in the legend Page 9 of 13 Page 9 of 13 Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 dengue seropositivity from 1982 to 2009 across all age groups [20]. The age-standardized seroprevalence in adults aged 18–79 years dropped from 63.1 % in 2004 to 54.4 % in 2010 [17]. Importantly, the sharpest reduction in seropositivity was seen in less than 20 years old age group, indicating the low level of exposure to dengue among young generations. The overall seroprevalence among individuals aged 16–25 years was 16.1 % during 2009-10 periods [20]. Moreover, the rapid growth of non-resident population (from 19 % in 2005 to 29 % in 2014) in the country due to its reliance on foreign labor is also likely to have contributed to the expansion of sus- ceptible proportion. However, recent data demonstrating a reducing trend of seroprevalence among Singapore resident population [20] and a low level (0.01; 95 % CI: 0.0085–0.012) of the force of infection since 1993 (un- published data, EHI) suggests that the gradual increase in magnitude of dengue case burden could partially be contributed by the improved rate of case detection in re- cent years. locations (10 of 34) were at high risk of transmission (index ≥12.0 % based on unpublished data, EHI) in 2013 and declined to 8.8 % (3 of 34) in 2014. Drivers of the epidemic Therefore, the long term absence or low prevalence of a particular serotype in an area increases the proportion of population susceptible to infections by the respective serotype. This may explain the dominant emergence of DENV-1 after eight years in Singapore. The last DENV-1 epidemic in the country was in 2005 [14]. The lower prevalence of antibodies reactive to DENV-1 than that of DENV-2 across mul- tiple age groups in a cross-sectional study among healthy individuals aged 16–60 years during 2009-10 also testified that a higher proportion of individuals, es- pecially those within the age group of 16–30 years, re- mains susceptible to DENV-1 than DENV-2 [20]. Epidemic response In Singapore, the control of vector-borne diseases, in- cluding dengue, is carried out by NEA, and the clinical management and surveillance is overseen by the Minis- try of Health (MOH) [25]. The close coordination be- tween the two sectors is therefore imperative to tackle vector-borne diseases. While surveillance of vector- borne diseases rides on the national disease surveillance programme at MOH, daily communication of surveil- lance data enables prompt vector control response by NEA. Therefore, several activities were jointly initiated by the MOH and NEA to manage the epidemic crisis. 1). Enhanced case surveillance measures to increase the diagnostic coverage and to enable prompt vector con- trol response. Due to the unprecedented increase in cases during the early phase of the epidemic, the medical community was appraised through MOH to have a high Furthermore, the seroprevalence data suggests that strict disease control measures since early 1970’s have widened the susceptible age range by reducing the risk of virus exposure. There has been a gradual reduction in Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 10 of 13 Page 10 of 13 Page 10 of 13 clinical suspicion of dengue among febrile cases. More- over, an existing network of general practitioners was encouraged to utilize a subsidized laboratory testing ser- vice provided by EHI, one of the public health laborator- ies under NEA, Singapore. seriousness of dengue situation in each vicinity. More than 400 individuals were trained as “dengue volunteers” in order to support the community outreach activities that continued throughout the epidemic through semi- nars, talks, roadshows and media. g p 2). Projection of case numbers to facilitate stockpiling of diagnostic reagents and to accommodate increased de- mand on the healthcare system. Based on the statistical model projection on weekly case numbers [23], arrange- ments were made to extend EHI diagnostic services dur- ing weekends and to stockpile diagnostic reagents several months ahead of the peak of the epidemic. The close communication between MOH and NEA on the weekly case projection also allowed hospitals to plan for increased demand on the healthcare system resulting from extra consultations and admissions. 5). Enhanced source reduction for mosquito breeding through an accelerated premise inspection programme. The entomological surveillance and enforcement activ- ities were also carried out in parallel. House-to-house checks were conducted by about 800 ground officers who targeted to check every home and surrounding areas for mosquito breeding. Epidemic response Apart from homes, daily checks also cover common ground areas, public areas and congregation areas for potential mosquito breeding spots. The inspection frequency was accelerated to cover all premises within the boundary of disease clusters dur- ing the epidemic period. A total of 8,781,935 checks was conducted in residential houses during 2013-14. Chem- ical larviciding activities were carried out in housing es- tates, industrial premises and public places on a regular basis and whenever Aedes breeding was detected through inspections. The public was encouraged to use pellets containing Bacillus thuringiensis in places such as roof gutters. Fogging activities were conducted in major clusters during the peak periods of transmission to reduce adult Aedes population density especially when the intensity of transmission sustained despite other source reduction efforts. Data on Aedes immatures col- lected through field inspections was used to gauge Aedes population density in a particular area. 3). Expansion of virus surveillance activities to facili- tate the resource allocation for targeted vector control. In order to achieve a better serotype and genotype coverage during the epidemic, the existing virus surveillance programme at EHI [15, 21] was extended to include samples from polyclinics, hospitals and private labora- tories through a joint initiative between NEA and MOH. During non-epidemic periods, the programme generally screens a relatively small proportion of reported cases received through an island wide general practitioner net- work as an assessment of impending epidemic risk [15, 21]. Consequently, serotype analysis was completed in more than 1/3 of total reported cases and the genotype coverage was doubled from 2013 to 2014. Assisted by geographical information system (GIS)-based plotting of patients’ residential locations, the serotype and genotype data provided a detailed understanding of the spatial and temporal distribution of virus strains within the country. Genetic fingerprinting of virus strains allowed monitor- ing of their spread between locations and facilitated tar- geted vector control in newly-introduced sites. The same approach was used to monitor the successful elimination of virus transmission in disease clusters. 6). Launch of gravitrap surveillance to monitor the fluctuations of adult Aedes population. The existing vec- tor population assessment based on Aedes breeding data was complemented by adult Aedes monitoring with Gravitraps in disease clusters as well as in 34 sentinel locations. 7). Guiding resource allocation for targeted vector con- trol based on a spatial risk map. Epidemic response An island wide risk map developed based on data pertaining to Aedes breed- ing, past dengue exposure, population density, circulating virus strains etc. was used to guide resource allocation for targeted vector control during the epidemic. 4). Early launch of the dengue campaign to promote community awareness of dengue situation and prevention of transmission. As the epidemic was looming, NEA brought forward the launch of “Do the Mozzie Wipeout” campaign to April in 2013. The campaign, which is usu- ally launched at the beginning of “dengue season” in the middle of each year, is part of the community outreach programme aimed at promoting awareness of dengue situation and inspiring action to prevent dengue. A colour-coded alert system was also launched through the Dengue Community Alert System to keep residents informed about active disease clusters in respective lo- calities. The banners coded green (no dengue clusters), yellow (dengue cluster with < 10 cases) and red (dengue cluster with ≥10 cases) were displayed to indicate the 8). Integrated vector management activities aligned with the whole-of-government’s effort. Integrated vector management activities in Singapore are closely aligned with the whole-of-government’s effort in establishing people, public and private (3P) partnership to develop innovative and sustainable initiatives to promote envir- onmental ownership in the community. NEA leads an Inter-Agency Dengue Task Force [14] comprising 27 stakeholders from the 3P sectors to coordinate nation- wide dengue control efforts. The Inter-Agency Dengue Task Force assisted the epidemic control by ensuring the planning and implementation of operational activities of Page 11 of 13 Page 11 of 13 Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 each agency to align with source reduction and vector control efforts carried out by NEA. epidemiologically linked by place (residential or work- place/school) within a radius of 150 meters and with their onset of illnesses within a 14 day period. Informa- tion was uploaded onto the Geographical Information Systems (GIS) server in order to identify areas of active dengue transmission. All those activities were aimed at gaining a clear pic- ture of case burden and Aedes population over time in order to monitor the intensity and the potential sustain- ability of disease transmission. In addition to NEA’s source reduction efforts on the ground, the control programme relied on the community participation to minimize human-vector contact and Aedes breeding at residential premises. Virological data collection g The circulating DENV populations were monitored through a virus surveillance programme jointly con- ducted by the MOH and NEA. Briefly, blood samples from suspected dengue patients who sought treatment at general practitioners, public/private hospitals and polyclinics were tested for the evidence of DENV by using either NS1 antigen or polymerase chain reaction (PCR) assays. The serotype and genotype analyses were performed on a weekly basis to provide a timely update on the composition and distribution of DENV popula- tions to facilitate the prioritizing of resource allocation for dengue control operations. The screening also in- cluded individual field-caught mosquitoes that were positive for DENV NS1 antigen. Viral RNA from mos- quitoes was extracted as described previously [28]. DENV-positive sera and mosquito specimens were fur- ther analysed to determine the serotype of DENV by using a real time reverse-transcription PCR (RT-PCR) assay as described elsewhere [26]. At EHI, one of the public health laboratories in Singapore, all DENV- positive sera that failed serotype screening by the real time RT-PCR assay were subjected to a modified semi- nested conventional PCR assay [29, 30]. Conclusions The epidemic resurgence of dengue fever in Singapore in 2013 was multi-factorial. The emergence of a new strain of DENV-1 genotype III, expansion of the susceptible human population, favourable conditions for mosquito breeding and a widely-distributed vector population collectively contributed to sustain the virus transmission during the epidemic. A multi-pronged approach backed by the epi- demiological, virological and entomological understanding facilitated the resource planning and community aware- ness that paved way to moderate dengue transmission through an integrated vector management approach. For the genotype surveillance, E gene of DENV was PCR amplified and sequenced using serotype-specific primers as described previously [29]. Nucleotide se- quences were assembled using the Lasergene package version 8.0 (DNASTAR Inc., Madison, WI, USA). Con- tiguous sequences were aligned using BioEdit v7.0.5 software [31]. Phylogenetic analysis of E gene sequences was performed in MEGA6 program [32] using the maximum-likelihood method based on the general time reversible model with gamma distribution and invariant sites. The robustness of the original tree was tested with 1000 bootstrap replications. Case data collection It is mandatory for medical practitioners and clinical la- boratories to notify all clinically-suspected and laboratory- confirmed dengue cases and deaths to MOH within 24 hours of detection. The notification information in- cluded demographic data, travel history, clinical data, dates of onset of illness and diagnosis. Notified cases were investigated to determine whether the infections were au- tochthonous or imported. Patients with no travel history to a dengue endemic area outside of Singapore within the seven days prior to onset of illness were defined as autoch- thonous cases. Any epidemiological link among cases was established whenever possible. The laboratory confirm- ation of clinically-suspected cases is achieved either through NS1 antigen detection or viral RNA detection by PCR [26, 27]. Only the laboratory-confirmed cases are of- ficially reported. Epidemic response Even though it was difficult to dir- ectly measure the impact of those activities on the dis- ease transmission, a comparison between the weekly trends of Aedes house index and case data showed that cases generally fluctuated in parallel to mosquito popu- lation variations (Fig. 6). The intense control efforts could bring down the case burden steeply from the peaks through the suppression of vector population (Fig. 6). Authors’ information Not applicable. Authors’ information Not applicable. Funding The study was funded by the NEA, Singapore. The funding sources of this study had no role in the study design, data collection, data analysis, data interpretation, writing of the report, or in the decision to submit the paper for publication. Abbreviations 3P, people, public and private; DENV, Dengue virus; DHF, dengue hemorrhagic fever; DSS, dengue shock syndrome; E, Envelope; EHI, Environmental Health Institute; HDB, Housing Development Board; MOH, Ministry of Health; NEA, National Environment Agency; NS1, Non-structural protein 1; PCR, polymerase chain reaction; RNA, Ribose Nucleic Acid; RT-PCR, Reverse Transcription PCR Availability of data and materials E gene sequences (n = 59) of DENV used in the phylogenetic analyses included in the current study are available in National Centre for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov) under the accession numbers KJ806814, KR779787 and from KX224256 to KX224312. Phylogenetic tree files can be accessed under submission ID 19274 in TreeBase database (https://www.treebase.org/). Ethics approval and consent to participate All sera used to generate E gene sequences described in the present study were obtained from patients after obtaining the written informed consent. The use of consented patient sera for the genotyping purposes was approved by the Institutional Review Board of National Environment Agency, Singapore (IRB 003.1). On the other hand, case data presented in this study was obtained from national disease surveillance statistics which is available in public domains (https://www.moh.gov.sg/content/moh_web/home/ statistics/infectiousDiseasesStatistics/weekly_infectiousdiseasesbulletin.html). The spatial distribution of dengue cases in Singapore was generated using the kernel density tool in the spatial analyst toolbox of ArcGIS 10.1 ArcMap software (ESRI, CA, USA) based on a search radius of 400 m. Case density values were classified into four classes of < 25th (2 cases/km2), 25th-50th (16-25 cases/km2), 51st-75th (56- 61 cases/km2) and more than 75th (217 cases/km2) Acknowledgements h k ll f We thank colleagues from EHI, NEA and Communicable Diseases Division, MOH, Singapore who contributed to various aspects of data collection. Data analysis Incidence rates were calculated using all laboratory- confirmed cases reported to MOH based on the estimated mid-year population obtained from the Singapore Depart- ment of Statistics, Ministry of Trade and Industry. The house index and Ae. aegypti house index were calculated based on Aedes breeding data collected by ground officers through routine premise inspections. The Ae. aegypti house indices from 2012 to 2014 were compared by using the paired two sample t test in R software version 3.1.1 [33]. The Gravitrap aegypti index was calculated based on adult Ae. aegypti data generated through the Gravitrap sentinel system. The index expresses the percentage of Gravitraps that caught at least one Ae. aeygpti in a par- ticular location. The relationship between the diagnostic rate and total number of reported cases was determined by calculating the Spearman’s Correlation Coefficient in R software version 3.1.1 [33]. The weekly national serotype data and weekly EHI genotype data from 2007 to 2014 were used to estimate the historical genotype proportions. In order to obtain smooth estimates of genotype propor- tions over time, we used Bayesian approach, assuming multinomial distribution of serotypes and genotypes, and an auto-correlated prior distribution for logarithm trans- formed proportions. We then sampled from the posterior distribution to get the Bayesian estimates of the weekly genotype proportions, which were used together with weekly national case count to calculate the weekly cases attributed to each genotype as well the cumulative case count. The analysis was done using R software version 3.1.1 [33]. Only the genotypes dominant during the epi- demic years have been plotted in the graph. Entomological data collection As part of active preventive surveillance, premise checks for larval breeding were undertaken by approximately 800 ground officers from the Environmental Public Health Operations, NEA, with an aim of checking every home and surrounding areas every three to six months. Such source reduction and vector surveillance efforts were enhanced in vicinities with high risk of DENV transmission, such as dengue clusters. Aedes immatures collected by field officers were morphologically identified Details of cases were sent promptly to NEA to deter- mine the clustering of cases and to conduct site visits for further investigations to aid vector control opera- tions. A dengue cluster is defined as two or more cases Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Page 12 of 13 Page 12 of 13 to the species level at EHI. The number and type of breeding places as well as mosquito identification data was recorded in a common database for operational pur- poses. The mosquito identification data was also used for enforcement activities. In addition, adult Aedes popu- lation monitoring was conducted with about 3,000 Gravitraps at 34 sentinel locations weekly and within se- lected cluster areas on ad hoc basis. The design and de- ployment of Gravitraps have previously been described [24]. For all adult samples collected from clusters and sentinel locations, the abdomens of trapped Aedes mos- quitoes were pooled into groups of five and screened for DENV by using a Dengue NS1 antigen assay as previ- ously described [28]. quantiles, using the quantile classification method, so that higher densities were shown in darker tones of pink and vice versa. Authors’ contributions NLC and JC supervised the case and virus surveillance programme. NLC and CSC supervised the entomological surveillance and GY, LL, JR and PLO supervised the epidemiological data analysis. YLL supervised the diagnostics and patient recruitment. JR and LL performed the epidemiological and entomological data analysis. CK and CL generated the sequence data. CK and HCH performed sequence analyses. HCH, JR, CSC, GY, PLO, JC and NLC wrote the manuscript. All authors read and approved the final manuscript. Competing interests p g The authors declare that they have no competing interests. Author details 1 1Environmental Health Institute, National Environment Agency, 11, Biopolis Way, #06-05-08, Singapore 138667, Singapore. 2Communicable Diseases Division, Ministry of Health, College of Medicine Building, 16 College Road, Singapore 169854, Singapore. 3National Public Health Laboratory, Ministry of Health, College of Medicine Building 16 College Road, Singapore 169854, Singapore. 4School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, Singapore 637551, Singapore. Page 13 of 13 Page 13 of 13 Hapuarachchi et al. BMC Infectious Diseases (2016) 16:300 Received: 5 December 2015 Accepted: 1 June 2016 References Diversity and evolution of the envelope gene of dengue virus type 1. Virology. 2002;303:110–9. 32. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol. 2013;30:2725–9. 33. RCoreTeam. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2014. gy 7. Holmes EC, Twiddy SS. The origin, emergence dengue virus. Infect Genet Evol. 2003;3:19–28. gy 7. Holmes EC, Twiddy SS. The origin, emergence and evolutionary genetics of dengue virus. Infect Genet Evol. 2003;3:19–28. 7. Holmes EC, Twiddy SS. The origin, emergence and dengue virus. Infect Genet Evol. 2003;3:19–28. 8. Gubler DJ. Dengue and dengue hemorrhagic fever. Clin Microbiol Rev. 1998;11:480–96. 9. WHO. Dengue: Guidelines for diagnosis, treatment, prevention and control. Geneva: World Health Organization; 2009. 10. Chan KL, Ng SK, Chew LM. The 1973 dengue haemorrhagic fever outbreak in Singapore and its control. Singapore Med J. 1977;18:81–93. 11. Goh KT. Dengue–a re-emerging infectious disease in Singapore. Ann Acad Med Singapore. 1997;26:664–70. 12. Chan YC, Ho BC, Chan KL. Aedes aegypti (L.) and Aedes albopictus (Skuse) in Singapore City. 5. Observations in relation to dengue haemorrhagic fever. Bull World Health Organ. 1971;44:651–7. 13. Goh KT. Changing epidemiology of dengue in Singapore. Lancet. 1995;346: 1098. 14. Koh BK, Ng LC, Kita Y, Tang CS, Ang LW, Wong KY, James L, Goh KT. The 2005 dengue epidemic in Singapore: epidemiology, prevention and control. Ann Acad Med Singapore. 2008;37:538–45. 15. Lee KS, Lai YL, Lo S, Barkham T, Aw P, Ooi PL, Tai JC, Hibberd M, Johansson P, Khoo SP, Ng LC. Dengue Virus Surveillance for Early Warning, Singapore. Emerg Infect Dis. 2010;16(5):847–49. 16. Ler TS, Ang LW, Yap GS, Ng LC, Tai JC, James L, Goh KT. Epidemiological characteristics of the 2005 and 2007 dengue epidemics in Singapore - similarities and distinctions. Western Pac Surveill Response J. 2011;2:24–9. 17. Ang LW, Cutter J, James L, Goh KT. Seroepidemiology of dengue virus infection in the adult population in tropical Singapore. Epidemiol Infect. 2015;143:1585–93. 18. Yew YW, Ye T, Ang LW, Ng LC, Yap G, James L, Chew SK, Goh KT. Seroepidemiology of dengue virus infection among adults in Singapore. Ann Acad Med Singapore. 2009;38:667–75. 19. Goh KT, Yamazaki S. Serological survey on dengue virus infection in Singapore. Trans R Soc Trop Med Hyg. 1987;81:687–9. 19. Goh KT, Yamazaki S. Serological survey on dengue virus infection in Singapore. Received: 5 December 2015 Accepted: 1 June 2016 26. Lai YL, Chung YK, Tan HC, Yap HF, Yap G, Ooi EE, Ng LC. Cost-effective real-time reverse transcriptase PCR (RT-PCR) to screen for Dengue virus followed by rapid single-tube multiplex RT-PCR for serotyping of the virus. J Clin Microbiol. 2007;45:935–41. 27. Pok KY, Lai YL, Sng J, Ng LC. Evaluation of Nonstructural 1 Antigen Assays for the Diagnosis and Surveillance of Dengue in Singapore. Vector Borne Zoonotic Dis. 2010. 28. Tan CH, Wong PS, Li MZ, Vythilingam I, Ng LC. Evaluation of the Dengue NS1 Ag Strip(R) for detection of dengue virus antigen in Aedes aegypti (Diptera: Culicidae). Vector Borne Zoonotic Dis. 2011;11:789–92. 29. Koo C, Nasir A, Hapuarachchi HC, Lee KS, Hasan Z, Ng LC, Khan E. Evolution and heterogeneity of multiple serotypes of Dengue virus in Pakistan, 2006-2011. Virol J. 2013;10:275. 30. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol. 1992;30:545–51. 31. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acid Symp Ser. 1999;41:95–8. 32. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol. 2013;30:2725–9. 33. RCoreTeam. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2014. 26. Lai YL, Chung YK, Tan HC, Yap HF, Yap G, Ooi EE, Ng LC. Cost-effective real-time reverse transcriptase PCR (RT-PCR) to screen for Dengue virus followed by rapid single-tube multiplex RT-PCR for serotyping of the virus. J Clin Microbiol. 2007;45:935–41. 27. Pok KY, Lai YL, Sng J, Ng LC. Evaluation of Nonstructural 1 Antigen Assays for the Diagnosis and Surveillance of Dengue in Singapore. Vector Borne Zoonotic Dis. 2010. References Trans R Soc Trop Med Hyg. 1987;81:687–9. 20. Low SL, Lam S, Wong WY, Teo D, Ng LC, Tan LK. Dengue Seroprevalence of Healthy Adults in Singapore: Serosurvey Among Blood Donors, 2009. Am J Trop Med Hyg. 2015;93:40–5. 20. Low SL, Lam S, Wong WY, Teo D, Ng LC, Tan LK. Dengue Seroprevalence of Healthy Adults in Singapore: Serosurvey Among Blood Donors, 2009. Am J Trop Med Hyg. 2015;93:40–5. 21. Lee KS, Lo S, Tan SS, Chua R, Tan LK, Xu H, Ng LC. Dengue virus surveillance in Singapore reveals high viral diversity through multiple introductions and in situ evolution. Infect Genet Evol. 2012;12:77–85. 21. Lee KS, Lo S, Tan SS, Chua R, Tan LK, Xu H, Ng LC. Dengue virus surveillance in Singapore reveals high viral diversity through multiple introductions and in situ evolution. Infect Genet Evol. 2012;12:77–85. References 1. Murray NE, Quam MB, Wilder-Smith A. Epidemiology of dengue: past, present and future prospects. Clin Epidemiol. 2013;5:299–309. 1. Murray NE, Quam MB, Wilder-Smith A. Epidemiology of dengue: past, present and future prospects. Clin Epidemiol. 2013;5:299–309. 28. Tan CH, Wong PS, Li MZ, Vythilingam I, Ng LC. Evaluation of the Dengue NS1 Ag Strip(R) for detection of dengue virus antigen in Aedes aegypti (Diptera: Culicidae). Vector Borne Zoonotic Dis. 2011;11:789–92. 2. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, Drake JM, Brownstein JS, Hoen AG, Sankoh O, et al. The global distribution and burden of dengue. Nature. 2013;496:504–7. 2. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, Drake JM, Brownstein JS, Hoen AG, Sankoh O, et al. The global distribution and burden of dengue. Nature. 2013;496:504–7. 29. Koo C, Nasir A, Hapuarachchi HC, Lee KS, Hasan Z, Ng LC, Khan E. Evolution and heterogeneity of multiple serotypes of Dengue virus in Pakistan, 2006-2011. Virol J. 2013;10:275. 3. Calisher CH, Karabatsos N, Dalrymple JM, Shope RE, Porterfield JS, Westaway EG, Brandt WE. Antigenic relationships between flaviviruses as determined by cross-neutralization tests with polyclonal antisera. J Gen Virol. 1989;70(Pt 1):37–43. 3. Calisher CH, Karabatsos N, Dalrymple JM, Shope RE, Porterfield JS, Westaway EG, Brandt WE. Antigenic relationships between flaviviruses as determined by cross-neutralization tests with polyclonal antisera. J Gen Virol. 1989;70(Pt 1):37–43. 30. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol. 1992;30:545–51. 4. Simmons CP, Farrar JJ, Nguyen VV, Wills B. Dengue. N Engl J Med. 2012;366: 1423–32. 4. Simmons CP, Farrar JJ, Nguyen VV, Wills B. Dengue. N Engl J Med. 2012;366: 1423–32. 5. Rico-Hesse R. Molecular evolution and distribution of dengue viruses type 1 and 2 in nature. Virology. 1990;174:479–93. 5. Rico-Hesse R. Molecular evolution and distribution of dengue viruses type 1 and 2 in nature. Virology. 1990;174:479–93. 31. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acid Symp Ser. 1999;41:95–8. 6. Goncalvez AP, Escalante AA, Pujol FH, Ludert JE, Tovar D, Salas RA, Liprandi F. Diversity and evolution of the envelope gene of dengue virus type 1. Virology. 2002;303:110–9. 6. Goncalvez AP, Escalante AA, Pujol FH, Ludert JE, Tovar D, Salas RA, Liprandi F. 26. Lai YL, Chung YK, Tan HC, Yap HF, Yap G, Ooi EE, Ng LC. Cost-effective real-time reverse transcriptase PCR (RT-PCR) to screen for Dengue virus followed by rapid single-tube multiplex RT-PCR for serotyping of the virus. J Clin Microbiol. 2007;45:935–41. 27. Pok KY, Lai YL, Sng J, Ng LC. Evaluation of Nonstructural 1 Antigen Assays for the Diagnosis and Surveillance of Dengue in Singapore. Vector Borne Zoonotic Dis. 2010. 28. Tan CH, Wong PS, Li MZ, Vythilingam I, Ng LC. Evaluation of the Dengue NS1 Ag Strip(R) for detection of dengue virus antigen in Aedes aegypti (Diptera: Culicidae). Vector Borne Zoonotic Dis. 2011;11:789–92. 29. Koo C, Nasir A, Hapuarachchi HC, Lee KS, Hasan Z, Ng LC, Khan E. Evolution and heterogeneity of multiple serotypes of Dengue virus in Pakistan, 2006-2011. Virol J. 2013;10:275. 30. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol. 1992;30:545–51. 31. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acid Symp Ser. 1999;41:95–8. 32. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol. 2013;30:2725–9. 33. RCoreTeam. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2014. 29. Koo C, Nasir A, Hapuarachchi HC, Lee KS, Hasan Z, Ng LC, Khan E. Evolution and heterogeneity of multiple serotypes of Dengue virus in Pakistan, 2006-2011. Virol J. 2013;10:275. Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: 22. Ng LC, Chem YK, Koo C, Mudin RN, Amin FM, Lee KS, Kheong CC. 2013 Dengue Outbreaks in Singapore and Malaysia Caused by Different Viral Strains. Am J Trop Med Hyg. 2015;92:1150–5. • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit and we will help you at every step: 23. Shi Y, Liu X, Kok SY, Rajarethinam J, Liang S, Yap G, Chong CS, Lee KS, Tan SS, Chin CK, et al. Three-Month Real-Time Dengue Forecast Models: An Early Warning System for Outbreak Alerts and Policy Decision Support in Singapore. Environmental health perspectives. 2015. doi:10.1289/ehp.1509981. 24. Lee C, Vythilingam I, Chong CS, Abdul Razak MA, Tan CH, Liew C, Pok KY, Ng LC. Gravitraps for management of dengue clusters in Singapore. Am J Trop Med Hyg. 2013;88:888–92. 25. Ng LC, Vythilingum, I (Ed.). Vectors of Flaviviruses and Strategies for Control, in Molecular Virology and Control of Flaviviruses. Poole, United Kingdom: Caister Academic Press; 2011. 25. Ng LC, Vythilingum, I (Ed.). Vectors of Flaviviruses and Strategies for Control, in Molecular Virology and Control of Flaviviruses. Poole, United Kingdom: Caister Academic Press; 2011.
https://openalex.org/W2594807754	https://diposit.ub.edu/dspace/bitstream/2445/163397/1/668864.pdf	English	null	Singlet ground states in compounds with a [MnIII4O<sub>2</sub>]<sup>8+</sup> core due to broken degeneration	New journal of chemistry	2,017	public-domain	10,013	Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 2 3 Luis Escriche-Tur,ab Belén Albela,b Mercè Font-Bardiac and Montserrat Corbellaad 4 5 6 7 8 9 10 11 12 13 14 a Departament de Química Inorgànica i Orgànica (Secció inorgànica), Universitat de Barcelona, C/ 15 Martı´ i Franque`s 1-11, 08028 Barcelona, Spain. 16 E-mail: luis.escrichetur@gmail.com , montse.corbella@qi.ub.es 17 b Laboratoire de Chimie, ENS de Lyon, Universite´ de Lyon, 46 Alle´e d’Italie, 69364 Lyon Cedex 07, 18 France 19 c Cristal·lografia, Mineralogia i Dipo`sits Minerals, Universitat de Barcelona, Martı´ i Franque`s s/n, 20 08028 Barcelona, Spain 21 d Institud de Nanocie`ncia i Nanotecnologia de la Universitat de Barcelona (IN2UB), Av. Joan XXIII 22 s/n, 08028 Barcelona, Spain 23 24 25 26 27 28 29 . 30 31 Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 2 3 Luis Escriche-Tur,ab Belén Albela,b Mercè Font-Bardiac and Montserrat Corbellaad 4 5 6 7 8 9 10 11 12 13 14 a Departament de Química Inorgànica i Orgànica (Secció inorgànica), Universitat de Barcelona, C/ 15 Martı´ i Franque`s 1-11, 08028 Barcelona, Spain. 16 E-mail: luis.escrichetur@gmail.com , montse.corbella@qi.ub.es 17 b Laboratoire de Chimie, ENS de Lyon, Universite´ de Lyon, 46 Alle´e d’Italie, 69364 Lyon Cedex 07, 18 France 19 c Cristal·lografia, Mineralogia i Dipo`sits Minerals, Universitat de Barcelona, Martı´ i Franque`s s/n, 20 08028 Barcelona, Spain 21 d Institud de Nanocie`ncia i Nanotecnologia de la Universitat de Barcelona (IN2UB), Av. Joan XXIII 22 s/n, 08028 Barcelona, Spain 23 24 25 26 27 28 29 . 30 31 Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 2 3 Luis Escriche-Tur,ab Belén Albela,b Mercè Font-Bardiac and Montserrat Corbellaad 4 5 6 7 8 9 10 11 12 13 14 a Departament de Química Inorgànica i Orgànica (Secció inorgànica), Universitat de Barcelona, C/ 15 Martı´ i Franque`s 1-11, 08028 Barcelona, Spain. 16 E-mail: luis.escrichetur@gmail.com , montse.corbella@qi.ub.es 17 b Laboratoire de Chimie, ENS de Lyon, Universite´ de Lyon, 46 Alle´e d’Italie, 69364 Lyon Cedex 07, 18 France 19 c Cristal·lografia, Mineralogia i Dipo`sits Minerals, Universitat de Barcelona, Martı´ i Franque`s s/n, 20 08028 Barcelona, Spain 21 d Institud de Nanocie`ncia i Nanotecnologia de la Universitat de Barcelona (IN2UB), Av. Joan XXIII 22 s/n, 08028 Barcelona, Spain 23 24 Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 ABSTRACT: 32 33 Two new tetranuclear compounds with a formula [MnIII4 (m-O)2(m-4- 34 RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO or tBu and m = 0 or 1, were synthesised 35 and studied structurally and magnetically. The core of these compounds comprises a central Mn2O2 36 rhombus to which two terminal ions are attached – one to each oxo bridge. There are two types of 37 bridges that alternately bind the central and terminal ions, those having a triple (m-O)(m-RCOO)2 or a 38 double (m-O)(m-RCOO) bridge. The fit of the magnetic data of analogous compounds has so far been 39 performed considering two different magnetic interactions, that between central ions (J1) and those 40 between terminal and central ions (Jct), leading to ground states with ST = 2 or 3, or to five energetically 41 degenerate ground states with ST = 0–4, depending on the J1/Jct ratio. In contrast, the compounds 42 presented herein show an isolated ST = 0 ground state, and it was necessary to distinguish the two types 43 of magnetic interactions between central and terminal ions (J2 and J3) to achieve a good fit of the 44 experimental data. The differentiation of these interactions causes a spin state redistribution: the 45 degeneration of ST = 0–4 breaks and the states with ST a 0 become unstable as J2 and J3 become more 46 different. Nevertheless, the assignment of these states to a particular spin configuration was 47 unachievable because the composition of these states changes upon decreasing the J3/J2 ratio. The 48 importance of considering the relative orientation of Jahn–Teller axes is also highlighted. 49 50 33 Two new tetranuclear compounds with a formula [MnIII4 (m-O)2(m-4- 34 RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO or tBu and m = 0 or 1, were synthesised 35 and studied structurally and magnetically. The core of these compounds comprises a central Mn2O2 36 rhombus to which two terminal ions are attached – one to each oxo bridge. There are two types of 37 bridges that alternately bind the central and terminal ions, those having a triple (m-O)(m-RCOO)2 or a 38 double (m-O)(m-RCOO) bridge. Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 The fit of the magnetic data of analogous compounds has so far been 39 performed considering two different magnetic interactions, that between central ions (J1) and those 40 between terminal and central ions (Jct), leading to ground states with ST = 2 or 3, or to five energetically 41 degenerate ground states with ST = 0–4, depending on the J1/Jct ratio. In contrast, the compounds 42 presented herein show an isolated ST = 0 ground state, and it was necessary to distinguish the two types 43 of magnetic interactions between central and terminal ions (J2 and J3) to achieve a good fit of the 44 experimental data. The differentiation of these interactions causes a spin state redistribution: the 45 degeneration of ST = 0–4 breaks and the states with ST a 0 become unstable as J2 and J3 become more 46 different. Nevertheless, the assignment of these states to a particular spin configuration was 47 unachievable because the composition of these states changes upon decreasing the J3/J2 ratio. The 48 importance of considering the relative orientation of Jahn–Teller axes is also highlighted. 49 50 50 INTRODUCTION 51 52 In the past few years many tetranuclear Mn compounds have been synthesised and characterised either 53 to mimic the water oxidizing center1 or to study the ground-state spin frustration that is characteristic of 54 these kinds of compounds.2,3 Moreover, such clusters possess large numbers of unpaired electrons, 55 making them attractive as precursors for other magnetic materials.4 Some of the first compounds 56 synthesised contained a [Mn4O2]6+/7+/8+ core, where the metals could be arranged either in a planar or 57 a non-planar (‘‘butterfly’’) fashion (Fig. 1), and the Mn oxidation states are MnII 2MnIII 2 , MnIIMnIII 58 3 or MnIII 4 .2,3,5–17 59 The variability of the ground-state spin and the presence of spin frustration in these tetranuclear 60 compounds have been profoundly studied for both MnIII 4 and mixed-valence compounds. The 61 resulting ground state depends on the relative magnitude of the magnetic interactions between the 62 central Mn ions ( Jcc or J1) and those between central and terminal ions ( Jct), both being 63 antiferromagnetic. In particular, MnIII 4 compounds may display a ground state with ST = 2 or 3, or 64 have five energetically degenerate ground states with ST = 0–4, depending on the J1/Jct ratio. Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 65 The analysis of the magnetic data for these compounds is rather challenging, since there are five Mn 66 Mn interactions and the presence of MnIII ions may lead to substantial zero-field splitting (ZFS) that 67 makes such properties more difficult to understand. In fact, among several examples found in the 68 literature, the analyses were performed applying several approximations3,17 or without analysing the 69 data completely, especially in the low temperature range.2,5,8 70 In this work we present the synthesis and crystal structures of two new [MnIII 4 O2]8+ compounds with 71 a general formula [Mn4(m-O)2(m-4- RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO (1) 72 or tBu (2) and m = 0 or 1. The crystal structure of compound 1 could be determined without any 73 complication, as we obtained high-quality single-crystals. However, the crystals obtained for 2 were 74 unfortunately poorly diffracting, and only the formula and approximate structural parameters of 2 could 75 be obtained. We also report an in depth study of the magnetic properties and the influence of the relative 76 magnitude of the Mnc Mnt interactions on the resulting spin state distribution. The inclusion of 77 the axial anisotropy parameter (DMn) and the consideration of the relative disposition of the Jahn–Teller 78 axes of the MnIII ions enabled us to completely fit the magnetic data and to estimate the approximate 79 In this work we present the synthesis and crystal structures of two new [MnIII 4 O2]8+ compounds with 71 a general formula [Mn4(m-O)2(m-4- RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO (1) 72 or tBu (2) and m = 0 or 1. The crystal structure of compound 1 could be determined without any 73 complication, as we obtained high-quality single-crystals. However, the crystals obtained for 2 were 74 unfortunately poorly diffracting, and only the formula and approximate structural parameters of 2 could 75 be obtained. We also report an in depth study of the magnetic properties and the influence of the relative 76 magnitude of the Mnc Mnt interactions on the resulting spin state distribution. The inclusion of 77 the axial anisotropy parameter (DMn) and the consideration of the relative disposition of the Jahn–Teller 78 axes of the MnIII ions enabled us to completely fit the magnetic data and to estimate the approximate 79 values of the ZFS of the MnIII ions. [Mn4(l-O)2(l-4-tBuC6H4COO)6(H2O)2(phen)2][Mn4(l-O)2(l-4- 101 [Mn4(l O)2(l 4 tBuC6H4COO)6(H2O)2(phen)2][Mn4(l O)2(l 4 101 tBuC6H4COO)6(CH3CN)2(phen)2](ClO4)4 (2). 4-tBuC6H4COOH (1.75 mmol, 0.31 g) and 102 Mn(ClO4)2 6H2O (0.8 mmol, 0.29 g) were dissolved in acetonitrile (10 mL). Then, solid NBu4MnO4 103 (0.20 mmol, 0.082 g) was added to the previous solution in small portions for 1–2 minutes while, almost 104 simultaneously, an acetonitrile solution (10 mL) of 1,10-phenanthroline (phen) (0.50 mmol, 0.10 g) was 105 added, also in small portions. Finally, solid NaClO4 H2O (10.6 mmol, 1.3 g) was added. The resulting 106 dark red solution (total volumeB25 mL) was stirred for 15 min and shortly afterward filtered to separate 107 any possible residue. After leaving the solution undisturbed for three weeks, dark red crystals were 108 isolated by filtration, washed with ether and dried under vacuum. Single-crystals were obtained under 109 the same conditions but using less NaClO4 H2O (9.8 mmol, 1.2 g). Yield: 16%. Anal. calcd for 110 C92H99Cl2Mn4N5O23 (average formula, referred to one Mn4 unit) (M.W. = 1933.45 g mol 1) (%): 111 C, 57.15; H, 5.16; N, 3.62. Found (%): C, 58.10; H, 5.32; N, 3.50. Selected IR data (cm 1): 3434 (br), 112 3075 (w), 2960 (m), 2906 (w), 2869 (w), 1597 (s), 1555 (s), 1521 (s), 1462 (w), 1383 (s), 1310 (w), 113 1269 (w), 1193 (w) 1101 (s), 1015 (w), 854 (m), 786 (m), 722 (m), 652 (m), 623 (m), 601 (m), 548 (w), 114 479 (m). 115 tBuC6H4COO)6(CH3CN)2(phen)2](ClO4)4 (2). 4-tBuC6H4COOH (1.75 mmol, 0.31 g) and 102 Mn(ClO4)2 6H2O (0.8 mmol, 0.29 g) were dissolved in acetonitrile (10 mL). Then, solid NBu4MnO4 103 (0.20 mmol, 0.082 g) was added to the previous solution in small portions for 1–2 minutes while, almost 104 simultaneously, an acetonitrile solution (10 mL) of 1,10-phenanthroline (phen) (0.50 mmol, 0.10 g) was 105 added, also in small portions. Finally, solid NaClO4 H2O (10.6 mmol, 1.3 g) was added. The resulting 106 dark red solution (total volumeB25 mL) was stirred for 15 min and shortly afterward filtered to separate 107 any possible residue. After leaving the solution undisturbed for three weeks, dark red crystals were 108 isolated by filtration, washed with ether and dried under vacuum. Single-crystals were obtained under 109 the same conditions but using less NaClO4 H2O (9.8 mmol, 1.2 g). Yield: 16%. Anal. calcd for 110 C92H99Cl2Mn4N5O23 (average formula, referred to one Mn4 unit) (M.W. Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 = 1769.59 g mol 1) (%): C, 54.30; H, 3.70; N, 3.17. Found (%): C, 53.67; H, 3.73; N, 3.17. 97 Selected IR data (cm 1): IR (cm 1): 3446 (br), 3070 (w), 2931 (w), 2836 (w), 1602 (s), 1559 (s), 1507 98 (m), 1457 (w), 1420 (m), 1380 (s), 1359 (s), 1314 (m), 1256 (s), 1170 (s) 1086 (m), 1025 (m), 852 (w), 99 790 (m), 749 (w), 667 (w), 621 (m), 504 (w), 437 (w). 100 Singlet ground states in compounds with a [MnIII 4 O2]8+ core due to broken degeneration† 1 80 81 EXPERIMENTAL SECTION 82 83 Synthesis 84 All manipulations were performed under aerobic conditions. Reagents and solvents were obtained from 85 commercial sources and used without further purification. NBu4MnO4 was prepared as described in the 86 literature.18 Caution! Perchlorate salts of compounds containing organic ligands are potentially 87 explosive. Only small quantities of these compounds should be prepared. 88 [Mn4(l-O)2(l-4-MeOC6H4COO)7(phen)2]ClO4 (1). 4-MeOC6H4- COOH (2.89 mmol, 0.44 g) and 89 Mn(ClO4)2 6H2O (1.32 mmol, 0.48 g) were dissolved in acetonitrile. Then, solid NBu4MnO4 (0.33 90 mmol, 0.12 g) was added to the previous solution in small portions for 1–2 min while, almost 91 simultaneously, 10 mL acetonitrile solution of 1,10-phenanthroline (phen) (0.83 mmol, 0.16 g) was 92 added, also in small portions. The resulting dark solution (total volume B20 mL) was stirred for 10 93 minutes and dried with a rotary evaporator. The resulting black oil was dissolved in a CH3CN: EtOH 94 (10 : 10 mL) mixture and filtered to separate any possible residue. Dark red crystals were obtained after 95 a week of slow evaporation at roomtemperature. Yield: 25%. Anal. calcd for C80H65ClMn4N4O27 96 (M.W. = 1769.59 g mol 1) (%): C, 54.30; H, 3.70; N, 3.17. Found (%): C, 53.67; H, 3.73; N, 3.17. 97 Selected IR data (cm 1): IR (cm 1): 3446 (br), 3070 (w), 2931 (w), 2836 (w), 1602 (s), 1559 (s), 1507 98 (m), 1457 (w), 1420 (m), 1380 (s), 1359 (s), 1314 (m), 1256 (s), 1170 (s) 1086 (m), 1025 (m), 852 (w), 99 790 (m), 749 (w), 667 (w), 621 (m), 504 (w), 437 (w). 100 [Mn4(l-O)2(l-4-MeOC6H4COO)7(phen)2]ClO4 (1). 4-MeOC6H4- COOH (2.89 mmol, 0.44 g) and 89 Mn(ClO4)2 6H2O (1.32 mmol, 0.48 g) were dissolved in acetonitrile. Then, solid NBu4MnO4 (0.33 90 mmol, 0.12 g) was added to the previous solution in small portions for 1–2 min while, almost 91 simultaneously, 10 mL acetonitrile solution of 1,10-phenanthroline (phen) (0.83 mmol, 0.16 g) was 92 added, also in small portions. The resulting dark solution (total volume B20 mL) was stirred for 10 93 minutes and dried with a rotary evaporator. The resulting black oil was dissolved in a CH3CN: EtOH 94 (10 : 10 mL) mixture and filtered to separate any possible residue. Dark red crystals were obtained after 95 a week of slow evaporation at roomtemperature. Yield: 25%. Anal. calcd for C80H65ClMn4N4O27 96 (M.W. [Mn4(l-O)2(l-4-tBuC6H4COO)6(H2O)2(phen)2][Mn4(l-O)2(l-4- 101 = 1933.45 g mol 1) (%): 111 C, 57.15; H, 5.16; N, 3.62. Found (%): C, 58.10; H, 5.32; N, 3.50. Selected IR data (cm 1): 3434 (br), 112 3075 (w), 2960 (m), 2906 (w), 2869 (w), 1597 (s), 1555 (s), 1521 (s), 1462 (w), 1383 (s), 1310 (w), 113 1269 (w), 1193 (w) 1101 (s), 1015 (w), 854 (m), 786 (m), 722 (m), 652 (m), 623 (m), 601 (m), 548 (w), 114 479 (m). 115 tBuC6H4COO)6(CH3CN)2(phen)2](ClO4)4 (2). 4-tBuC6H4COOH (1.75 mmol, 0.31 g) and 102 118 Physical characterisation 119 C, H and N analyses were carried out by the ‘‘Centres Cientı´fics i Tecnolo`gics’’ of the Universitat de 120 Barcelona. Infrared spectra were recorded on KBr pellets in the 4000– 400 cm 1 range using a Thermo 121 Nicolet Avatar 330 FTIR spectrometer. Magnetic measurements were performed on microcrystalline 122 samples in a Quantum Design MPMS XL5 SQUID Magnometer at the ‘‘Unitat de Mesures 123 Magne`tiques’’ (Universitat de Barcelona). Magnetic susceptibility was measured between 2 and 300 K 124 and with a magnetic field of 0.02 T. The fit of the experimental magnetic data was performed by 125 minimizing the function R = P[(wMT)exp (wMT)calcd]2/P[(wMT)exp]2. 126 127 Single-crystal X-ray crystallography 128 The data collection for compounds 1 and 2 was performed on a Bruker Apex-II diffractometer at 100 K, 129 equipped with a graphite monochromatic Mo Ka radiation (l = 0.71073 Å). Unit-cell parameters were 130 determined from B9500 reflections and refined by the least-squares method. Several thousand 131 reflections (161 200 for 1 and 84 790 for 2) were collected using the F- and o-scan. Data were corrected 132 for absorption effects using the multi-scan method (SADABS).19 Table S1 (ESI†) contains the 133 crystallographic data collection and structure refinement details. The structures were solved by direct 134 methods and refined by full-matrix leastsquares using SHELXL-2016/6,20 run by the Wingx21,22 and 135 ShelXle23 user interfaces, respectively. Non-hydrogen atoms were refined anisotropically, whereas 136 hydrogen atoms were computed and refined with isotropic thermal parameters riding on their respective 137 carbon or oxygen atoms. Particularly, solvent hydrogen atoms interacting with neighbours were placed 138 in ideal positions. 139 Compound 1 1/2 EtOH 5/4 CH3CN 1/4H2O crystallises in triclinic space group P1. The asymmetric 140 unit consists of two conformational isomers of the [Mn4(m-O)2(m-4-MeOC6H4COO)7(phen)2]+ 141 complex, two perchlorate anions and some solvent molecules (H2O, CH3CN and EtOH). [Mn4(l-O)2(l-4-tBuC6H4COO)6(H2O)2(phen)2][Mn4(l-O)2(l-4- 101 A total of 142 2240 parameters were refined in the final refinement on F2 using 75 restraints. 143 Single-crystals of compound 2 were isolated and mounted on the diffractometer. However, the crystals 144 of this compound were very thin and gave highly poor statistics (Rint = 0.253). Then, the structure could 145 not be completely refined and, therefore, it was not deposited in the CCDC database. Even so, the Q 146 peaks could be assigned to all atoms and isotropically refined. This compound crystallises in the triclinic 147 space group P21/c. The asymmetric unit consists of a half of the [Mn4(m-O)2(m-4- 148 tBuC6H4COO)6(H2O)2(phen)2]2+ and of the [Mn4(m-O)2(m-4- 149 tBuC6H4COO)6(CH3CN)2(phen)2]2+ complexes, two perchlorate anions, and two acetonitrile 150 molecules. The whole complexes are generated by an inversion centre situated in the middle of them. 151 Physical characterisation 119 C, H and N analyses were carried out by the ‘‘Centres Cientı´fics i Tecnolo`gics’’ of the Universitat de 120 Barcelona. Infrared spectra were recorded on KBr pellets in the 4000– 400 cm 1 range using a Thermo 121 Nicolet Avatar 330 FTIR spectrometer. Magnetic measurements were performed on microcrystalline 122 samples in a Quantum Design MPMS XL5 SQUID Magnometer at the ‘‘Unitat de Mesures 123 Magne`tiques’’ (Universitat de Barcelona). Magnetic susceptibility was measured between 2 and 300 K 124 and with a magnetic field of 0.02 T. The fit of the experimental magnetic data was performed by 125 minimizing the function R = P[(wMT)exp (wMT)calcd]2/P[(wMT)exp]2. 126 127 The data collection for compounds 1 and 2 was performed on a Bruker Apex-II diffractometer at 100 K, 129 equipped with a graphite monochromatic Mo Ka radiation (l = 0.71073 Å). Unit-cell parameters were 130 determined from B9500 reflections and refined by the least-squares method. Several thousand 131 reflections (161 200 for 1 and 84 790 for 2) were collected using the F- and o-scan. Data were corrected 132 for absorption effects using the multi-scan method (SADABS).19 Table S1 (ESI†) contains the 133 crystallographic data collection and structure refinement details. The structures were solved by direct 134 methods and refined by full-matrix leastsquares using SHELXL-2016/6,20 run by the Wingx21,22 and 135 ShelXle23 user interfaces, respectively. Non-hydrogen atoms were refined anisotropically, whereas 136 hydrogen atoms were computed and refined with isotropic thermal parameters riding on their respective 137 carbon or oxygen atoms. Particularly, solvent hydrogen atoms interacting with neighbours were placed 138 in ideal positions. 139 Compound 1 1/2 EtOH 5/4 CH3CN 1/4H2O crystallises in triclinic space group P1. The asymmetric 140 unit consists of two conformational isomers of the [Mn4(m-O)2(m-4-MeOC6H4COO)7(phen)2]+ 141 complex, two perchlorate anions and some solvent molecules (H2O, CH3CN and EtOH). A total of 142 2240 parameters were refined in the final refinement on F2 using 75 restraints. 143 Compound 1 1/2 EtOH 5/4 CH3CN 1/4H2O crystallises in triclinic space group P1. The asymmetric 140 unit consists of two conformational isomers of the [Mn4(m-O)2(m-4-MeOC6H4COO)7(phen)2]+ 141 complex, two perchlorate anions and some solvent molecules (H2O, CH3CN and EtOH). A total of 142 2240 parameters were refined in the final refinement on F2 using 75 restraints. 143 Single-crystals of compound 2 were isolated and mounted on the diffractometer. Physical characterisation 119 However, the crystals 144 of this compound were very thin and gave highly poor statistics (Rint = 0.253). Then, the structure could 145 not be completely refined and, therefore, it was not deposited in the CCDC database. Even so, the Q 146 peaks could be assigned to all atoms and isotropically refined. This compound crystallises in the triclinic 147 space group P21/c. The asymmetric unit consists of a half of the [Mn4(m-O)2(m-4- 148 tBuC6H4COO)6(H2O)2(phen)2]2+ and of the [Mn4(m-O)2(m-4- 149 tBuC6H4COO)6(CH3CN)2(phen)2]2+ complexes, two perchlorate anions, and two acetonitrile 150 molecules. The whole complexes are generated by an inversion centre situated in the middle of them. 151 152 RESULTS AND DISCUSSION 154 155 Synthesis 156 Both tetranuclear compounds 1 and 2 were obtained from the reaction of comproportionation between 157 MnII and MnO4 in the presence of a substituted derivative of benzoic acid and 1,10-phenanthroline 158 (phen), leading to compounds with general formula [Mn4(m-O)2(m-4- 159 RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO (1) or tBu (2), L is a monodentate ligand, 160 and m = 0 or 1. 161 The mother liquors of 1 and 2 were left to slowly evaporate, but no solid was obtained from either of 162 them, indicating that both compounds are highly soluble in acetonitrile. Thus, they needed to be 163 crystallised by mixing the mother solution with absolute ethanol (1) or by adding a huge excess of ClO4 164 anions (2). For compound 1, crystallisation from different CH3CN: EtOH mixtures was tried, but no 165 differences were observed changing the CH3CN/EtOH volume ratio (from 0.3 to 0.7). In all cases, the 166 reaction yield was about 25% considering the stoichiometry of the reaction above. In contrast, we had 167 some difficulties finding the optimum conditions to obtain compound 2 in appreciable yield. Its 168 crystallisation needed to be assisted by the addition of about 22 equivalents of ClO4 anions, leading 169 to 2 in 16% yield. If the amount of ClO4 anions used is decreased, crystallisation becomes very slow 170 and ineffective. For instance, it took nearly three months to obtain only a couple of crystals of 2 using 171 11.5 equivalents of ClO4 . Physical characterisation 119 172 173 Description of structures 174 The crystal structures of these compounds consist of two cationic complexes perchlorate anions and 175 RESULTS AND DISCUSSION 154 155 Synthesis 156 Both tetranuclear compounds 1 and 2 were obtained from the reaction of comproportionation between 157 MnII and MnO4 in the presence of a substituted derivative of benzoic acid and 1,10-phenanthroline 158 (phen), leading to compounds with general formula [Mn4(m-O)2(m-4- 159 RC6H4COO)7 m(L)2m(phen)2](ClO4)1+m, where R = MeO (1) or tBu (2), L is a monodentate ligand, 160 and m = 0 or 1. 161 The mother liquors of 1 and 2 were left to slowly evaporate, but no solid was obtained from either of 162 them, indicating that both compounds are highly soluble in acetonitrile. Thus, they needed to be 163 crystallised by mixing the mother solution with absolute ethanol (1) or by adding a huge excess of ClO4 164 anions (2). For compound 1, crystallisation from different CH3CN: EtOH mixtures was tried, but no 165 differences were observed changing the CH3CN/EtOH volume ratio (from 0.3 to 0.7). In all cases, the 166 reaction yield was about 25% considering the stoichiometry of the reaction above. In contrast, we had 167 some difficulties finding the optimum conditions to obtain compound 2 in appreciable yield. Its 168 crystallisation needed to be assisted by the addition of about 22 equivalents of ClO4 anions, leading 169 to 2 in 16% yield. If the amount of ClO4 anions used is decreased, crystallisation becomes very slow 170 and ineffective. For instance, it took nearly three months to obtain only a couple of crystals of 2 using 171 11.5 equivalents of ClO4 . 172 173 Description of structures 174 The crystal structures of these compounds consist of two cationic complexes, perchlorate anions and 175 molecules of solvent. Simplifications of these structures are represented in Fig. 2, respectively. Fig. S1 176 (ESI†) contains the fully labelled crystal structures of 1 and 2. The most relevant structural parameters 177 are listed in Tables S2 and S3 (ESI†). It is worth remembering that the crystal structure of 2 could not be 178 fully refined, and just approximate values of some structural parameters are provided. 179 Both compounds contain a [MnIII 4 (m3-O)2]8+ core comprising two central (Mnc) and two terminal 180 (Mnt) ions. There is an Mnc2Ob2 rhombus to which the Mnt ions are attached – one to each m3- O2 181 ligand. Physical characterisation 119 The Mnc Mnc distances are B2.85 (1) or B2.88 (2) Å, and the Mnc–Ob–Mnc angles are in 182 the 96–991 range. There are two types of bridges that alternately bind the Mnc and Mnt ions, those 183 having a triple (m-O)(m-RCOO)2 or a double (m-O)(m- RCOO) bridge. The Mnc Mnt distances 184 where Mn ions are linked by a triple bridge (B3.25–3.31 Å) are shorter than those of the double bridge 185 (B3.35–3.41 Å). Similarly, the Mnc–Ob–Mnt angles corresponding to the triple bridges (B120–1261) 186 are smaller than those corresponding to the double bridges (B126–1321). All these structural parameters 187 are in accord with those of compounds with a [Mn4O2]6+/7+/8+ core and carboxylate bridges.2,3,5– 188 17,24 189 Description of structures 174 The crystal structures of these compounds consist of two cationic complexes, perchlorate anions and 175 molecules of solvent. Simplifications of these structures are represented in Fig. 2, respectively. Fig. S1 176 (ESI†) contains the fully labelled crystal structures of 1 and 2. The most relevant structural parameters 177 are listed in Tables S2 and S3 (ESI†). It is worth remembering that the crystal structure of 2 could not be 178 fully refined, and just approximate values of some structural parameters are provided. 179 The crystal structures of these compounds consist of two cationic complexes, perchlorate anions and 175 molecules of solvent. Simplifications of these structures are represented in Fig. 2, respectively. Fig. S1 176 (ESI†) contains the fully labelled crystal structures of 1 and 2. The most relevant structural parameters 177 are listed in Tables S2 and S3 (ESI†). It is worth remembering that the crystal structure of 2 could not be 178 fully refined, and just approximate values of some structural parameters are provided. 179 Both compounds contain a [MnIII 4 (m3-O)2]8+ core comprising two central (Mnc) and two terminal 180 (Mnt) ions. There is an Mnc2Ob2 rhombus to which the Mnt ions are attached – one to each m3- O2 181 ligand. The Mnc Mnc distances are B2.85 (1) or B2.88 (2) Å, and the Mnc–Ob–Mnc angles are in 182 the 96–991 range. There are two types of bridges that alternately bind the Mnc and Mnt ions, those 183 having a triple (m-O)(m-RCOO)2 or a double (m-O)(m- RCOO) bridge. The Mnc Mnt distances 184 where Mn ions are linked by a triple bridge (B3.25–3.31 Å) are shorter than those of the double bridge 185 (B3.35–3.41 Å). Similarly, the Mnc–Ob–Mnt angles corresponding to the triple bridges (B120–1261) 186 are smaller than those corresponding to the double bridges (B126–1321). All these structural parameters 187 are in accord with those of compounds with a [Mn4O2]6+/7+/8+ core and carboxylate bridges.2,3,5– 188 17,24 189 In both crystal structures there are six carboxylate ligands linking Mnc with the Mnt ions. Four of these 190 ligands are approximately in the same plane of the central Mnc2O2 rhombus, whereas the other ones are 191 perpendicular to this rhombus. Furthermore, compound 1 has an additional carboxylate bridge that links 192 the two Mnc ions. Description of structures 174 In contrast, these seventh positions in 2 are instead occupied by monodentate ligands 193 that lie on the opposite sides of the Mnc2Ob2 rhombus. Then, while the cationic complexes of 2 display 194 different monodentate ligands, one having two molecules of CH3CN and the second one of H2O, those 195 of 1 are just conformational isomers. 196 The seventh carboxylate ligand in 1 causes some other differences in the structural parameters of the 197 metallic core. For instance, while the Mnc2Ob2 rhombus is completely planar in 2, that of 1 is slightly 198 twisted, with an Mnc–Ob–Ob–Mnc angle of B1681. Concerning the Mn4 arrangements, in 2 the four 199 Mn ions are in the same plane, one of the Mnt ions being below the Mnc2Ob2 rhombus and the other 200 one above. On the other hand, both Mnt ions in 1 are placed above the central rhombus, resulting in a 201 butterfly-like arrangement (Fig. 1). 202 All Mn atoms in compounds 1 and 2 have an elongated pseudo-octahedral geometry, displaying Jahn– 203 Teller distortion, as expected for MnIII ions. Jahn–Teller axes in Mnc ions are almost parallel, whereas 204 they are nearly perpendicular to those of Mnt ions (Fig. 3). Considering the z axes in the direction of the 205 Jahn–Teller axes and unfairly assigning the x and y axes, approximate values of lengths of the 206 octahedron axes can be found with the addition of the Mn–ligand distances (see Table S4, ESI†). To 207 quantify the distortions of the coordination octahedra, the procedure described in our previous work was 208 followed,25 whose results are listed in Table S4 (ESI†). All Mn ions display an elongated distortion (z 209 axes are longer than x and y axes) and an almost inappreciable rhombic distortion (x and y axes are very 210 similar in length). The Mnc ions (with D = 10–18%) aremuchmore elongated than the Mnt ones (with D 211 = 7.8–11%). 212 Magnetic properties 214 5): the magnetic 226 interaction between the Mn ions bound with a double oxo bridge, J1 (Mnc Mnc); two 227 centralterminal Mn ion interactions (Mnc Mnt), J2 and J3; and the magnetic interactions between 228 terminal Mn ions, J4 (Mnt Mnt). The Heisenberg spin Hamiltonian (H) considered is 229 230 (1) 231 232 (1) considering that DMn improved the fits of the experimental data substantially the entire curves were 244 reproduced much better. The rhombic zero-field splitting parameter was considered negligible (EMn = 245 0), in accord with the low rhombic distortion of the MnIII ions of these compounds (ro2%, see Table S4, 246 ESI†). The relative orientation of the Jahn–Teller axes is an important factor that should be taken into 247 account when the axial anisotropy of the Mn ions is appreciable. Indeed, we previously reported that this 248 parameter may have an effect on the overall magnetic properties,25 especially in the low temperature 249 range. In the tetranuclear compounds herein described, the Jahn–Teller axes are almost parallel between 250 the central Mn ions and nearly orthogonal between the central and terminal ions (Fig. 3). 251 Hence, the experimental data were fitted (300–2 K) using the PHI program27 and taking into account 252 the axial zero-field splitting (DMn) and the relative orientation of the Jahn–Teller axes of the MnIII 253 ions. The best fits correspond to g = 2.01, 2J1 = 45.5 cm 1, 2J2 = 15.1 cm 1, 2J3 = 4.4 cm 1, 254 and DMn = 3.5 cm 1 with R = 7.9 10 5 for 1; and to g = 2.01, 2 J1 = 43.0 cm 1, 2J2 = 14.7 255 cm 1, 2J3 = 8.2 cm 1, and DMn = 3.6 cm 1 with R = 6.1 10 5 for 2. Magnetic properties 214 The inclusion 239 of the DMn parameter, but still keeping J2 = J3, did not provide better fits. Therefore, the fits were 240 performed considering J2 a J3, which gave more appropriate reproductions of the experimental curves. 241 However, the shape of the maxima in the wM versus T plots was not completely reproduced in these 242 latter fits indicating that the inclusion of ZFS parameters could provide better results. Indeed, 243 considering that DMn improved the fits of the experimental data substantially the entire curves were 244 reproduced much better. The rhombic zero-field splitting parameter was considered negligible (EMn = 245 0), in accord with the low rhombic distortion of the MnIII ions of these compounds (ro2%, see Table S4, 246 ESI†). The relative orientation of the Jahn–Teller axes is an important factor that should be taken into 247 account when the axial anisotropy of the Mn ions is appreciable. Indeed, we previously reported that this 248 parameter may have an effect on the overall magnetic properties,25 especially in the low temperature 249 range. In the tetranuclear compounds herein described, the Jahn–Teller axes are almost parallel between 250 the central Mn ions and nearly orthogonal between the central and terminal ions (Fig. 3). 251 Hence, the experimental data were fitted (300–2 K) using the PHI program27 and taking into account 252 the axial zero-field splitting (DMn) and the relative orientation of the Jahn–Teller axes of the MnIII 253 ions. The best fits correspond to g = 2.01, 2J1 = 45.5 cm 1, 2J2 = 15.1 cm 1, 2J3 = 4.4 cm 1, 254 and DMn = 3.5 cm 1 with R = 7.9 10 5 for 1; and to g = 2.01, 2 J1 = 43.0 cm 1, 2J2 = 14.7 255 cm 1, 2J3 = 8.2 cm 1, and DMn = 3.6 cm 1 with R = 6.1 10 5 for 2. The DMn values are 256 consistent with elongated MnIII ions, which is expected to show moderate and negative DMn.28–31 257 Even though the ground states of these compounds have S = 0 the zero field splitting of the first excited 258 Four different Mn Mn exchange pathways may be considered (shown in Fig. Magnetic properties 214 Magnetic susceptibility (wM) data were recorded for compounds 1 and 2 from 300 to 2 K. wMT versus 215 T and wM versus T plots for 1 and 2 are shown in Fig. 4. Note that the molecular weight of 2 was 216 referred to one Mn4 unit, considering an average formula between the two entities. The wMT value at 217 room temperature is 8.3 cm3 mol K, which is quite below the expected value for four MnIII ions (12.0 218 cm3 mol K). wMT decreases with temperature almost linearly until B90 K and below this temperature it 219 decreases drastically to B0.4 cm3 mol K at 2 K, indicating a strong antiferromagnetic behaviour. wM 220 versus T plots for both compounds show nearly superimposable graphs between 300 and 80 K, but they 221 differ below this temperature. While the wM versus T plot of compound 1 displays a maximum at 9 K 222 (wM E 0.14 cm3 mol), the one of 2 is situated at 6 K (wM E 0.27 cm3 mol). This difference can also be 223 seen in the wMT versus T plots. The presence of these maxima suggests that both compounds have a 224 ground state with S = 0. 225 Four different Mn Mn exchange pathways may be considered (shown in Fig. 5): the magnetic 226 interaction between the Mn ions bound with a double oxo bridge, J1 (Mnc Mnc); two 227 centralterminal Mn ion interactions (Mnc Mnt), J2 and J3; and the magnetic interactions between 228 terminal Mn ions, J4 (Mnt Mnt). The Heisenberg spin Hamiltonian (H) considered is 229 230 (1) 231 232 where S1 = S2 = S3 = S4 = 2. Several fits of the experimental data were performed, screening different 233 values of the magnetic coupling constants (results shown in Fig. S2, ESI†). In all of them, the interaction 234 between terminal Mn ions was considered negligible (J4 = 0). In fact, this interaction, with a Mn 235 Mn distance of B5.6 Å, is expected to be of comparable magnitude to intermolecular interactions. 236 Firstly, the experimental data were fitted considering J2 = J3 to get an average value of the central- 237 terminal interactions, as usually performed for this kind of compound with a [Mn4O2]m+ core (m = 6– 238 8);3,5,8,17,24,26 but the shape of the curve could not be reproduced with any of these fits. Magnetic properties 214 We also tried to assign J2 and J3 by 280 comparing the structural parameters of other compounds with a [MnIII 4 O2]8+ core and carboxylate 281 bridges (Tables S6 and S7, ESI†), but no clear assignment could be done. 282 The strongest magnetic coupling constant ( J1) is unambiguously assigned to the double-oxo bridge, 263 since a strong antiferromagnetic interaction is expected for this subunit and it is consistent with that 264 observed in analogous compounds (Table S5, ESI†). The assignment of J2 and J3 to the triple (m-O)(m- 265 RCOO)2 or double (m-O)(m-RCOO) bridges appears much more challenging; for analogous systems 266 they are indeed considered too similar to be differentiated, obtaining just an average value. However, its 267 distinction was necessary in order to achieve a good fit of the experimental data for 1 and 2 (Fig. S3, 268 ESI†). Aiming to their assignment, we compared the structural parameters of these subunits and 269 dinuclear MnIII compounds with (m-O)- (m-R0COO) or (m-O)(m-R0COO)2 bridges. Only two 270 examples of dinuclear MnIII compounds with a double (m-O)(m-R0COO) bridge were found, in which 271 the magnetic coupling constants are quite different (2J = 19.5 and +1.33 cm 1 for H = 272 2JS1S2).32,33 However, both compounds display compressed MnIII octahedral (D E 10%, r E 1.5– 273 6%), contrary to 1 and 2, making them non-comparable. The magnetic properties of dinuclear MnIII 274 compounds with a triple (m-O)(m-R0COO)2 bridge have been extensively studied in our precedent 275 study.25 The value of the magnetic interaction of this triple-bridged subunit may be approximately 276 predicted using the magneto-structural correlation presented therein. However, these subunits in 277 compounds 1 and 2, with D E 7.8–18% and r o 4%, are very different from those in the dinuclear MnIII 278 compounds, which always show a significant degree of rhombicity (r = 3.5–5.4%). Hence, the 279 assignment of these bridging blocks to certain values is very risky. We also tried to assign J2 and J3 by 280 comparing the structural parameters of other compounds with a [MnIII 4 O2]8+ core and carboxylate 281 bridges (Tables S6 and S7, ESI†), but no clear assignment could be done. 282 283 Magnetic properties 214 The DMn values are 256 consistent with elongated MnIII ions, which is expected to show moderate and negative DMn.28–31 257 Even though the ground states of these compounds have S = 0, the zero-field splitting of the first excited 258 states may be of importance for the wM versus T plot when these states are populated at low 259 temperature, as we reported previously.25 Accordingly, the energy level distribution calculated for these 260 compounds with the parameters obtained from the fits (omitting DMn) revealed that the ground states 261 have S = 0 and that there are several low-lying excited states with S a 0 (see below). 262 The strongest magnetic coupling constant ( J1) is unambiguously assigned to the double-oxo bridge, 263 since a strong antiferromagnetic interaction is expected for this subunit and it is consistent with that 264 observed in analogous compounds (Table S5, ESI†). The assignment of J2 and J3 to the triple (m-O)(m- 265 RCOO)2 or double (m-O)(m-RCOO) bridges appears much more challenging; for analogous systems 266 they are indeed considered too similar to be differentiated, obtaining just an average value. However, its 267 distinction was necessary in order to achieve a good fit of the experimental data for 1 and 2 (Fig. S3, 268 ESI†). Aiming to their assignment, we compared the structural parameters of these subunits and 269 dinuclear MnIII compounds with (m-O)- (m-R0COO) or (m-O)(m-R0COO)2 bridges. Only two 270 examples of dinuclear MnIII compounds with a double (m-O)(m-R0COO) bridge were found, in which 271 the magnetic coupling constants are quite different (2J = 19.5 and +1.33 cm 1 for H = 272 2JS1S2).32,33 However, both compounds display compressed MnIII octahedral (D E 10%, r E 1.5– 273 6%), contrary to 1 and 2, making them non-comparable. The magnetic properties of dinuclear MnIII 274 compounds with a triple (m-O)(m-R0COO)2 bridge have been extensively studied in our precedent 275 study.25 The value of the magnetic interaction of this triple-bridged subunit may be approximately 276 predicted using the magneto-structural correlation presented therein. However, these subunits in 277 compounds 1 and 2, with D E 7.8–18% and r o 4%, are very different from those in the dinuclear MnIII 278 compounds, which always show a significant degree of rhombicity (r = 3.5–5.4%). Hence, the 279 assignment of these bridging blocks to certain values is very risky. Spin states distribution 284 Then, the distinction between these two magnetic interactions would 306 be highly likely unachievable. 307 However, this degeneration breaks and all states increase in energy as the J3/J2 ratio decreases. For 300 J3/J2 values between B0.6–0.7, the lowest spin states will be mixed and spin frustration is then 301 plausible. When J3/J2 is below 0.6, the spin ground state is ST = 0. Then, compounds 1 and 2, with 302 respective J3/J2 ratios of B0.3 andB0.6, have an ST = 0 ground state. Having this ground state is a key 303 point for the distinction between J2 and J3. Indeed, if the J3/J2 ratio was 40.8, the corresponding system 304 would display an ST = 3 ground state and wMT values would not approach zero at low temperature, as 305 similarly observed when J3 = J2. Then, the distinction between these two magnetic interactions would 306 be highly likely unachievable. 307 The separation between the ground and first excited states in these compounds is rather different: for 2 308 the three first excited states (with ST = 1, 2, and 3) are at the most at 10 cm 1 above the ground state, 309 whereas they are much more separated for 1 (at 3.5– 27 cm 1 above the ground state). As J1 and J2 310 values are very similar for both compounds, the cause of these different separations must rely on J3, the 311 value of that of 2 is twice the one of 1. This difference is also responsible for the different degree of 312 interaction observed in the wMT versus T and wM versus T plots, where the stronger interaction of 1 is 313 confirmed. Nonetheless, it is surprising that the compound showing a smaller \|J3\| value has a stronger 314 interaction. This fact can be also explained with the energy distribution of the excited states: the ST = 0 315 ground state in 1 is far more isolated than that in 2, maximizing the decrease of the wMT values upon 316 cooling. 317 A deeper analysis of the energy levels as a function of the J3/J2 ratio may provide useful information 318 concerning the spin configuration of the states. However, a non-intuitive and complex result was 319 obtained from this analysis. Each one of these states, commonly named eigenstates, is the result of the 320 combination of several configurations of single-ion spin moments. Spin states distribution 284 The magnetic properties of compounds with a similar structure to those of 1 and 2 were reported in the 285 literature.3,5,6,17,24 As commented above, J2 and J3 were not distinguished and an average value of 286 the Mnt Mnc interaction (Jct) was provided. These compounds usually show different spin ground 287 states depending on the J1/Jct ratio. The most common ground state is (ST, Scc, Stt) = (3, 1, 4) for J1/Jct 288 = 2.5–4.9.3,6,8,17 When J1/Jct44.9, then five energetically degenerate states, (n, 0, n) with n = 0–4, 289 become the ground state,5,9,24 corresponding to two noninteracting Mnt ions.3 In the lower limit, when 290 J1/Jct o 2.5 the ground state would be (2, 2, 4). In contrast, the compounds (1 and 2) presented herein 291 display an isolated (0, 0, 0) ground state. The explanation of this fact lies in the differentiation of the two 292 types of interactions between the terminal and central ions, J2 and J3 (following the diagram shown in 293 Fig. 5). It is important to remember that there are two different Mnc Mnt magnetic interaction 294 pathways: those consisting of a double (m-O)(m-R0COO) bridge and those having a triple (m-O)(m- 295 R0COO)2 bridge. Fig. 6 shows energy of the first spin states versus J3/J2 plots for hypothetical 296 compounds with a [MnIII 4 O2]8+ core, all energies being referred to the lower ST = 0 state. When J3 297 and J2 are equal (J3/J2 = 1), the spin ground state has ST = 3 and there are several states with ST = 0–4 298 that are energetically degenerated. 299 However, this degeneration breaks and all states increase in energy as the J3/J2 ratio decreases. For 300 J3/J2 values between B0.6–0.7, the lowest spin states will be mixed and spin frustration is then 301 plausible. When J3/J2 is below 0.6, the spin ground state is ST = 0. Then, compounds 1 and 2, with 302 respective J3/J2 ratios of B0.3 andB0.6, have an ST = 0 ground state. Having this ground state is a key 303 point for the distinction between J2 and J3. Indeed, if the J3/J2 ratio was 40.8, the corresponding system 304 would display an ST = 3 ground state and wMT values would not approach zero at low temperature, as 305 similarly observed when J3 = J2. Spin states distribution 284 These configurations are known as 321 basic elements. Moreover, the composition of these eigenstates changes upon decreasing the J3/J2 ratio. 322 Then, the assignment of the states to a particular configuration is unachievable. As an example, Fig. 6 323 also shows the percentage composition (calculated with the PHI27 program) of the three most relevant 324 (ST, Scc, Stt) basic elements in which the Hamiltonian is constructed for the ST = 0 and ST = 3 325 eigenstates that are lowest in energy. These eigenstates were not chosen arbitrarily, since the most 326 common ground states are those having ST = 3 (for J3/J2 4 0.7) and ST = 0 (for J3/J2 o 0.6) according 327 to the diagram of energy shown in Fig. 6. The spin configurations that represent these basic elements are 328 represented in Fig. 7. 329 As may be observed, when J3/J2 = 1, the ST = 0 and the ST = 3 states mainly correspond to spin 330 configurations with (ST, Scc, Stt) = (0, 0, 0) and (3, 1, 4), respectively, with contributions of at least 331 B80%. However, the composition of these states changes upon decreasing the J3/J2 ratio and the 332 assignment to a single spin configuration becomes unachievable. This fact is also observed for the rest 333 of the eigenstates included in Fig. 6 except for the ST = 4 state, whose highest contribution is never 334 lower than 70% and corresponds to the (4, 0, 4) configuration. 335 336 CONCLUSIONS 337 338 The reaction between Mn(ClO4)2 and Bu4NMnO4 in the presence of benzoic acid derivatives 4- 339 MeOC6H4COOH (1) or 4-tBuC6H4- COOH (2) and 1,10-phenantroline (phen) led to the formation of 340 two new tetranuclear compounds with a [MnIII 4 O2]8+ core, which comprises a central Mn2O2 341 rhombus to which two terminal Mn ions are attached – one to each m3- O2 ligand. The crystal 342 structures of these compounds revealed that the Mn ions are arranged in a butterfly (1) or in a Mn4 343 planar (2) fashion. There are two types of magnetic interactions between central and terminal ions: those 344 consisting of a double (m-O)(m-R0COO) bridge and those having a triple (m-O)(m-R0COO)2 bridge. 345 The MnIII ions in these compounds display a significant elongated distortion along the Jahn–Teller axes 346 and a negligible rhombic distortion. Spin states distribution 284 Moreover, the Jahn–Teller axes in the central Mn ions are almost 347 parallel, while they are nearly perpendicular to those of the terminal Mn ions. 348 The magnetic measurements revealed that these compounds show an ST = 0 ground state. There are 349 three different magnetic interaction pathways: one between central ions ( J1) and two between central 350 and terminal ions ( J2 and J3). The distinction between J2 and J3 was crucial to obtain a good fit of the 351 experimental data. However, the assignment of J2 and J3 to a particular bridging block was not 352 achieved. The inclusion of the axial anisotropy parameter (DMn) and the consideration of the relative 353 orientation of the Jahn–Teller axes led tomuch better fits. A deep analysis of the energy levels as a 354 function of the J3/J2 ratio (from 1.0 to 0.0) provided very useful information. The distribution of energy 355 levels changes completely with the J3/J2 ratio, having an ST = 0 ground state when J3/J2o0.6. 356 Nevertheless, the assignment of the states to a particular configuration was unachievable because the 357 composition of these states changes upon decreasing the J3/J2 ratio. 358 359 CONCLUSIONS 337 338 The reaction between Mn(ClO4)2 and Bu4NMnO4 in the presence of benzoic acid derivatives 4- 339 MeOC6H4COOH (1) or 4-tBuC6H4- COOH (2) and 1,10-phenantroline (phen) led to the formation of 340 two new tetranuclear compounds with a [MnIII 4 O2]8+ core, which comprises a central Mn2O2 341 rhombus to which two terminal Mn ions are attached – one to each m3- O2 ligand. The crystal 342 structures of these compounds revealed that the Mn ions are arranged in a butterfly (1) or in a Mn4 343 planar (2) fashion. There are two types of magnetic interactions between central and terminal ions: those 344 consisting of a double (m-O)(m-R0COO) bridge and those having a triple (m-O)(m-R0COO)2 bridge. 345 The MnIII ions in these compounds display a significant elongated distortion along the Jahn–Teller axes 346 and a negligible rhombic distortion. Moreover, the Jahn–Teller axes in the central Mn ions are almost 347 parallel, while they are nearly perpendicular to those of the terminal Mn ions. Spin states distribution 284 348 CONCLUSIONS 337 338 The reaction between Mn(ClO4)2 and Bu4NMnO4 in the presence of benzoic acid derivatives 4- 339 MeOC6H4COOH (1) or 4-tBuC6H4- COOH (2) and 1,10-phenantroline (phen) led to the formation of 340 two new tetranuclear compounds with a [MnIII 4 O2]8+ core, which comprises a central Mn2O2 341 rhombus to which two terminal Mn ions are attached – one to each m3- O2 ligand. The crystal 342 structures of these compounds revealed that the Mn ions are arranged in a butterfly (1) or in a Mn4 343 planar (2) fashion. There are two types of magnetic interactions between central and terminal ions: those 344 consisting of a double (m-O)(m-R0COO) bridge and those having a triple (m-O)(m-R0COO)2 bridge. 345 The MnIII ions in these compounds display a significant elongated distortion along the Jahn–Teller axes 346 and a negligible rhombic distortion. Moreover, the Jahn–Teller axes in the central Mn ions are almost 347 parallel, while they are nearly perpendicular to those of the terminal Mn ions. 348 The magnetic measurements revealed that these compounds show an ST = 0 ground state. There are 349 three different magnetic interaction pathways: one between central ions ( J1) and two between central 350 and terminal ions ( J2 and J3). The distinction between J2 and J3 was crucial to obtain a good fit of the 351 experimental data. However, the assignment of J2 and J3 to a particular bridging block was not 352 achieved. The inclusion of the axial anisotropy parameter (DMn) and the consideration of the relative 353 orientation of the Jahn–Teller axes led tomuch better fits. A deep analysis of the energy levels as a 354 function of the J3/J2 ratio (from 1.0 to 0.0) provided very useful information. The distribution of energy 355 levels changes completely with the J3/J2 ratio, having an ST = 0 ground state when J3/J2o0.6. 356 Nevertheless, the assignment of the states to a particular configuration was unachievable because the 357 composition of these states changes upon decreasing the J3/J2 ratio. 358 Nevertheless, the assignment of the states to a particular configuration was unachievable because the 357 composition of these states changes upon decreasing the J3/J2 ratio. 358 360 ACKNOWLEDGEMENTS 361 362 This work was supported by the Ministerio de Ciencia e Innovacio´n of Spain (project CTQ2012-30662 363 and CTQ2015-63614-P). L. E. thanks the University of Barcelona for the APIF fellowship. Spin states distribution 284 364 365 This work was supported by the Ministerio de Ciencia e Innovacio´n of Spain (project CTQ2012-30662 363 and CTQ2015-63614-P). L. E. thanks the University of Barcelona for the APIF fellowship. 364 1 S.Mukhopadhyay, S.K.Mandal, S. Bhaduri andW. H. Armstrong, Chem. Rev., 2004, 104, 3981– 368 4026. 369 2 B. Albela, M. S. El Fallah, J. Ribas, K. Folting, G. Christou and D. N. Hendrickson, Inorg. 370 Chem., 2001, 40, 1037–1044. 371 3 J. B. Vincent, C. Christmas, H. R. Chang, Q. Li, P. D. W. Boyd, J. C. Huffman, D. N. 372 J. B. Vincent, C. Christmas, H. R. Chang, Q. Li, P. D. W. Boyd, J. C. Huffman, D. N. Hendrickson and G. Christou, J. Am. Chem. Soc., 1989, 111, 2086–2097. 373 G. Christou, D. Gatteschi, D. N. Hendrickson and R. Sessoli, MRS Bull., 2000, 25, 66–71. 4 G. Christou, D. Gatteschi, D. N. Hendrickson and R. Sessoli, MRS Bull., 2 374 5 C. Can˜ada-Vilalta, J. C. Huffman and G. Christou, Polyhedron, 2001, 20, 375 C. Can˜ada-Vilalta, J. C. Huffman and G. Christou, Polyhedron, 2001, 20, 1785–1793. 6 M. J. Knapp, D. N. Hendrickson, V. A. Grillo, J. C. Bollinger and G. Christou, Angew. Chem., 376 Int. Ed. Engl., 1996, 35, 1818–1820. 377 Z. Chen, Y. Ma, F. Liang and Z. Zhou, J. Organomet. Chem., 2008, 693, 646–654. 8 Y.-S. Ma, H.-C. Yao, W.-J. Hua, S.-H. Li, Y.-Z. Li and L.-M. Zheng, Inorg. Chim. Acta, 2007, 379 360, 1645–1650. 380 T. Taguchi, M. R. Daniels, K. A. Abboud and G. Christou, Inorg. Chem., 2009, 48, 9325–933 10 R. Ruiz, C. Sangregorio, A. Caneschi, P. Rossi, A. B. Gaspar, J. A. Real and M. C. Mun˜oz, 382 Inorg. Chem. Commun., 2000, 3, 361–367. 383 J. Pistilli andR. H. Beer, Inorg. Chem. Commun., 2002, 5, 206–210. 12 T. Tamane, T. 11 J. Pistilli andR. H. Beer, Inorg. Chem. Commun., 2002, 5, 206–210. 12 T. Tam 384 Tsubomura and K. Sakai, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1996, 52, 777– 385 779. 386 13 S. Wang, M. S. Wemple, J. Yoo, K. Folting, J. C. Huffman, K. S. Hagen, D. N. Hendrickson 387 and G. Christou, Inorg. Chem., 2000, 39, 1501–1513. 388 g, p , , g, , g , and G. Christou, Inorg. Chem., 2000, 39, 1501–1513. 388 14 G. Aromı´, S. Bhaduri, P. Artu´s, K. Folting and G. Christou, Inorg. Chem., 2002, 41, 805–817. 389 15 N. E. Spin states distribution 284 Chakov, L. N. Zakharov, A. L. Rheingold, K. A. Abboud and G. Christou, Inorg. Chem., 390 2005, 44, 4555–4567. 391 16 M. Hołyn´ska and S. Dehnen, Z. Kristallogr. - Cryst. Mater. 2011, 226, 769–772. 392 17 E. Libby, J. K. McCusker, E. A. Schmitt, K. Folting, D. N. Hendrickson and G. Christou, Inorg. 393 Chem., 1991, 30, 3486–3495. 394 18 T. Sala and M. V. Sargent, J. Chem. Soc., Chem. Commun., 1978, 253–254. 395 19 Bruker, SADABS, version 2008/1, Bruker AXS Inc., Madison, Wisconsin, US 396 20 G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 3–8. 397 21 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838. 398 21 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838. 398 L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849–854. 22 L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849–854. 399 23 C. B. Hu¨bschle, G. M. Sheldrick and B. Dittrich, J. Appl. Crystallogr., 2011, 44, 1281–1284. 400 24 G. Aromı´, S. Bhaduri, P. Artu´s, J.C.Huffman,D. N. Hendrickson and G. Christou, Polyhedron, 401 2002, 21, 1779–1786. 402 25 L. Escriche-Tur, M. Font-Bardia, B. Albela and M. Corbella, Dalton Trans., 2016, 45, 11753– 403 11764. 404 26 U. Bossek, H. Hummel, T. Weyhermu¨ller, K. Wieghardt, S. Russell, L. van der Wolf and U. 405 Kolb, Angew. Chem., Int. Ed. Engl., 1996, 35, 1552–1554. 406 27 N. F. Chilton, R. P. Anderson, L. D. Turner, A. Soncini and K. S. Murray, J. Comput. Chem., 407 2013, 34, 1164–1175. 408 28 R. Bocˇa, Coord. Chem. Rev., 2004, 248, 757–815. 29 S. Gomez-Coca, E. Cremades, N. 409 R. Bocˇa, Coord. Chem. Rev., 2004, 248, 757–815. 29 S. Gomez-Coca, E. Cremades, N. 28 R. Boc a, Coord. Chem. Rev., 2004, 248, 757–815. 29 S. Gomez-Coca, E. Cremades, N. 409 Aliaga-Alcalde and E. Ruiz, J. Am. Chem. Soc., 2013, 135, 7010–7018. 410 Aliaga-Alcalde and E. Ruiz, J. Am. Chem. Soc., 2013, 135, 7010–7018. 410 30 C. Duboc, D. Ganyushin, K. Sivalingam, M.-N. Collomb and F. Neese, J. Phys. Chem. A, 2010, 411 114, 10750–10758. 412 31 G. Aromı´, J. Telser, A. Ozarowski, L.-C. Brunel, H.-M. Stoeckli-Evans and J. Krzystek, Inorg. 413 Chem., 2005, 44, 187–196. 414 32 K. J. Oberhausen, R. J. O’Brien, J. F. Richardson, R. M. Buchanan, R. Costa, J. M. Latour, H. L. 415 Tsai and D. N. Hendrickson, Inorg. Chem., 1993, 32, 4561–4565. 416 33 N. Arulsamy, J. Glerup, A. Spin states distribution 284 Hazell, D. J. Hodgson, C. J. McKenzie and H. Toftlund, Inorg. 417 Chem., 1994, 33, 3023–3025. 418 419 Legends to figures 420 421 Figure. 1 Schematic representation of two arrangements for the [Mn4(m3-O)2] core. Label code: Mnc = 422 central Mn ion, Mnt = terminal Mn ion, and 423 Ob = bridging oxygen atom. 424 425 Figure. 2 Crystal structures of the cationic complexes of 1 (only one of the subunits) and 2. The 4- 426 RC6H4COO groups have been omitted for better clarity. Colour code: MnIII, brown; C, grey; N, blue; 427 O, red. 428 429 Figure. 3 Schematic representation of the cationic complexes of 1 and 2. Green bold lines correspond to 430 the Jahn–Teller axes. L could be H2O or CH3CN. 431 432 Figure. 4 wMT versus T and wM versus T (inset) plots for compounds 1 (red) and 2 (blue). The solid 433 lines correspond to the best fit to the experimental data. The molecular weight of 2 was referred to one 434 Mn4 unit, considering an average formula between the two entities 435 436 Figure 5. Schematic representation of the possible Mn Mn exchange pathways in compounds 1 437 and 2. 438 439 Figure 6 (a) Energy at zero field for the first spin states as a function of J3/J2 ratio for a hypothetical 440 [MnIII 4O2]8+ compound with 2J1 = 45.6 cm 1, J2 = 15 cm 1, DMn = 0, and variable J3 values. 441 The coloured lines correspond to the most relevant states. (b) Percentage composition of the three most 442 relevant (ST, Scc, Stt) basic element in which the Hamiltonian is contracted for the eigenstates with ST 443 = 0 and ST = 3 that are lowest in energy. J3/J2E0.3 for 1 and 0.6 for 2. 444 445 Figure 7 Possible spin configuration representing the most relevant (ST, Scc, Stt) spin states that 446 configure the two lowest eigenstates with ST = 0 and ST = 3. 447 448 449 Figure 7 Possible spin configuration representing the most relevant (ST, Scc, Stt) spin states that 446 configure the two lowest eigenstates with ST = 0 and ST = 3. Spin states distribution 284 447 449 FIGURE 1 450 451 452 453 FIGURE 1 450 451 452 453 FIGURE 1 450 451 452 453 FIGURE 450 451 FIGURE 1 451 452 453 FIGURE 2 FIGURE 2 456 FIGURE 3 FIGURE 3 FIGURE 3 459 460 459 460 460 461 FIGURE 4 FIGURE 4 464 465 466 465 466 467 468 66 467 468 FIGURE 5 469 470 471 472 FIGURE 5 FIGURE 5 FIGURE 5 469 470 471 71 72 472 473 FIGURE 6 474 475 476 6 477 477 FIGURE 7 478 479 480 481 482 483 482 483
https://openalex.org/W2160373801	https://bmcpediatr.biomedcentral.com/counter/pdf/10.1186/1471-2431-13-146	English	null	Validation study of the Chinese Early Development Instrument (CEDI)	BMC pediatrics	2,013	cc-by	6,380	© 2013 Ip et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: The Early Development Instrument (EDI) is a comprehensive instrument used to assess school readiness in preschool children. This study was carried out to evaluate the psychometric properties of the Chinese version of the EDI (CEDI) in Hong Kong. Methods: One hundred and sixty-seven children were purposefully sampled from kindergartens in two districts with very different socioeconomic statuses. The CEDI was assessed for concurrent validity, internal consistency and test-retest reliability. The developmental vulnerability identified using the CEDI scores was further examined in relation to the socioeconomic status of the district and family. Results: The CEDI displayed adequate internal consistency, with Cronbach’s alphas ranging from 0.70 to 0.95 on its five domains. Concurrent validity was supported by moderate and significant correlations (0.25 to 0.49) on the relevant domains between the CEDI and a comparable measure. The level of test-retest reliability was good, with a kappa statistic of 0.89. In general, girls outperformed boys, particularly in the social, emotional and communication/ general knowledge domains. After controlling for the uneven distribution of sex, children from socioeconomically disadvantaged districts and families were found to be at greater risk of developmental vulnerability than their more advantaged counterparts. Conclusion: The evidence gathered in this study supports the CEDI’s use as a valid and reliable instrument in assessing school readiness and identifying developmentally vulnerable children in Chinese populations. Its preliminary findings on the socioeconomic gradients of child development suggest that the CEDI is a promising tool for leveraging evidence-based, context-sensitive policies and practices to foster the development of all children. Keywords: Early Development Instrument, Early child development, Validity, Chinese population, Socioeconomic gradient Keywords: Early Development Instrument, Early child development, Validity, Chinese population, Socioeconomic gradient Validation study of the Chinese Early Development Instrument (CEDI) Patrick Ip1, Sophia Ling Li1, Nirmala Rao2, Sharon Sui Ngan Ng3, Winnie Wai Sim Lau1 and Chun Bong Chow1 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Open Access Correspondence: patricip@hku.hk 1Department of Paediatrics and Adolescent Medicine, Queen Mary Hospital, The University of Hong Kong, Pok Fu Lam, Hong Kong, China Full list of author information is available at the end of the article Participants and procedures In 2011, four Chinese-speaking kindergartens were ran- domly selected from Hong Kong Island (HKI) and Yuen Long District (YL), two major administrative districts in Hong Kong with dramatically different economic levels. HKI is a wealthy district with median monthly family income of around US$4240, which is 33.2% higher than the population average, whereas the corresponding figure for YL is around US$2680, 15.7% below the popu- lation average [16]. Ethical approval for this study was granted by the ethics committee of the University of Hong Kong. Consistent with the EDI, the CEDI is also made up of 103 items assessing five developmental domains: a) physical health and wellbeing, b) social competence, c) emotional maturity, d) language and cognitive develop- ment, and e) communication skills and general know- ledge. The domain scores range from 0 to 10, and the items reflect children’s developmental milestones rather than specific curriculum goals. Children who score in the bottom 10th percentile in at least one of the five do- mains are deemed vulnerable in terms of school readi- ness, indicating that the problems they have within a given developmental area are likely to interfere with their success in school. The most recent evidence from the longitudinal study in Australia suggested that the vulnerability at school entry predicts the literacy and nu- meracy outcomes throughout primary school years [18]. All four kindergartens contacted agreed to join the study. With the approval of their principals, all third- year kindergarten (K3) children and their teachers and parents were invited to participate. In total, 175 children were contacted, and 167 K3 children were assessed with both the Chinese Early Development Instrument (CEDI) and the Hong Kong Early Child Development Scale (HKECDS). Informed written consent was obtained from the parents of all participating children. Of these chil- dren, 15 from each district were then re-assessed with the CEDI by the same teacher four weeks later to evalu- ate the instrument’s test-retest reliability. The teacher who was most familiar with each child was invited to rate him or her with the CEDI. To minimize measure- ment errors introduced by different raters, all of the teachers were trained beforehand in the instrument’s use. This training took the form of two three-hour work- shops with home exercises assigned in between. Measures and variables Chinese early development instrument (CEDI) Chinese early development instrument (CEDI) The CEDI was translated from English into Traditional Chinese with the permission of the EDI authors [4] using the back-translation method to translate and adapt the assessment instrument in a trans-cultural context [17]. A bilingual local university faculty member special- izing in early childhood education translated the original English-language version into traditional Chinese. An- other bilingual faculty member from the same depart- ment then translated it back into English separately. Local experts in child development, including university faculty, pediatricians, kindergarten teachers and educa- tion experts, were consulted on the local relevance of the instrument’s items, as well as the appropriateness and accuracy of their wording. Three items referring to English letters within the language and cognitive devel- opment domain required further modification to fit the context of the learning and use of Chinese characters. The finalized CEDI was then sent to the EDI authors at the Offord Centre for Child Studies (in Hamilton, ON, Canada) for review, and their approval was subsequently obtained. The aim of this study was to examine the internal consistency, concurrent validity and test-retest reliability of the Chinese Early Development Instrument (CEDI). The CEDI data were also analyzed in relation to socio- economic indicators to explore the potential existence of socioeconomic disparities in child development among preschoolers in a Chinese community. Participants and procedures The teachers were given a Chinese version of the CEDI teacher’s guide, which is a comprehensive and user- friendly reference book on the instrument’s use, coding Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 2 of 8 and interpretations of items in the local context. The HKECDS results were assessed by a separate team of trained research assistants with no knowledge of the CEDI results, and the children’s parents were asked to complete a family questionnaire (FQ). The completed CEDI, HKECDS and FQ were collected by the research team. and there is also evidence of its predictive validity for primary school performance when administered during kindergarten [9]. Although the EDI is reliable at the individual level, one of its strengths is to allow the aggregation of indi- vidual data to the group or community level, thus permitting examination of the role of socioeconomic in- equalities in child development from multiple perspec- tives [10-13]. Mapping the socioeconomic inequality patterns in early child development can aid communities and countries in forming universal and targeted policies to improve outcomes for all children [14,15]. Background assesses school readiness. It covers five major develop- mental domains, including physical health and well- being, social competence, emotional maturity, language and cognitive development, and communication skills and general knowledge. Early childhood development is the foundation of hu- man and community development [1]. The early years of life are a critical developmental period for both resili- ence and vulnerability [2]. School readiness among pre- school children has become an important concern for educators, academics and policy-makers [3]. Rather than focus on standard test scores and cognitive capabilities, the Early Development Instrument (EDI), which was developed in Canada by Janus and Offord in 2007 [4], is a comprehensive teacher-completed instrument that Research shows the EDI to be a valid, reliable and stable measure [5-7], and to have small to moderate levels of association with other teacher-reported mea- sures. Studies show its internal consistency to be high, ranging from 0.84 to 0.96, and inter-rater reliability to be satisfactory, ranging from 0.53 to 0.80. Janus et al. (2007) reported the test-retest correlation of the EDI ad- ministered twice to the same group of children within a reasonable period of time to be high (0.82 to 0.94) [8], * Correspondence: patricip@hku.hk 1Department of Paediatrics and Adolescent Medicine, Queen Mary Hospital, The University of Hong Kong, Pok Fu Lam, Hong Kong, China Full list of author information is available at the end of the article Family questionnaire (FQ) Information on the socioeconomic background of the participating children was obtained from their parents using the FQ, a self-developed, pre-tested questionnaire. Maternal education was measured with a single item on a scale ranging from 1 to 7, with higher scores representing higher education levels. In analysis of this study, maternal education was divided into three categories: junior second- ary education and below was defined as “low”, senior sec- ondary education to an associate degree as “medium” and a Bachelor’s degree and above as “high”. Family income was measured with one item soliciting total monthly family in- come on a scale ranging from 1 to 10 (from < HK$4000 to > HK$80,000) (US$1 ≈HK$7.8). With reference to Hong Kong’s family income distribution in 2011 [16], family in- come was further categorized into four context-meaningful groups: < $8000 was deemed the lowest 10th percentile of family income distribution; $8000 ~ < $20,000 was below the population median ($22,000); 20,000 ~ < 80,000 cov- ered the population median and the majority of the top half; and >= $80,000 was the highest 10th percentile. Characteristics of subjects Of the 167 children who participated in the study, seven were excluded from analysis, four of them because of a special needs designation and three because of missing data on one or more domains. In view of the wide age range of the remaining 160 children (5.43 to 7.31 years), we further restrained our analysis to children born in 2005, which resulted in 151 children in the same age co- hort. Table 1 summarizes the subjects’ characteristics. Sixty-six (43.7%) children were from HKI (the wealthy district) and 85 (56.3%) from YL (the poor district). Be- cause the children in the HKI kindergartens were pre- dominately female, the sex distribution of our sample is severely imbalanced, with two-thirds of the subjects girls. The samples from the two districts also differed significantly in terms of the socioeconomic indicators of maternal education and family income. Hong Kong early child development scale (HKECDS) The HKECDS is a direct assessment of child develop- ment (at 3–6 years) that was developed in Hong Kong and shown to display satisfactory psychometric qualities and excellent cultural and contextual appropriateness [19]. The scale contains 95 items in eight domains: a) Page 3 of 8 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 personal, social and self-care; b) language development; c) pre-academic learning; d) cognitive development; e) gross motor; f) fine motor; g) physical fitness, health and safety (knowledge about); and h) self and society. Compared to the CEDI domain structure, the HKECDS places greater emphasis on knowledge and learning and less on social and emotional assessment. Therefore, we expected the conceptually comparable domains between the two measures (CEDI with HKECDS) to be: a) the physical and well-being domain with gross and fine motor ability; b) language and cognitive development with language, pre-academic learning, and cognitive development; and c) communication skills and general knowledge with language and cognitive development and self and society. However, none of the HKEDCS domains specifically matches the social and emotional domain of the CEDI. The concurrent validity of the CEDI was assessed by its correlations with the HKECDS. using the kappa statistic (k). The relationship between child development vulnerabilities and socioeconomic indicators (district, family income and maternal educa- tion) was measured by the adjusted odds ratios from logistic regressions after controlling for sex. Statistical analysis was performed using SPSS (version 17), and p<0.05 was considered statistically significant. Internal consistency Cronbach’s α, a measure of internal consistency, ranges from 0.70 to 0.95 for the five CEDI domains (Table 2). Concurrent validity Based on the partial correlations with sex controlled, Table 3 highlights the two strongest correlations with the HKECDS for each CEDI domain. As expected, the Table 1 Major socioeconomic characteristics of subjects N (%) HKI (%) YL (%) p-value Male 52 (34.4) 10 (15.2) 42 (49.4) Females 99 (65.6) 56 (84.8) 43 (50.6) < 0.001 Maternal Education^ Low 13 (8.8) 0 13 (15.9) < 0.001 Medium 95 (64.2) 29 (43.9) 66 (80.5) High 40 (27.0) 37 (56.1) 3 (3.7) Family Income (HK$) < 0.001 < 8000 18 (12.2) 0 18 (22.2) 8000 ~ < 20,000 49 (33.3) 2 (3.0) 47 (58.0) 20,000 ~< 80,000 60 (40.8) 45 (68.2) 15 (18.5) >= 80,000 > 80,000 20 (13.6) 19 (28.8) 1 (1.2) Note: ^ Maternal education was categorized into three levels: low = junior secondary education and below; medium = senior secondary education to associate degree; and high = Bachelor’s degree and above. The unequal sample size is due to missing data. Table 1 Major socioeconomic characteristics of subjects Note: ^ Maternal education was categorized into three levels: low = junior secondary education and below; medium = senior secondary education to associate degree; and high = Bachelor’s degree and above. The unequal sample size is due to missing data. Relationship with socioeconomic status of district and family District Comparison of the socioeconomic status of the two communities in which the participating kindergartens were located showed a significantly higher proportion of children from the socioeconomically disadvantaged dis- trict, YL (42.4%), to display developmental vulnerability in at least one of the CEDI domains relative to their HKI counterparts (16.3%). After adjusting for the uneven distribution of sex in our sample, the excessive risk of vulnerability for the YL children still remained signifi- cant (aOR = 4.46, 95% CI: 1.74-11.41; p < 0.005). physical health and well-being domain correlates best with gross and fine motor, language and cognitive development with pre-academic learning and language development, and communication and general know- ledge with language and cognitive development. Because no HKECDS domain specifically measures social and emo- tional development, the social competence and emotional maturity domains of the CEDI were found to correlate best with gross motor and language development. Data analysis f h Because of the uneven sex distribution between the sam- pled districts, with many more girls in the HKI sample than the YL sample, statistical adjustment was adopted in the following analyses wherever appropriate. Concurrent valid- ity was assessed using the partial correlations between the CEDI and HKECDS domain scores, with sex controlled. Because the two instruments differed in their conceptual structure of child development measurement, the two best correlation coefficients were highlighted in the correl- ation matrix. Internal consistency was calculated using Cronbach’s α for each of the five CEDI domains. The test-retest reliability of the two scales was determined Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 4 of 8 Table 2 Summary of domain scores and internal consistency with Cronbach’s α Domain Items Mean sd Min Max Cronbach’s α 1. Physical health and well-being 13 8.77 1.19 3.46 10.00 0.70 2. Social competence 26 8.04 1.71 2.69 10.00 0.95 3. Emotional maturity 30 7.91 1.33 3.67 10.00 0.91 4. Language and cognitive development 26 8.97 1.52 3.20 10.00 0.90 5. Communication skills and general knowledge 8 8.07 2.07 1.88 10.00 0.91 Table 2 Summary of domain scores and internal consistency with Cronbach’s α and communication/general knowledge domains [8]. As shown in Table 4, 28.5% of the children in our study were found to be developmentally vulnerable in at least one CEDI domain, and 13.9% in at least two. Further, significantly more boys than girls (46.2% boys versus 19.2% girls) were identified as vulnerable (p < 0.05) in at least one developmental domain. Family income I i i Investigation of the relationship between a vulnerable classification in one or more developmental domains and family income revealed a decreasing gradient (see Table 5 and Figure 1), indicating that children from poorer families are at greater risk of developmental vul- nerability than those from relatively wealthy families. After taking the uneven sex distribution into account, the gradient trend between vulnerability and family in- come remained, as shown in the decreasing adjusted odds ratio with increasing family income in Table 5, although the relationship was no longer statistically significant because of the reduced sample size. Reliability The test-retest reliability of the CEDI after a four-week interval was analyzed in 30 participants using the kappa statistic (k). The kappa coefficient was 0.89, thus demon- strating the instrument’s stability over time. Vulnerability Th ff The cut-offs for vulnerability derived from our sample in Hong Kong are largely comparable with the Canadian normative references in the physical, social and emo- tional domains, but higher in the language/cognitive Table 3 Partial correlations between CEDI and HKECDS domain scores, with sex as the control variable CEDI P S E L/C C/G HKECDS Gross Motor .30 .41* .31*** .13 .26* Fine Motor .25 .32* .11 .25** .23* Language Development .16 .35* .32* .47* .37* Pre-academic Learning .22* .22* .30* .49* .19* Cognitive Development .17 .27 .27 .39* .28 Personal, Social, Self-care Environment .19* .19* .20* .32*** .21* Self and Society .21* .32* .26 .43* .27 Physical Fitness, Health and Safety (Knowledge about…) .09 .20* .24 .41* .23 Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge; p < .001, p < .01 and p < .05. Table 3 Partial correlations between CEDI and HKECDS d i ith th t l i bl Table 3 Partial correlations between CEDI and HKECDS domain scores, with sex as the control variable Table 3 Partial correlations between CEDI and HKECDS domain scores, with sex as the control variable CEDI P S E L/C C/G HKECDS Maternal education Across all of the CEDI domains, a decreasing gradient can be seen in the mean of the domain scores with ma- ternal education level (see Table 6). The lowest mean scores were found in the group of children whose mothers had a junior secondary level of education or less, whereas the highest scores were found in the group whose mothers held a Bachelor’s degree or higher aca- demic qualification. A similar decreasing gradient with maternal education level was also found in the propor- tion of children identified as vulnerable in one or more developmental domains (see Table 6 and Figure 2). After controlling for the effect of sex, the gradient trend be- tween vulnerability and maternal education remained significant (p<0.05), as illustrated in the decreasing ad- justed odds ratios with higher maternal education shown in Table 6. Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge; *p < .001, p < .01 and p < .05. Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 5 of 8 Page 5 of 8 Table 4 Cut-offs for vulnerability and distribution by sex CEDI P S E L/C C/G Total Cut-offs HK 6.9567 5.3846 6.0000 6.4400 5.0000 Canada 7.0833 5.5769 6.0000 5.7692 4.3750 Vulnerability N (%) 16 (10.6) 15 (9.9) 16 (10.6) 16 (10.6) 23 (15.2) 43 (28.5) Male 9 (17.3) 12 (23.1) 10 (19.2) 8 (15.4) 14 (26.9) 24 (46.2) Female 7 (7.1) 3 (3.0)** 6 (6.1)* 8 (8.1) 9 (9.1)* 19 (19.2)* Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge; *p < .001, p < .01 and p < .05. Table 4 Cut-offs for vulnerability and distribution by sex Development Scale (HKECDS) [19], a direct assess- ment of early childhood development developed locally in Hong Kong. After controlling for sex, strong and significant correlations remained between the CEDI and HKECDS in the expected domains. The correl- ation coefficients (0.25 to 0.49) were comparable to those reported between the EDI and direct child-based assessment, which ranged from 0.34 to 0.49 [21]. Maternal education The moderate correlations in the current study were expected because the comparison was between a teacher evaluation (CEDI) and direct child-based assess- ment (HKECDS) across a wide range of differently catego- rized domains. Stronger correlations have been reported in studies comparing the EDI with other teacher-rated mea- sures [21], although such comparisons are often subject to the problem of shared method variance [22]. y g; p ; ; g g d C/G=Communication and General Knowledge; #odds ratio estimated in the logistic regression with sex adjusted. te: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and d C/G C i i d G l K l d # dd i i d i h l i i i i h dj d Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = and C/G=Communication and General Knowledge; #odds ratio estimated in the logistic regression with sex adjusted. Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G=Communication and General Knowledge; #odds ratio estimated in the logistic regression with sex adjusted. Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C and C/G=Communication and General Knowledge; #odds ratio estimated in the logistic regression with sex adjusted. Note: Low = a maternal education level of junior secondary and below; medium = senior secondary education or a higher certificate or diploma; and high = a Bachelor’s degree and above. For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge. #Odds ratio estimated in the logistic regression with sex included; p < .05, p < .01 and p < .005. Discussion Figure 1 Developmental vulnerability versus family income. and vice versa [23,24]. Similarly, children experiencing delayed language development are likely to find it more difficult to acquire appropriate social and emotional skills [25,26]. relationship between socioeconomic disadvantage and developmental vulnerability revealed the children from an underprivileged district (Yuen Long [YL]) and family (as measured by family income, and maternal education) to be at greater risk of vulnerability in one or more developmental domains. Observations with the CEDI in the Hong Kong Chinese popula- tion are consistent with EDI observations in Western societies [11,13]. In this study, vulnerability was defined according to our Hong Kong sample rather than using Canadian normative data. Although doing so undoubtedly intro- duced bias, given the small sample size and non- representative sampling structure, the value of using cut-offs from a local sample is that there are recog- nized differences between the Canadian and Hong Kong Chinese populations with regard to the cultural and developmental context of preschool children, including societal expectations, parenting and the kindergarten environment. Further examination of the Discussion This study examined the internal consistency, concur- rent validity and reliability of the Chinese Early Develop- ment Instrument (CEDI), which was adapted from the EDI [4]. CEDI is a population tool to assess children’s development at aggregate level and it is not mean to assess children’s school readiness at the individual level. The preliminary evidence obtained therein supports the CEDI’s use as a valid and reliable measure of early child development and school readiness in Chinese populations. The internal consistency of the five CEDI domains ranged from 0.70 to 0.95, which is comparable with that of the EDI domains [4]. As a Cronbach’s α ranging be- tween 0.70 to 0.90 is generally considered good [20], we can conclude that the CEDI domains demonstrate an adequate level of internal consistency. The test-retest re- liability of the CEDI was also found to be good (0.89). These two psychometrical properties of the CEDI are largely comparable with those of the EDI when used in English-speaking countries [7,8]. In addition, although lacking pre-specified correlates, the social and emotional domains of the CEDI were found to correlate best with the gross motor and lan- guage domains of the HKECDS. Children with advanced motor development may display more constructive en- gagement in early activities, and thus have a better chance of acquiring key social and emotional abilities, The CEDI’s concurrent validity was established through comparison with the Hong Kong Early Child Table 5 Descriptive statistics for the CEDI domain scores and vulnerability by family income level Family income CEDI domain scores – mean (sd) Vulnerability aOR# P S E L/C C/G N (%) [95% CI] < 8000 8.57 7.74 7.61 8.27 7.43 8 4.88 (1.21) (1.73) (1.30) (1.88) (2.77) (44.4) [.82, 28.89] 8000 ~ < 20,000 8.36 7.80 7.68 8.75 8.10 16 2.36 (1.41) (1.72) (1.10) (1.42) (2.21) (32.6) [.45, 12.35] 20,000 ~ < 80,000 9.09 8.17 7.95 9.18 8.46 14 1.98 (0.97) (1.72) (1.52) (1.43) (1.81) (23.3) [.39, 9.34] >= 80,000 8.73 8.56 8.41 9.62 7.47 2 1 (1.11) (1.54) (1.14) (0.85) (1.62) (10.0) Note: For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G=Communication and General Knowledge; #odds ratio estimated in the logistic regression with sex adjusted. Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 6 of 8 Figure 1 Developmental vulnerability versus family income. Limitations This study suffered several limitations. First, its main limitation lies in recruitment of the sample. Though kindergartens were randomly selected from Hong Kong Island (HKI) and Yuen Long (YL), the HKI Table 6 Descriptive statistics for CEDI domain scores and vulnerability by maternal education level Maternal education CEDI domain score – mean (sd) Vulnerability aOR# P S E L/C C/G N (%) [95% CI] Low 8.38 (1.10) 7.88 7.65 8.35 7.88 18 5.39 (1.72) (1.19) (1.98) (2.46) (37.5) [1.14, 25.46] Medium 8.85 8.10 7.91 9.18 8.11 23 3.41* (1.26) (1.75) (1.40) (1.22) (1.91) (25.8) [1.09, 10.73] High 9.29 8.18 8.66 9.58 8.58 0 1 (0.89) (1.47) (0.80) (0.73) (1.48) (0) Note: Low = a maternal education level of junior secondary and below; medium = senior secondary education or a higher certificate or diploma; and high = a Bachelor’s degree and above. For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge. #Odds ratio estimated in the logistic regression with sex included; p < .05, p < .01 and *p < 005 tive statistics for CEDI domain scores and vulnerability by maternal education level Note: Low = a maternal education level of junior secondary and below; medium = senior secondary education or a higher certificate or diploma; and high = a Bachelor’s degree and above. For the CEDI domains: P = Physical Health and Well-being; S = Social Competence; E = Emotional Maturation; L/C = Language and Cognitive Development; and C/G = Communication and General Knowledge. #Odds ratio estimated in the logistic regression with sex included; p < .05, p < .01 and *p < .005. Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 7 of 8 Page 7 of 8 Figure 2 Developmental vulnerability versus maternal education level. Figure 2 Developmental vulnerability versus maternal education level. in 23 countries, this validation study opens up the exciting possibility of placing Chinese children’s devel- opment on an international scale for comparison. sample included significantly more girls than boys, and the reverse was true for the YL sample, because of the natural sex composition of the kindergartens recruited. Competing interests The authors declare that they have no competing interests. Competing interests The authors declare that they have no competing interests. Acknowledgements We thank the developer of the EDI, Dr. Magdalena Janus and the Offord Centre for Child Studies at McMaster University for allowing us to use the EDI in this study, all of the parents and lovely children who participated in this study, as well as the principals and teachers of the four participating kindergartens, and Dr. R. Christopher Sheldrick for his advice on the manuscript. Authors’ contributions PI designed the study, interpreted the data and wrote the manuscript. SLL analyzed the data and drafted the manuscript. NR participated in preparation of assessment tools and interpretation of data. SSNN participated in training of teachers and preparation of assessment tools. WWSL participated in training of teachers and data collection. CBC participated in research design and data interpretation. All authors read and approved the final manuscript. Disclosure of funding h k d b d The work described in this paper was fully supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. 746111). Limitations Although this severe sex imbalance may not have posed serious harm in testing the psychometric prop- erties of the CEDI, as the EDI factor structure has been reported stable between boys and girls [4], quantitative interpretations of the coefficients should be made with caution. To account for the compound effect between the imbalanced sex distribution and the difference in the socioeconomic status of the two districts, analyses between the CEDI and other factors were statistically adjusted for sex. Second, the EDI is intended for use and interpretation at the group level, whereas the current validation of the CEDI was conducted at the individual level. Third, because of the relatively small sample size, confirmatory factor analysis was not conducted in this study. Fourth, the CEDI question- naire was completed by the kindergarten teacher who was most familiar with each tested child, and we did not repeat the test with a different teacher; therefore, inter-rater reliability was not assessed. Competing interests The authors declare that they have no competing interests. Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 Page 8 of 8 Ip et al. BMC Pediatrics 2013, 13:146 http://www.biomedcentral.com/1471-2431/13/146 25. Cohen NJ, Menna R, Vallance DD, Barwick MA, Im N, Horodezky NB: Language, social cognitive processing, and behavioral characteristics of psychiatrically disturbed children with previously identified and unsuspected language impairments. J Child Psychol Psyc 1998, 39(6):853–864. 26. Beck L, Kumschick IR, Eid M, Klann-Delius G: Relationship between language competence and emotional competence in middle childhood. Emotion 2012, 12(3):503–514. doi:10.1186/1471-2431-13-146 Cite this article as: Ip et al.: Validation study of the Chinese Early Development Instrument (CEDI). BMC Pediatrics 2013 13:146. 25. Cohen NJ, Menna R, Vallance DD, Barwick MA, Im N, Horodezky NB: Language, social cognitive processing, and behavioral characteristics of psychiatrically disturbed children with previously identified and unsuspected language impairments. J Child Psychol Psyc 1998, 39(6):853–864. References 1. Shonkoff JP, Phillips DA: From neurons to neighborhoods: the science of early childhood development. Washington, DC, USA: National Academies Press; 2000. 1. Shonkoff JP, Phillips DA: From neurons to neighborhoods: the science of early childhood development. Washington, DC, USA: National Academies Press; 2000. 2. Fergusson DM, Horwood LJ: Resilience to childhood adversity: results of a 21-year study. In Resilience and vulnerability : adaptation in the context of childhood adversities. Edited by Luthar SS. Cambridge: Cambridge University Press; 2003:130–156. 2. Fergusson DM, Horwood LJ: Resilience to childhood adversity: results of a 21-year study. In Resilience and vulnerability : adaptation in the context of childhood adversities. Edited by Luthar SS. Cambridge: Cambridge University Press; 2003:130–156. 26. Beck L, Kumschick IR, Eid M, Klann-Delius G: Relationship between language competence and emotional competence in middle childhood. Emotion 2012, 12(3):503–514. 26. Beck L, Kumschick IR, Eid M, Klann-Delius G: Relationship between language competence and emotional competence in middle childhood. Emotion 2012, 12(3):503–514. 3. Blair C: School readiness - integrating cognition and emotion in a neurobiological conceptualization of children’s functioning at school entry. Am Psychol 2002, 57(2):111–127. 3. Blair C: School readiness - integrating cognition and emotion in a neurobiological conceptualization of children’s functioning at school entry. Am Psychol 2002, 57(2):111–127. doi:10.1186/1471-2431-13-146 Cite this article as: Ip et al.: Validation study of the Chinese Early Development Instrument (CEDI). BMC Pediatrics 2013 13:146. 4. Janus M, Offord DR: Development and psychometric properties of the early development instrument (EDI): a measure of children’s school readiness. Can J Behav Sci 2007, 39(1):1–22. 4. Janus M, Offord DR: Development and psychometric properties of the early development instrument (EDI): a measure of children’s school readiness. Can J Behav Sci 2007, 39(1):1–22. 5. Guhn M, Goelman H: Bioecological theory, early child development and the validation of the population-level early development instrument. Soc Indic Res 2011, 103:193–217. 6. Forer B, Zumbo BD: Validation of multilevel constructs: validation methods and empirical findings for the EDI. Soc Indic Res 2011, 103(2):231–265. 6. Forer B, Zumbo BD: Validation of multilevel constructs: validation methods and empirical findings for the EDI. Soc Indic Res 2011, 103(2):231–265. 7. Janus M, Brinkman SA, Duku EK: Validity and psychometric properties of the early development instrument in Canada, Australia, United States, and Jamaica. Soc Indic Res 2011, 103(2):283–297. 7. Janus M, Brinkman SA, Duku EK: Validity and psychometric properties of the early development instrument in Canada, Australia, United States, and Jamaica. References Soc Indic Res 2011, 103(2):283–297. 8. Janus M, Brinkman S, Duku E, Hertzman C, Santos R, Sayers M, Schroeder J, Walsh C: The early development instrument: a population-based measure for communities (a handbook on development, properties, and use). Ontario, CA: Offord Centre for Child Studies; 2007. 8. Janus M, Brinkman S, Duku E, Hertzman C, Santos R, Sayers M, Schroeder J, Walsh C: The early development instrument: a population-based measure for communities (a handbook on development, properties, and use). Ontario, CA: Offord Centre for Child Studies; 2007. 9. Lloyd JEV, Hertzman C: From kindergarten readiness to fourth-grade assessment: longitudinal analysis with linked population data. Soc Sci Med 2009, 68(1):111–123. 10. Cushon JA, Vu LTH, Janzen BL, Muhajarine N: Neighborhood poverty impacts children’s physical health and well-being over time: evidence from the early development instrument. Early Educ Dev 2011, 22(2):183–205. 11. Hertzman C, Bertrand J: Children in poverty and the use of early development instrument mapping to improve their worlds. Paediatr Child Health 2007, 12(8):687–692. 12. Kohen D, Oliver L, Pierre F: Examining the effects of schools and neighbourhoods on the outcomes of kindergarten children in Canada. Int J Speech-Lang Pa 2009, 11(5):404–418. Int J Speech-Lang Pa 2009, 11(5):404–418. 13. Carpiano RM, Lloyd JE, Hertzman C: Concentrated affluence, concentrated disadvantage, and children’s readiness for school: a population-based, multi-level investigation. Soc Sci Med 2009, 69(3):420–432. 14. Noble KG, McCandliss BD, Farah MJ: Socioeconomic gradients predict individual differences in neurocognitive abilities. Dev Sci 2007, 10(4):464–480. 15. Bradley RH, Corwyn RF: Socioeconomic status and child development. Annu Rev Psychol 2002, 53:371–399. 16. Hong Kong population census. [http://www.census2011.gov.hk] 17. Hambleton RK: Issues, designs, and technical guidelines for adapting tests into multiple languages and cultures. In Adapting educational and psychological tests for cross-cultural assessment. Edited by Hambleton RK, Merenda PF, Spielberger C. London: L.E.A; 2005:3–38. 18. Brinkman S, Gregory T, Harris J, Hart B, Blackmore S, Janus M: Associations between the early development instrument at age 5, and reading and numeracy skills at ages 8, 10 and 12: a prospective linked data study. Child Indic Res. published online on 14 April 2013. DOI 10.1007/s12187-013-9189-3. 19. Rao N, Sun J, Ng SSN, Ma K, Becher Y, Lee D, Lau C, Zhang L, Chow CB, Ip P: The Hong Kong early child development scale: a validation study. Child Indic Res 2013, 6(1):115–135. Author details 1 1Department of Paediatrics and Adolescent Medicine, Queen Mary Hospital, The University of Hong Kong, Pok Fu Lam, Hong Kong, China. 2Faculty of Education, The University of Hong Kong, Pok Fu Lam, Hong Kong, China. 3Department of Early Childhood Education, Hong Kong Institute of Education, Tai Po, Hong Kong, China. In conclusion, the evidence presented herein shows the CEDI to be a psychometrically sound measurement tool for early child development and the assessment of school readiness in Chinese populations. As the EDI has gained significant international popularity in the past decade, with successful adaptation and application Received: 17 July 2013 Accepted: 20 September 2013 Published: 23 September 2013 References Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit 20. Streiner DL, Norman GR: Health measurement scales: a practical guide to their development and use. Oxford, UK: University Press; 1995. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: Submit your next manuscript to BioMed Central and take full advantage of: 21. Brinkman SA, Silburn S, Lawrence D, Goldfeld S, Sayers M: Investigating the validity of the Australian early development index. Early Educ Dev 2007, 18(3):427–451. • Convenient online submission 22. Donaldson SI, Grant-Vallone EJ: Understanding self-report bias in organizational behavior research. J Bus Psychol 2002, 17(2):245–260. 23. Wilson A, Piek JP, Kane R: The mediating role of social skills in the relationship between motor ability and internalizing symptoms in pre- primary children. Infant Child Dev 2013, 22(2):151–164. 24. Whittingham K, Fahey M, Rawicki B, Boyd R: The relationship between motor abilities and early social development in a preschool cohort of children with cerebral palsy. Res Dev Disabil 2010, 31(6):1346–1351.
https://openalex.org/W4328130045	https://ars.copernicus.org/articles/20/93/2023/ars-20-93-2023.pdf	English	null	Towards Steering Magnetic Nanoparticles in Drug Targeting Using a Linear Halbach Array	Advances in radio science	2,023	cc-by	8,976	1 Introduction Abstract. Magnetic nanoparticles offer numerous promising biomedical applications, e.g. magnetic drug targeting. Here, magnetic drug carriers inside the human body are directed towards tumorous tissue by an external magnetic ﬁeld. How- ever, the success of the treatment strongly depends on the amount of drug carriers, reaching the desired tumor region. This steering process is still an open research topic. In this paper, the previous study of a linear Halbach array is ex- tended by an additional Halbach array with different mag- netization angles between two adjacent magnets and inves- tigated numerically using COMSOL Multiphysics. The Hal- bach arrays are arranged with permanent magnets and gen- erate a relatively large region of a moderately homogeneous, high magnetic ﬁeld while having a strong gradient. This re- sults in a strong magnetic force, trapping many particles at the magnets. Afterwards, to avoid particle agglomeration, the Halbach array is ﬂipped to its weak side. Therefore, the magnetic ﬂux density, its gradient and the resulting magnetic force are computed for the different Halbach arrays with dif- ferent constellations of magnetization directions. Since the calculation of the gradient can lead to high errors due to the used mesh in Comsol, the gradient was derived analyt- ically by investigating two different ﬁtting functions. Over- all, the array with a 90◦shifted magnetization performs best, changing the magnetic sides of the array easily and deﬂecting more particles. Besides, the results revealed that the magnetic force dominates directly underneath the magnets compared to the other existing forces on the SPIONS. Summarized, the results depict that the magnetic force and, thus, the region where the particles are able to get washed out, can be ad- justed using low-cost permanent magnets. In the last decades, the interest in steering superparamagnetic iron-oxide nanoparticles (SPIONs) to a desired location, par- ticularly in the context of biomedical engineering, has grown rapidly (Nacev et al., 2012). SPIONs are supporting a great variety of diagnostic methods, e.g. as a contrast agent in mag- netic resonance imaging (MRI; Vogel et al., 2020) or ultra- sound (Fink et al., 2021), as well as therapeutic issues like in magnetic drug targeting (MDT; Alexiou et al., 2006) or hyperthermia (Mues et al., 2021). In MDT for example, the chemotherapeutic agents are bound to the SPIONs and in- jected into the cardiovascular system. Due to their magnetic properties, SPIONs can be pulled through the vessels into the tumorous tissue. Towards Steering Magnetic Nanoparticles in Drug Targeting Using a Linear Halbach Array Angelika S. Thalmayer1,z, Samuel Zeising1, Maximilian Lübke1, and Georg Fischer1 1Institute for Electronics Engineering, Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), 91058 Erlangen, Germany zFor this work, Angelika S. Thalmayer received the Young Scientist Award from the German URSI member committee at the Kleinheubacher Tagung 2021. orrespondence: Angelika S. Thalmayer (angelika.thalmayer@fau.de) Correspondence: Angelika S. Thalmayer (angelika.thalmayer@fau.de) Received: 31 January 2022 – Accepted: 8 July 2022 – Published: 21 March 2023 Received: 31 January 2022 – Accepted: 8 July 2022 – Published: 21 March 2023 1 Introduction This enables a local cancer treatment in a desired target volume, while at the same time side effects for the patient are reduced (Tietze et al., 2013). The resulting high efﬁciency of a MDT treatment was proven in an animal study by Tietze et al. (2013). Nevertheless, the success of a MDT treatment strongly de- pends on the accumulation and therefore, on the ability to steer the SPIONs to the tumorous tissue. However, the re- quired guiding is a complex problem, since it depends on various multiphysical parameters like the velocity proﬁle of the blood ﬂow, the gradient of the applied magnetic ﬁeld as well as the properties of the nanoparticles themselves. The SPIONs should have a diameter of < 200 nm to be able to penetrate into tissue and get not destroyed by the body’s im- mune system (Zaloga et al., 2014). Since the magnetic force on the particles is proportional to the volume of one SPION, the forces are very weak due to their small size (Nacev et al., 2012). Furthermore, the particles need a coating with lipidic Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 © Author(s) 2022. This work is distributed under the Creative Commons Attribution 4.0 License. 2.1 Linear Halbach array conﬁguration In the literature, Halbach arrays are primarily utilized for electromotors (Zhang et al., 2011). However, in those se- tups, the Halbach array is usually coaxially arranged like in Shen and Zhu (2013). For separating particles, there is few research and also mostly in a coaxial arrangement (Blümler, 2021). However, Shiriny and Bayareh (2020), Stevens et al. (2021), Ijiri et al. (2013), and Kang et al. (2016) exemplary utilized a linear Halbach array for separating magnetic par- ticles or cells. The results showed a stronger gradient and magnetic force compared to an array where the magnets were arranged with an alternating magnetization direction. How- ever, their arrays were static, whereas in the proposed study, a linear Halbach array was utilized as an adjustable magnetic ﬁeld source. Thus, the single magnets of the Halbach array are arranged to have a 90◦shifted magnetization compared to their direct neighbors. This leads to a one-sided magnetic ﬁeld (see Fig. 1). By rotating the single magnets by π/2 as depicted in Fig. 1, the strong and the weak side of the ar- ray can be changed. For the proposed conﬁguration, Hilton and McMurry (2012) evaluated the torque N = m×B (Jack- son, 1998) for the proposed conﬁguration. They ﬁgured out that for a practical realization a mechanical stabilization is mandatory. In practice, however, it is not always possible to place the magnets surrounding the vessel. Thus, Thalmayer et al. (2021) introduced a conﬁguration with a linear adjustable Halbach array (proposed by Hilton and McMurry, 2012), which is placed parallel to the vessel, for steering SPIONs through an Y-shaped bifurcation. The Halbach array contains a strong and a weak magnetic side, which can be switched by rotating the single magnets. This leads to an adjustable magnetic force. The basic idea is, having at ﬁrst the strong magnetic side, and, thus, the strong magnetic force facing the vessel in order to pull as many particles as possible to- wards the magnets. Afterwards, the magnets are rotated so that the weak side points towards the vessel and, thus, the trapped particles are washed out by the ﬂuid. In this work, the results of the previous conference pa- per (Thalmayer et al., 2021) are extended and an additional Halbach conﬁguration is investigated and compared to the primary adjustable Halbach array. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 94 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Figure 1. Magnetization pattern and ﬁeld distribution of the ad- justable Halbach array. The magnetic ﬁeld can be shifted from be- low to above the magnets by changing ϕ from 0 to π/2 (Reprinted with permission from Thalmayer et al., 2021; © IEEE). acids like lauric acid to be biocompatible and not agglom- erating in blood (Zaloga et al., 2014), which decreases the possible size of the magnetic part inside the particles even more. The main parameter for guiding a particle swarm through the cardiovascular system is the applied magnetic ﬁeld. Here, the distance from the magnet between the SPIONs plays a crucial role, since the generated magnetic ﬁeld decays ap- proximately with 1/R3 and the magnetic force on one SPION depends on the magnitude and the gradient of the mag- netic ﬁeld. This leads to a trade-off between the magnetic ﬁeld chosen too weak to attract enough SPIONs or one too strong, resulting in a scene, where most of the SPIONs get stuck inside the vessel underneath the magnet (Thalmayer et al., 2020). To overcome this problem, Park et al. (2020), Hoshiar et al. (2017), Kim et al. (2021) and many other au- thors proposed setups containing two or more electromag- nets (EMs), which were placed at the opposite side of a ves- sel and switched on and off alternately. Nevertheless, EMs require a power supply in kW range, induce heat and have a weaker magnetic ﬁeld (especially for the same volume) com- pared to rare earth permanent magnets (e.g. NdFeB; Bjørk et al., 2010, and Alnaimat et al., 2020). In contrast, perma- nent magnets yield the disadvantage that their magnetic ﬁeld is constant and cannot be regulated or switched on and off. However, by using arrays of permanent magnets, the mag- netic ﬂux density can be adjusted by mechanical operations, like rotating the magnets (Bjørk et al., 2010). In the litera- ture, Halbach conﬁgurations are well-established for this pur- pose, since their pattern produces the strongest magnetic ﬂux density (Sakuma and Nakagawara, 2021). Baun and Blüm- ler (2017) developed a coaxial arrangement of two Halbach arrays. It was further developed and is able to steer SPIONs precisely by rotating the arrays (Blümler, 2021). Figure 1. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Magnetization pattern and ﬁeld distribution of the ad- justable Halbach array. The magnetic ﬁeld can be shifted from be- low to above the magnets by changing ϕ from 0 to π/2 (Reprinted with permission from Thalmayer et al., 2021; © IEEE). density on a discrete mesh. This is done with two different functions. The corresponding ﬁtting functions are addition- ally evaluated regarding the obtained errors and compared. All results were generated numerically using COMSOL Multiphysics®. 2 Concepts and theory To evaluate the precise steering of the SPIONs, two different arrangements of Halbach arrays are proposed in this section. Furthermore, the magnetic ﬂux density of a Halbach array and the resulting forces acting on the particles in a MDT sce- nario, namely magnetic and hydrodynamic drag force, will be introduced. lications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V. Published by Copernicus Publications on behalf of the URSI Landesausschuss in der Bundesrepublik Deutschland e.V. almayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 95 Figure 2. Magnetization pattern and ﬁeld distribution of the ex- tended Halbach array. Here, the two sides cannot be changed by a simultaneous rotation of all single magnets. to the analytic solution of Sim and Ro (2020) to the analytic solution of Sim and Ro (2020) B = c1y + c2 c3y2 + c4 · p c5y2 + c6 , (1) (1) where ci with i ∈{1,2,...,6} are arbitrary constants. Here, B is assumed to only have a component in the y-direction. The second ﬁtting function is based on an exponential ansatz B = k1e−k2y, (2) B = k1e−k2y, (2) Figure 2. Magnetization pattern and ﬁeld distribution of the ex- tended Halbach array. Here, the two sides cannot be changed by a simultaneous rotation of all single magnets. which was already used in Thalmayer et al. (2021). Here, k1 and k2 correspond to arbitrary constants. The constants ci and ki are ﬁtted for both, the Halbach and the extended Halbach array for the strong and the weak magnetic side. The ﬁtting is done based on the simulation results from Comsol, by using a least square (LS) algorithm for every x. which was already used in Thalmayer et al. (2021). Here, k1 and k2 correspond to arbitrary constants. The constants ci and ki are ﬁtted for both, the Halbach and the extended Halbach array for the strong and the weak magnetic side. The ﬁtting is done based on the simulation results from Comsol, by using a least square (LS) algorithm for every x. 2.4 Magnetic force The resulting magnetic force F mag on one SPION at a posi- tion r is deﬁned by (Jackson, 1998) F mag (r) = ∇(m(r) · B (r)). (3) (3) Thereby, B [T] is the magnetic ﬂux density at the location of the SPIONs with a magnetic moment m [A m2]. The mag- netic moment m is given by the integral of the magnetiza- tion M over the particle’s volume V (Jackson, 1998). Usu- ally, the magnetization M = Msat ·L(ξ) is assigned to be ho- mogeneous within a SPION and increases by a monotonic Langevin-function L(ξ) (Baun and Blümler, 2017). ξ is di- rectly proportional to B, whereas Msat corresponds to the sat- uration magnetization. Since one SPION is only composed of a single magnetic domain, it shows no hysteretic behavior (Cullity and Graham, 2008). Alternatively to the magnetiza- tion M, the SPIONs can be characterized by their magnetic susceptibility χ (Baun and Blümler, 2017). In the proposed extended array, the magnetic ﬁeld of the extended Halbach array is augmented and attenuated on one side, respectively, like in the normal Halbach array in Fig. 1. However, in case of the extended Halbach array the strong and the weak side cannot be changed by a simultaneous rota- tion of all single magnets. Therefore, the mechanical rotation of the strong and the weak magnetic side is signiﬁcantly more challenging. From Eq. 3 and the assumption of a homogeneous magne- tization, it follows that the magnitude of the magnetic force \|F mag\| depends on the volume V , the magnetization M and the gradient of the magnetic ﬂux density G = ∇B: \|F mag\| = mG = V M G. (4) (4) 2.2 Extended Halbach array In extension to our previous published work (Thalmayer et al., 2021) of analyzing the adjustable Halbach array, an ex- tended Halbach array is investigated for its performance on particle deﬂection. The magnets of this extended array have a 45◦shifted magnetization compared to their direct neighbors according to the setup displayed in Fig. 2. In the literature, Halbach arrays are used with several magnetization angles between two adjacent magnets (Zhang et al., 2011, or Shen and Zhu, 2013), however, as aforementioned the arrays are usually coaxial and not linear arranged. For the use case of particle deﬂection, to the best of the authors’ knowledge cur- rently only linear Halbach arrays with a magnetization shift of 90◦between each other were investigated. Therefore, the performance on the SPION deﬂection for an array with a 45◦ shifted magnetization, is investigated. 2.3 Magnetic ﬂux density of a Halbach array The magnitude of \|F mag\| for the application of MDT is usu- ally in the range between 10−11 to 10−25 N (Nacev et al., 2012). Most approaches in the literature use numerical methods or programs such as COMSOL Multiphysics® (Bjørk and In- singa, 2018; Ijiri et al., 2013) for determining the magnetic ﬂux density B of a Halbach array. However, there are also an- alytic solutions (Peng and Zhou, 2013; Shen and Zhu, 2013; Zhang et al., 2020; Sim and Ro, 2020). For particle deﬂection the gradient of B is needed additionally. Calculating the gra- dient numerically often leads to high errors due to the mesh- wise computation of B, which is computed with Comsol in this work. Therefore, B is ﬁtted with two variable functions for further investigations and, thus, the gradient can be de- rived analytically. The ﬁrst function is adapted in accordance 2.1 Linear Halbach array conﬁguration For the calculation of the magnetic force, the magnetic ﬂux density and its gradient are computed via ﬁtting the simulation results, which is neces- sary due to high error peaks by deriving the magnetic ﬂux https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 95 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 95 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Arra 96 Figure 3. Proposed geometry model used in the 2D Comsol sim- ulations. The magnetic ﬁeld is analyzed at three positions, marked as red numbers. Whereas the measuring positions “1” and “3” are located underneath the fourth and the last magnet, respectively, “2” is located symmetrical in the center of the array at the gap between two adjacent magnets. the distribution of the SPIONs were investigated for discrete values with a step of π/20, whereas in case of the extended Halbach array only the strong and the weak side as depicted in Fig. 2 were investigated. The magnetic ﬂux density was studied at the three positions “1”–“3” in detail, which are de- noted in red in Fig. 3. Figure 3. Proposed geometry model used in the 2D Comsol sim- ulations. The magnetic ﬁeld is analyzed at three positions, marked as red numbers. Whereas the measuring positions “1” and “3” are located underneath the fourth and the last magnet, respectively, “2” is located symmetrical in the center of the array at the gap between two adjacent magnets. 3 Results and discussion In this section the simulation results will be presented and discussed. At ﬁrst, the magnetic ﬂux density of the different Halbach arrays at the three aforementioned evaluation posi- tions will be investigated in detail. Afterwards, their perfor- mance on deﬂecting the SPIONs in the given scenario will be evaluated. For further investigating the magnetic force F mag on the SPIONs, the gradient of the magnetic ﬂux density has to be derived. In this work, this is done by ﬁtting the sim- ulation results of B for every x-position with the two ﬁt- ting functions and determining the analytical derivative af- terwards. The errors of the two ﬁtting functions will be com- pared and the gradient of B evaluated. Finally, the magnetic force F mag and the drag force F drag will be investigated in detail and limitations of the proposed work will be explained. (like buoyant or lift force) on the SPIONs can be neglected (Thalmayer et al., 2020). The velocity proﬁle u(y) is as- sumed to be laminar. Thus, the proﬁle is parabolic and can be expressed by u(y) = umax " 1 − y Rv 2# , (6) (6) where umax = 2umean. For blood, umax varies between 0.5 mm s−1 to 40 cm s−1 (Nacev et al., 2012). For a predominant movement of the SPIONs towards the magnets, the magnitude of the magnetic force F mag has to be greater than the magnitude of F drag: 3.1 Evaluation of the magnetic ﬂux density In Fig. 4, the isolines of the magnetic ﬂux density B of the adjustable Halbach array are depicted for ϕ = 0,π/4,π/2 . It can be observed, that the Halbach array has a strong and a weak magnetic side, which can be switched by rotating ϕ from 0 to π/2. The maximum value of the magnetic ﬂux den- sity B is the strongest for ϕ = π/4 and the weakest for ϕ = 0 and π/2. The maximum values are located between the ﬁrst and second (or last and second last) magnet as can be seen in Fig. 4. However, the maximum ﬂux densities differ only in 0.04 T. Furthermore, for ϕ = 0, B is approx. constant in the x-direction in the region of the vessel, whereas for ϕ = π/4, the magnetic ﬂux density is symmetric on both sides of the array. \|F mag\| !≥\|F drag\| ⇔V M G !≥6πηRhu. (7) \|F mag\| !≥\|F drag\| ⇔V M G !≥6πηRhu. (7) (7) In our proposed setup (see Fig. 3), F mag and F drag can be assumed to be perpendicular, as the changes of the magnetic ﬂux density in the x-direction can be neglected. In conse- quence, G becomes G = 9ηu 2R2pM (8) (8) for an equilibrium of F mag and F drag. For sake of simplicity, the hydrodynamic radius Rh of one SPION is assumed to be equal to the magnetic particle radius Rp. for an equilibrium of F mag and F drag. For sake of simplicity, the hydrodynamic radius Rh of one SPION is assumed to be equal to the magnetic particle radius Rp. In Fig. 5, the isolines of the magnetic ﬂux density B for the weak and the strong side of the extended Halbach array are illustrated. Here, the maximum value of the ﬂux density is slightly increased compared to the adjustable Halbach ar- ray and is located between the second and third (or third and second last) magnet. However, in comparison to the ﬁeld of Fig. 4, the ﬂux density is not that compact and also less ho- mogeneous in x-direction. 2.5 Hydrodynamic drag force The (hydrodynamic) drag force on one spherical SPION is F drag = 6πηRhu (5) (5) F drag = 6πηRhu according to Baun and Blümler (2017), where η is the ﬂuid’s dynamic viscosity [Pa s]. Rh and u correspond to the hydro- dynamic radius and the resulting velocity of one SPION, re- spectively. Since a low ﬂow regime is assumed, other forces according to Baun and Blümler (2017), where η is the ﬂuid’s dynamic viscosity [Pa s]. Rh and u correspond to the hydro- dynamic radius and the resulting velocity of one SPION, re- spectively. Since a low ﬂow regime is assumed, other forces https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 Adv. Radio Sci., 20, 93–104, 2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array er: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 97 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Table 1 Fixed simulation parameters Table 1. Fixed simulation parameters. Table 1. Fixed simulation parameters. Figure 5. Isolines of the magnetic ﬂux density B for the strong and the weak side of the extended Halbach array. Figure 4. Isolines of the magnetic ﬂux density B of the adjustable Halbach array for various rotation angles ϕ (Reprinted with permis- sion from Thalmayer et al., 2021; © IEEE). Figure 5. Isolines of the magnetic ﬂux density B for the strong and the weak side of the extended Halbach array. Figure 4. Isolines of the magnetic ﬂux density B of the adjustable Halbach array for various rotation angles ϕ (Reprinted with permis- sion from Thalmayer et al., 2021; © IEEE). smaller the distance d between the magnets is, the more ho- mogeneous is B. However, this results in drawbacks, namely more needed space and a stronger torque. Thus, the number of magnets and d is restricted to 8 and 0.25 cm, respectively. smaller the distance d between the magnets is, the more ho- mogeneous is B. However, this results in drawbacks, namely more needed space and a stronger torque. Thus, the number of magnets and d is restricted to 8 and 0.25 cm, respectively. The behavior of the magnetic ﬂux density B at the three evaluation positions, is illustrated in Fig. 6. Both, the ad- justable and the extended Halbach array, for ϕ = 0 and ϕ = π/2 and the strong and weak side are analyzed, respectively. Directly at the magnets, the ﬂux density is approximately in the same magnitude for both arrays, however, B decays faster for the weak magnetic sides. At the distance of 0.5 cm, where the vessel starts, B is approx. 3 times greater for the strong side compared to the weak side. Overall, it can be seen that for the strong side (ϕ = 0) and for the weak side (ϕ = π/2) the magnetic ﬂux density is stronger for the adjustable Hal- bach. Moreover, the results in Fig. 6 depict, that B is stronger signiﬁcant changes occur at the outer magnets. Thus, the std is evaluated excluding the area under the two outer magnets, resulting in a std being 50 times smaller than the mean B. 2.6 Simulation model To investigate the performance of the particle steering, the linear Halbach array and the extended Halbach array were built up as a 2D model in COMSOL Multiphysics® 5.6. Each array consists of 8 circular magnets and a Y-shaped geometry with a 30◦bifurcation representing the vessel. The setup is depicted in Fig. 3. In the simulations, the “Magnetic Fields, No Currents (mfnc)”, “Laminar Flow (spf)” and the “Par- ticle Tracing (fpt)” modules were used. The ﬁxed simula- tion parameters are summarized in Table 1. Here, the ap- plied magnetic parameters correspond to typical values for NdFeB, whereas the ﬂuid was set to water. In case of the ad- justable Halbach array, the magnetization direction can be ro- tated over ϕ as shown in Fig. 1. Here, the magnetic ﬁeld and The variation of B(y) along the x-direction is evaluated by computing its standard deviation (std) for every y inside the vessel over the length beneath the magnets for ϕ = 0 and ϕ = π/2 for the adjustable and for the strong and weak side for the extended Halbach array, respectively. In case of the adjustable Halbach array, for both ϕ, the std is by a factor of 10 smaller than the average value of B. However, the most https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 97 Table 1. Fixed simulation parameters. Category Symbol Value Unit Label Br 1.2 T remanent ﬂux density magnet Rm 1 cm radius d 0.25 cm distance between magnets L 30 cm length vessel Rv 0.5 cm radius d 0.5 cm distance between magnets and vessel wall umean 1 cm s−1 average velocity ρ 2200 kg m−3 density N 100 1 number of simulated SPIONs particles Rp 250 nm radius χ 9 1 susceptibility I [0.2,0.8] cm inlet range Figure 4. Isolines of the magnetic ﬂux density B of the adjustable Halbach array for various rotation angles ϕ (Reprinted with permis- sion from Thalmayer et al., 2021; © IEEE). Figure 5. Isolines of the magnetic ﬂux density B for the strong and the weak side of the extended Halbach array. smaller the distance d between the magnets is, the more ho- mogeneous is B. However, this results in drawbacks, namely A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Table 1. Fixed simulation parameters. https://doi.org/10.5194/ars-20-93-2023 2.6 Simulation model Category Symbol Value Unit Label Br 1.2 T remanent ﬂux density magnet Rm 1 cm radius d 0.25 cm distance between magnets L 30 cm length vessel Rv 0.5 cm radius d 0.5 cm distance between magnets and vessel wall umean 1 cm s−1 average velocity ρ 2200 kg m−3 density N 100 1 number of simulated SPIONs particles Rp 250 nm radius χ 9 1 susceptibility I [0.2,0.8] cm inlet range A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 3.2 Evaluation of SPION distribution For the adjustable Halbach array, the propagation of the SPIONs for a ﬁxed rotation angle ϕ of the magnetization di- rection was investigated. The distribution of the SPIONs to the branches is depicted in Fig. 7. Without the magnet (labeled “none” in Fig. 7), the SPIONs split equally between the upper and lower branch. With the Halbach array, Fig. 7 depicts that for the strong side of the ar- ray (ϕ = 0), only 27 % of the SPIONs are in the lower branch. However, 31 % of the SPIONs get trapped by the magnet and, thus, remain in the vessel. For the weak side (ϕ = π/2), 46 % of the SPIONs are in the lower branch, while 6 % get trapped. For the extended Halbach array (not plotted in Fig. 7) in case of the strong side 49 %, 37 % and 14 % take the upper, lower branch and get trapped by the magnet, respectively. In case of the weak side facing the vessel, 52 % and 48 % of the SPIONs take the upper and lower branch, respectively. Overall, it can be seen that the inﬂuence of the magnetic ﬁeld on the SPIONs is stronger for the adjustable Halbach array. The reason can be found in the homogeneity of the magnetic ﬁeld and its strong gradient as described in the section before. This leads to a strong and approximately constant magnetic force over a large area all underneath the magnets, resulting in a signiﬁcant effect on the particles. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 98 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array at positions underneath the magnets (position 1 and 3). Nev- ertheless, apart from position 2 at the strong side of the ar- rays, B decays for all functions approx. exponentially with the distance from the magnets. Table 2. Comparison of the mean ﬁtting errors of the the exponen- tial Eq. (2) and the analytic Eq. (1) function for different rotation angles ϕ for the adjustable Halbach array and for the strong and weak side of the extended Halbach array. Table 2. Comparison of the mean ﬁtting errors of the the exponen- tial Eq. (2) and the analytic Eq. (1) function for different rotation angles ϕ for the adjustable Halbach array and for the strong and weak side of the extended Halbach array. Equation ϕ/side Position εmean [mT] εmean,rel [%] 1 12.9 3.85 (2) 0 2 11.4 4.02 3 15.9 5.91 1 5.6 5.02 (2) π/2 2 3.8 4.27 3 12.2 7.78 1 3.0 0.90 (1) 0 2 4.3 1.51 3 1.0 0.38 1 1.6 1.46 (1) π/2 2 1.3 1.50 3 1.2 0.79 1 23.5 7.17 (2) strong 2 19.0 7.55 3 19.6 8.10 1 3.0 20.1 (2) weak 2 3.7 15.4 3 15.6 18.6 1 1.8 0.51 (1) strong 2 10.6 4.17 3 0.9 0.26 1 10.6 81.1 (1) weak 2 6.0 17.4 3 3.2 3.53 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Similar results were archived for the extended Halbach array. However, here the std was only approximately a factor of 5 (full length) and 20 (excluding outer magnets) smaller. By decreasing the angle between the magnetization direction of two adjacent magnets even more, the results become worse. Thus, it can be seen, that the magnetic ﬁeld is more homoge- neous for the adjustable Halbach array with a magnetization shift of 90◦between two neighboring magnets. Further in- vestigations show, that the more magnets are used and the The behavior of the magnetic ﬂux density B at the three evaluation positions, is illustrated in Fig. 6. Both, the ad- justable and the extended Halbach array, for ϕ = 0 and ϕ = π/2 and the strong and weak side are analyzed, respectively. Directly at the magnets, the ﬂux density is approximately in the same magnitude for both arrays, however, B decays faster for the weak magnetic sides. At the distance of 0.5 cm, where the vessel starts, B is approx. 3 times greater for the strong side compared to the weak side. Overall, it can be seen that for the strong side (ϕ = 0) and for the weak side (ϕ = π/2) the magnetic ﬂux density is stronger for the adjustable Hal- bach. Moreover, the results in Fig. 6 depict, that B is stronger https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 3.3 Evaluation of the gradient and ﬁtting of the magnetic ﬂux density Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 99 Figure 6. Magnetic ﬂux density B over y for the different evaluation positions for ϕ = 0 and ϕ = π/2 of the adjustable Halbach array (a, b) and for the strong and weak magnetic side of the extended Halbach array (c, d). The black vertical lines corresponds to the vessel walls. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 99 Figure 6. Magnetic ﬂux density B over y for the different evaluation positions for ϕ = 0 and ϕ = π/2 of the adjustable Halbach array (a, b) and for the strong and weak magnetic side of the extended Halbach array (c, d). The black vertical lines corresponds to the vessel walls. Figure 7. Distribution of the SPIONs for different rotation angles ϕ of the magnetization direction for the adjustable Halbach array. The blue and the red color correspond to the SPIONs taking the lower and upper branch, respectively; yellow represents the number of SPIONs trapped by the magnets. For the case “none”, no magnet was considered in the simulation (Reprinted with permission from Thalmayer et al., 2021; © IEEE). the vessel, M of the particles has to be determined ﬁrst. This is done by evaluating Fmag,y and the gradient G of the mag- netic ﬂux density at a ﬁxed position in the simulation. From Eq. (4) and with Fmag,y = 8.38·10−13 N and G = 32 T m−1, M is calculated to M ≈4 · 105 A m−1. This is in good ac- cordance to the results of Alexiou et al. (2006) and Wei and Wang (2018). For reasons of simpliﬁcation, the magnetiza- tion is assumed to be constant. Based on this, Fmag reached values from 5.1·10−14 to 2.1·10−12 N in the area underneath the magnets. The strongest and weakest Fmag was calculated for ϕ = 0 and ϕ = π/2, respectively. Figure 7. Distribution of the SPIONs for different rotation angles ϕ of the magnetization direction for the adjustable Halbach array. The blue and the red color correspond to the SPIONs taking the lower and upper branch, respectively; yellow represents the number of SPIONs trapped by the magnets. For the case “none”, no magnet was considered in the simulation (Reprinted with permission from Thalmayer et al., 2021; © IEEE). Additionally, Fmag,max and Fmag,mean were calculated for both arrays. 3.3 Evaluation of the gradient and ﬁtting of the magnetic ﬂux density the strong side small errors result in smaller relative errors than for the weak side. Additionally, due to the numerical calculation on the mesh of B, there are discretization errors in the solution (compare Fig. 6). Since the ﬁtting errors were conducted by subtraction the ﬁtted results from the simulated one, this discretization can lead to high (especially relative) ﬁtting errors. Since B is calculated numerically on a mesh in Comsol, the numerical calculation of the gradient of B is quite challeng- ing. Thus, by determining the derivative, there are a lot of irregularities and additional noise on it. Therefore, in this work, the derivative is determined analytically using the re- sults of the ﬁtting functions (Eqs. 1 and 2). The mean ﬁtting error for both equations and both arrays is summarized in Table 2. More detailed tables are listed in Appendix A. Applying the ﬁtting functions, the gradient of B can be determined. Since the exponential approach (Eq. 2) and it derivative are less complex and perform sufﬁcient enough, it is used in the following. The resulting gradients for both Halbach arrays are depicted in Fig. 8. The gradient directly next to the magnet is stronger for the two weaker sides of the arrays. However, the gradient decreases very fast there, whereas the gradient of the two strong sides stays high over a larger distance. Moreover, the gradient is stronger directly under the magnets (positions “1” and “3”) and, therefore, the magnetic force on the particles is also stronger. Overall, the ﬁtting performs quite well for both functions over the whole evaluation domain. By considering only the results inside the vessel, the errors even decrease, since the greatest errors occur directly at the boundary of the magnets. The ﬁtting errors are smaller for the analytic ﬁtting function Eq. (1). However, this is obvious, since six arbitrary con- stants are used in contrast to the exponential ﬁtting function where only two were considered. In general, the ﬁtting re- sults perform better for the adjustable Halbach array than for the extended one. Furthermore, the errors are smaller for the strong side of the adjustable and extended Halbach array. This can especially be seen in the relative errors. The rea- son can be found in the magnitude of the magnetic ﬁeld: for https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 A. S. 3.3 Evaluation of the gradient and ﬁtting of the magnetic ﬂux density Furthermore, the ratio Fd−m = Fdrag/Fmag is evaluated for every x and y. The results are listed in Table 3. Fd−m has a minimum of 0, since Fdrag = 0 N at the vessel wall, and maximum values of 236 and 647 for the adjustable Halbach array, respectively. For ϕ = 0, Fd−m is smaller, since Fmag on the SPIONs is stronger in this case. For the extended Halbach array, the ratio is even greater (608 and 1454), il- lustrating that the magnetic force on the particles is much smaller for this array. 3.4 Evaluation of Fmag and Fdrag To steer the particles, the main idea is to attract the SPIONs by a strong magnetic ﬁeld and then rotate the array to get the particles washed out by the (hydrodynamic) drag force Fdrag. For this purpose, the magnetic force Fmag and Fdrag are calculated underneath the magnets. Fdrag reaches values from 0 N at the boundary to 9.4 · 10−11 N at the center of the vessel. Before Fmag can be investigated at every position of Besides, the main moving direction is given by Fd−m. At a ratio of > 100, the inﬂuence of the magnetic force on the SPIONs is assumed to be negligible. For this, a boundary dis- tance yb [m] was calculated, which corresponds to the dis- tance from the upper vessel for Fd−m ≤100 (see Table 3). https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 100 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Figure 8. Gradient of the magnetic ﬂux density B for the different evaluation positions of the adjustable Halbach array (a) and the extended Halbach array (b). The solid lines correspond to ϕ = 0 and the strong side and the dashed lines to ϕ = π/2 and the weak side of the extended Halbach array, respectively. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array ring Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 100 Figure 8. Gradient of the magnetic ﬂux density B for the different evaluation positions of the adjustable Halbach array (a) and the extended Halbach array (b). The solid lines correspond to ϕ = 0 and the strong side and the dashed lines to ϕ = π/2 and the weak side of the extended Halbach array, respectively. Table 3. Comparison of the results for different rotation angles ϕ for the adjustable Halbach array and for the strong and weak side of the extended Halbach array. tion within the SPIONs is assumed to be constant. How- ever, it can be observed from measurements that the parti- cles form chains parallel to the magnetic ﬁeld lines (My- rovali et al., 2021). This changes the magnetic ﬁeld distri- bution and, therefore, the magnetic force on the particles. 4 Conclusions In this extended paper, an adjustable linear Halbach array and an extended version for steering magnetic nanoparticles were investigated numerically by using COMSOL Multiphysics®. Both Halbach arrays provide a strong and a weak magnetic side, which can be switched by rotating the single magnets. In case of the adjustable Halbach array, this can be done by a simultaneous rotation of all single magnets around 90◦clock- wise or counterclockwise, respectively, whereas in case of the extended array the switching is much more complex. The magnetic ﬁeld, its gradient and the force on magnetic parti- cles were analyzed for different rotation angles of the mag- 3.4 Evaluation of Fmag and Fdrag ϕ/side Fmag,mean [N] Fmag,max [N] Fd−m,max yb [mm] 0 5.1 · 10−13 2.1 · 10−12 236 [2,2.6] π/2 3.4 · 10−13 1.0 · 10−12 647 1.8 strong 2.8 · 10−13 6.1 · 10−13 608 [0.6,1.4] weak 2.3 · 10−13 8.9 · 10−13 1454 1.4 Moreover, for the magnetic force F mag only the compo- nent in the y-direction is taken into account, since B is dom- inant in this direction. However, of course B has a compo- nent and gradient in the x-direction, too. The particles are accelerated before one magnet of the array and decelerated after it. Furthermore, this also contributes to agglomeration of the particles directly underneath the magnets, which again changes the magnetic ﬁeld distribution. The ratio Fd−m is also illustrated in Fig. 9 exemplary for the strong and the weak side of the adjustable Halbach array. To make the boundary yb visible, Fd−m is set equal to 0 for all values < 100. Another limitation of this study is caused by the permanent magnets. They cannot be switched off like an electromagnet. Therefore, there will be always a remaining magnetic ﬁeld and, thus, a magnetic force towards the magnets, even for the weak side. In consequence, the magnetic array has to have a certain distance to the vessel, which, however, results in the decrease of the magnetic force for the strong side. As expected, yb is greater for ϕ = 0. For ϕ = π/2, the boundary distance is quite constant, whereas in the results of ϕ = 0 the inhomogeneity of B can be seen. The minima and maxima have a distance of 2.25 cm, which corresponds to the distance of the magnets in the array. This is in good accordance with the results of the gradient (compare Fig. 8), which is predominantly underneath the magnets. Moreover, the adjustable Halbach array also performs better than the ex- tended Halbach array, since yb is smaller, which ﬁts also to the results of the gradient. 3.5 Limitations of this study It is worth mentioning, that the proposed study has some lim- itations: Within the simulations, the particles were consid- ered separately. This means, that the interaction between the particles among each other were neglected. Also the change of the magnetic ﬁeld due to the particles propagating through it, is not taken into account. Furthermore, the magnetiza- https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 101 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 101 Figure 9. Ratio of F drag/F mag of the adjustable Halbach arry. For ratios smaller than 100, the ratio is set to 0 to make the boundary yb visible. Figure 9. Ratio of F drag/F mag of the adjustable Halbach arry. For ratios smaller than 100, the ratio is set to 0 to make the boundary yb visible F mag of the adjustable Halbach arry. For ratios smaller than 100, the ratio is set to 0 to make the boundary y Figure 9. Ratio of F drag/F mag of the adjustable Halbach arry. For ratios smaller than 100, the ratio is set to 0 to make the boundary yb visible. nets as well as for the strong and weak magnetic side of the extended Halbach array. For the calculation of the gradient, two different ﬁtting functions were compared. The results reveal that overall the adjustable Halbach array performs better than the extended one in both, changing the magnetic force easily and deﬂecting the particles. In general it can be concluded that the magnetic force is the strongest directly underneath the magnets and by rotating the single magnets, the region where the drag force is able to wash the particles out, can be adjusted. In future research, the simu- lation model will be expanded and every second magnet of the adjustable Halbach array will be replaced by an electro- magnet. Therefore, the conﬁguration has to be optimized to design a proper electromagnet needing less space and power. By that, the strong and weak magnetic side can be switched by switching the direction of the current. With this, an elec- tromagnetic ﬁeld of travelling waves similar to a linear motor drive will be designed for steering the particles to a desired region. Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array This paper was edited by Lars Ole Fichte and reviewed by two anonymous referees. Author contributions. AST initiated the research project and devel- oped the method. AST, SZ and ML conducted the simulation and carried out the visualization. AST interpreted the results. GF was supervisor. AST wrote the original draft. All authors were involved in editing and reviewing the manuscript. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 102 Table A1. Overview of the ﬁtting results for the exponential Eq. (2) and the analytic Eq. (1) for the different evaluation positions of the adjustable Halbach array. All absolute ﬁtting errors ε are given in mT and the relative ones (in brackets) in %. For the last two columns, only the results inside the vessel were considered. Equation Number Position εmean,y εmax,y εmax, ϕ εmean,y in vessel εmax,y in vessel 1 14.0 (7.37) 62.2 (20.3) 17.1 (10.2) 9.4 (4.73) 18.2 (9.80) (2) 2 4.1 (2.11) 15.5 (6.51) 11.4 (4.27) 3.1 (1.40) 5.6 (2.11) 3 17.4 (10.6) 68.8 (31.8) 19.7 (11.4) 11.6 (6.61) 23.4 (13.8) 1 2.1 (1.16) 5.4 (5.56) 3.0 (1.46) 1.5 (0.74) 2.4 (1.42) (1) 2 1.6 (0.88) 5.5 (3.37) 6.8 (2.29) 1.3 (0.62) 2.8 (1.25) 3 1.2 (0.59) 8.7 (2.10) 2.9 (1.72) 0.8 (0.46) 1.7 (0.92) Table A2. Overview of the ﬁtting results for the exponential Eq. (2) and the analytic Eq. (1) for the different evaluation positions of the extended Halbach array. All absolute ﬁtting errors ε are given in mT and the relative ones (in brackets) in %. For the last two columns, only the results inside the vessel were considered. Equation Number Side Position εmean εmax εmean in vessel εmax in vessel strong 1 23.5 (7.17) 114.3 (17.0) 17.2 (5.35) 33.1 (9.14) (2) strong 2 19.0 (7.55) 85.3 (41.7) 16.3 (5.72) 26.6 (9.01) strong 3 19.6 (8.10) 69.9 (23.1) 14.0 (5.46) 27.2 (8.84) weak 1 3.0 (20.1) 21.6 (108) 2.6 (33.0) 5.5 (108) (2) weak 2 3.7 (15.4) 8.8 (17.0) 2.5 (7.32) 5.1 (25.9) weak 3 15.6 (18.6) 52.5 (58.1) 9.9 (11.2) 17.9 (30.3) strong 1 1.8 (0.51) 9.2 (1.70) 1.4 (0.43) 2.9 (0.79) (1) strong 2 10.6 (4.17) 53.5 (26.2) 8.8 (3.20) 17.7 (5.96) strong 3 0.9 (0.26) 8.4 (1.33) 0.4 (0.19) 1.3 (0.5) weak 1 10.6 (81.1) 47.6 (157) 9.9 (84.5) 14.5 (157) (1) weak 2 6.0 (17.4) 21.0 (88.1) 5.4 (14.2) 8.7 (24.5) weak 3 3.2 (3.53) 14.3 (15.9) 2.9 (3.2) 4.7 (5.73) Special issue statement. This article is part of the special issue “Kleinheubacher Berichte 2021”. Code and data availability. The simulation data and MATLAB- code are available from the corresponding author upon request. Code and data availability. The simulation data and MATLAB- code are available from the corresponding author upon request. Review statement. Appendix A: Fitting Errors The errors for the ﬁtting of the magnetic ﬂux density with the exponential Eq. (2) and the analytic Eq. (1) are listed in detail in Table A1 for the adjustable Halbach array and in Table A2 for the extended Halbach array, respectively. Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array cles in Blood Vessels, IEEE Access, 8, 106714–106725, https://doi.org/10.1109/ACCESS.2020.3000329, 2020. cles in Blood Vessels, IEEE Access, 8, 106714–106725, https://doi.org/10.1109/ACCESS.2020.3000329, 2020. Bjørk, R. and Insinga, A. R.: A topology optimized switchable per- manent magnet system, J. Magn. Magn. Mater., 465, 106–113, https://doi.org/10.1016/j.jmmm.2018.05.076, 2018. Peng, J. and Zhou, Y.: Modeling and Analysis of a New 2-D Halbach Array for Magnetically Levi- tated Planar Motor, IEEE T. Magn., 49, 618–627, https://doi.org/10.1109/TMAG.2012.2210435, 2013. Bjørk, R., Bahl, C., Smith, A., and Pryds, N.: Com- parison of adjustable permanent magnetic ﬁeld sources, J. Magn. Magn. Mater., 322, 3664–3671, https://doi.org/10.1016/j.jmmm.2010.07.022, 2010. Sakuma, H. and Nakagawara, T.: Optimization of rotation patterns of a mangle-type magnetic ﬁeld source using covariance ma- trix adaptation evolution strategy, J. Magn. Magn. Mater., 527, 167752, https://doi.org/10.1016/j.jmmm.2021.167752, 2021. Blümler, P.: Magnetic Guiding with Permanent Magnets: Concept, Realization and Applications to Nanoparticles and Cells, Cells, 10, 2708, https://doi.org/10.3390/cells10102708, 2021. Cullity, B. D. and Graham, C. D.: Introduction to Magnetic Ma- terials, 1st edn., John Wiley & Sons, Inc, Hoboken, NJ, USA, https://doi.org/10.1002/9780470386323, 2008. Shen, Y. and Zhu, Z.-Q.: General analytical model for calculat- ing electromagnetic performance of permanent magnet brush- less machines having segmented Halbach array, IET Electr. Syst. Transp., 3, 57–66, https://doi.org/10.1049/iet-est.2012.0055, 2013. Fink, M., Rupitsch, S. J., Lyer, S., and Ermert, H.: Quantitative Determination of Local Density of Iron Oxide Nanoparticles Used for Drug Targeting Employing Inverse Magnetomotive Ultrasound, IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 68, 2482–2495, https://doi.org/10.1109/TUFFC.2021.3068791, 2021. Shiriny, A. and Bayareh, M.: On magnetophoretic separation of blood cells using Halbach array of magnets, Meccanica, 55, 1903–1916, https://doi.org/10.1007/s11012-020-01225-y, 2020. Sim, M.-S. and Ro, J.-S.: Semi-Analytical Modeling and Analysis of Halbach Array, Energies, 13, 1252, https://doi.org/10.3390/en13051252, 2020. Hilton, J. E. and McMurry, S. M.: An adjustable linear Halbach array, J. Magn. Magn. Mater., 324, 2051–2056, https://doi.org/10.1016/j.jmmm.2012.02.014, 2012. Stevens, M., Liu, P., Niessink, T., Mentink, A., Abelmann, L., and Terstappen, L.: Optimal Halbach Conﬁguration for Flow- through Immunomagnetic CTC Enrichment, Diagnostics, 11, 1020, https://doi.org/10.3390/diagnostics11061020, 2021. Hoshiar, A. K., Le, T.-A., Amin, F. U., Kim, M. O., and Yoon, J.: Studies of aggregated nanoparticles steering during magnetic- guided drug delivery in the blood vessels, J. Magn. Magn. Mater., 427, 181–187, https://doi.org/10.1016/j.jmmm.2016.11.016, 2017. Thalmayer, A., Zeising, S., Fischer, G., and Kirchner, J.: Investi- gation of Particle Steering for Different Cylindrical Permanent Magnets in Magnetic Drug Targeting, in: Proceedings of 7th In- ternational Electronic Conference on Sensors and Applications (MDPI), Basel, Switzerland, 13–30 November 2020, p. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array 8269, https://doi.org/10.3390/ecsa-7-08269, 2020. Ijiri, Y., Poudel, C., Williams, P. S., Moore, L. R., Orita, T., and Zborowski, M.: Inverted Linear Halbach Array for Separation of Magnetic Nanoparticles, IEEE T. Magn., 49, 3449–3452, https://doi.org/10.1109/TMAG.2013.2244577, 2013. Jackson, J. D.: Classical electrodynamics, 3rd edn., John Wiley & Sons, New York, USA, ISBN 978-0-47130-932-1, 1998. Thalmayer, A. S., Zeising, S., Fischer, G., and Kirchner, J.: Steering Magnetic Nanoparticles by Utilizing an Adjustable Linear Halbach Array, in: 2021 Kleinheubach Confer- ence, Miltenberg, Germany, 28–30 September 2021, 1–4, https://doi.org/10.23919/IEEECONF54431.2021.9598436, 2021. Kang, J. H., Driscoll, H., Super, M., and Ingber, D. E.: Application of a Halbach magnetic array for long-range cell and particle sep- arations in biological samples, Appl. Phys. Lett., 108, 213702, https://doi.org/10.1063/1.4952612, 2016. Kim, Y., Chae, J. K., Lee, J.-H., Choi, E., Lee, Y. K., and Song, J.: Free manipulation system for nanorobot cluster based on com- plicated multi-coil electromagnetic actuator, IEEE T. Magn., 43, 2956–2958, https://doi.org/10.1109/TMAG.2007.893798, 2021. Tietze, R., Lyer, S., Dürr, S., Struffert, T., Engelhorn, T., Schwarz, M., Eckert, E., Göen, T., Vasylyev, S., Peukert, W., Wiekhorst, F., Trahms, L., Dörﬂer, A., and Alexiou, C.: Efﬁcient drug-delivery using magnetic nanoparticles–biodistribution and therapeutic ef- fects in tumour bearing rabbits, Nanomed.-Nanotechnol., 9, 961– 971, https://doi.org/10.1016/j.nano.2013.05.001, 2013. Mues, B., Bauer, B., Roeth, A. A., Ortega, J., Buhl, E. M., Radon, P., Wiekhorst, F., Gries, T., Schmitz-Rode, T., and Slabu, I.: Nano- magnetic Actuation of Hybrid Stents for Hyperthermia Treat- ment of Hollow Organ Tumors, Nanomaterials-Basel, 11, 618, https://doi.org/10.3390/nano11030618, 2021. Vogel, P., Kampf, T., Herz, S., Ruckert, M. A., Bley, T. A., and Behr, V. C.: Adjustable Hardware Lens for Traveling Wave Magnetic Particle Imaging, IEEE T. Magn., 56, 1–6, https://doi.org/10.1109/TMAG.2020.3023686, 2020. Myrovali, E., Papadopoulos, K., Iglesias, I., Spasova, M., Farle, M., Wiedwald, U., and Angelakeris, M.: Long-Range Ordering Ef- fects in Magnetic Nanoparticles, ACS Appl. Mater. Inter., 13, 21602–21612, https://doi.org/10.1021/acsami.1c01820, 2021. Wei, W. and Wang, Z.: Investigation of Magnetic Nanoparticle Mo- tion under a Gradient Magnetic Field by an Electromagnet, J. Nanomater., 2018, 1–5, https://doi.org/10.1155/2018/6246917, 2018. Nacev, A., Komaee, A., Sarwar, A., Probst, R., Kim, S. H., Emmert-Buck, M., and Shapiro, B.: Towards Control of Mag- netic Fluids in Patients: Directing Therapeutic Nanoparticles to Disease Locations, IEEE Contr. Syst. Mag., 32, 32–74, https://doi.org/10.1109/MCS.2012.2189052, 2012. Zaloga, J., Janko, C., Nowak, J., Matuszak, J., Knaup, S., Eber- beck, D., Tietze, R., Unterweger, H., Friedrich, R. References Alexiou, C., Diehl, D., Henninger, P., Iro, H., Rockelein, R., Schmidt, W., and Weber, H.: A High Field Gradient Magnet for Magnetic Drug Targeting, IEEE T. Appl. Supercon., 16, 1527– 1530, https://doi.org/10.1109/TASC.2005.864457, 2006. Competing interests. The contact author has declared that none of the authors has any competing interests. Alnaimat, F., Karam, S., Mathew, B., and Mathew, B.: Magnetophoresis and Microﬂuidics: A Great Union, IEEE Nanotechnol. Mag., 14, 24–41, https://doi.org/10.1109/MNANO.2020.2966029, 2020. Disclaimer. Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Baun, O. and Blümler, P.: Permanent magnet system to guide super- paramagnetic particles, J. Magn. Magn. Mater., 439, 294–304, https://doi.org/10.1016/j.jmmm.2017.05.001, 2017. Adv. Radio Sci., 20, 93–104, 2023 Adv. Radio Sci., 20, 93–104, 2023 Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 103 https://doi.org/10.5194/ars-20-93-2023 Adv. Radio Sci., 20, 93–104, 2023 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array P., Duerr, S., Heimke-Brinck, R., Baum, E., Cicha, I., Dörje, F., Oden- bach, S., Lyer, S., Lee, G., and Alexiou, C.: Development of a lauric acid/albumin hybrid iron oxide nanoparticle system with improved biocompatibility, Int. J. Nanomed., 9, 4847–4866, https://doi.org/10.2147/IJN.S68539, 2014. p g Park, M., Le, T.-A., Eizad, A., and Yoon, J.: A Novel Shared Guidance Scheme for Intelligent Haptic Inter- action Based Swarm Control of Magnetic Nanoparti- Adv. Radio Sci., 20, 93–104, 2023 https://doi.org/10.5194/ars-20-93-2023 104 A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array ng, Z. X., Yang, M. L., Zhang, Y., Tuo, J. Y., and cal Prediction of Halbach Array Permanent Mag- Considering Finite Tooth Permeability, IEEE T. 10, https://doi.org/10.1109/TMAG.2020.2982844, Zhang, X., Li, Y. G., Cheng, H., and Liu, H. K.: Analysis of the Planar Magnetic Field of Linear Permanent Magnet Halbach Array, Appl. Mech. Mater., 66–68, 1336–1341, https://doi.org/10.4028/www.scientiﬁc.net/AMM.66-68.1336, 2011. A. S. Thalmayer: Towards Steering Magnetic Nanoparticles Using a Linear Halbach Array Zhang, H. S., Deng, Z. X., Yang, M. L., Zhang, Y., Tuo, J. Y., and Xu, J.: Analytical Prediction of Halbach Array Permanent Mag- net Machines Considering Finite Tooth Permeability, IEEE T. Magn., 56, 1–10, https://doi.org/10.1109/TMAG.2020.2982844, 2020. Adv. Radio Sci., 20, 93–104, 2023
https://openalex.org/W4367151717	https://link.springer.com/content/pdf/10.1007/978-3-031-28839-5_120.pdf	English	null	Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX	Lecture notes in mechanical engineering	2,023	cc-by	3,700	E.-Lexus Thornton1,2(B), Avery Hartley1,2, Daniel Caudill1,2, Martin Gamesu1,2, Julius Schoop1,2, and Fazleena Badurdeen1,2 1 Institute for Sustainable Manufacturing, University of Kentucky, Lexington, KY 40506, USA eth231@uky.edu 2 Department of Mechanical Engineering, University of Kentucky, Lexington, KY 40506, USA 1 Institute for Sustainable Manufacturing, University of Kentucky, Lexington, KY 40506, USA eth231@uky.edu 2 Department of Mechanical Engineering, University of Kentucky, Lexington, KY 40506, USA 2 Department of Mechanical Engineering, University of Kentucky, Lexington, KY 40506, U Abstract. With the increasing decline in the environment and natural resources it is important to ﬁnd new ways in which manufacturing can increase the sustain- ability of the world. This study seeks to compare LimeX, a state-of-the-art paper and plastic alternative primarily made of limestone, with conventional paper and plastic materials. For a better understanding of LimeX as a material, a brief inves- tigation of the mechanical and chemical properties will be performed through experimentation and analysis. In addition, this paper aims to evaluate the sus- tainability of LimeX through the analysis of metrics relating to the triple bottom line (environmental impacts, economic impacts and societal impacts) to evaluate and compare the sustainability performance of LimeX products with conventional paper and plastic products. The previously developed Product Sustainability Index (ProdSI) will be adapted and used in this study to conduct the sustainability evalu- ation. Major ﬁndings will include results from an experimental analysis of LimeX via SEM and EDS, and LimeX material property measurements via tensile test- ing and density measurements. In addition, there will also be a comparison of the sustainability performance of conventional paper and LimeX using a simpli- ﬁed ProdSI. The study found that LimeX was marginally more sustainable than paper, but this evaluation could change with information from the development of a life-cycle analysis report on the material. Keywords: LimeX · Product sustainability index · Paper and plastic alternative © The Author(s) 2023 H. Kohl et al. (Eds.): GCSM 2022, LNME, pp. 1082–1090, 2023. https://doi.org/10.1007/978-3-031-28839-5_120 Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX E.-Lexus Thornton1,2(B), Avery Hartley1,2, Daniel Caudill1,2, Martin Gamesu1,2, Julius Schoop1,2, and Fazleena Badurdeen1,2 1 Introduction There has long been a focus on the development of an alternative for conventional paper and plastic to improve global sustainability [1]. The manufacturing of synthetic oil-based plastics such as polypropylene (PP), polyethylene (PE) and polystyrene (PS) has led to an insurmountable amount of waste and CO2 emissions accumulating. This study aims to introduce and investigate a recently developed paper/plastic alternative called LimeX, a material primarily made from limestone (CaCO3) with a plastic additive added to act as a binder for the material, to determine its sustainable performance in comparison to that of conventional paper and plastic. In recent research [2–5], the Product Sustainability © The Author(s) 2023 H. Kohl et al. (Eds.): GCSM 2022, LNME, pp. 1082–1090, 2023. https://doi.org/10.1007/978-3-031-28839-5_120 Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX 1083 Index (ProdSI) has been instrumental in successfully evaluating and improving the sus- tainability of products. The ProdSI centers around three basic parameters of economic prosperity, environmental protection and societal well-being which are referred to as the triple bottom line (TBL) [3]. Fiksel et al. determined a list of product sustainability indicators to assess the TBL aspects of products [6]. Then a new and improved frame- work for the evaluation of product design for sustainability was presented by Jawahir et al. [7]. Although subjective, sustainability evaluations taking the lifecycle of a product into account could lead to a more effective evaluation of the total lifecycle sustainability performance [7]. Finally, Shuaib et al., [8] used the metrics from [7], together with new metrics for TBL criteria, to develop the ProdSI and applied it to compare the sustainabil- ity performance of two different generations of consumer electronics products [7]. The metrics highlighted in [7, 8] include waste, extraction, emissions, material recovery and efﬁciency and recyclability in terms of the environmental impacts; material, production, injury, warranty, labor and raw material costs in terms of economics; and ergonomics, ethics, functionality, safety and health in terms of societal impacts. In this study, similar metrics are highlighted to evaluate the sustainable performance of LimeX products. 2 Experimental Analysis of LimeX A series of experiments were performed to investigate the information provided by the company that manufactures LimeX, TBM [9]. Part of the investigation into LimeX involved conﬁrming that its composition was primarily limestone, as indicated by the manufacturer. In addition, an evaluation of the strength of LimeX paper was conducted to determine whether it was more durable than conventional paper. The measurements obtained from the analysis were used to inform the ProdSI-based sustainability analysis. The mechanical properties of the products play a role in how the product functions and the feasibility of multi-life-cycle applications of LimeX. 2.2 Material Property Testing A 5 kN Instron Tensile Testing machine and Bluehill Software were utilized to conduct semi-destructive tensile tests that measured the strength and Young’s modulus of four LimeX paper samples with differing thicknesses of 150, 200, 300, and 400 µm. An extension rate of 7.5 mm/min was applied. The experiment was repeated three times per thickness. Table 2 shows the material properties found for LimeX paper. The experi- mental tensile strength values match the values provided by TBM, but the experimental values of Young’s modulus differ from the values provided by approximately 20% [9]. This error may be ascribed to a difference in the material properties due to changes made by manufacturing processes. To conclude, the strength and toughness values indicate that LimeX paper is stronger and more durable than conventional paper when comparing the tensile strength test to a bursting test [10]. However, conventional paper is anisotropic and dependent on ﬁber direction leading to a measurement of tensile strength in units per length rather than area. 2.1 Scanning Electron Microscopy and Energy-Dispersive X-ray Spectroscopy Analysis of LimeX Pellets Furthermore, there is a small portion of silica and aluminum found in the LimeX pellet. These are thought to be imbedded during the polishing process. Fig. 3. EDS Micrographs showing anisotropy of LimeX’s chemical composition. Fig. 3. EDS Micrographs showing anisotropy of LimeX’s chemical composition. 2.1 Scanning Electron Microscopy and Energy-Dispersive X-ray Spectroscopy Analysis of LimeX Pellets Initial electron microscopy characterization of the LimeX surface topography was performed with a FEI Helios NanoLab 660 microscope. Subsequently, the ﬁrst few micrographs were used to determine if the LimeX pellets were porous. Fig. 1. Electron Microscopy Micrographs of LimeX Pellet. Fig. 1. Electron Microscopy Micrographs of LimeX Pellet. 1084 E.-L. Thornton et al. Becauseofthenon-conductivenatureoftheLimeX,thereweredifﬁcultiesinimaging caused by the charging of electrons on the surface. However, this charging led to some insights in the composition of LimeX. Charging, instead of random placement, appeared to be localized in speciﬁc areas of the sample seen in Fig. 1. This was likely due to inhomogeneity in composition. Additionally, the charging of electrons made it difﬁcult to capture extremely magniﬁed images as the charging increases as the beam is localized. A lower voltage along with a circular backscatter detector was utilized in combating charging. After initial analysis, the pellets were imbedded into a conductive, graphite puck to prevent charging issues by giving the electrons a path to disperse more quickly. Grinding away the surface and making a planar sample also reveals that the interior surface of the pellet is not porous. Moreover, EDS results (presented in Fig. 3) show anisotropy of chemical composition suggesting LimeX is non-homogeneous. Fig. 2. Electron Microscopy Micrograph used for EDS. Fig. 2. Electron Microscopy Micrograph used for EDS. The voltage was increased from the previous 2kV to 10 kV to allow for better penetration of electrons and more surface detail when capturing images of the pellets inside the graphite puck. The compositional maps in Fig. 2 show that the material is in-homogeneous with very little calcium being present on the surface compared to the amount of carbon and oxygen present on the surface. What’s more, the magnesium present suggests the mined limestone contained dolomite (CaMg(CO3)2) and not just the typical calcite (CaCO3) found in limestone. The complete chemical composition can be seen in Table 1 below. Conventional paper is primarily made of cellulose. Table 1. Chemical composition of LimeX pellets. Element Weight % Signal C 72.28 0.11 O 24.14 0.10 Ca 2.64 0.04 Mg 0.65 0.2 Si 0.18 0.2 Al 0.11 0.2 Table 1. Chemical composition of LimeX pellets. Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX 1085 Furthermore, there is a small portion of silica and aluminum found in the LimeX pellet. These are thought to be imbedded during the polishing process. 3 LimeX Sustainability Evaluation Product sustainability evaluations for product design as researched by Shuaib et al. and Jawahir et al. [3, 8], and additively manufactured products as researched by Zhang et al. and Happuwatte et al. [4, 11] have been developed and have been used to make substantial progress in the assessment of product sustainability. While product sustainability eval- uations provide beneﬁts through a rigorous sustainability analysis framework, they are also very subjective due to the weighing and normalizing of individual metrics [11]. The subjectivity involved in sustainability evaluations can lead to different results depend- ing on the biases. Although this is the case, product sustainability assessment methods 1086 E.-L. Thornton et al. Table 2. Experimental tensile test results for LimeX paper. Thickness (µm) LimeX - Average Tensile Strength (MPa) Average Young’s Modulus (MPa) Average Max. Load (N) 150 13.46 1949.05 13.82 200 14.38 2181.57 18.70 300 14.1 2,117.71 27.49 400 14.31 2,029.93 37.21 St. Dev 0.42 101.58 10.30 Table 2. Experimental tensile test results for LimeX paper. have been shown to accurately reﬂect changes in sustainability performance for different generations of products as shown by [3]. For context, there is a ﬁve-level hierarchical structure in the construction of a ProdSI [5]. The ﬁve levels, top to bottom, include: ProdSI, sub-indices, clusters, sub-clusters, and individual metrics. The sub-indices rep- resent key performance areas relevant to each TBL aspect. The clusters are divided into sub-clusters that address elements relating to the clusters they fall under [5]. Finally, the sub-clusters are divided into individual, measurable metrics [5]. The measured values are then converted into dimensionless scores. Afterwards the dimensionless scores gen- erated are aggregated to determine the ProdSI score. Once the ProdSI score is obtained, the entire process is reviewed for errors and the sustainability performance is deter- mined. Because this paper focused on the introduction of a new paper/plastics product, performing a product sustainability evaluation is a necessary step in investigating its sustainability performance. Due to the lack of access to data for all different metrics, this paper utilized a simpliﬁed ProdSI to investigate the performance of LimeX products. 3.2 Weighing and Aggregation Weights are assigned to metric based on the equal weight method, given by [5]. However, not every individual cluster and sub-cluster has a quantiﬁable unit due to the lack of infor- mation on LimeX, and so a simpliﬁed approach was taken in which positive and negative values were assigned based on the information obtained for each individual metric. 11 . For example, if the development of LimeX plastic is safer than the development of conventional plastic (no quantiﬁable measurement), the score for the LimeX plastic safety metric would be slightly higher than the conventional plastic safety metric based on the conclusions of the study taken. The quantiﬁable measurements can be scored directly based on the measured values and what’s being measured. This is explained in depth in the next section. ProdSI = 1 3(Ec + Ev + Sc) (1) Ec + Ev + Ec = 1 3 3 i=1wiCi + 1 3 7 i=4wiCi + 1 3 10 i=8wiCi (2) ProdSI = 1 3(Ec + Ev + Sc) (1) (1) Ec + Ev + Ec = 1 3 3 i=1wiCi + 1 3 7 i=4wiCi + 1 3 10 i=8wiCi (2) (2) Here, Ec, Ev, Sc, wi, and Cj are the sub-index score for economic impact, sub-index score for environmental impact, sub-index score for societal impact, weight of the ith metric and score of jth metric, respectively. 3.1 Selection of Individual Metrics Firstly, the individual metrics chosen in this study for sustainability evaluation were determined based on elements related to the manufacturing of paper and plastic products. The ProdSI application was simpliﬁed (not concentrating on clusters and sub-cluster) to only identify metrics that will provide a general framework to measure performance of the two products under review towards achieving sustainability goals. To start, the key metrics in determining environmental sustainability were identiﬁed as energy use, waste generated, and resource efﬁciency. Next, the economic impacts sub-index is assessed based primarily on the initial investments, costs, beneﬁts and losses. Finally, the societal impacts sub-index was assessed by primarily focusing on safety and health impacts, functional performance, product end-of-life (EOL) management, and product impact. These individual metrics were chosen based on each focus and had to be a quantiﬁable value with a unit or have shown a clear advantage/disadvantage in terms of sustainability. 1087 Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX 4 Sustainability Results and Discussion Sub-Index LimeX Metric Score Paper Metric Score LimeX SubIndex Score Paper Sub-Index LimeX ProdSI Score Score Paper ProdSI Score Env 7,6,7 4.5,4,5 6.67 4.5 Econ 6,5,6 5,7,8 5.67 6.67 6.78 6.22 Soc 7,6,7,4 6.5,6,3,7 8 7.5 Table 4. ProdSI Computation. 4.1 Environmental Impacts The energy consumption for LimeX is less than that of paper because LimeX products use more renewable resources and less non-renewable resources than conventional paper and plastic [13]. Secondly, the waste generation metric in the environmental section of the ProdSI was given a score of 6 and 4 because LimeX products emit less CO2 than conventional paper and plastic [14]. Lastly, there is the material use metric. The score given for material use was a 7 and 5 because to date, a higher percentage of LimeX products are recycled than conventional paper and plastics [14]. 4 Sustainability Results and Discussion As stated in the introduction, the aim of this paper was to investigate LimeX products and calculate a ProdSI to evaluate its sustainability. The ProdSI will be simpliﬁed due to the lack of life cycle data available for the material. The individual metrics generated by the ProdSI will be utilized to quantitatively evaluate and compare its environmental, societal, and economic impacts with that of conventional paper and plastic. This section will include a breakdown of each metric, sub-index and ProdSI score. The physical quantiﬁcation of individual metrics cannot be grouped together directly, therefore the metrics must be converted into a single normalized scale [3]. In this study, rather than normalizing measured values for the individual metrics [3], a score from 0–10 where 0, 2, 4, 6, 8, and 10 are the worst case, bad, below average, average good, and best respectively was utilized to place the individual metrics on a dimensionless scale. The scores are based on subjective factors speciﬁcally pertaining to the importance of each metric in terms of each sub-index. A standard normalization method for the conversion does not exist [12]. In this context, the score of each metric was determined based on the value of each measurement, based on the advantage and/or disadvantage of each measured value pertaining to the overall sustainability, and how relevant the metric is to the sustainability performance. The individual metrics were weighed equally for simplicity. Equation 1 and 2 were used to compute the sub-index and ProdSI scores (Table 3 and 4). 1088 E.-L. Thornton et al. Table 3. ProdSI Metrics selected for analysis. Sub-Index Metric 1 Metric 2 Metric 3 Metric 4 Env Energy use Waste Generation Material Use Econ Costs Beneﬁts/Losses Investments Soc Safety/Health Function Production EOL Product Impact Table 4. ProdSI Computation. Sub-Index LimeX Metric Score Paper Metric Score LimeX SubIndex Score Paper Sub-Index LimeX ProdSI Score Score Paper ProdSI Score Env 7,6,7 4.5,4,5 6.67 4.5 Econ 6,5,6 5,7,8 5.67 6.67 6.78 6.22 Soc 7,6,7,4 6.5,6,3,7 8 7.5 Table 3. ProdSI Metrics selected for analysis. Table 3. ProdSI Metrics selected for analysis. Sub-Index Metric 1 Metric 2 Metric 3 Metric 4 Env Energy use Waste Generation Material Use Econ Costs Beneﬁts/Losses Investments Soc Safety/Health Function Production EOL Product Impact Table 4. ProdSI Computation. Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX 1089 Evaluating the Sustainability of Paper and Plastic Substitute Material LimeX 5 Conclusion The ProdSI method adapted from previous methods in this paper compares the paper and plastic substitute material LimeX to currently used conventional paper/plastic products. Results from the simpliﬁed ProdSI indicate the sustainability performance of LimeX, producing a ProdSI of 6.78, was better than conventional paper and plastics, with a ProdSI of 6.22. However, since LimeX products are relatively new there has been little to no evidence of a LimeX lifecycle and discarding of LimeX products in the US. Results from the experimental analysis in Sect. 2 indicate that LimeX paper is stronger than conventional paper and that the material is primarily limestone as claimed by TBM [9]. Further research is required to validate long-term sustainability of LimeX. In future work, a life-cycle analysis of LimeX will be conducted in the US. 4.3 Societal Impacts In the ﬁnal sub-index section, safety and health metric in the societal impacts section of the ProdSI was given a score of 7 and 6.5 because the operator safety/health for LimeX products was reported to be higher and better than for conventional paper. Afterwards, the functional performance metric in the societal impacts section of the ProdSI was given a score of 6 and 6 because LimeX is stronger than conventional paper. Lastly, the product EOL management metric in the societal impacts section of the ProdSI was given a score of 7 and 3 because LimeX products are much more recyclable than conventional paper and plastics. Ultimately, the product social impact metric in the product societal impacts section was given a score of 4 and 7 because LimeX products are not visible globally, being relatively new. 4.2 Economic Impacts The cost metric in the economic impacts section of the ProdSI was given a score of 6 and 5 because the cost of conventional paper was more than the cost of LimeX products and because there are ﬂuctuations in the cost of the oil-based resin used in conventional plastic where this is not the case for LimeX products. Additionally, the beneﬁts/losses metric in the economic impacts section of the ProdSI was given a score of 5 and 7 because there are already many recycling facilities for conventional paper and plastic in the US, but there are no recycling facilities for LimeX products. Although LimeX does not use trees in its lifecycle and uses very little water, it does require the mining of limestone. Finally, the initial investment metric in the economic impacts section of the ProdSI was given a score of 6 and 8 because very little investments are required for conventional paper/plastics products that are already the norm whereas there is no infrastructure for LimeX. References 1. Song, J., et al.: Biodegradable and compostable alternatives to conventional plastics. Philos. Trans. R. Soc. B: Biol. Sci. 364(1526), 2127–2139 (2009) 1. Song, J., et al.: Biodegradable and compostable alternatives to conventional plastics. Philos. Trans. R. Soc. B: Biol. Sci. 364(1526), 2127–2139 (2009) 2. Huang, A., Badurdeen, F.: Sustainable manufacturing performance evaluation: Integrating product and process metrics for systems level assessment. Procedia Manuf. 8, 563–570 (2017) 2. Huang, A., Badurdeen, F.: Sustainable manufacturing performance evaluation: Integrating product and process metrics for systems level assessment. Procedia Manuf. 8, 563–570 (2017) 3. Shuaib, M., et al.: Product sustainability index (ProdSI) a metrics-based framework to evaluate the total life cycle sustainability of manufactured products. J. Ind. Ecol. 18(4), 491–507 (2014) 4. Zhang, X. et al.: A metrics-based methodology for establishing product sustainability index (ProdSI) for manufactured products. In: Dornfeld, D., Linke, B. (eds) Leveraging Technology for a Sustainable World, p. 435-441. Springer, Berlin (2012). https://doi.org/10.1007/978-3- 642-29069-5_74 5. Zhang, X., et al.: On improving the product sustainability of metallic automotive components by using the total life-cycle approach and the 6R methodology (2013). https://doi.org/10. 14279/depositonce-3753 5. Zhang, X., et al.: On improving the product sustainability of metallic automotive components by using the total life-cycle approach and the 6R methodology (2013). https://doi.org/10. 14279/depositonce-3753 6. Fiksel, J., McDaniel, J., Spitzley, D.: Measuring product sustainability. J. Sustain. Prod. De 7–18 (1998) 6. Fiksel, J., McDaniel, J., Spitzley, D.: Measuring product sustainability. J. Sustain. Prod. Des. 7–18 (1998) 7. Shuaib, M., et al.: 15.6 Sustainability evaluation using a metrics-based Product Sustainability Index (ProdSI) methodology–a case study of a consumer electronics product 7. Shuaib, M., et al.: 15.6 Sustainability evaluation using a metrics-based Product Sustainability Index (ProdSI) methodology–a case study of a consumer electronics product 1090 E.-L. Thornton et al. 8. Jawahir, I., et al.: Total life-cycle considerations in product design for sustainability: a frame- work for comprehensive evaluation. In: Proceedings of the 10th International Research/Expert Conference, Barcelona, Spain, Citeseer (2006) p ( ) 9. TBM 10. Paper Properties. Paperonweb. Properties of Paper 7 Dec 2022. https://www.paperonweb. com/paperpro.htm 11. Hapuwatte, B., et al.: Total life cycle sustainability analysis of additively manufactured products. Procedia CIRP 48, 376–381 (2016) 12. Böhringer, C., Jochem, P.E.: Measuring the immeasurable—a survey of sustainability indices. Ecol. Econ. 63(1), 1–8 (2007) 13. Narita, N., Sagisaka, M., Inaba, A.: Life cycle inventory analysis of CO2 emissions manufacturing commodity plastics in Japan. Int. J. Life Cycle Assess. References 7(5), 277–282 (2002) 14. Chasan, E.E.G.B.: There’s Finally a Way to Recycle the Plastic in Shampoo and Yogurt Packaging P&G’s technology solves polypropylene’s smell problem (2019) Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
https://openalex.org/W4391220877	http://asiacall-acoj.org/index.php/journal/article/download/54/29	English	null	Translation in Language Teaching - The Need for Redefinition of Translation	AsiaCALL online journal	2,024	cc-by	8,002	AsiaCALL Online Journal Received: 04/10/2023 ISSN 1936-9859; https://asiacall-acoj.org Revision: 19/01/2024 Vol. 15, No. 1, 2024 Accepted: 24/01/2024 pp. 19-33 Online: 25/01/2024 Translation in Language Teaching - The Need for Redefinition of Translation Nguyen Thi Thu Huong1* 1 The University of Foreign Language Studies, the University of Da Nang, Viet Nam * Corresponding author’s email: ntthhuong@ufl.udn.vn * https://orcid.org/0000-0001-8032-2701 https://doi.org/10.54855/acoj.241512 ®Copyright (c) 2023 Nguyen Thi Thu Huong Abstract AsiaCALL Online Journal Received: 04/10/2023 ISSN 1936-9859; https://asiacall-acoj.org Revision: 19/01/2024 Vol. 15, No. 1, 2024 Accepted: 24/01/2024 pp. 19-33 Online: 25/01/2024 Translation in Language Teaching - The Need for Redefinition of Translation Nguyen Thi Thu Huong1* 1 The University of Foreign Language Studies, the University of Da Nang, Viet Nam * Corresponding author’s email: ntthhuong@ufl.udn.vn * https://orcid.org/0000-0001-8032-2701 https://doi.org/10.54855/acoj.241512 ®Copyright (c) 2023 Nguyen Thi Thu Huong Abstract AsiaCALL Online Journal Received: 04/10/2023 ISSN 1936-9859; https://asiacall-acoj.org Revision: 19/01/2024 Vol. 15, No. 1, 2024 Accepted: 24/01/2024 pp. 19-33 Online: 25/01/2024 Translation in Language Teaching - The Need for Redefinition of Translation Nguyen Thi Thu Huong1* 1 The University of Foreign Language Studies, the University of Da Nang, Viet Nam * Corresponding author’s email: ntthhuong@ufl.udn.vn * https://orcid.org/0000-0001-8032-2701 https://doi.org/10.54855/acoj.241512 ®Copyright (c) 2023 Nguyen Thi Thu Huong Abstract Translation has been long excluded from language classes due to its association with the drawbacks of the Grammar Translation Method (GTM). However, in recent years, translation has made a comeback with its positive impact on students’ language performance. The effect of translation on the learner’s language knowledge has been finite compared with other language activities. The paper presents a critical analysis of research on translation in language teaching. It is found that the limited results of studies on the impact of translation may involve linguistic focus. Generally, most of the studies did not focus on the meaningfulness of linguistic items, which can only be achieved by placing them in context. In other words, the linguistic view of translation is still common in studies supporting translation. The redefinition of translation should incorporate communicative aspects of translation and the function of translation in a communicative context. The literature-based paper reinstates the importance of a communicative view of translation in language teaching, and it makes proposals related to the inclusion of translation in language learning classes. Keywords: Translation, language learning, communicative view AsiaCALL Online Journal ISSN 1936-9859; https://asiacall-acoj.org Vol. 15, No. 1, 2024 pp. CITATION\| Nguyen, T. T. H. (2024). Translation in Language Teaching - The Need for Redefinition of Translation. AsiaCALL Online Journal, 15(1), 19-33. DOI: https://doi.org/10.54855/acoj.241512 AsiaCALL Online Journal Received: 04/10/2023 ISSN 1936-9859; https://asiacall-acoj.org Revision: 19/01/2024 Vol. 15, No. 1, 2024 Accepted: 24/01/2024 pp. 19-33 Online: 25/01/2024 Translation in Language Teaching - The Need for Redefinition of Translation Nguyen Thi Thu Huong1* 1 The University of Foreign Language Studies, the University of Da Nang, Viet Nam * Corresponding author’s email: ntthhuong@ufl.udn.vn * https://orcid.org/0000-0001-8032-2701 https://doi.org/10.54855/acoj.241512 ®Copyright (c) 2023 Nguyen Thi Thu Huong Abstract 19-33 Received: 04/10/2023 Revision: 19/01/2024 Accepted: 24/01/2024 Online: 25/01/2024 Translation in Language Teaching - The Need for Redefinition of Translation Nguyen Thi Thu Huong1* 1 The University of Foreign Language Studies, the University of Da Nang, Viet Nam * Corresponding author’s email: ntthhuong@ufl.udn.vn * https://orcid.org/0000-0001-8032-2701 https://doi.org/10.54855/acoj.241512 ®Copyright (c) 2023 Nguyen Thi Thu Huong Abstract Translation in Language Teaching - The Need for Redefinition of Translatio Nguyen Thi Thu Huong1* g g g Nguyen Thi Thu Huong1* Translation has been long excluded from language classes due to its association with the drawbacks of the Grammar Translation Method (GTM). However, in recent years, translation has made a comeback with its positive impact on students’ language performance. The effect of translation on the learner’s language knowledge has been finite compared with other language activities. The paper presents a critical analysis of research on translation in language teaching. It is found that the limited results of studies on the impact of translation may involve linguistic focus. Generally, most of the studies did not focus on the meaningfulness of linguistic items, which can only be achieved by placing them in context. In other words, the linguistic view of translation is still common in studies supporting translation. The redefinition of translation should incorporate communicative aspects of translation and the function of translation in a communicative context. The literature-based paper reinstates the importance of a communicative view of translation in language teaching, and it makes proposals related to the inclusion of translation in language learning classes. Keywords: Translation, language learning, communicative view Introduction The translation was well-known in language teaching in the 18th and 19th centuries through the GTM, which prioritized students' literature reading skills and their grammatical knowledge over oral skills. During the 20th century, a variety of teaching methods stressed oral skills, and translation lost favor in second-language pedagogy. In many classrooms, translation activities associated with the GTM disappeared. More recently, however, the role of translation in language classes has been reconfigured (Beecroft, 2013; Bonyadi, 2003; Carreres, 2014; Carreres & Noriega-Sánchez, 2011; Cook, 2010; Leonardi, 2010; Leonardi & Salvi, 2016). A myriad of studies have shown that translation can be an effective means of developing students’ language knowledge and skills (Belpoliti & Plascencia, 2013; Pariente-Beltran, 2013). In the meantime, translation can be considered a worthwhile skill in itself when learners use it in their daily and professional lives (Cook, 2010; Duff, 1989; Klein-Braley, 1996). Some authors, including Carreres (2014) and Leonardi (2010), also argue that translation should be regarded as the "fifth skill" in language learning in addition to speaking, listening, reading, and writing. Carreras (2014) indicates that there is a blurred line between translation as a means (to ACOJ- ISSN 1936-9859 AsiaCALL Online Journal Vol. 15; No. 1; 2024 help students improve their language learning) and an end (students knowing how to translate) in language teaching. A number of studies conducted in undergraduate language programs have attempted to enhance students' knowledge of and skills in Translation (Translation as an end) (see Section 2.2.2). While there is insufficient research related to the shift from the traditional view of translation to the more communicative one, this paper attempts to explain this tendency. Specifically, this paper which reviews relevant journal articles and books describes a resurgence of translation in language teaching and re-examines how the role of translation is viewed and how it is taught in different contexts, which calls for the need to redefine translation in language teaching. Literature review Language teaching and translation teaching Pedagogical Translation vs. Translation Pedagogy Pedagogical Translation vs. Translation Pedagogy A number of scholars (Davies & Kiraly, 2006; Holmes, 1994, 2004; Vermes, 2010) have agreed on a clear demarcation between pedagogical translation and translation pedagogy as they differ, mainly in function and audience. Firstly, the former kind of translation is a means of enhancing foreign language knowledge and skills and is a well-known concept in language teaching, while the latter equips students with knowledge, skills, and principles of translation as a profession and relates to professional translator training. Secondly, students who do pedagogical translation do not have to deal with any client or reader who needs to understand the target language (TT); the teacher, however, may use translation to assess students’ comprehension and/or translation ability. This kind of translation is not a communicative activity in which a translator should fulfill the task of linguistic and cultural mediation, taking into account the needs of their readers or clients. Some researchers advocate that translation in language teaching should reflect the principles of professional translation and enable learners to translate rather than just develop L2 competence (Carreres, 2014; Carreres & Noriega-Sánchez, 2011; Colina, 2002; Cook, 2010; Duff, 1989; Klein-Braley, 1996; Suparmin, 2003). Cook (2010, p. xx) argues that language learners need to be able to translate because translation is "part of everyday bilingual language use" both personally and professionally. Klein-Braley (1996, p. 24) adds that the aim of language courses "must be to enable all-round language professionals to tackle translation themselves for in- house and informal purposes, and also-and importantly-to supervise the translation of texts for public and formal purposes". Similarly, according to Duff (1989), who sees that translation is natural and necessary in the real world, when learners are introduced to a variety of oral and written text types, registers, and styles in translation, they will also develop the skills necessary in second language acquisition and the multilingual world. The separation of translation as a means to teach languages and translation being taught as an end or a skill/profession is a result of diverging views about translation. Advocates of pedagogical Translation see translation as a linguistic activity in which learners learn language features by comparing and contrasting the first language (L1) and the second language (L2). On the other hand, through translation pedagogy, learners should learn to translate an ST, keeping in mind the target reader and the purpose of the translation. This means that translation 20 Nguyen Thi Thu Huong Vol. Pedagogical Translation vs. Translation Pedagogy 15; No. 1; 2024 https://asiacall-acoj.org is seen as an act of communication in translation pedagogy. In an article titled "Translation as a means and as an end: Reassessing the divide", Carreres (2014) maintains that if we consider translation as a means to language learning and as a skill as two independent activities, we seem to acknowledge the "disconnection of translation from its natural goal of communication, which characterized the grammar-translation method" (p. 130). Carreres argues that “translation as a means is at its most effective and stimulating when learning objectives and pedagogical design are brought as close as possible to the realities of professional translation – that is, to translation as an end in itself” (p. 130). Meanwhile, Carreres indicates that the training of professional translators should also take into account the improvement of learners' knowledge of the languages involved. In agreement with Carreres, who sees no difference between translating in language teaching and translation education, Colina and Lafford (2018) argue that the dichotomies, including pedagogical translation vs. translation pedagogy and language learning vs. translation learning, have "prevented fruitful interaction between these fields [translation studies and language acquisition], as the areas of overlap have been minimi(s)ed by prescriptivist and over-simplistic approaches" (p. 3). Therefore, I maintain that teaching translation in a language teaching context should not be only limited to pedagogical Translation (Translation as a means), but teaching translation should promote language learners’ ability to translate (Translation as an end), adopting the view of translation as a communicative activity. The argument is supported throughout the rest of the paper. It will be first consolidated by a description of the similarities between language learning and translating, given the two activities are naturally part of communication. The next section compares language learning and translating before presenting the impact of language learning on translating. Language learning and translating For instance, in comprehending a job advertising text, it is necessary for learners to assess the situation (e.g., the author, our background knowledge, and our reading expectations) before planning available resources (e.g., knowledge of vocabulary and the subject matter relevant to job descriptions) and adopting different reading strategies to achieve the goal of comprehension. In response to the advert, learners may write a letter of job application. Similar processes (i.e., assessment, planning, and execution) are involved. Learners should assess the communicative situation (e.g., who the reader is—the employer and what the purpose of the communication is—applying for a job) so that they can plan and execute relevant linguistic features and conventions of an application letter to succeed in applying for the job. Similar phases of assessment, planning, and execution take place in the act of translating. In translating an advert, translators or translation students first assess the situation of the source text or ST (available from ST features such as author and place of publication) before planning and executing their available language competencies (linguistic and extralinguistic knowledge) to understand the ST. Then they go on to the production of the translation or the TT, which also involves assessment (the context of the TT is considered), planning, and execution (linguistic choices are considered and decided on depending on the target reader and the context of the translated advert). Despite such similarities, language learning and translating differ in that the language learner directly produces a speech act in a communicative context, while the translator produces a speech act that is dependent on the context of its translation. In other words, both activities are focused on functional language use or functions of language. In language learning, comprehension may not necessarily be followed by production (reading the job advert is independent of writing a job application). If "production does follow comprehension, it is a reaction with a different Message, not a reformulation of the same Message" (Gile, 2009, p. 106). The production of a job application letter, which follows the act of reading, has a different message from the advert text. In the meantime, in translating, "comprehension and production follow each other systematically and act on the same message" (Gile, 2009, p. 106). In other words, the main difference between the two activities lies in the mediation of the ST in translating. Language learning and translating Language learning and translating Language learning and translating both emphasize learners' ability to communicate (Carreres & Noriega-Sánchez, 2011; Colina, 2002). Colina (2002) relates the concept of communicative language competence by Savignon (1972, 1983) and Lee and VanPatten's (1995) to communicative translation competence by Kiraly (1995). She maintains that both activities aim at improving learners' ability to express, interpret, and negotiate meaning in communicative situations. In other words, communicative language competence is achieved by interacting with the input rather than learning formal aspects of the language, while communicative translation competence enables learners to act appropriately in communicative translation tasks. In the same vein, Carreres (2014) referred to the concept of communicative competence developed by Canale (1983) and Canale and Swain (1980), including: grammatical competence (the knowledge of linguistic aspects of languages), sociolinguistic competence (understanding of the social contexts and cultures), discourse competence (knowledge of text types and text type conventions) and strategic competence (strategies to enhance communication). Carreres affirms its closeness with translation competence, particularly as defined by Kelly (2005, pp. 32-33): “macrocompetence that comprises the different capacities, skills, knowledge and even attitudes that professional translators possess and which are involved in translation as an expert activity”. Kelly’s notion includes aspects such as communicative and textual competence in at least two languages and cultures, cultural and intercultural competence, subject area competence, and 21 ACOJ- ISSN 1936-9859 AsiaCALL Online Journal Vol. 15; No. 1; 2024 strategic competence (e.g., problem identification and solving and self-assessment). Kelly also refers to professional and instrumental competence (use of documentary resources including dictionaries), psycho-physiological competence (e.g., self-confidence and memory), and interpersonal competence (e.g., teamwork and negotiation skills). The strategic processes of language learning and translating are also similar. According to Bachman (1990), a language learning activity includes assessment, planning, and execution. In the assessment phase, language learners identify the situation in which communication takes place and determine the type of language competencies needed to achieve the goal of communication. The planning component allows them to retrieve relevant items from language competence, including linguistic and extralinguistic knowledge. Then they execute the plan in an appropriate way to fulfill the communication goal(s) (neurological and psychological processes involved). Impact of language learning on translation Given the analogy between language learning and translating, language learners are expected to assess, plan, and execute the translation task in a communicative way. However, in practice, they lack the awareness of pragmatic features or the situations of communication to perform a communicative translation task. Colina (2002) and Pariente-Beltran (2013) explain that while many language programs are oriented to communicative practices, the GTM is still prevalent, and learners' communicative competence (particularly sociolinguistic, discursive, and strategic competencies) is not prioritized. As a result, learners lack an awareness of the communicative situation or pragmatic features in achieving the communication goal(s). Even when being taught by CLT methods, language learners usually have deeply rooted misconceptions about translation, and they consider translation as an activity in which they mainly achieve linguistic equivalence and/or grammatical correctness between the ST and the TT (Colina, 2002). Paying inadequate attention to ST and TT features, language learners may not properly understand ST meanings constrained by the author's intention or text type; neither do they produce a translation that meets the requirements of the TT. Influenced by their language learning experience and/or their pre-existing conceptions about translation, language learners lack an awareness of the translation process as firmly supported by various studies on the behaviours of language learners and translators in Translation (Alves, 2005; Jonasson, 1998; Lörscher, 2005; Okonska & Kościałkowska-Okońska, 2013; Olk, 2001; Tirkkonen-Condit, 2005). Language learners often try to reproduce the ST as closely as possible, which does not always fulfill the communicative function of the translation. Unlike professionals who consider the situation of text and specific linguistic features in their top-down approach, language learners rely more on bottom-up processing, which is focused on the local level of words or phrases (Okonska & Kościałkowska-Okońska, 2013, pp. 227-228). They often rely on a heavy use of dictionaries to check the meanings of these lexical items (González- Davies, 2004). In other words, language learners take the “form-oriented approach in that they produce translations mainly by an exchange of language signs” (Lörscher, 2005, p. 605). This downplays the meaning of the translated text. They rarely use "sense-oriented procedures" (p. 605), which the translator frequently adopts to avoid the distortion of sense and violations of TL norms. Monitoring or checking the output with reference to the text type and discursive features is also lacking in their translation processes. and cultures to facilitate target readers’ comprehension of the ST. While the noticeable difference between language learning and translating involves the presence and the importance of the ST in Translation, the two activities can inform each other. It is assumed that language learners who have communicative language competence and an awareness of pragmatic features in communicative contexts are expected to perform well in translation. However, this is not always the case. Therefore, it is of great importance to understand language learners' behaviors in translation to apply appropriate pedagogical intervention in translation teaching (Colina, 2002). Language learning and translating The translator not only reads the text for their own comprehension, but they must also send the ST content and message(s) to target readers. Functioning as both the receiver and the sender of the ST, the translator must understand the ST and mediate differences in languages 22 Nguyen Thi Thu Huong Vol. 15; No. 1; 2024 https://asiacall-acoj.org and cultures to facilitate target readers’ comprehension of the ST. Impact of language learning on translation In general, language learners' lack of awareness of texts and discursive features means that they disregard contextual factors in translating ST micro elements. Therefore, in seeking an effective method in teaching translation, language learners should be encouraged to avoid a total reliance on literal 23 ACOJ- ISSN 1936-9859 AsiaCALL Online Journal Vol. 15; No. 1; 2024 translation, unlearn their misconceptions about translation, and develop an understanding of the translation process. The discussion of the kinship between language learning and translating demonstrates the need to see translation as communication in language teaching and emphasizes language learners' ability to translate. This would, to some extent, resolve the differences between pedagogical translation and translation pedagogy. The urge to view translation as a communicative act will also be illustrated in the following review of translation in language teaching. Specifically, I elucidate how the status of translation has evolved in language teaching and how the issue of teaching translation has been discussed in language teaching in the next sections. Translation in language teaching Long-term neglect of translation in language teaching The translation was first used to teach languages in the 18th and 19th centuries with the rise of the GTM; it went out of favor after the advent of subsequent language teaching methods, such as the Direct Method, the Audiolingual Method, and the Communicative Approach. The GTM promoted the students' ability to read classical literature rather than speak the target language (Richards & Rodgers, 2001). In the GTM, grammar rules were presented in the L1 in a graded sequence and were exemplified through the translation of isolated sentences. Lists of decontextualized vocabulary items, along with their corresponding equivalents, were also memorized. With the advent of language teaching methods that prioritized speech and oral practice, translation disappeared from language classes. The Direct Method advocated that L2 learning should resemble the learning of the L1 as much as possible. Therefore, teachers were advised not to include translation in their language class due to its lack of focus on oral language and the excessive use of isolated and artificial sentence structures (Newson, 1998). Similarly, translation was excluded from the Audiolingual Method, which was based on the structuralist linguistic paradigm and behaviorism. Particular dialogues and structures were learned by imitation, repetition, and memorization. Impact of language learning on translation The L1 was considered to be a source of interference with L2 acquisition because, according to behaviorism, new L2 habits are highly influenced by old L1 habits. If there is heavy use of L1 in L2 language classes, there will be L1 interference or a harmful effect on students' acquisition of native-like language proficiency (Cook, 2010; Lado, 1964; Newson, 1998). In the 1970s, when the Communicative Approach focused on functional categories or the purpose of communication rather than language structures, the translation of isolated sentences played a minimal part in enabling learners to express meanings and develop their' communicative competence (Laviosa, 2014). In other words, translation was only considered acceptable when it facilitated the learning process or enabled students to understand class instructions. In fact, translation was not considered a communicative activity in communicative language teaching. Leonardi and Salvi (2016) summarise the limitations of translation teaching associated with the GTM: its excessive reliance on learners’ L1, its neglect of oral language, and its use of artificial sentences to illustrate grammatical points. With these focuses, the learners’ acquisition of native-like competence and their development of communicative competence are hindered. 24 Vol. 15; No. 1; 2024 Nguyen Thi Thu Huong https://asiacall-acoj.org Malmkjær (1998) lists other arguments against translation in language teaching, claiming that translation - is independent of the four skills which define language competence: reading, writing, speaking and listening - is radically different from the four skills - is radically different from the four skills - takes up valuable time which could be used to teach these four skills i l - takes up valuable time which could be used to teach these four skills - is unnatural - misleads students into thinking that expressions in two languages correspond one-to-one - prevents students from thinking in a foreign language - misleads students into thinking that expressions in two languages correspond one-to-o - prevents students from thinking in a foreign language - prevents students from thinking in a foreign language - produces interference - is a bad test of language skills - is only appropriate for training translators. (p. 6) Many of the above-mentioned arguments are challenged by studies reviewed in Section 2.2.2. The major reason for the objection of translation in language teaching can be somewhat accounted for by its connection with the GTM (Carreres & Noriega-Sánchez, 2011; House, 2008; Tsagari & Phlōros, 2013; Vermes, 2010; Widdowson, 2003). Impact of language learning on translation There have been “equally fallacious interpretations of the translation task as the common attempt of finding lexical and structural correspondences among L1 and L2 (grammar-translation)” (Tsagari & Phlōros, 2013, p. vii). The role of the view of translation will be discussed in the next section, which describes the revival of translation in language teaching. Resurgence of Translation and the need for redefinition Resurgence of Translation and the need for redefinition The author did not emphasize the contextual factors in students' translation, even though they briefly referred to social contexts in the use of taboos. Hummel (2010) experimented with three conditions: French L1 to English L2 translation, English to French Translation, and a rote-copying task. All the students were first given 15 unknown English words. Students of the first two conditions translated the English words into their L1, while those in the last group copied the words. After that, all the students were provided with booklets that contained L2 equivalents for the English words. The first two groups were asked to translate the provided sentences that contained the words into the L2 and L1, respectively, while the rote-copying group just copied the L2 sentences that contained the English words. Without previously having been informed that they would be tested on these words, all the students were then asked to recall the introduced words in a follow-up test. Hummel (2010) noticed that the three conditions all increased the learners' short-term lexical recall, but rote copying had a more significant impact. The author explains that the students in translation groups put more effort into translating the sentences in time-consuming tasks than in remembering the words, while the students copying sentences had more time to concentrate on memorizing them. This implies that the author adopted the perception of translation as a linguistic activity that focuses on contrasting isolated ST and TT items. Korošec (2013) examined the role of Translation in English learning of first-year Slovene students. An experimental group was taught with the use of L1 and Translation. The control group was taught without L1 or Translation (there is no information about the teaching method used for this group). The teacher in the experimental group used the students' L1 to explain grammatical rules and provided them with homework in which they translated lexical items (e.g., nouns), sentences, or shorter texts that targeted particular structures from L1 into L2. The students then discussed these exercises in class. Even though the researcher mentioned discussions of appropriate solutions in contexts, there was not a clear elaboration on which factors they should focus on. The study revealed the usefulness of L1 and Translation to students' learning of linguistic knowledge. However, no significant differences in test results between the two groups were found. Resurgence of Translation and the need for redefinition After decades of being downplayed in language teaching due to its association with the GTM, Translation has been reinstated in language teaching (Cook, 2010; House, 2008; Laviosa, 2014). At the least, it does no harm to language development and does not interfere with L2 acquisition (Duff, 1989; Harvey, 1996; Malmkjaer, 1998). For example, Duff (1989) maintains that translation enables students to be aware of the L1 influence on L2 and to deal with problems caused by the interference. It enhances learners’ linguistic knowledge (Belpoliti & Plascencia, 2013; Hummel, 2010; Korošec, 2013; Pariente-Beltran, 2013; Valdeón, 2015), and develops their language skills (D’Amore, 2014; Kim, 2011; Lee, 2013). Translation enables students to develop confidence and self-esteem, and it is favorably regarded by both students and teachers (Canga-Alonson & Rubio-Goitia, 2016; Dagilienė, 2012; Fernández-Guerra, 2014; Kelly & Bruen, 2015; Kokkinidou & Spanou, 2013; Murtisari, 2016). The matter of concern is not whether translation is useful or not, but rather, how to use it for what purpose and to what effect. I explore these issues further by examining recent empirical studies on translation in language teaching. Belpoliti and Plascencia (2013) experimented with English-to-Spanish translation activities in which students were asked to read and analyze two newspaper articles (the original text in English and the Translation in Spanish). Groups of students found five pairs of equivalents in the two texts. The groups then looked for metaphors in the translated text and checked if they had equivalents in the English text. The students' work was collected, and they discussed the exercises about problems in translation (e.g., word search) with their instructor. They finally 25 ACOJ- ISSN 1936-9859 AsiaCALL Online Journal Vol. 15; No. 1; 2024 summarized the vocabulary learned. The authors found that learners improved their lexical awareness, but the effect was finite. summarized the vocabulary learned. The authors found that learners improved their lexical awareness, but the effect was finite. Valdeón (2015) asked students to render two episodes of a British sitcom into Spanish. The author noticed that the students’ translations of taboos in the texts were close to the original. Valdeón found that the students were able to understand L1 and L2 differences in the use of taboos, and they were cautious in treating them in translation. Valdeón did not have a clear explanation for the students' caution except for his explanation that taboos were common in Spanish. Resurgence of Translation and the need for redefinition The author attributed the limited results to methodological flaws (e.g., statistical measurements on small samples and tests implemented long after treatment) rather than their limited focus on contextual factors of translation exercises. Lee (2013) investigated whether translation enhanced students' scores on reading. The selected English and non-English-majors translated two passages of an English L2 reading passage into Chinese L1 before answering five multiple-choice questions in reading comprehension (three questions were concerned with the translated paragraphs). Even though the study suggested the benefits of translation in promoting students' text comprehension, the students did not have 26 Nguyen Thi Thu Huong Vol. 15; No. 1; 2024 https://asiacall-acoj.org correct answers for "text-based" or "inference-based" questions related to the meanings of words and phrases or general ideas expressed in paragraphs. This can be explained by students' lack of attention to textual features. In fact, in analyzing students' solutions, the researcher mentioned the application of the professional assessment criteria—accuracy and expression that focused on "student performance in accuracy, understanding of the original, expression, and language competence in using their mother tongue to represent the original meaning" (p. 15). This indicated the author's lack of emphasis on a communicative view of translation and contextualized translations. Therefore, these ideas could not have been engendered in the students. Another study by D’Amore (2014) on the role of translation in reading combines reading comprehension and translation in the form of contrastive analysis to improve students' English L2 reading skills in Mexico. Students discussed their comprehension of English texts before they translated them into Spanish. In class discussions, the students read aloud sentences they had translated. D'More explains that by reading aloud their translations, the students could improve the coherence of Spanish versions. D’Amore also indicates that literal translation highlighted structural differences. The author criticized the translation of isolated sentences as advocated in the GTM and referred to authentic texts and communicative, collaborative translation tasks. However, how communicative tasks were implemented was not described. Exercises that used authentic texts mainly focused on matching extracts with their book covers. In fact, D’Amore employed translation mainly as a tool to learn English vocabulary and grammar. Raw data indicated that during semesters from 2009 to 2013, more students achieved a higher level of reading (from B1 to B2+ and C1 as described in the European English proficiency framework). In a study that investigated the impact of translation on students’ writing skill, E. Resurgence of Translation and the need for redefinition Kim (2011) required students to translate verbatim sentences of their own writing in English L2 into Korean L1. Based on the students’ reflections after translating their own English writing and that of their peers’ into L1 or Korean, Kim concludes (with reference to anecdotal data) that translation enables the students to “look at their own writing more objectively, which they failed to do without the aid of their first language” (p. 158) and understand that “grammatical accuracy is vital for successful written communication” (p. 158). In other words, E. Kim employed translation as a linguistic activity to assist students in learning English. In the above studies in support of translation in language teaching, translation has been used as a contrastive analysis activity in which sentences in the L1 and the L2 are compared and contrasted. The purpose of such an activity has been to enable students to learn isolated vocabulary items and grammatical sentential structures rather than focusing on contextual features of translation tasks or treating translation as a skill by itself. Most studies have prioritized students' memory of linguistic items to their functional use of language. The effect of translation on the learner's language knowledge has been finite compared with other language activities. Authors of some of these studies have not presented sound methodologies (e.g., results obtained from raw data or anecdotal data) and they did not have satisfactory explanations for the limited results of their studies. Generally, most of the studies did not focus on the meaningfulness of linguistic items which can only be achieved by placing them in contexts. In 27 ACOJ- ISSN 1936-9859 AsiaCALL Online Journal Vol. 15; No. 1; 2024 other words, the linguistic view of translation is still common in studies supporting translation, which is similar to those against translation. In fact, House (2008) claims that a linguistic view of translation is held among those who plead against and for the use of translation in language teaching. other words, the linguistic view of translation is still common in studies supporting translation, which is similar to those against translation. In fact, House (2008) claims that a linguistic view of translation is held among those who plead against and for the use of translation in language teaching. Linguistic approaches to translation (based on contrastive linguistics) do not consider the contextual use of words and sentences. Resurgence of Translation and the need for redefinition It is important to note that translation is not the same as contrastive linguistics. Emphasizing the need to understand the differences between the two fields, House (2008) explains how "langue" or the language system differs from "parole" or concrete utterances in texts and insists that translation is performed at the level of parole rather than langue. This means that translators must choose among various target equivalents. According to House, "(w)hile contrastive linguistics tends to focus on the language system, translation is concerned with the reali(s)ation of that system in acts of communication" (p. 136). In fact, translation tasks based on contrastive analysis do not offer learners opportunities to improve their ability to achieve communicative goals in using a L2 language, as identified by Colina (2002), who mentions the drawback of associating translation with contrastive analysis as follows: Translation is often associated with contrastive analysis since translation exercises are used to learn structural differences between languages and to test the student's knowledge of these differences. Students exposed to behaviorist/contrastive methodologies learn about the language rather than how to use it: there is no creation or exchange of meaningful information. (p. 3) Colina (2002) explains that the GTM and audiolingual methodologies based their language learning theories on structuralism or behaviorism which focuses on linguistic forms rather than communicative aspects of language. Colina points out that due to the "formalist, non- communicative views of language" (p. 4), translation for the purpose of language learning, which is different from professional translation activity, has been frowned on by communicative language teachers. Notably, avoiding the contrastive or behaviorist approach to learning, an increasing number of scholars who recognize the synergies between the two fields suggest that translation in language teaching should be defined from the perspective of communication (Carreres & Noriega- Sánchez, 2011; Leonardi & Salvi, 2016). "Translating means mediating a message between two different linguistic and cultural communities, and the same applies to language learning" (Leonardi & Salvi, 2016, p. 336). In other words, both language learning and translation should aim at getting the message across over linguistic and cultural differences. Obviously, the need to see translation as a communicative activity is emphasized. Many recent studies on translation in language teaching, including Phat (2022), have emphasized the role of contextual factors and communicative aspects in teaching translation. Conclusion Translation as a means of language teaching has had a resurgence in recent decades. Concurrently, there has been a demand for training in translation as an end or a skill by itself to meet the need for translation services. The paper reported that the dominant view of translation among teachers and learners in language courses in language programs was a linguistic one. When compared to other language exercises, translation has a limited impact on the learner's linguistic understanding. Some of these studies' authors have not provided sound techniques (such as conclusions derived from raw data or anecdotal data) or adequate justifications for the study's restricted findings. The meaningfulness of linguistic items—which can only be attained by placing them in contexts—was generally not the focus of most studies. Due to the limited view of translation and similarities between translation and language learning, a number of scholars have advocated for translation both as a means and an end in language teaching and indicated similarities between language teaching and translation studies based on communicative views of language and translation. In tertiary language programs, it is necessary to broaden students’ view of language in English classes so that by the time they enter translation courses, they will have an understanding that a range of factors influences all language use. Language and translation learning should be driven by a common communicative view of language and translation, and the two should be mutually informative. Students will then be less focused on the idea of language as a linguistic code than they are now. This will facilitate the teaching and learning of translation in subsequent translation courses. It is the role of a language program to enable students to broaden their views of language and translation. Language teachers can be first introduced to diverse evidence for the resurgence of translation in language teaching so that they consider adopting a translation component in language classes. They should then be made aware of the role of a communicative view of translation in the teaching of translation before discussing the selection of texts for translation and the planning of effective translation activities in language teaching. More empirical research is needed to justify the usefulness of a communicative view of translation in yielding satisfactory results in teaching a language as a means and an end. Resurgence of Translation and the need for redefinition Generally, translation has made a comeback in language teaching, as evidenced by a myriad of studies on the use of translation in language teaching. However, the impact of translation in language teaching is still limited partly due to the adopted purely linguistic view of translation. Similarities between language learning and translation and discussions on the communicative aspects of the two types of activities can justify the proposal to define translation from the 28 Vol. 15; No. 1; 2024 Vol. 15; No. 1; 2024 Nguyen Thi Thu Huong https://asiacall-acoj.org Conclusion Studies that adopt this view of translation can be carried out in a larger scope and in greater depth. Comparison of the results of translation with those of other language learning activities in different contexts can be useful in enhancing the role of translation in language learning and teaching. This may help promote language proficiency in Vietnam, where foreign languages, particularly English, are used for different purposes (Truong & Le, 2022). 29 ACOJ- ISSN 1936-9859 Vol. 15; No. 1; 2024 AsiaCALL Online Journal References Alves, F. (2005). Bridging the gap between declarative and procedural knowledge in the training of translators: Meta-reflection under scrutiny. Meta: Journal des Traducteurs/Translators’ Journal, 50(4). doi: 10.7202/019861ar Bachman, L. F. (1990). Fundamental considerations in language testing. Oxford: Oxford University Press. Beecroft, R. (2013). From intercultural speaker to intercultural writer: Towards a new understanding of translation in foreign language teaching. In D. Tsagari & G. Phlōros (Eds.), translation in language teaching and assessment (pp. 155-172). Newcastle Upon Tyne: Cambridge Scholars Publishing. Belpoliti, F., & Plascencia, A. (2013). Translation techniques in the Spanish for heritage learners’ classroom: Promoting lexical development. In D. Tsagari & G. Floros (Eds.), translation in language teaching and assessment (pp. 65-92). Newcastle Upon Tyne: Cambridge Scholars Publishing. Bonyadi, A. (2003). Translation: Back from Siberia. Translation Journal, 7(3), 1983. Brender, A. S. (2002). The effectiveness of teaching articles to (-Art) ESL students in writing classes using consciousness-raising methods. (Doctoral dissertation, Temple University, America). Retrieved from ProQuest Dissertations and Theses. Canale, M. (1983). From communicative competence to communicative language pedagogy. Language and Communication, 1(1), 1-47. Canale, M., & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics, I(1), 1-47. doi: 10.1093/applin/I.1.1 Canga-Alonso, A., & Rubio-Goitia, A. (2016). Students' reflections on pedagogical Translation in Spanish as a foreign language. Tejuelo, 23(1), 132-157 Carreres, Á. (2014). Translation as a means and as an end: Reassessing the divide. The Interpreter and Translator Trainer, 8(1), 123-135. doi: 10.1080/1750399X.2014.908561 Carreres, Á., & Noriega-Sánchez, M. (2011). Translation in language teaching: Insights from professional translator training. The Language Learning Journal, 39(3), 281-297. doi: 10.1080/09571736.2011.567356 Colina, S. (2002). Second language acquisition, language teaching, and translation studies. Translator: Studies in Intercultural Communication, 8(1), 1-24. Colina, S., & Lafford, B. A. (2018). Translation in Spanish language teaching: The integration of a “fifth skill” in the second language curriculum. Journal of Spanish Language Teaching, 1-14. doi: 10.1080/23247797.2017.1407127 Cook, G. (2010). Translation in language teaching: An argument for reassessment. Oxford: Oxford University Press. 30 https://asiacall-acoj.org Nguyen Thi Thu Huong Vol. 15; No. 1; 2024 D’Amore, A. M. (2014). The role of translation in language teaching: Back to GT in ELT?. In Y. Cui & W. Zhao (Eds.), Handbook of research on teaching methods in language translation and interpretation (pp. 118-135). Hershey, PA: IGI Global. Dagilienė, I. (2012). Translation as a learning method in English language teaching. Studies About Languages, 21, 124-129. References doi: 10.5755/j01.sal.0.21.1469 Davies, M., & Kiraly, D. (2006). Translation: Pedagogy. In K. Brown (Ed.), Encyclopedia of Language and Linguistics (2nd ed., pp. 81-85). Cambridge, UK: Elsevier Ltd. Duff, A. (1989). Translation. Oxford: Oxford University Press. Fernández-Guerra, A. B. (2014). The usefulness of translation in foreign language learning: Students’ attitudes. IJ-ELTS, 2(1), 153-170. Gile, D. (2009). Basic concepts and models for interpreter and translator training (Rev. ed.). Amsterdam: John Benjamins Publishing. González-Davies, M. (2004). Multiple voices in the translation classroom: Activities, tasks and projects. Amsterdam: John Benjamins Publishing. Harvey, M. (1996). A translation course for French-speaking students. In P. Sewell & I. Higgins (Eds.), Teaching translation in universities: Present and future perspectives (pp. 45-65). London, UK: Assn. for French Language Studies & Centre for Information on Language Teaching & Research. Holmes, J. S. (1994). Translated! Papers on literary translation and translation studies (Vol. 7.). Amsterdam: Rodopi. House, J. (2008). Using translation to improve pragmatic competence. In E. Alco’n Soler & A. Martinez-Flor (Eds.), Investigating pragmatics in foreign language learning, teaching and testing, (pp. 3-25). Bristol, PA: Multilingual Matters. Hummel, K. M. (2010). Translation and short-term L2 vocabulary retention: Hindrance or help? Language Teaching Research, 14(1), 61-74. doi: 10.1177/1362168809346497 Jonasson, K. (1998). Degree of text awareness in professional vs. non-professional translators. In A. Beylard-Ozeroff, J. Králová & B. Moser-Mercer (Eds.), Translators’ strategies and Kelly, D. (2005). A handbook for translator trainers: A guide to reflective practice. Manchester, UK: St. Jerome Publishing. Kelly, N., & Bruen, J. (2015). Translation as a pedagogical tool in the foreign language classroom: A qualitative study of attitudes and behaviours. Language Teaching Research, 19(2), 150-168. doi: 10.1177/1362168814541720 Kim, E. (2011). Using translation exercises in the communicative EFL writing classroom. ELT Journal, 65(2), 154-160. doi: 10.1093/elt/ccq039 Klein-Braley, C. (1996). Teaching translation: A brief for the future. In P. Sewell & I. Higgins (Ed.), Teaching translation in universities: Present and future perspectives (pp. 15-30). London, UK: AFLS & CILT. Klein-Braley, C. (1996). Teaching translation: A brief for the future. In P. Sewell & I. Higgins (Ed.), Teaching translation in universities: Present and future perspectives (pp. 15-30). London, UK: AFLS & CILT. 31 ACOJ- ISSN 1936-9859 Vol. 15; No. 1; 2024 AsiaCALL Online Journal Kokkinidou, A., & Spanou, K. (2013). The didactic use of translation in foreign language teaching: A practical example. In D. Tsagari & G. Phlōros (Eds.), translation in language teaching and assessment (pp. 173-190). References Newcastle upon Tyne, UK: Cambridge Scholars Publishing. Korošec, M. K. (2013). Teaching grammar through translation. In D. Tsagari & G. Phlōros (Eds.), translation in language teaching and assessment (pp. 23-40). Newcastle upon Tyne, UK: Cambridge Scholars Publishing. Lado, R. (1964). Language teaching: A scientific approach. New York, NY: McGraw-Hill. Laviosa, S. (2014). Translation and language education: Pedagogic approaches explored. New Lee, T. (2013). Incorporating translation into the language classroom and its potential impacts upon L2 learners. In D. Tsagari & G. Floros (Eds.), translation in language teaching and assessment (pp. 3-22). Newcastle Upon Tyne, UK: Cambridge Scholars Publishing. Leonardi, V. (2010). The role of pedagogical translation in second language acquisition: From theory to practice. Bern: Peter Lang. Leonardi, V., & Salvi, R. (2016). Language pedagogy and translation studies: Towards a (re)definition of translation. In Y. Gambier & L. van Doorslaer (Eds.), Border Crossings: Translation Studies and other disciplines (pp. 331-348). Amsterdam: John Benjamins Lörscher, W. (2005). The translation process: Methods and problems of its investigation. Meta: Journal des Traducteurs/Translators’ Journal, 50(2), 597-608. doi: 10.7202/011003ar Malmkjær, K. (1998). Introduction: Translation and language teaching. In K. Malmkjaer (Ed.), Translation & language teaching: Language teaching & translation. Manchester, UK: St. Jerome Publishing. Murtisari, E. T. (2016). Translation skill in language learning/teaching: EFL learners’ point of view. Studies About Languages (29), 102-113. Newson, D. (1998). Translation and foreign language learning. In K. Malmkjær (Ed.), Translation and language teaching: Language teaching and translation (pp. 63-68). Okonska, E. K., & Kościałkowska-Okońska, E. (2013). Researching translation competence: The expert problem. In K. Pįatkowska (Ed.), Correspondences and contrasts in foreign language pedagogy and translation studies (pp. 219-232). New York, NY: Springer. Olk, H. M. (2001). The translation of cultural references: An empirical investigation into the translation of culture-specific lexis by degree-level language students (Doctoral dissertation, University of Kent, Canterbury, England). Phat, D. D. (2022). On the Shift of Parts of Speech in Prepositional Meaning in English- Vietnamese Translation. Dinh, DP (2022). On the Shift of Parts of Speech in 32 Vol. 15; No. 1; 2024 Nguyen Thi Thu Huong https://asiacall-acoj.org Prepositional Meaning in English-Vietnamese Translation. AsiaCALL Online Journal, 13(1), 166-179. Prepositional Meaning in English-Vietnamese Translation. AsiaCALL Online Journal, 13(1), 166-179. Pariente-Beltran, B. (2013). Evaluating translation as an explicit instruction tool to improve L2 written skills: An empirical study. (Doctoral dissertation, University of Massachusetts Amherst, America). Retrieved from ProQuest, UMI Dissertations Publishing. Richards, J. C., & Rodgers, T. S. References (2001). Approaches and methods in language teaching. Cambridge, UK: Cambridge University Press. Savignon, S. J. (1972). Communicative competence: An experiment in foreign language teaching. Philadelphia: Center for Curriculum Development. Savignon, S. J. (1983). Communicative competence: Theory and classroom practice: Texts and contexts in second language learning. Reading, MA: Addison-Wesley. Suparmin, S. (2003). Translation in the teaching of a foreign language as a means and as an end. TEFLIN Journal, 14(1), 73-81. Tirkkonen-Condit, S., & Laukkanen, J. (1996). Evaluations: A key towards understanding the affective dimension of translational decisions. Meta: Journal des Traducteurs/Translators’ Journal, 41(1), 45-59. doi: 10.7202/002360ar Truong, V. N. K., & Le, Q. T. (2022). Utilizing YouTube to Enhance English Speaking Skill: EFL Tertiary Students’ Practices and Perceptions. AsiaCALL Online Journal, 13(4), 7- 31. Tsagari, D., & Phlōros, G. (2013). Preface. In D. Tsagari & G. Phlōros (Eds.), translation in language teaching and assessment (pp. vii-xii). Newcastle Upon Tyne, UK: Cambridge Scholars Publishing. Valdeón, R. A. (2015). The (ab)use of taboo lexis in audiovisual Translation: Raising awareness of pragmatic variation in English-Spanish. Intercultural Pragmatics, 12(3), 363-385. doi: 10.1515/ip-2015-0018. Vermes, A. (2010). Translation in foreign language teaching: A brief overview of pros and cons. Eger Journal of English Studies, 10, 83-93. Widdowson, H. G. (2003). Defining issues in English language teaching: Oxford, UK: Oxford University Press. Dr. Huong Thi Thu Nguyen is a lecturer and a Vice Dean of the Faculty of English, the University of Foreign Language Studies - The University of Da Nang (UFLS-UD). Her research interests include translation, translation in language teaching, teaching translation, translation strategies, reader’s reactions, and second language acquisition. Biodata Dr. Huong Thi Thu Nguyen is a lecturer and a Vice Dean of the Faculty of English, the University of Foreign Language Studies - The University of Da Nang (UFLS-UD). Her research interests include translation, translation in language teaching, teaching translation, translation strategies, reader’s reactions, and second language acquisition. 33
https://openalex.org/W2124753257	https://asp-eurasipjournals.springeropen.com/counter/pdf/10.1186/s13634-015-0193-2	English	null	Nonlinear joint transmit-receive processing for coordinated multi-cell systems: centralized and decentralized	EURASIP Journal on Advances in Signal Processing	2,015	cc-by	11,817	RESEARCH Open Access © 2015 Hu et al.; licensee Springer. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. Abstract This paper proposes a nonlinear joint transmit-receive (tx-rx) processing scheme for downlink-coordinated multi-cell systems with multi-stream multi-antenna users. The nonlinear joint tx-rx processing is formulated as an optimization problem to maximize the minimum signal-to-interference noise ratio (SINR) of streams to guarantee the fairness among streams of each user. Nonlinear Tomlinson-Harashima precoding (THP) is applied at transmitters, and linear receive processing is applied at receivers, to eliminate the inter-user interference and inter-stream interference. We consider multi-cell systems under two coordinated modes: centralized and decentralized, corresponding to systems with high- and low-capacity backhaul links, respectively. For the centralized coordinated mode, transmit and receive processing matrices are jointly determined by the central processing unit based on the global channel state information (CSI) shared by base stations (BSs). For the decentralized coordinated mode, transmit and receive processing matrices are computed independently based on the local CSI at each BS. In correspondence, we propose both a centralized and a decentralized algorithm to solve the optimization problem under the two modes, respectively. Feasibility and computational complexity of the proposed algorithms are also analyzed. Simulation results prove that the proposed nonlinear joint tx-rx processing scheme can achieve user fairness by equalizing the bit error rate (BER) among streams of each user and the proposed scheme outperforms the existing linear joint tx-rx processing. Moreover, consistent with previous research results, performance of the proposed centralized nonlinear joint tx-rx processing scheme is proved to be better than that of the decentralized nonlinear joint tx-rx processing. Keywords: Coordinated multi-cell; Centralized coordinated; Decentralized coordinated; Joint transmit-receive processing; Nonlinear precoding; Tomlinson-Harashima precoding (THP) Nonlinear joint transmit-receive processing for coordinated multi-cell systems: centralized and decentralized Zhirui Hu1,2, Chunyan Feng1, Tiankui Zhang1, Qin Niu1 and Yue Chen3 Zhirui Hu1,2, Chunyan Feng1, Tiankui Zhang1, Qin Niu1 and Yue Chen3 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 DOI 10.1186/s13634-015-0193-2 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 DOI 10.1186/s13634-015-0193-2 1 Introduction multi-user multiple-input multiple-output (MU-MIMO) systems [4,5]. The joint transmit-receive (tx-rx) process- ing can be used to further improve the downlink performance of MU-MIMO systems by optimizing the transmit precoding and receive filter matrices jointly. According to the processing of the transmit precoding, the joint tx-rx processing technology can be divided into two types, linear and nonlinear schemes. Coordinated multi-cell is a promising technology to reduce inter-cell interference and increase user data rate, which has been considered as one of the potential technologies for LTE Advanced [1,2]. To fully utilize the advantage of coordinated multi-cell technology, it is essential to manage the multi-user interference (MUI) within the coordinated area appropriately as it is directly related to the achievable spectrum efficiency [3]. Precod- ing is a well-known technique for MUI mitigation in The coordinated multi-cell technology can be imple- mented in a centralized or decentralized mode based on the backhaul capacity of the systems. The centralized co- ordinated mode can achieve higher data rate at the cost of high-capacity backhaul links in order to enable base sta- tions (BSs) to share their channel state information (CSI) (defined as local CSI) and data. Hence, the centralized * Correspondence: huzhirui74@163.com 1Beijing Key Laboratory of Network System Architecture and Convergence, Beijing University of Posts and Telecommunications, Beijing 100876, China 2School of Communication Engineering, Hangzhou Dianzi University, Hangzhou 310018, China Full list of author information is available at the end of the article * Correspondence: huzhirui74@163.com 1Beijing Key Laboratory of Network System Architecture and Convergence, Beijing University of Posts and Telecommunications, Beijing 100876, China 2School of Communication Engineering, Hangzhou Dianzi University, Hangzhou 310018, China Full list of author information is available at the end of the article Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 2 of 14 has a much lower complexity [5]. For the centralized coordinated mode, the tx-rx processing scheme was de- signed to minimize the S-MSE in [15] and maximize the SINR in [16], wherein both should be solved by an itera- tive method, resulting in high computational complexity. The schemes with low complexity were proposed and derived a closed-form solution based on minimum aver- age bit error rate (BER) in [17], minimum mean square error (MMSE) in [18], or ZF criterion in [19]. In [18], the receive processing matrix was firstly computed by CSI. 1.1 Prior art Linear joint tx-rx processing algorithms have been widely studied for coordinated multi-cell systems under centralized mode [9-14]. In [9], block diagonalization (BD) precoding was designed to maximize the weighted sum rate of all users. The tx-rx processing optimization with the criterion of minimizing the S-MSE was pre- sented in [10-12], and the authors of [13] proposed a weighted S-MSE minimization algorithm by considering the channel gain as the weight factor. In [14], the energy efficiency was considered in the tx-rx processing design. A new criterion of maximizing weighted sum energy efficiency was formulated, and the optimization problem was solved by an iterative algorithm. For the decentra- lized coordinated mode, D. Gesbert and R. Holakouei, et al. studied the decentralized linear precoding tech- niques for the system with single-antenna users recently [20-22]. In [20], a distributed precoding scheme based on zero-forcing (ZF) criterion (defined as DZF) and sev- eral centralized power allocation approaches was pro- posed. In [21,22], a characterization of the optimal linear precoding strategy was derived. Distributed virtual SINR (DV-SINR) precoding approaches, where each BS bal- ances the ratio between signal gain at the intended user and the interference caused by other users, had been pro- posed for the particular case of two users in [21] and gen- eralized for multi-user in [22]. The DV-SINR scheme was illustrated to satisfy the optimal precoding characterization and outperform DZF. Previous work did not consider fairness among streams of each user in the coordinated multi-cell system with multi-stream multi-antenna users. It is essential to study the fairness for nonlinear scheme, as unfairness is an in- herent character of THP and the worst performance de- termines the whole performance of the user [24]. 1 Introduction Then,the transmit processing matrix and receive weight coefficient were computed based on MMSE. In [19], the algorithm decomposed the MU-MIMO channel into par- allel independent single user MIMO (SU-MIMO) chan- nels, and then, closed-form expressions of transmit and receive processing matrices were derived to optimize the performance of each user. The above research works on nonlinear tx-rx processing were all developed for the centralized coordinated mode. The relevant works for the decentralized coordinated mode are relatively fewer. A decentralized nonlinear precoding, ZF-THP, was proposed in [23] but can only be applied for the system with a single user. To the best of our know- ledge, for the system with multi-stream multi-antenna users, the tx-rx processing solutions under decentra- lized coordination mode have not been addressed in the literature. approach is limited to systems with sufficient backhaul capacity [6,7]. The decentralized coordinated mode does not require BSs to share their local CSI, and the precoding or tx-rx processing is conducted at each BS [8]. This approach has less requirement on backhaul link capacity at some loss on the data rate in comparison to the central- ized coordinated mode. In recent years, relevant works on joint tx-rx process- ing in coordinated multi-cell systems have been widely studied under either centralized [9-19] or decentralized mode [20-23]. For designing the nonlinear joint tx-rx processing, many different optimal objectives have been considered, such as minimizing the sum mean square error (S-MSE) or maximizing the SINR; yet, the fairness among the streams of each user has not been solved for the coordinated multi-cell systems with multi-stream multi-antenna users. 1.2 Contributions h In this paper, a nonlinear joint tx-rx processing scheme is proposed to improve fairness among streams of each user with multi-antenna. The nonlinear joint tx-rx processing is formulated as an optimization problem to maximize the minimum SINR of streams. The perform- ance of the proposed scheme is evaluated under both centralized and decentralized coordinated modes. Two algorithms for solving the optimization problem are derived. The main work of this paper can be summarized as follows. Nonlinear joint tx-rx processing scheme is developed for a coordinated multi-cell system with multi-stream multi-antenna users under two coordinated modes, centralized and decentralized mode. Compared with the linear joint tx-rx processing schemes, the nonlinear joint tx-rx processing schemes are more complex but can obtain more system gain, which have gained much attention recently. Most re- search about the nonlinear precoding focus on Tomlinson- Harashima precoding (THP), as it can achieve approximate performance with the optimal dirty paper coding but Two algorithms, the centralized and the decentralized algorithms, are proposed to Two algorithms, the centralized and the decentralized algorithms, are proposed to solve the Page 3 of 14 Page 3 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 between the nth BS and the kth user, whose entries are independent and identically distributed (i.i.d.) complex Gaussian variables with zero mean and unit variance. In Equation 1, the second term on the right-hand side is MUI, and nk e CN 0; σ2Inr ð Þ is the additive white Gaussian noise variable. optimization problem, and both of them derive the closed-form solutions. The algorithms guarantee the fairness among the streams of each user, which not only boost the performance of each user, but bring much convenience to the modulation/demodulation and coding/decoding procedures. Each user decodes the desired data by multiplying with the receive processing matrix. The received data of the kth user is given as: The remainder of this paper is organized as follows. Section 2 presents the coordinated multi-cell system model. The proposed nonlinear joint tx-rx processing scheme is described in detail in Section 3. A perform- ance analysis of the proposed algorithms is developed in Section 4. Simulation results and conclusions are presented in Section 5 and Section 6, respectively. 1.2 Contributions h ~yk ¼ Rkyk ¼ RkX N n¼1 Hk nxk n þ Rk X K t¼1;t≠k X N n¼1 Hk nxt n þ ~nk ¼ RkHkxk þ X K t¼1;t≠k RkHkxt þ ~nk ð2Þ 1.3 Notation We use uppercase boldface letters to denote matrices and lowercase boldface to denote vectors. The operators (⋅)T, (⋅)H, (⋅)†, E(⋅), and Tr(⋅) stand for transpose, Hermit- ian, Moore-Penrose pseudo-inverse, expectation, and the trace of a matrix, respectively. diag(⋅) and blockdiag(⋅) denote diagonal and block diagonal matrix. I and 0 are the identity and the all-zero matrix, respectively, with appropriate dimensions. ‖ ⋅‖F represents the Frobenius norm of a matrix. [⋅]i : j,k : l denotes the submatrix com- prised of row i through row j and column k through column l of a matrix. where Rk ∈Clknr denotes the receive processing matrix of the kth user. ñk = Rknk is the equivalent received noise vector at the kth user. Let y ¼ y1T ; ⋯; yKT T represent the received signal of the K users. Equation 2 can be expressed as: ~y ¼ Ry ¼ R X N n¼1 Hnxn þ ~n ¼ RHx þ ~n ð3Þ ð3Þ where R = blockdiag(R1, ⋯, RK) is a L × Knr matrix. H ¼ H1T; ⋯; HKT h iT ∈CKnrNt is the global CSI between BSs and K users. Hn ¼ H1T n ; ⋯; HKT n h iT ∈CKnrnt de- notes the nth local CSI between the nth BS and K users. where R = blockdiag(R1, ⋯, RK) is a L × Knr matrix. H ¼ H1T; ⋯; HKT h iT ∈CKnrNt is the global CSI between BSs and K users. Hn ¼ H1T n ; ⋯; HKT n h iT ∈CKnrnt de- notes the nth local CSI between the nth BS and K users. 2 System model where Hk ¼ Hk 1; ⋯; Hk N is the global CSI between BSs and the kth user and Hk n ∈Cnrnt denotes the local CSI 0 Page 4 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 4 of 14 Page 4 of 14 where sk is comprised of the lk data streams for the kth user. In THP, feedback matrix Bn is a unit lower triangular matrix, For the centralized coordinated mode, it is assumed that all BSs exchange their local CSI, and the tx-rx processing matrices are jointly designed at the central processing unit. The system can be seen as a virtual MU-MIMO system with Nt = Nnt transmit antennas. On the contrary, for the decentralized coordinated mode, BSs do not share their CSI, and every BS only has knowledge of local CSI between itself and K users. Therefore, the tx-rx processing matrices are independently designed at each BS. Bn ¼ B1;1 n 0 ⋯ 0 B2;1 n B2;2 n ⋱ ⋮ ⋮ ⋱ ⋱ 0 BK;1 n ⋯ BK;K−1; n BK;K n 2 664 3 775 ð5Þ Bn ¼ B1;1 n 0 ⋯ 0 B2;1 n B2;2 n ⋱ ⋮ ⋮ ⋱ ⋱ 0 BK;1 n ⋯ BK;K−1; n BK;K n 2 664 3 775 ð5Þ ð5Þ where Bk;k n is a unit lower triangular matrix with lk × lk size. un is the output data of THP. Therefore, the lth data stream of un is interfered by the first (l-1) data streams; in other words, the lth(l = 2, ⋯, L) element ul n in un is a linear combination of sj(j ≤l). Assume M-ary square constellation is employed to s. To ensure that the real and the imaginary parts of ul n are constrained into − ﬃﬃﬃﬃﬃ M p ; ﬃﬃﬃﬃﬃ M p , modulo 2 ﬃﬃﬃﬃﬃ M p operation mod2 ﬃﬃﬃﬃ M p ⋅ð Þ is introduced. The output data of THP is expressed as: 3 Nonlinear joint transmit-receive processing algorithm In this section, we present nonlinear joint tx-rx process- ing algorithms for a coordinated multi-cell system under two different coordinated modes. The algorithm structure is firstly shown. Then, we formulate the optimization problem, aiming at maximizing the minimum SINR of streams to guarantee the fairness among the streams of each user. Finally, the algorithms for different coordinated modes are proposed. un ¼ mod2 ﬃﬃﬃﬃ M p I−Bn ð Þun þ s ½ ¼ I−Bn ð Þun þ s þ dn ð6Þ ð6Þ 2 System model Consider a downlink coordinated multi-cell system, where N BSs cooperatively serve K users. Each BS and user is equipped with nt and nr antennas, re- spectively. All BSs share user data and cooperatively transmit the data to an intended user. Each BS trans- mits L ¼ XK k¼1lk data streams to K users, where lk is the number of transmitted data streams for user k. x ¼ XK k¼1xk denotes the transmit signal at BSs, and xn ¼ XK k¼1xk n is the transmit signal at the nth BS. ~n ¼ ~n1T ; ⋯; ~nKT T is the combination of the receive i t th K noise at the K users. We assume that BSs' transmit power for every user is P. Therefore, the total transmit power of BSs is KP. De- note xk ¼ xkT 1 ; ⋯; xkT N h iT , where xk n denotes the prepro- cessed signal transmitted by the nth BS for user k, satisfying Tr xkxkH n o ¼ P. The received signal of the kth user is: Define Λ k; t ð Þ ¼ RkXN n¼1Hk nxt n ¼ RkHkxt. The rate of the kth user is given by Define Λ k; t ð Þ ¼ RkXN n¼1Hk nxt n ¼ RkHkxt. The rate of the kth user is given by rk ¼ log2 I þ Λ k; k ð ÞΛ k; k ð ÞH ~nk ~nkH þ XK t¼1;t≠kΛ k; t ð ÞΛ k; t ð ÞH ð4Þ ð4Þ yk ¼ X N n¼1 Hk nxk n þ X K t¼1;t≠k X N n¼1 Hk nxt n þ nk ¼ Hkxk þ X K t¼1;t≠k Hkxt þ nk ð1Þ Then, the system sum rate can be obtained by r ¼ XK k¼1rk. Then, the system sum rate can be obtained by r ¼ XK k¼1rk. Then, the system sum rate can be obtained by r ¼ XK k¼1rk. ð1Þ The coverage of N-coordinated BSs is defined as one coordinated area. We mainly focus on the interference within the coordinated area. The interference from other coordinated areas is ignored in this paper, which can be eliminated by inter-cell interference coordination tech- nology [25] or interference alignment technology [26]. 3.2 Problem formulation ð16Þ From Equation 1, it is noticed that the received signal of every user is influenced by MUI. In order to liberate every user from MUI, the relative matrices in this algorithm are designed to satisfy ZF criterion: We assume that Fk is represented as Fk ¼ FkF k. Then, k h i Fk ¼ Vk 0 ∈C Nt Nt− Xk−1 i¼1li h i is named as the transmit space matrix and F k is the transmit diversity matrix with Nt− Xk−1 i¼1li h i lk size. RHFCB−1 ¼ W ð11Þ ð11Þ RHFCB−1 ¼ W where W ¼ diag ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τl1 ð Þ p Il1; ⋯; ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlK ð Þ p IlK . The SINR of the kth user can be obtained as: h i The above analysis is suitable for the kth(k > 1) user. Since the first user is not limited by Equation 15, we use F1 ¼ INt. Fk can be formed as: The above analysis is suitable for the kth(k > 1) user. Since the first user is not limited by Equation 15, we use ρk 1; ⋯; ρk lk ¼ P τlk 1 σ2 n;1 ; ⋯; 1 σ2 n;lk ! ð12Þ F1 ¼ INt. Fk can be formed as: F1 ¼ INt. Fk can be formed as: ð12Þ Fk ¼ INt; k ¼ 1 Vk 0; 1 < k≤K ð17Þ ð17Þ In order to guarantee the fairness among streams of each user, we investigate the tx-rx processing matrices design to maximize the minimum SINR for each stream of each user, which is formulated as follows: With Fk and THP, the proposed algorithm decom- poses the MU-MIMO channel into parallel independent SU-MIMO channels [19]. We can comprehend this as max B;F;G min k ρk 1; ⋯; ρk lk s:t: RHFCB−1 ¼ W a ð Þ SiBnei ¼ 0i; i ¼ 1; ⋯; L b ð Þ Tr FkFkH n o ¼ P=τ cð Þ follows: for the kth user, Fk is designed to avoid the interference from users (k + 1, ⋯, K). Meanwhile, THP is used to eliminate the interference from the first (k −1) users. Therefore, user k will not suffer from MUI. 3.1 Algorithm structure The structure of the proposed algorithm is shown in Figure 1. In the proposed algorithms, nonlinear preprocessing is applied at transmitters; meanwhile, linear processing is applied at each receiver. where dn ¼ 2 ﬃﬃﬃﬃﬃ M p zI þ jzQ , zI and zQ are both integers. Define vn = s + dn, and then, un is written as: At the nth(n =1,…,N) transmitter, s ¼ s1T;⋯; sKT T ∈C L1 H At the nth(n =1,…,N) transmitter, s ¼ s1T;⋯; sKT T ∈C L1 denotes the modulated data vector satisfying E{ssH} = I, un ¼ B−1 n vn ð7Þ ð7Þ un ¼ B−1 n vn denotes the modulated data vector satisfying E{ssH} = I, (a) nth transmitter structure (b) receiver structure y R ~ y n H1 HN s mod Hn xn THP Precoder un s I-Bn mod Fn Figure 1 Structure of the transmit-receive processing: (a) transmitter structure; (b) receiver structure. (a) nth transmitter structure Hn xn THP Precoder un s I-Bn mod Fn (a) nth transmitter structure Hn xn THP Precoder un s I-Bn mod Fn mod (a) nth transmitter structure (b) receiver structure Figure 1 Structure of the transmit-receive processing: (a) transmitter structure; (b) receiver structure. 15:10 Page 5 of 14 Page 5 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 There is a power enhancement of τ = M/(M −1) due to THP, i.e., E unuH n ¼ τI [27]. The transmit signal at the nth BS is: IL. (c) is used to guarantee the power constraint and Fk ¼ FkH 1 ; ⋯; FkH N h iH . RHF 0 ¼ WB ð14Þ ~y ¼ R X N n¼1 HnFnB−1 n vn þ ~n ¼ RHFCB−1v þ ~n ð9Þ ð9Þ where F 0 ¼ FT 1 ; ⋯; FT N T. As B is a unit lower triangular matrix, the left side of Equation 14 should satisfy: where F = diag(F1, ⋯, FN) is a block diagonal matrix with Nt × NL size and CB−1 ¼ B−1T 1 ; ⋯; B−1T N h iT ∈C NLL. The user data will finally be obtained by modulo operation and demodulation. Obviously, the received noise power of lk streams of the kth user is: RtHtFk ¼ 0 t < k ð Þ RtHtFk ¼ 0 t < k ð Þ ð15Þ ð15Þ which reveals that Fk ¼ FkT 1 ; ⋯; FkT N h iT ∈C Ntlk k > 1 ð Þ lies in the null space of ⌣H k ¼ H1T; ⋯; H k−1 ð ÞT h iT , where Hi ¼ RiHi is the equivalent CSI of the ith user. Fk can be found by doing singular value decomposition (SVD) on ⌣ H k : which reveals that Fk ¼ FkT 1 ; ⋯; FkT N h iT ∈C Ntlk k > 1 ð Þ lies in the null space of ⌣H k ¼ H1T; ⋯; H k−1 ð ÞT h iT , where Hi ¼ RiHi is the equivalent CSI of the ith user. Fk can be found by doing singular value decomposition (SVD) on ⌣ H k : σ2 n;1; ⋯; σ2 n;lk ¼ σ2diag RkRkH ð10Þ ð10Þ ⌣ H k ¼ Uk Σk 0 Vk 1 Vk 0 H ð16Þ 3.3 Centralized algorithm xn ¼ Fnun ¼ X K k¼1 Fk nxk n ð8Þ In Equation 13, the relative matrices are entangled with each other. To solve this problem, we start from the ZF constraint. Every BS is assumed to have the same feed- back matrix, denoted as B, which will be determined at the central processing unit based on the global CSI. (a) in Equation 13 can be rewritten as ð8Þ where Fn ¼ F1 n; ⋯; FK n ∈CntL is the transmit process- ing matrix. At the receivers, the received signal in Equation 3 can be rewritten as: RHF 0 ¼ WB ð14Þ HkFk ¼ QkDkPkH ð20Þ where Qk∈C nrS and Pk∈C S Nt− Xk−1 i¼1li h i h k RHnFnB−1 n ¼ Wn; n ¼ 1; ⋯; N ð24Þ ð24Þ normal columns, and S is the rank of HkFk. Dk∈C SS is a lower triangular matrix, the diagonal elements of which satisfy: where Wn ¼ diag p1 nIl1; ⋯; pK n IlK satisfies XN n¼1Wn ¼ W. The receive processing matrix Rk(k = 1, ⋯, K) of each user is related to the transmit signals from N BSs. If Rk is computed at BSs, each BS can only decide it dependently as the local CSI of each BS is not exchanged. Generally, Rk derived at different BSs has different values, which is unreasonable. Otherwise, for each user, fre- quently interactive information with all coordinated BSs is required. It will largely increase the system computational complexity. Therefore, we firstly compute Rk(k = 1, ⋯, K) at users. Denote Hk ¼ Uk 1 Uk 0 ΣkVkH as the SVD of Hk, where Uk 1 ∈Cnrlk . Then Rk can be obtained by Rk ¼ Gk Uk 1 H , where Gk is a diagonal matrix for normalizing the received signal and will be determined at the BSs. For fre- quency division duplex system, user k can only feedback the equivalent local CSI Hk n ¼ Uk 1Hk n to the nth BS. There- fore, Equation 10 can be rewritten as: dk i;i ¼ λ k; i ¼ 1; ⋯; lk λk i ; i ¼ lk þ 1; ⋯; S ; ð21Þ ð21Þ where λk i is the ith largest positive singular value of Hk Fk, and λ k ¼ Ylk j¼1λk j 1=lk. Define Λk ¼ λ k −1 Ilk. Then Rk; F k and Bk,k are given by: where λk i is the ith largest positive singular value of Hk Fk, and λ k ¼ Ylk j¼1λk j 1=lk. Define Λk ¼ λ k −1 Ilk. Then Rk; F k and Bk,k are given by: Rk; F k and Bk,k are given by: Rk ¼ Λk Qk H :;1:lk Bk;k ¼ Λk Dk 1:lk;1:lk F k ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p P ½ :;1:lk ð22Þ ð22Þ Based on Equation 10 and Equation 22, the received noise power of every stream of the kth user is σ2 λ k −2 . 3.2 Problem formulation For any user k(k = 1, ⋯, K), Bk;k; F k and Rk satisfy: any user k(k = 1, ⋯, K), Bk;k; F k and Rk satisfy: ð13Þ RkHkFkF k ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p Bk;k; k ¼ 1; ⋯; K ð18Þ ð18Þ for k = 1, ⋯, K, n = 1, ⋯, N. (a) denotes ZF criterion. (b) denotes that Bn is the unit lower triangular matrix, where Si = [Ii, 0i × (L −i)] and ei is the ith column of where Bk,k is a unit lower triangular matrix with lk × lk size. Therefore, Bk;k; F k and Rk(k = 1, ⋯, K) can be Page 6 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 designed separately. For the kth user, Equation 13 can be reduced to: designed step-by-step, which starts by computing the transmit space matrix and the receive processing matrix of the first user, then computes the matrices for the sec- ond user by utilizing the receive processing matrix of the first user and so on. max Rk;Fk 2;Bk;k min ρk 1; ⋯; ρk lk s:t: RkHkFkF k ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p Bk;k a ð Þ Sk i Bk;kek i ¼ 0i; i ¼ 1; ⋯; lk b ð Þ Tr F k F kH ¼ P=τ cð Þ All of the matrices are designed at the central process- ing unit, and the receive processing matrices are trans- mitted to each user by downlink channel. The procedure of the proposed centralized algorithm is summarized in Table 1. ð19Þ 3.4 Decentralized algorithm where Sk i ¼ Ii; 0i lk−i ð Þ , and ek i is the ith column of Ilk. where Sk i ¼ Ii; 0i lk−i ð Þ , and ek i is the ith column of Ilk. The optimal solution of Equation 19 can be obtained from the generalized triangular decomposition of HkFk [19,28]: In this scenario, as BSs do not exchange their local CSI, each BS independently preprocesses the user data with the local CSI of itself. The data processed by each BS cannot be obtained by other BSs. In order to ensure that the user's receive signal is not interfered by MUI, relative matrices at each BS should satisfy the ZF criterion. Therefore, Equation 11 is reduced to: HkFk ¼ QkDkPkH HkFk ¼ QkDkPkH ð20Þ HkFk ¼ QkDkPkH HkFk ¼ QkDkPkH It can be found by doing SVD on ⌣ H k: which reveals that Fk n k > 1 ð Þ lies in the null space of ⌣ H k n ¼ H1T n ; ⋯; H k−1 ð ÞT n h iT . It can be found by doing SVD on ⌣ H k n: SVD on ⌣ H k n: ð33Þ ⌣ H k n ¼ Uk n Σk n 0 Vk n1 Vk n0 H ð29Þ ð29Þ where γn ¼ αk=pk n . The constrain condition (a) in Equation 33 can be rewritten as F k n ¼ γ−1 n H k† n Bk;k n . Com- bining with (c), γ2 np=τ ¼Tr H k† n Bk;k n H k† n Bk;k n H ( ) is ob- tained. The problem of Equation 33 can be rewritten as: where γn ¼ αk=pk n . The constrain condition (a) in Equation 33 can be rewritten as F k n ¼ γ−1 n H k† n Bk;k n . Com- bining with (c), γ2 np=τ ¼Tr H k† n Bk;k n H k† n Bk;k n H ( ) is ob- tained. The problem of Equation 33 can be rewritten as: where γn ¼ αk=pk n . The constrain condition (a) in Equation 33 can be rewritten as F k n ¼ γ−1 n H k† n Bk;k n . Com- bining with (c), γ2 np=τ ¼Tr H k† n Bk;k n H k† n Bk;k n H ( ) is ob- tained. The problem of Equation 33 can be rewritten as: We assume that Fk n is represented as Fk n ¼ Fk nF k n . We define Fk n ¼ Vk n0 ∈C nt nt− Xk−1 i¼1li h i , and F k n is a nt− Xk−1 i¼1li h i lk matrix. We assume that Fk n is represented as Fk n ¼ Fk nF k n . We define Fk n ¼ Vk n0 ∈C nt nt− Xk−1 i¼1li h i , and F k n is a nt− Xk−1 i¼1li h i lk matrix. ( ) tained. The problem of Equation 33 can be rewritten as: ( ) tained. HkFk ¼ QkDkPkH For any user k (k = 1, ⋯, K), Bk;k n ; F k n and Gk satisfy: (k = 1, ⋯, K), Bk;k n ; F k n and Gk satisfy: (k = 1, ⋯, K), Bk;k n ; F k n and Gk satisfy: Gk H k nF k n ¼ pk nBk;k n ð31Þ max B;F;G min k ρk 1; ⋯; ρk lk s:t: GHnFnB−1 n ¼ Wn a ð Þ G ¼ diag g1; ⋯; gL b ð Þ SiBnei ¼ 0i; i ¼ 1; ⋯; L cð Þ Tr Fk nFkH n n o ¼ p=τ d ð Þ XN n¼1Wn ¼ W e ð Þ ð26Þ Gk H k nF k n ¼ pk nBk;k n ð31Þ Therefore, Bk;k n ; F k n and Gk(k = 1, ⋯, K) can be designed separately. For the kth user, Equation 26 can be reduced to: max Bk;k n ;Fk n2;Gk min k ρk 1; ⋯; ρk lk s:t: Gk H k nFn k¼ pk nBk;k n a ð Þ Gk ¼ diag gk 1; ⋯; gk lk n o b ð Þ SkBk k k 0 i 1 l ð Þ max B;F;G min k ρk 1; ⋯; ρk lk s:t: GHnFnB−1 n ¼ Wn a ð Þ G ¼ diag g1; ⋯; gL b ð Þ SiBnei ¼ 0i; i ¼ 1; ⋯; L cð Þ Tr Fk nFkH n n o ¼ p=τ d ð Þ XN n¼1Wn ¼ W e ð Þ ð26Þ Gk H k nF k n ¼ pk nBk;k n Therefore, Bk;k n ; F k n a separately. For the kth u max Bk;k n ;Fk n2;Gk min k s:t: Gk H k nFn k Gk ¼ dia SkBk k k max B;F;G min k ρk 1; ⋯; ρk lk s:t: GHnFnB−1 n ¼ Wn a ð Þ G ¼ diag g1; ⋯; gL b ð Þ SiBnei ¼ 0i; i ¼ 1; ⋯; L cð Þ Tr Fk nFkH n n o ¼ p=τ d ð Þ XN n¼1Wn ¼ W e ð Þ ð26Þ ð31Þ Therefore, Bk;k n ; F k n and Gk(k = 1, ⋯, K) can be designed separately. HkFk ¼ QkDkPkH Therefore, the SINR of the kth user is: σ2 n;1; ⋯; σ2 n;lk ¼ σ2diag GkGkH ð25Þ ð25Þ ρk 1 ¼ ρk 2 ¼ ⋯¼ ρk lk ¼ p τlk ⋅ λ k 2 σ2 ð23Þ ð23Þ Table 1 Centralized nonlinear joint tx-rx processing algorithm 1 for k = 1:K 2 Compute Fk using Equation 17 3 Obtain Rk; F k, and Bk,k using Equation 22 4 Compute Fk by Fk ¼ FkF k 5 end 6 Compute B by B = W−1RHF' Table 1 Centralized nonlinear joint tx-rx processing algorithm Table 1 Centralized nonlinear joint tx-rx processing algorithm It can be seen that every stream of the kth user can achieve equal SINR. It can be seen that every stream of the kth user can achieve equal SINR. Note that for the computation of the transmit space matrix of the kth user Fk, we need to know the receive processing matrices of the first (k −1) users Rt(t < k) and that, for the computation of the receive processing matrix of the kth user Rk, we need to know the transmit space matrix of the kth user. Therefore, Fk and Rk are Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 7 of 14 Assume that every BS has equal transmit power p = P/N. Based on the above analysis, Equation 13 is equivalent to the following optimization problem: SU-MIMO channels. Define H k n ¼ Hk nFk n. For any user k SU-MIMO channels. Define H k n ¼ Hk nFk n. For any user k (k = 1, ⋯, K), Bk;k n ; F k n and Gk satisfy: SU-MIMO channels. Define H k n ¼ Hk nFk n. For any user k SU-MIMO channels. Define Hn ¼ HnFn. HkFk ¼ QkDkPkH For the kth user, Equation 26 can be reduced to: max Bk;k n ;Fk n2;Gk min k ρk 1; ⋯; ρk lk s:t: Gk H k nFn k¼ pk nBk;k n a ð Þ Gk ¼ diag gk 1; ⋯; gk lk n o b ð Þ Sk i Bk;k n ek i ¼ 0i; i ¼ 1; ⋯; lk cð Þ Tr F k nF kH n ¼ p=τ d ð Þ XN n¼1pk n ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p ; e ð Þ ð32Þ ð26Þ for k = 1, ⋯, K, n = 1, ⋯, N. In (a), Hn ¼ H1T n ; ⋯; HKT n h iT . In (b), gj(j = 1, 2,⋯, L) are diagonal elements of G. (d)(e) are used to guarantee the power constraint. In Equation 26, the relative matrices are entangled with each other. Similarly, to solve this problem, we start from the ZF constraint. Take the nth BS for example. (a) in Equation 26 can be rewritten as: ð32Þ where Sk i ¼ Ii; 0i lk−i ð Þ , and ek i is the ith column of Ilk . (d) is obtained because FkH n Fk n ¼ I. The optimal solution of Equation 32 is obtained when all lk streams attain equal SINR [29]. According to Equation 12 and Equation 25, it is equivalent to possess equal value for diagonal elements of Gk, expressed as Gk = αkI. Equation 32 is equivalent to: HnFn ¼ G−1WnBn ð27Þ ð27Þ As G and Wn are diagonal matrices and Bn is a unit lower triangular matrix, the left side of Equation 27 is a lower triangular matrix, i.e., min Bk;k n ;Fk n2;Gk αk 2 s:t: γn H k nFn k¼ Bk;k n ; a ð Þ Sk i Bk;k n ek i ¼ 0i; i ¼ 1; ⋯; lk b ð Þ Tr F k nF kH n ¼ p=τ cð Þ XN n¼1pk n ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p ; d ð Þ ð33Þ Ht nFk n ¼ 0 t < k ð Þ ð28Þ ð28Þ which reveals that Fk n k > 1 ð Þ lies in the null space of ⌣ H k n ¼ H1T n ; ⋯; H k−1 ð ÞT n h iT . HkFk ¼ QkDkPkH By combining (a) and (d) in Equation 33, Gk = αkI is obtained, where: Equation 33, Gk = αkI is obtained, where: 4.1 Feasibility analysis h In the MIMO system, in order to distinguish every transmit stream, the constraint that the number of transmit data streams is no more than the number of transmit and re- ceive antennas should be satisfied. For the centralized coor- dinated mode and the decentralized coordinated mode, the constraint on the number of transmit data streams is speci- fied as follows: αk ¼ ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ P= τlk ð Þ p XN n¼11=γn ð38Þ ð38Þ Finally, Bn is determined by Bn ¼ W−1 n RHnFn. k Finally, Bn is determined by Bn ¼ W−1 n RHnFn. k In this algorithm, Rk(k = 1, ⋯, K) are designed at the users, and other matrices are designed at the BSs. The design of matrices is independently performed at every BS. H k† n Bk;k n 2 F n ¼ 1; ⋯; N ð Þ should be transmitted to the kth(k = 1, ⋯, K) user by downlink channel for achiev- ing Gk at the kth user. The procedure of the proposed decentralized algorithm is described in Table 2. In this algorithm, Rk(k = 1, ⋯, K) are designed at the users, and other matrices are designed at the BSs. The design of matrices is independently performed at every Lemma 1: For the centralized coordinated mode, the number of transmit data streams are bounded by L ≤Nt, lk ≤nr; for the decentralized coordinated mode, the number of transmit data streams are bounded by L ≤nt, lk ≤nr. f y t, k r In the proposed centralized algorithm, the design of the transmit space matrix Fk k ¼ 1; ⋯; K ð Þ requires Nt− Xk−1 i¼1li > 0 . Furthermore, to guarantee that the optimization problem Equation 19 has solutions, S ≥lk is required. As the entries of HkFk are zero-mean com- plex Gaussian variables, the rank of HkFk is S ¼ min nr; Nt− Xk−1 i¼1li with a probability of 1. Therefore, S ≥lk is the necessary condition to carry out the algorithm. Base on Lemma 1, the necessary condition is satisfied to the centralized coordinated system. Therefore, the proposed centralized algorithm is feasible. 3.5 Remark 1 (applicability) It should be noted that the proposed two algorithms are also suitable for the system with a single-data stream transmitted for each user. Moreover, the proposed two algorithms both are applicable to the noncoordinated system. However, the centralized scheme is suggested to apply for the noncoordinated system, as the decentra- lized scheme is a suboptimal solution in this situation. nr; Nt− Xk−1 i¼1li HkFk ¼ QkDkPkH Obtain Uk 1 ¼ Uk H :;1:lk 3 end 4 for n = 1:N 5 for k = 1:K 6 Compute Fk n using Equation 30 7 Obtain lk columns of Bk;k n using Equation 37 8 Compute γn by γ2 n ¼ τ H k† n Bk;k n 2 F =p 9 Compute F k n by F k n ¼ γ−1 n H k† n Bk;k n 10 Compute Gk = αkI using Equation 38 and derive Rk by Rk ¼ Gk Uk 1 H 11 end 12 Compute Bn by Bn ¼ W−1 n RHnFn 13 end Table 2 Decentralized nonlinear joint tx-rx processing algorithm Let Li n i ¼ 1; ⋯; lk ð Þ represent the ith column of H k† n . Yi n is comprised of (i + 1, ⋯, lk) columns of H k† n , i.e., Yi n ¼ Liþ1 n ; ⋯; Llkn . Then, we can attain: H k† n Bk;k n ek i 2 ¼ Li n; Yi n 1 bi n 2 ð35Þ ð35Þ By differentiating of Equation 35 with respect to bi n and setting the result to zero, bi n is achieved by: bi n ¼ −YiHYi −1 YiHLi n; i ¼ 1; ⋯; lk−1 ð36Þ ð36Þ Therefore, the ith column of Bk;k n is obtained by: Bk;k n ek i ¼ 1; biT n h iT ; i ¼ 1 01 i−1 ð Þ; 1; biT n h iT ; i ¼ 2; ⋯; lk−1 ek lk; i ¼ lk 8 > > > < > > > : ð37Þ performance among streams of each user, which bring much convenience to the modulation/demodulation and coding/decoding procedures. In this section, we analyze the feasibility and the computational complexity of the proposed two algorithms. performance among streams of each user, which bring much convenience to the modulation/demodulation and coding/decoding procedures. In this section, we analyze the feasibility and the computational complexity of the proposed two algorithms. ð37Þ Then, Bk;k n is obtained by combining all columns Bk;k n ek i (i = 1, ⋯, lk). Therefore, we can derive γ2 n ¼τ H k† n Bk;k n 2 F =p . HkFk ¼ QkDkPkH The problem of Equation 33 can be rewritten as: min Bk;k n H k† n Bk;k n 2 F s:t: Sk i Bk;k n ek i ¼ 0i; i ¼ 1; ⋯; lk: ð34Þ The above analysis is suitable for the kth(k > 1) user. Since the first user is not limited by Equation 28, we use Fk n ¼ Int. Fk n can be achieved by: ð34Þ Fk n ¼ Int; k ¼ 1 Vk n0; 1 < k≤K ð30Þ ð30Þ Actually, Bk;k n ek i denotes the ith column of Bk;k n . The objective of Equation 34 is equivalent to minimizing 2 Actually, Bk;k n ek i denotes the ith column of Bk;k n . The objective of Equation 34 is equivalent to minimizing 2 H k† n Bk;k n ek i 2 for any i(i = 1, ⋯, lk). Similarly, with Fn1 and THP, the algorithm decom- poses the MU-MIMO channel into parallel independent Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 8 of 14 Table 2 Decentralized nonlinear joint tx-rx processing algorithm 1 for k = 1:K 2 Doing SVD on Hk. Obtain Uk 1 ¼ Uk H :;1:lk 3 end 4 for n = 1:N 5 for k = 1:K 6 Compute Fk n using Equation 30 7 Obtain lk columns of Bk;k n using Equation 37 8 Compute γn by γ2 n ¼ τ H k† n Bk;k n 2 F =p 9 Compute F k n by F k n ¼ γ−1 n H k† n Bk;k n 10 Compute Gk = αkI using Equation 38 and derive Rk by Rk ¼ Gk Uk 1 H 11 end 12 Compute Bn by Bn ¼ W−1 n RHnFn 13 end Table 2 Decentralized nonlinear joint tx-rx processing algorithm 1 for k = 1:K 2 Doing SVD on Hk. 5 Numerical results and discussions The complexity of other scalar computations can be ignored. Therefore, the complexity of the relative matrices designed for the kth user is O Xk−1 i¼1li 2 Nt þ n2 rNt þ Nlkn2 t þ l3 kNt . space of a Xk−1 i¼1li nt matrix with O Xk−1 i¼1li 2 nt complexity; a onetime multiplication of a lk × nt matrix and a nt nt− Xk−1 i¼1li h i matrix, the complexity of which is O lknt nt− Xk−1 i¼1li h i ; and lk-times computation of the Moore-Penrose pseudo-inverse of a nt− Xk−1 i¼1li h i lk−i ½ space of a Xk−1 i¼1li nt matrix with O Xk−1 i¼1li 2 nt complexity; a onetime multiplication of a lk × nt matrix and 4.3 Remark 2 (backhaul latency effect) For centralized coordinated mode, tx-rx processing matrices are jointly computed at the central processing unit and then reported to every BS through the backhaul link. The existing backhaul latency can affect the system performance. We ignore the backhaul latency effect in the paper and will study it in the future work. 4.2 Computational complexity For simplicity, the number of float point operations is used to measure the computational complexity of the proposed algorithms. In the proposed centralized algorithm, the design of the relative matrices for the kth user includes the follow- ing: a onetime multiplication of a lk −1 × nr matrix and a nr × Nt matrix, the complexity of which is O(lk −1nrNt); a onetime computation of the null space of a Xk−1 i¼1li Nt matrix with O Xk−1 i¼1li 2 Nt complexity; a one- time multiplication of a nr × Nt matrix and a Nt Nt− Xk−1 i¼1li h i matrix, the complexity of which is O nrNt Nt− Xk−1 i¼1li h i ; and a onetime computation of the singular value of a nr Nt− Xk−1 i¼1li h i matrix with O n2 r Nt− Xk−1 i¼1li h i complexity. Therefore, the complex- ity of the relative matrices designed for the kth user is O 4 Performance analysis From Equation 23 and Equation 38, it is noted that both of the proposed two algorithms can achieve equal SINR for every stream of the user. They guarantee the balance In the proposed decentralized algorithm, nt− Xk−1 i¼1li > 0 is required to guarantee the existence of the transmit space Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 9 of 14 Assume that the data streams for every user is equal, i.e., l1= ⋯=lK=l. Thus, the complexity of the proposed centralized al- gorithm is O KL2Nt þ Kn2 rNt þ KnrN2 t , and the decen- tralized algorithm is O KL2Nt þ Kn2 rNt þ NLn2 t þ Kl3Nt . matrix Fk n k ¼ 1; ⋯; K; n ¼ 1; ⋯; N ð Þ. Moreover, the solu- tion of optimization problem Equation 34 requires nt− Xk−1 i¼1li≥lk , which is satisfied in the decentralized coordi- nated system. Therefore, the proposed decentralized algorithm is feasible. matrix Fk n k ¼ 1; ⋯; K; n ¼ 1; ⋯; N ð Þ. Moreover, the solu- tion of optimization problem Equation 34 requires nt− Xk−1 i¼1li≥lk , which is satisfied in the decentralized coordi- nated system. Therefore, the proposed decentralized algorithm is feasible. 5 Numerical results and discussions This section presents some simulation results to evaluate the BER performance of the proposed two algorithms. We compare them with the following algorithms: the interference-free algorithm, the joint transit-receive pro- cessing algorithm proposed in [19], and the centralized BD (CBD) and decentralized BD (DBD). As the trad- itional BD cannot be directly applied in a decentralized manner, here in DBD, receive processing matrix is de- rived firstly based on the same method for receive pro- cessing matrix in the proposed decentralized algorithm, then the precoding matrix is derived based on the ZF criterion. For the system with a single stream transmit- ted for each user, i.e., lk = 1 (k = 1,…,K) system, we also compare the proposed decentralized algorithm with DZF [20] and DV-SINR [22]. Flat Rayleigh fading channels are considered in simulations. The elements of the chan- nels are i.i.d. complex Gaussian variables with zero mean and unit variance. In this simulation, a 64-QAM modu- lation scheme is employed in the simulation. The signal to noise ratio (SNR) is defined as SNR = P/(SMσ2), where M = 4 is the signal constellation size and S is the average number of the data streams transmitted for each user. Xk−1 i¼1li 2 Nt þ n2 rNt þ nrN2 t . In the proposed decentralized algorithm, every BS has the same computational complexity. For any BS, the design of the relative matrices for the kth user includes the fol- lowing: a onetime computation of the singular vector of a nr × nt matrix with O n2 rnt complexity; onetime multipli- cations of a lk × nr matrix and a nr × nt matrix, the complex- ity of which is O(lknrnt); a onetime computation of the null ity of which is O(lknrnt); a onetime computation of the null space of a Xk−1 i¼1li nt matrix with O Xk−1 i¼1li 2 nt complexity; a onetime multiplication of a lk × nt matrix and a nt nt− Xk−1 i¼1li h i matrix, the complexity of which is O lknt nt− Xk−1 i¼1li h i ; and lk-times computation of the Moore-Penrose pseudo-inverse of a nt− Xk−1 i¼1li h i lk−i ½ i ¼ 1; ⋯; lk ð Þ matrix, the complexity of which is O nt− Xk−1 i¼1li h iXlk i¼1 lk−i ½ 2 . 5.1 Balance BER performance among streams of each user Figure 2 verifies the balance performance among the streams of each user in the proposed algorithms. We consider a 3-cell coordinated system with nt = 6 transmit antennas and K = 3 users each equipped with nr = 3 re- ceive antennas. There are lk = 2 (k = 1,…,K) data streams transmitted for each user. The BER performance of two streams of the first user and the third user are shown. It can be seen that the two streams of any user achieve the approximately equal BER, not only for the centralized algorithm, but also for the decentralized algorithm. The simulation results are in accordance with the theor- etical analysis. a nt nt− Xk−1 i¼1li h i matrix, the complexity of which is O lknt nt− Xk−1 i¼1li h i ; and lk-times computation of the Moore-Penrose pseudo-inverse of a nt− Xk−1 i¼1li h i lk−i ½ i ¼ 1; ⋯; lk ð Þ matrix, the complexity of which is O nt− Xk−1 i¼1li h iXlk i¼1 lk−i ½ 2 . The complexity of other scalar computations can be ignored. Therefore, the complexity of the relative matrices designed for the kth user is O Xk−1 i¼1li 2 Nt þ n2 rNt þ Nlkn2 t þ l3 kNt . Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 10 of 14 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER first user−first stream first user−last stream last user−first stream last user−last stream first user−first stream first user−last stream last user−first stream last user−last stream proposed centralized algorithm proposed decentralized algorithm Figure 2 BER performance of each stream in the proposed two algorithms with N = 3, K = 3, nt = 6, nr = 3, lk = 2 (k = 1,…,K). Figure 2 BER performance of each stream in the proposed two algorithms with N = 3, K = 3, nt = 6, nr = 3, lk = 2 (k = 1,…,K). 5.2 BER performance comparison of different algorithms Figure 3 presents the BER performance comparison of the six algorithms. A 2-cell coordinated system with nt = 6 transmit antennas and K = 3 users each equipped with nr = 5 antennas is considered. 5.1 Balance BER performance among streams of each user The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). It is noticed that in the interference-free algorithm, only a single user is served by BSs. On the whole, centralized algorithms have better performance than decentralized algorithms, at the cost of information 5.2 BER performance comparison of different algorithms Figure 3 presents the BER performance comparison of the six algorithms. A 2-cell coordinated system with nt = 6 transmit antennas and K = 3 users each equipped with nr = 5 antennas is considered. The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). It is noticed that in the interference-free algorithm, only a single user is served by BSs. On the whole, centralized algorithms have better performance than decentralized algorithms, at the cost of information exchange among BSs. For the proposed algorithms, the centralized algorithm achieves about 7-dB gain related to the decentralized algorithm at BER = 10−3. Compared with the existing algorithms, when BER = 10−3, the proposed centralized algorithm has an approximately 5-dB gain to the algorithm in [19] and a 10-dB gain to CBD. Also, about a 10-dB gain is achieved by the proposed decentra- lized algorithm related to DBD at BER = 10−2. In Figure 4, we consider the BER performance of a 3- cell coordinated system, with nt = 6 transmit antennas 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER interference−free proposed centralized algorithm in [19] CBD proposed decentralized DBD Figure 3 BER performance of centralized and decentralized algorithms with N = 2, K = 3, nt = 6, nr = 5, lk = 2 (k = 1,…,K). 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER interference−free proposed centralized algorithm in [19] CBD proposed decentralized DBD Figure 3 BER performance of centralized and decentralized algorithms with N = 2, K = 3, nt = 6, nr = 5, lk = 2 (k = 1,…,K). Figure 3 BER performance of centralized and decentralized algorithms with N = 2, K = 3, nt = 6, nr = 5, lk = 2 (k = 1,…,K). Hu et al. 5.1 Balance BER performance among streams of each user EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 11 of 14 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER interference−free proposed centralized algorithm in [19] CBD proposed decentralized DBD Figure 4 BER performance of centralized and decentralized algorithms with N = 3, K = 3, nt = 6, nr = 3, lk = 2 (k = 1,…,K). Figure 4 BER performance of centralized and decentralized algorithms with N = 3, K = 3, nt = 6, nr = 3, lk = 2 (k = 1,…,K). performance gains among algorithms are different, as they are related with system configuration. and K = 3 users each equipped with nr = 3 antennas. The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). As mentioned in Figure 3, in the interference-free algorithm, only a single user is served by BSs. As can be seen from Figure 4, centralized algorithms have better BER performance than decentra- lized algorithms. The proposed centralized algorithm has a lower BER than the algorithm in [19] and CBD, and the proposed decentralized algorithm achieves bet- ter performance than DBD. Compared with Figure 3, the In Figure 5, the performance of the proposed decen- tralized algorithm for the system with a single stream transmitted for each user, i.e., lk = 1 (k = 1,…,K), is veri- fied and compared with the existing decentralized algo- rithms, DZF [20], and DV-SINR [22] in BER. A 3-cell coordinated system with nt = 6 transmit antennas and K = 5 users each equipped with nr = 3 antennas is considered. As can be seen from Figure 5, the proposed decentralized 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER proposed decentralized algorithm DBD DVSINR Figure 5 BER performance of different decentralized algorithms with single-stream users, N = 3, K = 5, nt = 6, nr = 3, lk = 1 (k = 1,…,K). 5.1 Balance BER performance among streams of each user 0 5 10 15 20 25 30 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER proposed decentralized algorithm DBD DVSINR Figure 5 BER performance of different decentralized algorithms with single-stream users, N = 3, K = 5, nt = 6, nr = 3, lk = 1 (k = 1,…,K). Figure 5 BER performance of different decentralized algorithms with single-stream users, N = 3, K = 5, nt = 6, nr = 3, lk = 1 (k = 1,…,K). rformance of different decentralized algorithms with single-stream users, N = 3, K = 5, nt = 6, nr = 3, lk = 1 (k = 1,…,K Page 12 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 12 of 14 algorithm has a lower BER than other algorithms. When BER = 10−3, it can achieve an approximately 7-dB gain compared with DZF, and a 5-dB gain compared with DV- SINR. nt = 6 transmit antennas and nr = 3 receive antennas. The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). As can be seen from Figure 7, the increased number of users enlarges the performance differences among the algorithms. The tx-rx processing matrices of each user, in CBD, are used to eliminate the interference of all other users. Differently, in the proposed centralized algorithm and the algorithm in [19], they are used to eliminate the interference of part of the other users, bringing in more space dimensions for the diversity gain. 5.3 The effect of the receive antennas and user's number to centralized algorithms Figure 6 illustrates the effect of the number of receive antennas to the proposed centralized algorithm, the al- gorithm in [19], and the CBD. We consider a 3-cell coor- dinated system with nt = 6 transmit antennas and K = 3 users. The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). As can be seen from Figure 6, the performance difference between the pro- posed centralized algorithm and the algorithm in [19] is increased with the number of the receive antennas. In the proposed centralized algorithm, the receive processing matrix is considered into the MU-MIMO channel de- composition. Compared with the algorithm in [19], the decomposed SU-MIMO channels have larger dimensions, which increases the system diversity gain and improves the system performance. With a larger number of the re- ceive antennas, the decomposed SU-MIMO channels have the same dimension in the proposed centralized algorithm but have smaller dimensions in the algorithm in [19]. Therefore, with increased number of the receive antennas, the proposed centralized algorithm can achieve more performance gain than the algorithm in [19]. 5.4 BER performance of the proposed algorithms in a noncoordinated system In Figure 8, we illustrate the performance of the proposed algorithms for the noncoordinated system with N = 1 and compare them with CBD. In this situation, CBD is equiva- lent to the traditional BD in a single-cell MIMO system. A MU-MIMO system, in which there are nt = 8 transmit antennas and K = 4 users each equipped with nr = 2 antennas, is considered. The number of the data streams transmitted to each user is set to 2, i.e., lk = 2 (k = 1,…,K). It is shown that the proposed algorithms can achieve better BER performance than BD, and that the proposed decentralized algorithm is only a suboptimal scheme for the noncoordinated system, as part of the receive pro- cessing matrix is not jointly derived with the transmit processing matrix. In this situation, the proposed centralized algorithm is verified to achieve lower BER than the proposed decentralized algorithm. It exhibits an approximately 6-dB gain over the decentralized scheme at BER = 10−2. In Figure 7, the effect of the number of users to the pro- posed centralized algorithm, the algorithm in [19] and CBD is illustrated. We consider a 3-cell coordinated system with 0 5 10 15 20 25 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER nr=6,K=3 nr=3,K=3 proposed centralized algorithm CBD algorithm in [19] Figure 6 The effect of the number of receive antennas nr to different centralized algorithms with N = 3, nt = 6, lk = 2 (k = 1,…,K). 0 5 10 15 20 25 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER nr=6,K=3 nr=3,K=3 proposed centralized algorithm CBD algorithm in [19] Figure 6 The effect of the number of receive antennas nr to different centralized algorithms with N = 3, nt = 6, lk = 2 (k = 1,…,K). Figure 6 The effect of the number of receive antennas nr to different centralized algorithms with N = 3, nt = 6, lk = 2 (k = 1,…,K). Hu et al. 5.4 BER performance of the proposed algorithms in a noncoordinated system EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Page 13 of 14 0 5 10 15 20 25 30 10 −6 10 −5 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER proposed centralized algorithm algorithm in [19] CBD nr=3,K=3 nr=3,K=6 Figure 7 The effect of the number of users K to different centralized algorithms with N = 3, nt = 6, lk = 2 (k = 1,…,K). Figure 7 The effect of the number of users K to different centralized algorithms with N = 3, nt = 6, lk = 2 (k = 1,…,K). References Z Keke, RC de Lamare, M Haardt, Multi-branch Tomlinson-Harashimaprecoding design for MU-MIMO systems: theory and algorithms. IEEE Trans Commun. 62(3), 939–951 (2014) 6. S Jing, D Tse, J Soriaga, J Hou, J Smee, R Padovani, Multicell downlink capacity with coordinated processing. EURASIP J. Wireless Commun.Netw 2008, 586878 (2008) 6. S Jing, D Tse, J Soriaga, J Hou, J Smee, R Padovani, Multicell downlink capacity with coordinated processing. EURASIP J. Wireless Commun.Netw 2008, 586878 (2008) 7. S Liyan, Y Chenyang, H Shengqian, The value of channel prediction in CoMP systems with large backhaul latency. IEEE Trans Commun. 61(11), 4577–4590 (2013) 7. S Liyan, Y Chenyang, H Shengqian, The value of channel prediction in CoMP systems with large backhaul latency. IEEE Trans Commun. 61(11), 4577–4590 (2013) 8. Papadogiannis, E Hardouin, D Gesbert, Decentralisingmulticell cooperative processing: a novel robust framework. EURASIP J. Wireless Commun.Netw 2009, 890685 (2009) 9. R Zhang, Cooperative multi-cell block diagonalization with per-basestation power constraints. IEEE J. Sel. Areas Commun. 28, 1435–1445 (2010) 10. S Shi, M Schubert, N Vucic, H Boche, MMSE optimization with per-base- station power constraints for network MIMO systems. Proceedings of the IEEE International Conference Communications (IEEE, Beijing, 2008), pp. 4106–4110 11. J Zhang, Y Wu, S Zhou, J Wang, Joint linear transmitter and receiver design for the downlink of multiuser MIMO systems. IEEE Commun Lett 9(11), 991–993 (2005) 12. RC de Lamare, Adaptive and iterative multi-branch MMSE decision feedback detection algorithms for multi-antenna systems. IEEE Trans. Wirel. Commun. 12(10), 5294–5308 (2013) 13. H Park, SH Park, HB Kong, I Lee, Weighted sum MSE minimization under per-BS power constraint for network MIMO systems. IEEE Commun. Lett 16(3), 360–363 (2012) Submit your manuscript to a journal and beneﬁ t from: 7 Convenient online submission 7 Rigorous peer review 7 Immediate publication on acceptance 7 Open access: articles freely available online 7 High visibility within the ﬁ eld 7 Retaining the copyright to your article Submit your next manuscript at 7 springeropen.com 14. S He, Y Huang, L Yang, B Ottersten, Coordinated multicell multiuser precoding for maximizing weighted sum energy efficiency. IEEE Trans. Signal Process. 62(3), 741–751 (2014) 15. M Wei, C Xiang, Z Ming, W Jing, Joint stream-wise THP transceiver design for the multiuser MIMO downlink. IEICE Trans. Commun. 92(1), 209–218 (2009) 16. W Hardjawana, B Vucetic, Y Li, Multi-user cooperative base station systems with joint precoding and beamforming. Competing interests The authors declare that they have no competing interests. 20. R Holakouei, A Silva, A Gameiro, Distributed versus centralized zero-forcing precoding for multicell OFDM systems. Proceedings of the IEEE Globe Communications ConferenceWorkshops (IEEE, Houston, 2011), pp. 188–193 Acknowledgements Thi k i d y g j g Construction Project and the Beijing Natural Science Foundation (4144079). 21. R Zakhour, D Gesbert, Distributed multicell-MISO precoding using the layered virtual SINR framework. IEEE Trans. Wireless Commun. 9(8), 2444–2448 (2010) Author details 1 1Beijing Key Laboratory of Network System Architecture and Convergence, Beijing University of Posts and Telecommunications, Beijing 100876, China. 2School of Communication Engineering, Hangzhou Dianzi University, 1Beijing Key Laboratory of Network System Architecture and Convergence, Beijing University of Posts and Telecommunications, Beijing 100876, China. 2School of Communication Engineering Hangzhou Dianzi University 22. E Bjornson, R Zakhour, D Gesbert, B Ottersten, Cooperative multicell precoding: rate region characterization and distributed strategies with instantaneous and statistical CSI. IEEE Trans. Signal Processing 58(8), 4298–4310 (2010) Hangzhou 310018, China. 3School of Electronic Engineering and Computer Science, Queen Mary University of London, London E1 4NS, UK. Hangzhou 310018, China. 3School of Electronic Engineering and Computer Science, Queen Mary University of London, London E1 4NS, UK. 23. X Zhao, H Xu, X Yang, Performance enhancement for CoMP based on power allocation and a modified ZF-THP. Proceedings of the IEEE Personal Indoor and Mobile Radio Communications (IEEE, Sydney, 2012), pp. 2309–2313 Received: 3 August 2014 Accepted: 5 January 2015 Received: 3 August 2014 Accepted: 5 January 2015 24. I Krikidis, B Ottersten, Diversity fairness in Tomlinson–Harashimaprecoded multiuser MIMO through retransmission. IEEE Signal Process Lett. 20(4), 375–378 (2013) References 25. G Boudreau, J Panicker, N Guo, R Chang, N Wang, S Vrzic, Interference coordination and cancellation for 4G networks. IEEE Commun. Mag. 47(4), 74–81 (2009) 1. M Sawahashi, Y Kishiyama, A Morimoto, D Nishikawa, M Tano, Coordinated multipoint transmission/reception techniques for LTE-advanced [Coordinated and Distributed MIMO]. IEEE Wireless Commun. 17(3), 26–34 (2010) 1. M Sawahashi, Y Kishiyama, A Morimoto, D Nishikawa, M Tano, Coordinated multipoint transmission/reception techniques for LTE-advanced [Coordinated and Distributed MIMO]. IEEE Wireless Commun. 17(3), 26–34 (2010) and Distributed MIMO]. IEEE Wireless Commun. 17(3), 26–34 (2010 26. J Tang, S Lambotharan, Interference alignment techniques for MIMO multi-cell interfering broadcast channels. IEEE Trans. Commun. 61(1), 164–175 (2013) 2. D Lee, H Seo, B Clerckx, E Hardouin, D Mazzarese, S Nagata, K Sayana, Coordinated multipoint transmission and reception in LTE-advanced: deployment scenarios and operational challenges. IEEE Commun. Mag. 50(2), 148–155 (2012) 2. D Lee, H Seo, B Clerckx, E Hardouin, D Mazzarese, S Nagata, K Sayana, Coordinated multipoint transmission and reception in LTE-advanced: deployment scenarios and operational challenges. IEEE Commun. Mag. 50(2), 148–155 (2012) 27. R.F.H.Fisher, Precoding and signal shaping for digital transmission (John Wiley & Sons Ltd, 2002) 3. D Gesbert, S Hanly, H Huang, S ShamaiShitz, O Simeone, W Yu, Multi-cell MIMO cooperative networks: a new look at interference. IEEE J. Sel. Areas Commun. 28(9), 1380–1408 (2010) 3. D Gesbert, S Hanly, H Huang, S ShamaiShitz, O Simeone, W Yu, Multi-cell MIMO cooperative networks: a new look at interference. IEEE J. Sel. Areas Commun. 28(9), 1380–1408 (2010) 28. Y Jiang, W Hager, J Li, The gene Comput. 77, 1037–1056 (2007) 28. Y Jiang, W Hager, J Li, The generalized triangular decomposition. Math. Comput. 77, 1037–1056 (2007) 29. A Wiesel, E Yonina, CS Shlomo, Linear precoding via conic optimization for fixed MIMO receivers. IEEE Trans. Signal Process. 54(1), 161–176 (2006) 4. K Karakayali, GJ Foschini, RA Valenzuela, R Yates, On the maximum common rate achievable in a coordinated network. Proceedings of the IEEE International Conference Communications (IEEE, Istanbul, 2006), pp. 4333–4338 4. K Karakayali, GJ Foschini, RA Valenzuela, R Yates, On the maximum common rate achievable in a coordinated network. Proceedings of the IEEE International Conference Communications (IEEE, Istanbul, 2006), pp. 4333–4338 5. Z Keke, RC de Lamare, M Haardt, Multi-branch Tomlinson-Harashimaprecoding design for MU-MIMO systems: theory and algorithms. IEEE Trans Commun. 62(3), 939–951 (2014) 5. 6 Conclusions unit. The proposed decentralized algorithm allows each BS design to transmit precoding in a decentralized manner, which alleviates the demand on the backhaul capacity. The analysis and simulation results show that the centralized algorithm achieves better performance than the decentra- lized algorithm. And, the proposed algorithms achieve better performance than the existing joint tx-rx processing algorithms and the decentralized linear precodings. Nonlinear joint tx-rx processing technology for a coordi- nated multi-cell system with multi-stream multi-antenna users has been studied. The capacity of the backhaul link determines different coordinated modes among BSs, in- cluding centralized and decentralized coordinated. The proposed centralized algorithm is proposed to derive the tx-rx processing matrices jointly at the central processing 0 5 10 15 20 25 30 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER proposed centralized algorithm proposed decentralized algorithm CBD Figure 8 BER performance of the proposed algorithms in noncoordinated multi-cell system with N = 1, K = 4, nt = 8, nr = 2, lk = 2 (k = 1,…,K). 0 5 10 15 20 25 30 10 −4 10 −3 10 −2 10 −1 10 0 SNR/dB BER proposed centralized algorithm proposed decentralized algorithm CBD Figure 8 BER performance of the proposed algorithms in noncoordinated multi-cell system with N = 1, K = 4, nt = 8, nr = 2, lk = 2 (k = 1,…,K). BER R performance of the proposed algorithms in noncoordinated multi-cell system with N = 1, K = 4, nt = 8, nr = 2, …,K). Figure 8 BER performance of the proposed algorithms in noncoordinated multi-cell system with N = 1, K = 4, nt = 8, nr = 2, lk = 2 (k = 1,…,K). Page 14 of 14 Hu et al. EURASIP Journal on Advances in Signal Processing (2015) 2015:10 Competing interests 19. L Sun, M Lei, Adaptive joint nonlinear transmit-receive processing for multi-cell MIMO networks. Proceedings of the IEEE Globe Communications Conference (IEEE, Anaheim, 2012), pp. 3766–3771 Competing interests The authors declare that they have no competing interests. References IEEE J Sel Top Signal Process 3(6), 1079–1093 (2009) 17. S Adão, H Reza, G Atílio, Power allocation strategies for distributed precodedmulticell based systems. EURASIP J. Wireless Commun. Netw 2011, 1 (2011) 17. S Adão, H Reza, G Atílio, Power allocation strategies for distributed precodedmulticell based systems. EURASIP J. Wireless Commun. Netw 2011, 1 (2011) 18. Y Sun, M Wu, M Zhao, C Xu, Transceiver designs using non-linear precoding for multiuser MIMO systems with limited feedback. Proceedings of the IEEE VehicularTechnology Conference (IEEE, Dresden, 2013), pp. 1–5 18. Y Sun, M Wu, M Zhao, C Xu, Transceiver designs using non-linear precoding for multiuser MIMO systems with limited feedback. Proceedings of the IEEE VehicularTechnology Conference (IEEE, Dresden, 2013), pp. 1–5
https://openalex.org/W1994822905	http://estudiosgallegos.revistas.csic.es/index.php/estudiosgallegos/article/download/217/220	es		A pequeña fidalguia rural e a parcería de gando: a casa da Fraga de San Xiao de Carballo-Friol, 1680-1800	Cuadernos de estudios gallegos	1,997	cc-by	8,288	A PEQUEÑA FIDALGUIA RURAL E A PARCERÍA DE GANDO: A CASA DA FRAGA DE SAN XIAO DE CARBALLO -FRIOL-, 1680-1800 Por ANTONIO PRESEDO GARAZO' I. INTRODUCCIÓN A chamada fidalguía, adxectivizada progresivamente de rural, é un dos temas preferidos pola investigación histórica modernista galega na actualidade. Un grupo heteroxéneo obxecto de interese na medida en que se foi demostrando a súa posición de privilexio nunha sociedade rural na súa base, como detentador —á vez que posuidor— do dominio útil dos medios de producción, cando non tamén do directo dominio. As monografías comarcáis realizadas ó longo da década de 1970 pola área de Historia Moderna da Faculdade de Xeografía e Historia de Santiago de Compostela, resaltan a presencia dun grupo social minoritario desde o punto de vista porcentual que sobresae por riba do resto da poboación, posuíndo unha destacada parte das terras ñas áreas obxecto de estudio, que acceden ó disfrute do seu usufructo nun contexto que, como era ' Bolseiro predoutoral do Departamento de Historia II —área de Historia Moderna— da Universidade de Santiago de Compostela, pola Consellería de Educación e Ordenación Universitaria da Xunta de Galicia. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 54 A PEQUEÑA FIDALGUIA RURAL... O da Galicia do Antigo Réxime, se circunscribía os recursos xerados a partir da economía agraria^. Deste xeito cobra importancia o concepto de hidalguía intermediaria^ —ou si se prefire o de señores medianeros"^—, que acada os seus logros mais salientables coa tese de R. Villares Paz^. Desde entón, o estudio desta elite do mundo rural está sendo tratado progresivamente desde diversas perspectivas de análise ñas que se venen primando fundamentalmente os componentes económicos e sociais, se ben é certo que circunscribindo as distintas análises a cuestións puntuáis que ás veces teñen relegado a un segundo plano elementos interpretativos de relevo que sería interesante desenvolver mais polo miúdo. Os logros acadados na investigación en relación con esta temática —derivados de análises concretas de casas fidalgas espalladas pola xeografía galega ó longo do Antigo Réxime—, permítennos dar hoxe en día unha visión de conxunto, da que o primeiro elemento a destacar e sen lugar a dúbidas o seu carácter heteroxéneo^. Unha heteroxeneidade extensible á consolidación dominial, pois non todos aqueles que queren chegar a fidalgo o fan do mesmo xeito. Todo o contrario. A realidade " En Trasdeza suponen na primeira metade do século XVIII o 1,86% do monto demográfico total; no Xallas non superan o 10,8% en 1708; no concello de Burón non chegan ó 5%; na Ulla representan o 0,79%; no Morrazo as porcentaxes móvense entre 1,14 e 1,3%; e na provincia Mondoñedo non superan o 3%. RODRIGUEZ FERRIERO, H.: La Tierra de Trasdeza. Una economía rural antigua, Santiago, 1973, p.83. BARREIRO MALLON, B.:La jurisdicción de Xallas en el siglo XVIIL Población, sociedad y economía, Santiago, 1973, p.570. SAAVEDRA, R: Economía rural antigua en la montaña lucense. El concejo de Burón, Santiago, 1979, p.78. REY CASTELAO, O.: Aproximación a la historia rural en la comarca de la Ulla (ss.XVIIy XVIII), Santiago, 1981, p. 159. ^ Posto en funcionamento por A. Eiras Roel en «Régimen subforal e hidalguía intermediaria», na introducción ó libro de M^.C. Quintans: El dominio de San Martín Finarlo ante la desamortización, Santiago, 1972, pp. 10-12. ^ Seguido polos historiadores economicistas galegos logo do traballo de J. GarcíaLombardero: La agricultura y el estancamiento económico de Galicia en La España del Antiguo Régimen, Madrid, 1973. ^ VILLARES, R.: La propiedad de la tierra en Galicia, 1500-1936, Madrid, 1982. ^ LEIROS DE LA PEÑA, P.: La Casa de Fonte Fiz (Contribución al estudio de la hidalguía gallega), Santiago, 1986, memoria de licenciatura inédita; DOMÍNGUEZ CASTRO, L.: Vinos, viñas e xentes do Ribeiro, Vigo, 1992; MIGUEZ, V. M.: Aproximación ao estudo da fidalguía galega a travesó do Marquesado de Hombreiro (1500-1800), A Coruña, 1993, memoria de licenciatura inédita; e PRESEDO, A.: Comportamentos económicos e sociais da pequeña fidalguía na Galicia interior, ss.XVI-XVIII, Santiago, 1995, memoria de licenciatura inédita. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 55 estructural das áreas escollidas para a sua consolidación como grupo dominante determina de antemán a composición dos sens futuros patrimonios. Deste xeito, hai individuos que constrúen os seus montos patrimoniais familiares a partir do usufructo do dominio útil de terras pertencentes a bieitos e cistercienses, que así mesmo subaforan á masa campesina^. Neste contexto o subforo xoga un papel de relevo para a consolidación dos patrimonios fidalgos. Noutros casos, estes individuos aspirantes ó privilexio da nobreza fanse con propiedades do campesinado que vive ó límite das súas posibilidades —debido á presencia maioritaria do minifundio como estructurador da explotación agraria— case que nun estado de subsistencia. Unha mala conxuntura, ou mesmo unha morte non esperada catapultan ó labrego á precariedade económica, obrigándoo a se desfacer de lotes de terra e outro tipo de bens que unha vez postos no mercado acaban ñas mans desta fidalguíal A heteroxeneidade tamén se fai extensible á hora de analizármo-la composición dos seus dominios. Tanto na análise das diversas calidades do terrádego dos seus patrimonios fundiarios, comprobando unha composición non uniforme, como á hora de adentrarnos no estudio da composición da renda, vése con facilidade a primacía da non concordancia e igualdade dos resultados. En determinadas áreas xeograficas do interior da provincia de A Coruña e de Lugo, o monte domina como calidade do terrádego por riba do prado e incluso do labradío^; mentres que no Ribeiro do Avia, as terras a vide convírtense no eixo que vertebra a explotación do patrimonio^^. E o mesmo se observa cando dirixímo-lo noso interese ' VILLARES, R.: Op. ciL, p.81 e ss., e PRESEDO, A.: Op. cit., p.54 e ss. « VILLARES, R. Op. cit., p.88 e ss.; BARREIRO MALLON, B.: «El dominio de la familia Porras y la evolución de las rentas agrarias en la Tierra de Santiago», in Obradoiro de Historia Moderna. Homenaje al Prof. Antonio Eiras Roel, Santiago, 1990, p.29 e ss.; MIGUEZ, VM.: «O dominio da Casa de San Fiz de Asma: evolución e inversión dun dominio fidalgo durante o Antigo Réxime», Cuadernos de Estudios Gallegos, t.XLII, 1995, pp.37-67; e PRESEDO, A.: «Da casa de labranza ó pazo: A pequeña fidalguía rural na Galicia interior no Antigo Réxime», Obradoiro de Historia Moderna, no prelo. 9 PRESEDO, A.: art. cit. '° L. Domínguez Castro así o confirma ó estudia-la familia Meruéndano, no Ribeiro do Avia, se ben é certo que no patrimonio familiar a extensión a monte -54.930m^- supera á da viña -39.767m^-, a racionalidade da explotación tende a dar preferencia a esta segunda calidade. Vid. DOMÍNGUEZ CASTRO, L.: Op. cit., p.37. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es Mapa I. Concello de Friol. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es Mapa IL S. Xiao de Carballo Mapa III. Lugar da Fraga (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 58 A PEQUEÑA FIDALGUIA RURAL... cara as rendas agrarias, comprobando que non so se cobran a partir de contratos agrarios que varían dunhas comarcas para outras, senón que a composición das mesmas non é nen moito menos uniforme a teor dos estudios realizados^'. Estamos diante dun colectivo que aspira a acadar unha posición destacada dentro do conxunto social, que unifica as súas estratexias unha vez supera a fase de vinculación dos seus bens. Así, a heteroxeneidade estructural observada nos seus dominios —tanto no que toca á sua consolidación como no que respecta á siía xestión— deixa paso a unha homoxeneidade que se plasma ñas estratexias familiares, tendentes a fortifica-la posición privilexiada do grupo en relación co resto da sociedade agraria, e tamén no xeito de vida no que as formas simbólicas de poder xogan un papel de primeira orde, igual que as reais. Non obstante, a existencia de determinados aspectos aínda non totalmente desenvolvidos pola investigación histórica, nos obrigan a estudiar con puntualidade cuestións das que non se conta con datos globalizadores. Un deles é a imbricación da fidalguía rural galega dos sáculos XVI-XIX/ 1 na economía protoindustrial, á que se ten referido V. M. Míguez Rodríguez á hora de estudia-la Casa de Quinta no Caurel en relación coa explotación da siderurxia do ferro na área da Galicia centro-orientaP^. E tamén a importancia acadada por este grupo como posuidor de cabanas gandeiras numéricamente importantes, estudiadas por J.M. Pérez García no momento anterior á fundación de vínculos'^ asunto tamén desenvolvido na nosa memoria de licenciatura'"^. E precisamente este aspecto gandeiro da economía fidalga, o que ha de vertebra-lo presente traballo, pois a Casa da Fraga sita no actual concello de Friol —na Galicia interior— " Esta heteroxeneidade da renda coincide coa panorámica analizada nos estratos mais elevados da nobreza galega no Antigo Réxime. Vid. BAZ VICENTE, M^.J.: El patrimonio de la Casa de Alba en Galicia en el siglo XIX, Lugo, 1991, pp. 105-106; GARCIA ACUÑA, M^.L.: «Unha primeira aproximación os ingresos do estado de Ribadavia no século XVIII», in Historia Nova III, Santiago, 1995, pp. 125-148. '- MIGUEZ, V.M.: «Algunas consideraciones al respecto de la hidalguía gallega a través de la Casa de San Fiz de Asma y agregadas (1500-1800)», Obradoiro de Historia Moderna, n°3, 1994, pp.204-205 e táboa 5 -p.210-. '^ PEREZ GARCIA, J.M.: «Niveles y transformaciones de la ganadería en el siglo XVn», Cuadernos de Estudios Gallegos, t.XXXIII, 1982, pp. 139-140. '4 PRESEDO, A.: Op. cit., p.l28 e ss. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 59 permítenos acceder a coñece-la racionalidade da especialización económica ó redor do sector gandeiro por parte dos Fraga, que habitaron ó longo da Modernidade e aínda en boa parte da Contemporaneidade o caserón afidalgado da parroquia friolense de San Xiao de Carballo^^. II. A CONSOLIDACIÓN DO DOMINIO Os Fraga construen o seu dominio na freguesía de San Xiao de Carballo ñas dúas derradeiras décadas do sáculo XVII, e o consolidan nos vinte primeiros anos do XVIir^. En concreto, acceden a unha serie de posesións pertencentes a determinados vecinos da parroquia friolense, que Helos venden por separado, permitindo que o investidor en cuestión —Gregorio da Fragas/ mozo^^— acceda á propiedade dunha serie de predios e lugares que pasan a conformar en menos de vinte anos o solar orixinario desta familia que chega tardíamente á fidalguía'^. Cinco operacións de compra-venda feitas entre 1685 e 1697 ñas que desembolsa unha cantidade 737 Rs. confirman o seu interese por establecerse en Carballo. Os bens adqueridos son, dunha banda pezas labradías —catro en total— vendidas por dous labregos da freguesía —Juan do Carregal e Domingo de la Iglesia— entre 1685 e 1690 ó precio uniforme de doce ducados por unidade^^; e doutra todo o lugar da Fraga que Gregorio da Fraga do Celeiro e a sua muller María de Buxán lie '^ A documentación empregada para a elaboración do presente traballo foi catalogada na nosa memoria de licenciatura. Para unha mellor comprensión do texto e facilita-la lectura, empregámo-la abreviatura de A.C.K para referirnos ó Arquivo da Casa de Fraga, A.H.RL. para o Arquivo Histórico Provincial de Lugo, e A.C.S. para o Arquivo da Catedral de Santiago. '^ Vid. os mapas I, II e III. '^ Vid. cadro xenealóxico no Apéndice I. "^ Esta chegada tardía coincide coa denominación dada por I. Dubert de «segunda hornada de hidalgos» para a Galicia Oriental. Vid. DUBERT, I.: «Espacio y comportamientos sociales en la Galicia de la época moderna», in Concepcións espaciáis e estratexias territorials na Historia de Galicia, Santiago, 1993, p.l53. '^ A.CF., na Partilla dos bens fincables de Gregorio da Fraga e a sua muller Dominga Fernández (maio de 1726), aparecen especificadas estas vendas: 1685, marzo,8; 1686, abril, 22; 1687, febreiro, 11; e 1690, abril, 5. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 60 A PEQUEÑA FIDALGUIA RURAL... venden en 1696 por dezanove ducados^^. Queda clara a existencia dun ritmo investidor que determina a chegada do especulador en primeiro termo a fincas separadas, mais fácilmente vendibles por parte da comunidade campesina unha vez que as circunstancias adversas fan mella inesperadamente no seo da economía doméstica; deixando nun segundo momento a compra dunha explotación racionalizada e centralizada —o lugar da Fraga, composto por cassas, cassares, prados, aguas de riega, arboles y plantíos, eredades labradías y montesías, molino y batan, y todo lo demás. A partir desta derradeira compra os Fraga deixan de especular coas propiedades agrarias e ponen as súas miras ñas actividades económicas de índole gandeiro. En 1707 casa o filio primoxénito de Gregorio da Fraga e Dominga Fernádez -Gerónimo-, e os seus pais o dotan coa mellora de tercio e quinto do remanente dos bens posuídos no lugar da Fraga, comprado en 1696 e considerablemente agrandado a teor dos datos derivados da partilla que se fai en 1726^^ Seguindo pautas de racionalidade que respostan a necesidades particulares dunha economía agraria mais dependente do valor que adquire na provincia de Lugo o concepto de casa^^, os Fraga deciden consolida-lo seu particular dominio a través do beneficiamento dun filio escollido premeditadamente polos patrucios para que rixa a nova unidade patrimonial, engrandecida na xeración que dictamina e designa o futuro da casa aspirante ó ennobrecemento, e en torno a cal se moverá o resto dos componentes do grupo familar. O momento decisivo en que queda escriturada a consolidación do dominio coincide eos derradeiros designios de Gregorio da Fraga, cando en 1717 decide face-lo seu testamento conxuntamente coa sua muller Dominga Fernández. Ante o inminente encontró con Deus como premio polos bos actos feitos nesta vida, deixa ben encarrilados os asuntos mundanos, aspirando ganar mais méritos que lie eviten unha estancia demasiado longa no Purgatorio^^. E o momento ideal, pois pode decidir por 2° A.C.F., Papéis soltos. ^' A.C.F., Partilla dos bens fincables de Gregorio da Fraga e a sua muller Dominga Fernández. ^^ SAAVEDRA, P.: «Casa y comunidad en la Galicia interior», in J.C. Bermejo (coord.): Parentesco, familia y matrimonio en la Historia de Galicia, Santiago, 1989, p.95 e ss.; e DUBERT, L: art. cit., pp.154-156. ^^ GONZALEZ LOPO, D.: «La mortaja religiosa en Santiago entre los siglos XVI y XIX», Compostellanum, v.XXXIV, 1989, p.271 e ss. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO G ARAZO 61 derradeira vez como patrucio da casa —é dicir, como cabeza rectora de todos aqueles que viven baixo o seu teito— o que ha de face-la sua familia logo de que inicie o seu particular tránsito cara ó Paraíso. E o fai pensando en incrementa-las propias posibilidades do grupo familiar para se reproducir socialmente, coa pretensión de aumenta-las particulares espectativas familiares ó participar no mercado marital con mellores posibilidades que lie permitan ir ascendendo xeración tras xeración nos estratos mais elebados da sociedade rural. Fundando ó vínculo coa condición de que tanto a el como a sua dona, Dominga Fernández, se les diga a cada uno su missa de pension en cada un año de siempre jamas por dia de San Gregorio, acada a solución perfecta. Primeiro, porque a misa dita a perpetuidade garante a salvación final do patrucio vinculeiro no Reino de Deus. E segundo, porque asegura a reproducción social das seguintes xeracións da casa ó sinalar unha serie de bens intocables que non poderán ser integrados no mercado: toda la parte de vienes raizes que avian comprado y adquerido de mano de Juan do Carre gal ..., para que la persona que viviere en cassa de los otorgantes los lleve y possea con dha. carga de missas, sin que sea visto poderlos vender, trocar ni enajenar a ninguna persona seglar ni eclesiástica, yglesia, ospital ni cofradía, sino que siempre anden juntos en el poseedor que fuere de dha. casa^'^. En definitiva, a idea —ou dito mais axeitadamente, o concepto— de casa que funciona como unha pina unificadora tendente a beneficiar a aquel que reúna as aptitudes exixidas polos patrucios vinculeiros, está deixando ben claro a primacía dun sistema hereditario no que o filio varón primoxénito está chamado "a aglutinar na sua figura os éxitos socioeconómicos acadados polas xeracións antecesoras da súa^^ O patrucio é unha peza mais dun plan deseñado xeneracionalmente pola mesma familia. Neste contexto, o pagamento dos dotes ós fillos non beneficiados evita a aparición de litixios destes contra o irmán favorecido pola mellora. Sempre, o filio designado para ocupa-la xefatura da casa é quen centrali- ^"^ A.C.F., Partilla dos bens fincables de Gregorio da Fraga e a súa muller Dominga Fernández. ^^ MIGUEZ, V.M.: «Algunas consideraciones al respecto de la hidalguía ...», art. cit., pp. 195-196; e PRESEDO, A.: Op. cit., p.l44 e ss. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es A PEQUEÑA FIDALGUIA RURAL.. 62 za no sen entorno a xestion dunha porcentaxe maior do patrimonio familiar, como se comproba co seguimento do reparto dos bens mobles fincados dos petrucios vinculeiros de 1726, na que Gerónimo da Fraga recibe unha porcentaxe sobradamente destacada dos bens sinalados por riba dos seus irmáns -Táboa 1-. TABOA1 Repartición dos bens mobles de Gregorio da Fraga e a súa muller Dominga Fernández feita en 1726, e a súa capitalización en Rs. Gerónimo María Gregorio Pedro Isabel utillaxe agrícola uteis cocina mobiliario menaxe doméstico gando 132 160 138 88 132 34 12 471 127 77 89 88 148,0 180,0 157,0 254,0 304,5 1.414 1.032 742 675 675 Fonte: Elaboración propia a partir do A.C.R, Partilla dos bens fincables de Gregorio da Fraga e a súa muller Dominga Fernández. Os pais recoñecen la mucha afición, cariño y afecto que thienen al dho. Gerónimo da Fraga, su hijo, y por averies asistido en todos sus achaques y enfermedades, pero lie exixen que case pasando a vivir y residir en este dho. lugar da Fraga, y en compañía del dho. Gregorio da Fraga [e] Dominga Fernandez..., J todos en el tiempo que vivieren estaren a un pan y cuchillo, xa que ata o momento xusto do pasamento cara ó Alen, os fillos -pese a ser futuros fidalgos- dependen dos vellos, sabedores de que foron eles quen construíron o patrimonio sobre o que se ha de asenta-lo vínculo que ennobreza á familia. Unha mellora propia de campesinos acaudalados que pon as bases para o proceso de enriquecemento dos Fraga coincidindo coa súa asimilación á pequeña fidalguía rural. Gerónimo da Fraga, con setenta anos en 1752^^, ' A.H.P.L., Catastro de Ensenada, Libro Personal de Legos, leg.3.343, s.f. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 63 aínda cumpre á perfección coas mandas testamentarias dispostas polos sens pais en 1717^^. Un neto seu —D. Francisco de Fraga— casa en 1784 con D^. Maria Josepha Francisca Várela Pardo y Taibo, filia dos patrucios da Casanova de S. Xoán de Golán, en Melide^^ quen recibe en 1805 a herencia fincable do vínculo melidiense; e outro —D. Manuel Ventura de Fraga— aparece como racioeiro da Catedral de Santiago de Compostela en 1756^9. m . ANALISE DO PATRIMONIO FAMILIAR III.l. A composición fundiaria do patrimonio Hai dúas cuestións que merecen o noso interese de inmediato: saber que distintas calidades de terra conforman o dominio dos Fraga no seu solar orixinario; e verificar se respostan a estratexias suxeridas por outros patrimonios coñecidos da pequeña fidalguía na Galicia interior. A partilla que se fai en 1726 dos bens fincables de Gregorio da Fraga e a sua muller Dominga Fernández, permítenos saber qué parte dos mesmos é adxudicada para o pago da mellora vincular redactada en 1717. ^^ No Catastro de Ensenada aparece perfectamente documentado este cumprimento das dúas misas fundadas: Como asimismo están sujetos [os bens de Gerónimo de Fraga] a dos misas rezadas por cuia limosna asimismo percive el párroco de estafra. quatro reales vellón, y ademas dello quatro ferrados de centeno. A.H.P.L., Catastro de Ensenada, Libro de Reales de Legos, Leg.3.344, f.93v. ^^ Os patrucios en cuestión son D. Juan Francisco Valeriano Várela y Aguiar e D^. María Luisa Pardo. A.CF., Libro Becerro redactado por D. Juan Francisco Valeriano Várela y Aguiar, no f.l58 aparece redactado o contido do dote pagado por eles para o casamento da sua filia co herdeiro da Casa da Fraga, observándose como ademáis do enxoval doméstico, certas rendas e unha egua, D'. María Josepha Francisca é dotada con 9.600 Rs. -^ A.C.S., Limpieza de Sangre. Tomo XI, Leg.740, expediente n%. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 64 A PEQUEÑA FIDALGUIA RURAL... TABOA 2 Bens que conforman o vínculo da Casa da Fraga en 1726^°. TERRAS n°fincas Has. labradío prado monte 13 11 8 4,77 0,78 6,32 % Rs. Rs./n° fincas 40,18 6,57 53,24 4.385 2.246 1.448 337,30 204,18 181,00 Fonte: Elaboración propia a partir do A.C.F, Partilla dos bens fincables de Gregoria da Fraga ... Unha ollada á táboa 2 serve perfectamente para averiguar qué características esenciais se destacan no patrimonio fundiario dos Fraga na freguesía de San Xiao de Carballo a principios do século XVIII. Do total das 11,87 Has., a extensión a monte ocupa o 53,24% con 6,32 Has., deixando ó labradío en segundo posto co 40,18% e ó prado en terceiro cun 6,57%. As fincas máis valoradas son as labradías con 337,30 Rs./ finca, logo o prado con 204,18 Rs./finca, e o monte en terceiro lugar con 181 Rs./finca. E dicir, a maior cantidade de calidade de hectáreas, menor taxación en reas por cada finca, e viceversa^ ^ III. 2. «La cria de ganados» Tanto os viaxeiros que dirixiron os seus pasos cara a Galicia, así como especialistas agrónomos, deixaron especificada desde fins do s.XVIII, ó longo do XIX e principios do XX, a inmensa riqueza deste reino no que se refire ó sector pecuario, definíndoo como bien de la industria del la- ^° Unha comparación eos datos derivados do Catastro de Ensenada confirman unha vez máis o ánimo ocultador por parte dos enquisados debido á finalidade fiscalizadora desta fonte. En 1726 a mellora vincular é pagada con 11,87 Has. no lugar da Fraga, mentres que en 1752 non temos nen tan sequera declarada a metade: 5,14 Has. A.H.P.L., Catastro de Ensenada, Libro de Reales de Legos, Leg.3.344, ff.91-94. ^' Estes resultados coinciden co que sabemos doutros patrimonios fidalgos tamén da Galicia interior. PRESEDO, A.: art. cit. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 65 brador'^. Unha orixinalidade que resposta as limitacions mesmas da explotación agraria, caracterizada pola excesiva división do terrádego, que obriga ó labrego a se surtir de bens complementarios que suplan ós xustísimos recursos xerados pola agricultura, sempre ó límite da supervivencia. Esta riqueza localízase con mais pulo na Galicia interior e orientaP^ respostando ás particulares peculiaridades da agricultuira das sociedades tradicionais nesta área, sempre mais retardataria á hora de introducilas innovacións nos sistemas agrarios que callan con mais facilidade na Galicia occidental. A explotación do monte ven ratifica-la complementariedade existente ñas sociedades rurais de Antigo Réxime entre a gandería e a agricultura, afirmándose a unha sobre a outra na medida en que as terras a monte producen o abono necesario para as agras labradías, mentres que o barbeito destas terras obriga a roza-lo monte, humanizado directamente polo home, e indirectamente polas súas cabanas gandeiras^"^. ^^ SANCHEZ, P.A.: La economía gallega en los escritos de ..., Vigo, 1973, p.ll5. Sirva de exemplo ilustrativo a testemuña de Francisco de Paula, cando di en 1850 que£'n la cria de ganados se aventajan los gallegos á los habitantes de todas las otras provincias, proporcionando á aquellos escelentes pastos por medio de prados naturales y artificiales. PAULA MELLADO, F. de: Recuerdos de un viaje por Galicia en 1850, A Coruña, 1987, p.3. ^^ A análise espacial do gando vacún para Galicia en 1750 é de 3,5 reses por vecino. Esta media espacial destaca con claridade ós concellos do interior -A Cápela, Curtis, Sobrado, As Pontes, Muras- con cifras que fan oscila-lo índice entre 7 e 8 pezas por unidade vecinal. Vid. DOPICO, F. et al.: «A distribucuión espacial do gando en Galicia segundo o Catastro de Ensenada», Revista Galega de Estudios Agrarios, n°5,1981, pp.3583. Esta diferenciación é aínda mais contundente de revisármo-las ratios de n° de reses/ vecino obtidos para dúas áreas de transición entre os dous modelos agrarios -o Xallas e A Ulla-, e os comparamos cunha comarca de montaña como A Fonsagrada. No Xallas, a media está ñas 3,9 reses/veciño, baixando na Ulla ata 1,8; en cambio, na Fonsagrada o índice sube ata 6,2. Vid. BARREIRO MALLON, B.:Op. cit., p.374; REY CASTELAO, O.: Op. cit., p. 119; e SAAVEDRA, P.: Op. cit., p.49. Esta superioridade do producto gandeiro por vecino —unidade de explotación— tamén é destacada para as terras altas por A. Eiras Roel, quen dá para estas áreas medias que oscilan entre os 177 animais como máximo e os 89 como mínimo, por riba das terras baixas. EIRAS ROEL, A.: «Concentración y condicionantes geográficos de la ganadería gallega en el s.XVIII», Estudios Geográficos, n° 170-173, 1983, pp.448-449. ^"^ Esta humanización do monte está desenvolvida en SAAVEDRA, R: «Los montes abiertos y los concejos rurales en Galicia en los siglos XVI-XVIIL aproximación a un problema». Cuadernos de Estudios Gallegos, t.XXXIII, 1982, p.l79 e ss.; BALBOA, X.: O monte en Galicia, Vigo, 1990; e REY CASTELAO, O.: Montes y política forestal en la Galicia del Antiguo Régimen, Santiago, 1995. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 66 A PEQUEÑA noALGUIA RURAL... Lucas Labrada magnifica así ó sector pecuario en 1804, situándoo incluso por riba da privatización dos comunais: El ganado en los países interiores^ a donde por falta de consumos y de comunicaciones sólo se cultivan los granos precisos para la subsistencia de sus habitantes, es una industria, la más útil y la que les proporciona mayores ventajas de las que sacarían en el rompimiento de las tierras incultas ..?^ E o coengo compostelán Pedro Antonio Sánchez deixa ben claro a fins do XVIII como en momentos decisivos á hora de tomar decisions no grupo doméstico, a ganancia obtida do gando pode relaxa-la trascendencia do decidido: Pero el ganado en todas partes se cría para extraer, y si no se extrae más, es porque halla obstáculos. Así es que se puede decir que en lugar de fábricas, este reino tiene por equivalente la industria del ganado. Este es el gran tesoro del labrador y el principal recurso en sus necesidades. De su venta saca casi cuanto necesita para su subsistencia. No tiene granos suficientes para pagar las pensiones cargadas sobre las tierras; su auxilio es el ganado. Tiene que satisfacer los diversos impuestos: acude al ganado. Ha de comprar el carro, la sal, el jubón, la ropa, el tabaco; ha de casar al hijo o la hija, ha de satisfacer los derechos parroquiales, etc.; su gran bolsillo, y casi único, consiste en las utilidades del ganado^^ Este gando non estaba distribuido entre a poboación dun xeito igualitario, tal como demostrou concienzudamente J.M Pérez García^^. Do conxunto de inventarios post-mortem por el revisados, se destaca un 10% de individuos que denomina campesinos ricos, adxectivizándoos de «importante burguesía ruraPS>. Participan en actividades especuladoras ñas ^^ LABRADA, L.: Descripción económica del Reino de Galicia, Vigo, 1971, p.206. ^^ SANCHEZ, RA.: Op. cit., p. 119. '' PEREZ GARCIA, J.M.: art. cit. ^^ As características necesarias para estar neste grupo é ter polo menos unha ducia de pezas de gando vacún, acadando a cifra de 20 con facilidade, tres ou catro ducias de ovellas, posuir équidos, e sacrificar unha media de seis cabezas de porcino ó ano. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 67 que invisten fortes sumas de cartos, e son o punto de partida para familias que acaban por se integrar na fidalguía. Seguindo estes parámetros, os resultados obtidos en S. Xiao de Carballo, sitúan os Fraga dentro deste grupo triplicando os valores medios da sondaxe de J.M. Pérez García, tal como se evidencia se reparamos no gando citado na partilla de 1726: - vacún: 31 vacas + 19 bois + 24 crías = 74 - equino: 1 égoa + 1 poldro = 2 - ovicápridos: 90 - colmeas de abellas: 27 A fin dalgunha destas reses era alimentar os habitantes da casa patrucial. Unha dieta mais rica en proteínas e grasas nun contexto de monotonía alimenticia^^ serve como elemento diferenciador entre os privilexiados que residen no caserón afidalgado e o resto da comunidade campesina. A mediados do XVIII os Fraga seguen a manter un posto de primacía na parroquia friolense no que toca ó sector pecuario. Da partilla feita en 1726 corresóndenlle a Gerónimo da Fraga 3 bois, 8 vacas e dúas crías de vacún, 18 ovicápridos e 4 colmeas de abellas. En 1752 declara 4 bois, 8 vacas e 16 crías de vacún, 69 ovicápridos, 36 porcos, 1 cabalo e 10 colmeas. Un incremento considerable nuns 25 anos que fai que o filio mellorado supere a media de gando por vecino en calqueira das distintas calidades en S. Xiao de Carballo, tal como se pode apreciar na táboa 3 e nos índices elaborados posteriormente. Se nos fixamos ñas parellas de bois posuídas por cada vecino, tan só seis deles —o 13,95%— posúen dúas, mentres que a maioría tan só declaran unha coa que de seguro fan as tarefas agrarias —41,86%—, cando non un só boi —2,32%—, ou nigún —41,86%— . Gerónimo da Fraga forma parte desta media ducia de labregos acomodados que posúen dúas parellas de bois, destacándose aínda mais de repararmos no número total de cabezas de gando vacún por cada vecino, onde xunto con Santiago do Espino, son os únicos que acadan o estadio de 26-30 reses de vacún/ vecino representando o 4,65% do conxunto vecinal, nun contexto no que sete vecinos da freguesía non teñen ningunha ^^ SAAVEDRA, E: La vida cotidiana en la Galicia del Antiguo Régimen, Barcelona, 1994, p.l29 e ss.; SOBRADO CORREA, H.: «Aproximación al consumo alimentario en el área rural gallega: El interior lucense (ss.XVII-XIX)», Obradoiro de Historia Moderna, n''^, 1994, pp.87-110. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 68 A PEQUEÑA HDALGUIA RURAL... res vacua —16,27%—, e o 27,90% —é dicir, 12 vecinos— non teñen mais de cinco pezas. Non é de extrañar pois que ó longo do s.XVIII e principios do XIX, os patrucios da Casa da Fraga se integren no comercio de gando, xa que a meirande parte destas reses están destinadas a lie reinvirtir interesantes ganancias en metálico. Aprovéitanse de que un irmán do patrucio recibe unha cantidade fixa anual de ingresos en cabezas de gando ovicáprido procedentes do cobro do décimo das dúas freguesías da comarca ñas que exerce a cura de ánimas: Sta. María de Prado e Sta. María de Ramelle"^^. Os anos de mais ingresos neste apartado son 1791 —24 cabezas—, 1796 —21—, 1799 —24— e 1803 —36—, e nunca os de rexistros menores chegan a suponer un trastorno trascendental na economía dos Fraga, pouco dados ó gasto desenfrenado no apartado suntuario. 40 j Décimos cobrados en ovicápridos por D. Francisco de Fraga, 1785-1804 cabritos cordeiros ano mort. sacrif. mort. n.vend. vend. sacrif. n.vend. vend. 5 1 1 2 1785 4 7 2 1 3 1786 3 2 1787 2 2 1 6 1 1 6 8 1 3 1788 3 2 4 2 2 1 1789 1790 6 2 3 1791 11 5 6 1792 2 1 3 1 4 1 2 3 1793 1 3 1794 2 2 6 3 3 1 1 3 1795 1 18 3 1796 3 4 1 1797 7 3 2 1798 5 3 3 1 14 1 4 3 5 2 1799 2 10 2 1800 6 4 1801 8 2 2 1802 1 18 1803 18 1804 4 2 74 32 40 Total 53 23 6 19 30 Fonte: Elaboración propia a partir do A.CF., Libro de Caja de D. Francisco de Fraga. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO 69 Mais arriscado é o negocio da parcería'^^ O rexistro das operacións económicas referidas ó gando vacún ào Libro de Caja de D. Francisco de Fraga, permítenos facer un seguimento exhaustivo da rendabilidade desta «industria»"^^. O patrucio da Casa da Fraga desembolsa anualmente unha cantidade variable de cartos na feira de Guimarei —sita no concello de '^' Este tipo de cesión do gando convirte os individuos que acceden á especulación gandeira en maioristas de reses, comprándoas, cedéndoas e vendéndoas en lotes significativos, coa intención de obter un rendemento en metálico. A extensión desta modalidade de cesión do gando está testimoniada polo Frei Martín Sarmiento: Es mui usado en Galicia, dar ganado a medias, o a la mitad de las ganancias. Pedro v.g. compra dos Bacas Jovencas, y se las da a Juan Labrador para que las críe, y alimente con sola la pensión de que el valor de las crías, dividido en dos partes, una sea para Pedro, u otra para el Labrador. Esto se llaman Bacas de a medias. El Labrador sin poner un ochavo, con solo su trabajo se interesa en el travajo de las Bacas. En la Leche, manteca y quesos. En el avono, y en la de crianza. Vid. DOPICO, R: A Ilustración e a sociedade galega, Vigo, 1978, p.l27. Comarcalmente, tal como adiantou P. Saavedra para a área da Fonsagrada, os resultados permiten falar de negocio á hora de nos referir á parcería de gando vacún, que se convirte na modalidade imperante en Galicia. Nesta área da Galicia oriental, as 1.001 cabezas de vacún postas en parcería suponen o 35,1% do total de gando rexistrado neste rexime contractual. Na Ulla, as 1.364 reses vacuas catalogadas nesta modalidade, permítenlle falar a O. Rey Castelao de exclusividade. En todo caso, a concentración destas cesións gandeiras en espacios controlados polos posuidores das reses cedidas, evidencian unha unidade parroquial, tal como anunciou J. M. Pérez García para O Saines. SAAVEDRA, P.: Economía rural antigua ..., Op. cit., p.53; REY CASTELAO, O.: Op. cit., p. 120; e PEREZ GARCIA, J.M.: Op. cit., p.220. Non obstante, non hai que esquecer que estructuralmente estáse asistindo en toda a franxa cantábrica da Peninsula Ibérica á consolidación dun destacado proceso de especialización pecuaria a partir da segunda metade do s.XVIII e ó longo do XIX, na que se evidencia unha crecente orientación que tende a prima-lo interese pola cabana gandeira vacua, a teor das recentes aportacións de R. Domínguez Martín, circunscribindo e ratificando a testemuña de Lucas Labrada cando dixo en 1804 que el ganado, lejos de haberse disminuido en Galicia, se aumentó en tanto cuanto ha crecido su consumo en el mismo Reino y su extracción para Portugal y Castilla. DOMÍNGUEZ MARTIN, R.: El campesinado adaptativo. Campesinos y mercado en el norte de España, 1750-1880, Santander, 1996, p.43 e ss.; e LABRADA, L.: Op. cit., pp.200-201. Vid. tamén RODRIGUEZ GALDO, M^.J. e CORDERO, X.: «Rentistas urbanos y capital usurario. La aparcería de ganado en Galicia en el siglo XVIII», Revista de Historia Económica, n°3, 1984, pp.287-294. "^^ Este legaxo do A.C.F. é de grande utilidade porque recolle minuciosamente, ás veces a xeito de diario, moitas das operacións económicas ñas que os Fraga invisten os seus recursos en metálico. Ademáis das transaccións e negocios gandeiros, pódese facer un seguimento do gasto no servicio doméstico, a compra de obxectos de luxo, e outros aspectos de relevo para o coñecemento da vida cotia deste grupo social no Antigo Réxime. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 70 A PEQUEÑA FIDALGUIA RURAL... Friol— para facerse cun número determinado de reses cedidas a posteriori en réxime de parcería a membros da comunidade vecinal. O gando se pon a medias, de tal xeito que as ganancias obtidas serán repartidas en partes iguais entre o fidalgo, que fai as veces de empresario, e o labrego parceiro"^^. As femias aportan unha cantidade de crías susceptibles de dous procesos: 1° ou ben acabar no mercado gandeiro, ou ben T" voltar ó fidalgo previo pago da porcentaxe que lie corresponde ó labrego parceiro despois de face-la taxa da peza en cuestión, para poñela logo a medias con outro campesino. Así, rexístrase un aumento no número de reses desde a fase de compra"^"^ á da súa posta en aparcería'^^ pois hai que ter en conta os animais nacidos non incorporados ó mercado gandeiro, que son postos de novo a medias. O contrato liquídase traía venda das reses agora incrementadas numéricamente debido os nacementos habidos das femias cedidas ñas parcerías, orixinando uns beneficios nos que participa o patrucio da Casa da Fraga, a quen vai corresponde-la metade do ingresado ñas vendas"^^. A maior número de parceiros, aumentan as posibilidades de acumular mais dividendos. A clave do entramado son as crías de gando vacún. É certo que os équidos son xunto eos bois os animais mais cotizados. Dos 10 équidos comprados entre 1779 e 1800, non se cede en parcería ningún, vendéndose 3 —a terceira parte do total comprado—. Trátase dun gando moi particular que representa prestixio para quen o posúe, e incluso a posibilidade de ^^ Véxase como exemplo o seguinte caso: « En 20 de Agosto de 1780 puse una baca marela buena de 7 años con un jato blanco de 5 meses a Domingo de Prado, carzelero vzo. de Friol, el jato a las ganancias, hes mansa la baca. Se vendió el jato en 100 rs., llebó su media ganancia. Parió esta baca una jata, se vendió en 9 Ds. Enfin de maio de 83 parió un jato más. Parió unajata, año 84. Se vendió la baca viexa en Guimarei, año de 84, en 140 rs. Se vendió el jato en 26 de agosto de 85 en 100 rs., llebó su media ganada. Lajata la comió e llebó el medio año, y así no quedó más de la baca», A.C.F., Libro de Caja de D. Francisco de Fraga, f.36. Estamos asistindo á aprición de transaccións entrelazadas supostamente disfrazadas dun certo cariz contractual, ñas que o maiorista é visto como un «amo» ou «señor», tal como fundamenta R. Domínguez Martín, Op. cit., p.202. '^ Vid. Táboa 4. '' Vid. Táboa 5. '"^ Vid. Táboa 6. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO G ARAZO 71 xerar riqueza a través de formas contractuais mais complexas'^. Cos bois sucede algo similar, pois aforran traballo humano ñas tarefas agrarias, especialmente á hora de prepara-las terras para bota-la sementé. Os 4 comprados péñense en parcería para logo venderse. De descontármo-lo gasto que supon a compra dos équidos -3.086 Rs. dos 6.549 totals- o resultado obtido do contraste entre os reas gastados ñas compras e os ganados traías vendas de gando -táboas 4 e 6-, verifícase unha ganancia de 1,45 Rs. por cada real investido na feira de Guimarei. Dato que cobra mais relevencia ó verificar que os Fraga obteñen con este negocio un 45,22% de beneficio ñas vendas realizadas, confirmando así a rendabilidade da parcería gandeira no referente ó sector vacún'^l Este sistema vólvese así mais dinámico e racional. Despois da posta en parcería das reses vacuas adquiridas, o número de vacas e o de crías femias aumenta de 22 pezas compradas a 51 rexistradas ñas vendas. A clave deste crecemento progresivo está na integración das pezas xoves no proceso reproductor —as xovencas—. Deste xeito, de cada camada de crías nacidas, se multiplican polo número de nacementos de femias habidas, as posibilidades de aumenta-lo monto de reses postas en parcería. De seren demasiadas, se venden; si son machos —xatos— pasan ó mercado, CO cal se obteñen uns beneficios repartibles equitativamente entre o parceiro e o señor. Cantas mais reproduccións, mais posibilidades de incrementa-la riqueza derivada das vendas por ambas partes"^^. Entre 1779 e 1800 cómpranse 8 vacas, mentres que son postas en aparcería 14, feito "^^ Sirvan de exemplo os negocios do escriban de Mondoñedo D. Blas de Rubiños testimoniados no seu Libro de Caixa. En 22-III-1775, a permuta dunha egua por un cabalo prodúcelle 400 Rs. En 8-1-1784, o crego de Recesende cómpralle una caballería, su color negro, esttrellada y calzada de los pies por 88 Rs. Arquivo da Casa de Terrafeita, Leg. 15, Libro de caixa de D. Blas de Rubiños comenzado en 1773, íf.2 e 29. "^^ Esta ganancia obténse de sustrae-lo importe do precio a que se compraron os 10 équidos en Guimarei -x- do total de reas -T- investidos na merca de gando polos Fraga ó longo do período estudiado. O resultado desta operación danos a cantidade de investimentos totals en ovicápridos e vacún -y-. Logo y = T-x = 6.549-3.086 = 3.463 Rs. O cociente entre os reas obtidos da venda do gando cedido en aparcería -Z- polo gasto y, darános o resultado f, que representa o valor en reas en que se recupera cada real investido anteriormente. Logo f = Z / y = 5.029 / 3.463 = 1,45 Rs. gañados/Rl. investido. "^^ Esta producción de reses vacuas xoves para o mercado é destacado por A. Eiras Roel cun índice de 1,2 para as áreas altas de Galicia, fronte ás terras baixas -0,6-. EIRAS ROEL, A.: art. cit., pp.161-162. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 72 A PEQUEÑA FIDALGUIA RURAL... so explicable a partir do aporte derivado da incorporación das xovencas ó mentado proceso reproductor^^. Un negocio racionalmente deseñado que depende da fertilidade dos animais comprados na feira de Guimarei e os nacementos das reses postas a medias, que so se desaxusta -e non moi chamativamente- polo gasto da compra dos 10 équidos da serie, compensable por outras vías contractuais que de inmediato se nos escapan, pero que non anula -logo do exposto- a rendabilidade positiva para as economías fidalgas da parcería de gando vacún. TABOA 3 Gando total existente en S. Xiao de Carballo en 1752 VACUN bois vacas crías 2 Antonio Bermúdez Antonio Sánchez Anxel López Agustín Fernández Andrés Gil Alonso da Fraga Antonia das Pardiñas 2 Antonia de Buxán 4 Bentura Ares 2 Bernardo do Espino Bartolomé de Espiñeira 2 Bentado Cavo OVICAPRIDOS ovellas cabras PORCINO ÉQUIDOS COLMEAS 2 10 3 2 3 5 4 8 3 4 12 15 19 5 2 3 2 4 3 5 7 5 8 8 7 24 27 22 14 5 6 9 10 31 4 14 1 5 1 11 1 2 3 4 4 2 ^° Os datos subminstrados polos Fraga na serie elaborada coinciden coa visión xeral definida para o mercado do gando vacún na Galicia do s.XVIII por A. Eiras Roel. No seu estudio citado ó longo do presente traballo, chega á conclusión de que das 340.000 reses vendidas anualmente en Galicia, o 82,35% das mesmas eran crías, mentres que as adultas acadaban a porcentaxe de 17,64%, é dicir, que por cada adulto vendido, eran postas no mercado 4,66 crías. En Friol, entre 1780 e 1796 os Fraga venden 55 reses procedentes de parcerías. O 69,09% das mesmas son crías, o 23,63% vacas que pariron polo menos dúas veces, e o 7,27% restante bois. Isto é, por cada vaca vendida, venden en Guimarei 2,92 crías. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDO GARAZO VACUN bois vacas OVICAPRIDOS crías ovellas cabras 73 PORCINO ÉQUIDOS COLMEAS Domingo de Bernai 2 4 2 5 7 4 3 8 6 8 51 44 7 3 16 3 4 8 2 2 2 3 11 3 1 12 1 11 48 18 11 12 8 10 6 Domingo da Iglesia filio' 2 8 6 8 6 7 9 16 3 25 32 51 8 5 9 18 2 5 4 7 3 3 6 6 2 1 1 1 5 1 2 35 31 16 32 13 8 30 17 8 14 36 5 1 8 13 4 8 Domingo Galbán Domingo do Cabo Domingo Ares 3 Domingo da Iglesia Domingo Várela Domingo do Carregal Domingo Mourente 1 1 17 Domingo da Riba Francisco do Cavo Gerónimo da Fraga Gregorio Vilela 2 4 2 Joseph Ferro Jacinto do Ferro Jacinto Quiroga Juan Cruzado 2 2 2 Juan de Pénelas Jacinto de Castro 1 Juan do Bernadal Juan de Castiñeiras 2 Joseph David 4 2 10 3 16 3 5 16 68 32 31 12 22 18 7 61 166 185 819 Lucas Ferro Manuel García Melchor Cavado María das Seixas Pedro Espino Pedro de Serén 4 2 4 2 Roque Sánchez Santiago do Espino 5 2 3 3 7 5 8 12 6 4 Silvestre Montecelo TOTAL 43 3 5 3 10 5 3 13 16 10 4 3 8 12 247 4 11 5 5 1 1 4 10 2 1 6 1 9 6 6 1 1 10 6 3 1 23 258 17 70 8 5 11 1 Fonte: Elaboración propia a partir do A.H.P.L., Catastro de Ensenada, Libro de Reales de Legos, leg.3.344, ff.91-94. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 74 A PEQUEÑA FIDALGUIA RURAL.. iices referidos á TABOA 3: - Media de animais/vecifio: - 1,41 bois, 3,80 vacas e 4,30 crías de vacún - 19,04 ovellas - 5,74 cabras - 6,00 porcos - 0,39 équidos - l,62colmeas - Parelias de bois/veciños: - 0 parellas: 18 vecinos - 0,5 parellas: 1 vecinos - 1 parellas: 18 vecinos - 2 parellas: 6 vecinos - Pezas de gando vacún/veciño: - 0 pezas: 7 vecs. - 1-5 pezas: 12 vecs. - 6-10 pezas: 4 vecs. -11-15 pezas: 9 vecs. - 16-20 pezas: 6 vecs. - 21-25 pezas: 3 vecs. - 26-30 pezas: 2 vecs. (41,86%) ( 2,32%) (41,86%) (13,95%) (16,27%) (27,90%) ( 9,30%) (20,93%) (13,95%) ( 6,97%) ( 4,65%) "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es ANTONIO PRESEDOGARAZO 75 TABOA 4 Gando comprado por D. Francisco de Fraga, 1779-1800 ÉQUIDOS 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 TOTAL VACUN bois vacas crías 2 2 1 2 7 1 2 2 4 1 1 2 OVICAPRIDOS 35 3 1 2 1 201 1.424 517 619 1.326 729 66 1 9 Rs. 1.425 4 8 14 35 6.307 Fonte: Elaboración propia a partir do A.C.F., Libro de Caja de Don Francisco de Fraga. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 76 A PEQUEÑA FIDALGUIA RURAL... TABOA 5 Gando cedido en parcería por D. Francisco de Fraga, 1779-1800 bois 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 TOTAL VACUN vacas crías 3 1 5 1 1 1 2 5 1 2 3 1 1 1 1 1 1 14 17 2 2 4 OVICAPRIDOS 63 63 Fonte: Elaboración propia a partir do A.CF., Libro de Caja de Don Francisco de Fraga. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 77 ANTONIO PRESEDO GARAZO TABOA 6 Gando vendido por D. Francisco de Fraga, 1779-1800 ÉQUIDOS bois 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 TOTAL VACUN vacas 1 1 2 2 2 OVICAPRIDOS Rs. 4 3 651 205 142 645 crías 3 1 1 5 8 1 1.215 50 1 1 3 1 143 3 5 1 3 1 2 241 197 428 411 88 187 66 360 13 38 5.029 Fonte: Elaboración propia a partir do A.CF., Libro de caja de Don Francisco de Fraga. "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es 78 A PEQUEÑA FIDALGUIA RURAL.. APÉNDICE I XENEALOXÍA DA FAMILIA FRAGA DE S. XIAO DE CARBALLO Gregorio C. da Fraga Dominga Fernández (+ca. 1717) Gregorio María Isabel Gerónimo da Fraga Fernández Fernández da Fraga C. C. Andrés García Pedro López C. María López Buxán das Seixas D. Francisco da Fraga 1784 D^ M^ Jpha. Francisca C. D. Francisco Várela y Aguiar de Fraga D^ Vicenta Fondevilla C. D. Gregorio C. D^ Dominga da Fraga Várela y Peña I D. Manuel Ventura D' Isabel de Fraga y Várela da Fraga D. Ángel José D^ Francisca de Fraga de Fraga y Várela C. 1808 y Seixas D.Juan López "CUADERNOS DE ESTUDIOS GALLEGOS", Tomo XLIV, Fascículo 109, Santiago 1997. (c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://estudiosgallegos.revistas.csic.es
https://openalex.org/W2144545083	https://europepmc.org/articles/pmc3989585?pdf=render	English	null	The myogenic and neurogenic components of the rhythmic segmentation motor patterns of the intestine	Frontiers in neuroscience	2,014	cc-by	3,532	Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacemaking et al., 1987). In 2006, two major reviews on control of motility only very brieﬂy mentioned segmentation: “The most basic small intestinal contractile pattern, seg- mentation, results from reciprocal inhibi- tion and dis-inhibition of adjacent circular muscle” (Hasler, 2006) and “Oscillation of membrane potential through the slow wave cycle results in periods of high and low open probability for Ca2+ chan- nels, and this naturally organizes the contractile pattern into a series of pha- sic contractions contributing to motility patterns such as peristalsis and segmen- tation” (Sanders et al., 2006). Recently, it was shown that spontaneous segmentation or decanoic acid-induced segmentation is associated with a waxing and waning of the slow wave activity that can occur promi- nently in the presence of nerve conduc- tion block (Huizinga et al., 2014). Evidence was provided that waxing and waning developed when low frequency rhythmic transient depolarizations originating from interstitial cells of Cajal (ICC) associ- ated with the deep muscular plexus (ICC- DMP) interacted with the omnipresent slow wave activity originating from ICC associated with the myenteric plexus (ICC- MP) through phase–amplitude coupling. That is, the phase of the low frequency activity modulated the amplitude of the higher frequency slow wave activity. Hence interacting myogenic electrical activities were seen to underlie the segmentation motor pattern. The segmentation motor pattern appeared to be associated with the induction of a low frequency compo- nent, causing minute rhythm clusters of contractions and a waxing and waning of the amplitudes of the individual con- tractions within a cluster (Huizinga et al., 2014). These clusters occurred promi- nently after decanoic acid in rats (Huizinga et al., 2014) or oleic acid in dogs (Ehrlein et al., 1987). In human small intestine studies, the post-prandial motility pattern can be quite variable but regular “cluster contractions” are often reported and seen to be mainly stationary (Hellstrom, 1995). In a study by Husebye (1999), the clusters were shown to have a minute rhythm, the contraction amplitudes within the clus- ters had a waxing and waning appearance with a frequency of ∼10/min, hence occur- ring at the slow wave frequency (Husebye, 1999; Gallego et al., 2014). OPINION ARTICLE published: 10 April 2014 doi: 10.3389/fnins.2014.00078 Jan D. Huizinga 1* and Ji-Hong Chen 2 1 Department of Medicine, Farncombe Family Digestive Health Research Institute, McMaster University, Hamilton, ON, Canada 2 Key Laboratory of Hubei Province for Digestive System Diseases, Department of Gastroenterology and Hepatology, Renmin Hospital of Wuhan University, Wuhan University Institute of Digestive and Liver Diseases, Wuhan, China Correspondence: huizinga@mcmaster ca Medicine, Farncombe Family Digestive Health Research Institute, McMaster University, Hamilton, ON, Canada 2 Key Laboratory of Hubei Province for Digestive System Diseases, Department of Gastroenterology and Hepatology, Renmin Hospital of Wuhan University, Wuhan University Institute of Digestive and Liver Diseases, Wuhan, China C d h i i @ OPINION ARTICLE published: 10 April 2014 doi: 10.3389/fnins.2014.00078 OPINION ARTICLE published: 10 April 2014 doi: 10.3389/fnins.2014.00078 Reviewed by: Miyako Takaki, Nara Medical University, Japan Miyako Takaki, Nara Medical University, Japan Marcel Jimenez, Universidad Autonoma de Barcelona, Spain www.frontiersin.org Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem The ICC-MP activity takes part in the orchestration of colonic low-frequency propulsive contractions (Huizinga et al., 2011; Chen et al., 2013b). In the human esophagus, rhythmic con- tractile activity occurs when nitrergic inhibition is removed (Chen et al., 2013a). Vagal stimulation can initiate slow wave activity in the intramuscular ICC (ICC- IM) of the stomach (i.e., provide the pacemaker component) which together with activity from ICC-MP generate the full slow wave activity (Hirst et al., 2002). The critical importance of the segmen- tation motor pattern makes it unlikely that only one mechanism exists to evoke it. Intraluminal nutrients in particular, in contrast to a non-caloric viscous meal, evoke segmental contractile activity in the dog intestine (Schemann and Ehrlein, 1986). Decanoic acid administered in this way in the guinea pig evokes a mix- ture of propulsion and segmentation (Ellis et al., 2013). Decanoic acid communicates with the epithelial layer and intralumi- nal decanoic acid-induced segmentation is inhibited by 5-HT3 and 5-HT4 antag- onists and CCK antagonists. Most likely, 5-HT and CCK, released from enterochro- mafﬁn cells, stimulate enteric sensory neu- rons which then triggers excitatory activity in the ENS resulting in the motor pat- terns observed. The stimulatory action of motor neurons on the smooth muscle— ICC networks will then result in neu- rally driven activity that in part is orches- trated by ICC and smooth muscle intrinsic activities. Can the myogenic and neurogenic hypotheses be reconciled? It is possible that control systems in the guinea pig are unique. However, there are arguments to be brought forward to suggest that slow waves may play a role in some if not all segmentation motor patterns of the guinea pig: The segmentation motor pat- tern is extremely rhythmic, 15–18 cycles per min according to ﬁgures shown in Gwynne and Bornstein (2007) which is within the range of the slow wave fre- quency of the guinea pig small intestine which is 11–19 cycles per min (Donnelly et al., 2001). The guinea pig small intes- tine is unique in that slow wave activity is not omnipresent but is induced by stim- uli such as distention. Trendelenburg, in 1917, already recognized that the rhythmic peristaltic activity induced by distention in the guinea pig small intestine can pro- ceed without a nervous conduction system (Trendelenburg, 2006). Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem When a nutri- ent solution was given to healthy volun- teers with or without 40 g/l ethanol, it was ethanol in particular that induced a low frequency component, the clustered con- tractions; the clusters occurred at 1 per 2–3 min and the individual contractions within the clusters at ∼12/min with wax- ing and waning amplitudes (Schmidt et al., 1997). Hence also in the human small intestine, the slow wave frequency, as well as an additional lower frequency compo- nent, are reported in the post-prandial intestinal motor activity. Almost all motor patterns in gut organs have as primary function the mixing of content. Although classical peristalsis is equated with propulsion, and this is cer- tainly the case in the esophagus, the predominant effect in all other organs is mixing and exposing the content opti- mally to the mucosal surface, because the propulsion ends somewhere and the content is moving back; only very rarely does propulsion end with evacuation of content from the body. The segmenta- tion motor pattern is different from peri- stalsis in that it contains only stationary or very short distance propagating con- tractions and hence is considered a spe- cialized motor pattern for mixing and absorption. The segmentation motor pat- tern was described and illustrated by Cannon in 1902 based on X-ray observa- tions and shown to be extremely rhyth- mic (Cannon, 1902). Alvarez was the ﬁrst to ﬁnd that the frequency of rhyth- mic segmenting contractions occurred at the frequency of a myogenic pacemaker and that in various regions of the intes- tine the frequency decreased in the same way as the pacemaker frequency decreased (Alvarez, 1914), suggesting a ﬁrm relation- ship between slow waves and segmenta- tion. In 1968, Code and co-workers also recognized the role of slow waves in seg- mentation (Code et al., 1968); the slow waves were thought to go in and out of excitable regions of smooth muscle ﬁbers. Ehrlein noted in 1987 that there were no electrical or mechanical features known that could distinguish peristaltic and segmental motor patterns (Ehrlein Although the segmentation rhythmic- ity has been associated with slow wave activity in many studies, other reports, primarily conducted using the guinea pig small intestine, explain segmenta- tion solely on the basis of enteric neu- ronal activity. Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem Important studies from Bornstein’s laboratory provided evidence April 2014 \| Volume 8 \| Article 78 \| 1 www.frontiersin.org www.frontiersin.org Intestinal segmentation motor patterns Huizinga and Chen for the hypothesis that cholinergic motor neurons acting on muscarinic receptors periodically activate the musculature and that inhibitory neurons surround this contraction to ﬁnalize the motor pat- tern (Gwynne et al., 2004; Gwynne and Bornstein, 2007). Another study states that CCK and 5-HT are critical medi- ators in the regulation of segmentation (Ellis et al., 2013). Segmentation has also been suggested to result from a reduced degree of synchrony of AH neuron activ- ity and sustained inhibition of the after- hyperpolarization (Ferens et al., 2007). musculature above threshold for L-type calcium channel activation. Hence, neu- ral activity is usually an active component of any segmentation motor activity in any organ of any species and therefore blockers of neural activity may inhibit segmenta- tion motor patterns. This does not negate an essential role for slow wave activity. may be associated with segmentation, the activity will be hyoscine sensitive. In many organs in most species, omnipresent slow wave activity exists that will take part in the orchestration of motor patterns once the smooth muscle is stimulated above thresh- old for contractile activity. However, it has become clear in recent years that in addi- tion to the omnipresent slow waves or, as is the case in the guinea pig intestine instead of omnipresent slow waves, slow wave activity occurs that depends on a stimu- lus. In the guinea pig, distention induces strong slow wave activity (Donnelly et al., 2001). In some guinea pig small intes- tine preparations, the distention-induced slow waves were abolished by atropine or TTX, in other preparations slow wave activity was not affected by these drugs, indicating that several distinct mecha- nisms can evoke the slow wave activity (Donnelly et al., 2001). In the rat and mouse colon, stimulus-dependent rhyth- mic transient depolarizations (similar to slow wave activity but called differently to avoid confusion and much lower in fre- quency) occur dependent on activation of L-type calcium channels and generated by a network of interstitial cells, the ICC-MP (Pluja et al., 2001). The rhythmic tran- sient depolarizations occur in addition to the omnipresent slow wave activity, which is generated by the primary pacemaker cells of the colon, the interstitial cells of Cajal associated with the submuscular plexus (ICC-SMP). Frontiers in Neuroscience \| Autonomic Neuroscience Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem Slow waves in the guinea pig intestine are likely originating in networks of ICC, which are prominent and appear similar to other animal mod- els (Zhou and Komuro, 1992; Burns et al., 1997; Lavin et al., 1998; Seki et al., 1998). How activation of mucosal 5-HT recep- tors on enteric sensory neurons can lead to different motor patterns is not clear. It is possible that certain motor patterns depend on stimulation of serotonergic and purinergic receptors (Galligan, 2004). It is also possible that in conjunction with mucosal activation, actions occur through the blood stream on motor neurons, ICC and/or smooth muscle cells. In the study of Gwynne and Bornstein (2007), segmental contractions were blocked by hyoscine (Gwynne and Bornstein, 2007). Does this prove that the segmentation is entirely neurogenic and that slow waves are not involved? No, because the expression of slow waves in tis- sues where slow waves have to be induced, the inductor often is the enteric nervous system. Hence, neural activity including muscarinic stimulation may induce ICC slow wave activity and provide excitation of the musculature at the same time. When muscarinic stimulation is involved in induction of slow wave activity, which then In summary, a number of studies have been published on the mechanisms under- lying the segmentation motor pattern. Segmentation in all cases is observed as short-lived contractions that are propagat- ing over a few mm, are often rhythmic in nature involving the whole intestine for a relatively long period or form clus- ters of segmental contractions for short periods. There is overwhelming evidence that the rhythmicity of segmentation is associated with ICC activity and recently, the classical segmentation motor pattern has been shown to occur in the presence Any hypothesis on the role of slow waves in contraction patterns has to rec- ognize that slow waves themselves do not provide forceful contractions, force- ful contractile activity usually depends on excitatory neural stimulation to generate general depolarization of the musculature to bring the slow waves that have propagated from ICC into the April 2014 \| Volume 8 \| Article 78 \| 2 Frontiers in Neuroscience \| Autonomic Neuroscience Intestinal segmentation motor patterns Huizinga and Chen of nerve conduction blockade. Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem However, under most experimental conditions, and in vivo, the enteric nervous system pro- vides an essential stimulus for the motor activity to develop and hence a variety of nerve conduction blockers or neural receptor blockers will inhibit segmenta- tion activity. Whether or not a motor pat- tern occurs in response to nutrients or luminal distention is often determined by the response of the ENS to the stimu- lus, including sensory and motor neurons. This is also the case for segmentation, and several components of the ENS have been shown to be involved. This neural activ- ity then works in concert with the ICC pacemaker activities to generate the motor pattern of segmentation. Motil. 11, 141–161. doi: 10.1046/j.1365-2982.1999. 00147.x Motil. 11, 141–161. doi: 10.1046/j.1365-2982.1999. 00147.x Donnelly, G., Jackson, T. D., Ambrous, K., Ye, J., Safdar, A., Farraway, L., et al. (2001). The myo- genic component in distention-induced peristalsis in the guinea pig small intestine. Am. J. Physiol. Gastrointest. Liver Physiol. 280, G491–G500. Lavin, S. T., Southwell, B. R., Murphy, R., Jenkinson, K. M., and Furness, J. B. (1998). Activation of neurokinin 1 receptors on interstitial cells of Cajal of the guinea-pig small intestine by sub- stance P. Histochem. Cell Biol. 110, 263–271. doi: 10.1007/s004180050288 Ehrlein, H. J., Schemann, M., and Siegle, M. L. (1987). Motor patterns of small intestine determined by closely spaced extraluminal transducers and vide- oﬂuoroscopy. Am. J. Physiol. 253, G259–G267. Pluja, L., Alberti, E., Fernandez, E., Mikkelsen, H. B., Thuneberg, L., and Jimenez, M. (2001). Evidence supporting presence of two pacemakers in rat colon. Am. J. Physiol. Gastrointest. Liver Physiol. 281, G255–G266. Ellis, M., Chambers, J. D., Gwynne, R. M., and Bornstein, J. C. (2013). Serotonin (5-HT) and cholecystokinin (CCK) mediate nutrient induced segmentation in guinea pig small intestine. Am. J. Physiol. Gastrointest. Liver Physiol. 304, G749–G761. doi: 10.1152/ajpgi.00358.2012 Sanders, K. M., Koh, S. D., and Ward, S. M. (2006). Interstitial cells of cajal as pacemakers in the gastrointestinal tract. Annu. Rev. Physiol. 68, 307–343. doi: 10.1146/annurev.physiol.68.040504. 094718 Ferens, D., Baell, J., Lessene, G., Smith, J. E., and Furness, J. B. (2007). Effects of modulators of Ca(2+)-activated, intermediate-conductance potassium channels on motility of the rat small intestine, in vivo. Neurogastroenterol. Motil. 19, 383–389. doi: 10.1111/j.1365-2982.2007.00898.x Schemann, M., and Ehrlein, H. J. (1986). Postprandial patterns of canine jejunal motility and transit of luminal content. Gastroenterology 90, 991–1000. Schmidt, T., Eberle, R., Pfeiffer, A., and Kaess, H. (1997). REFERENCES Alvarez, W. C. (1914). Functional variations in con- tractions of different parts of the small intestine. Am. J. Physiol. 35, 177–193. Hasler, W. L. (2006). “Small intestine motility,” in Physiology of the Gastrointestinal Tract, 4th Edn., ed L. R. Johnson (Amsterdam: Elsevier), 935–964. Burns, A. J., Herbert, T. M., Ward, S. M., and Sanders, K. M. (1997). Interstitial cells of Cajal in the guinea-pig gastrointestinal tract as revealed by c- Kit immunohistochemistry. Cell Tissue Res. 290, 11–20. doi: 10.1007/s004410050902 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Hellstrom, P. M. (1995). Motility of small intes- tine: a case for pattern recognition. J. Intern. Med. 237, 391–394. doi: 10.1111/j.1365-2796.1995. tb01192.x Hirst, G. D., Dickens, E. J., and Edwards, F. R. (2002). Pacemaker shift in the gastric antrum of guinea- pigs produced by excitatory vagal stimulation involves intramuscular interstitial cells. J. Physiol. 541, 917–928. doi: 10.1113/jphysiol.2002.018614 Cannon, W. B. (1902). The movements of the intestines studied by means of the Rontgen Rays. J. Med. Res. 7, 72–75. Received: 24 February 2014; paper pending published: 14 March 2014; accepted: 28 March 2014; published online: 10 April 2014. Received: 24 February 2014; paper pending published: 14 March 2014; accepted: 28 March 2014; published online: 10 April 2014. Citation: Huizinga JD and Chen J-H (2014) The myogenic and neurogenic components of the rhythmic segmentation motor patterns of the intestine. Front. Neurosci. 8:78. doi: 10.3389/fnins.2014.00078 This article was submitted to Autonomic Neuroscience, a section of the journal Frontiers in Neuroscience. Copyright © 2014 Huizinga and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is per- mitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Chen, J. H., Wang, X. Y., Liu, L. W., Yu, W., Yu, Y., Zhao, L., et al. (2013a). On the origin of rhythmic contractile activity of the esophagus in early acha- lasia, a clinical case study. Front. Neurosci. 7:77. ACKNOWLEDGMENTS The authors are supported by grants from the National Natural Science Foundation of China (NSFC) # 81170249 to Ji-Hong Chen and from the Canadian Institutes of Health Research (CIHR) # MOP12874 to Jan D. Huizinga. Parts of this work were presented at the symposium “The Enteric Nervous System, 30 years later” (ENS II 2014) held in Adelaide, Australia on the 1st and 2nd of February 2014, organized by Marcello Costa. Seki, K., Zhou, D. S., and Komuro, T. (1998). Immunohistochemical study of the c-kit express- ing cells and connexin 43 in the guinea-pig diges- tive tract. J. Auton. Nerv. Syst. 68, 182–187. doi: 10.1016/S0165-1838(97)00134-3 Galligan, J. J. (2004). Enteric P2X receptors as poten- tial targets for drug treatment of the irritable bowel syndrome. Br. J. Pharmacol. 141, 1294–1302. doi: 10.1038/sj.bjp.0705761 Trendelenburg, P. (2006). Physiological and pharmacological investigations of small intestinal peristalsis. Translation of the article “Physiologische und pharmakologische Versuche uber die Dunndarmperistaltik,” Arch. Exp. Pathol. Pharmakol. 81, 55–129, 1917. Naunyn Schmiedebergs Arch. Pharmacol. 373, 101–133. doi: 10.1007/s00210-006-0052-7 Gwynne, R. M., and Bornstein, J. C. (2007). Mechanisms underlying nutrient-induced seg- mentation in isolated guinea pig small intestine. Am. J. Physiol. Gastrointest. Liver Physiol. 292, G1162–G1172. doi: 10.1152/ajpgi.00441.2006 Gwynne, R. M., Thomas, E. A., Goh, S. M., Sjovall, H., and Bornstein, J. C. (2004). Segmentation induced by intraluminal fatty acid in isolated guinea-pig duodenum and jejunum. J. Physiol. 556, 557–569. doi: 10.1113/jphysiol.2003.057182 Zhou, D. S., and Komuro, T. (1992). Interstitial cells associated with the deep muscular plexus of the guinea-pig small intestine, with special reference to the interstitial cells of Cajal. Cell Tissue Res. 268, 205–216. doi: 10.1007/BF00318788 Keywords: interstitial cells of Cajal, segmentation, motility disorders, motility, enteric nervous system, pacem Effect of ethanol on postprandial duode- nojejunal motility in humans. Dig. Dis. Sci. 42, 1628–1633. doi: 10.1023/A:1018892827554 Gallego, D., Malagelada, C., Accarino, A., De Giorgio, R., Malagelada, J. R., Azpiroz, F., et al. (2014). Nitrergic and purinergic mechanisms evoke inhibitory neuromuscular transmission in the human small intestine. Neurogastroenterol. Motil. 26, 419–429. doi: 10.1111/nmo.12293 REFERENCES doi: 10.3389/fnins.2013.00077 Citation: Huizinga JD and Chen J-H (2014) The myogenic and neurogenic components of the rhythmic segmentation motor patterns of the intestine. Front. Neurosci. 8:78. doi: 10.3389/fnins.2014.00078 Huizinga, J. D., Chen, J. H., Zhu, Y. F., Pawelka, A., McGinn, R. J., Bardakjian, B. L., et al. (2014). The origin of segmentation motor activ- ity in the intestine. Nat. Commun. 5:3326. doi: 10.1038/ncomms4326 segmentation motor patterns of the intestine. Front. Neurosci. 8:78. doi: 10.3389/fnins.2014.00078 This article was submitted to Autonomic Neuroscience, a section of the journal Frontiers in Neuroscience. Chen, J. H., Zhang, Q., Yu, Y., Li, K., Liao, H., Jiang, L. S., et al. (2013b). Neurogenic and myo- genic properties of pan-colonic motor patterns and their spatiotemporal organization in rats. PLoS ONE 8:e60474. doi: 10.1371/journal.pone. 0060474 Huizinga, J. D., Martz, S., Gill, V., Wang, X.-Y., Jimenez, M., and Parsons, S. (2011). Two inde- pendent networks of interstitial cells of Cajal work cooperatively with the enteric nervous system to create colonic motor patterns. Front. Neurosci. 5:93. doi: 10.3389/fnins.2011.00093 Copyright © 2014 Huizinga and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is per- mitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Code, C. F., Szurszewski, J., Keith, A. K., and Smith, I. B. (1968). “A concept of control of gastrointestinal motility,” in Handbook of Physiology: Alimentary Canal, ed C. F. Code (Washington, DC: American Physiological Society), 2881–2896. Husebye, E. (1999). The patterns of small bowel motil- ity: physiology and implications in organic dis- ease and functional disorders. Neurogastroenterol. April 2014 \| Volume 8 \| Article 78 \| 3 www.frontiersin.org
https://openalex.org/W4384655490	https://zenodo.org/records/8164499/files/D6B_1_2_HyDelta_Tweede_Tranche_Ventilation_in_Gas_Stations_NL.pdf	Dutch	null	D6B.1A & D6B.1B Inventarization, modeling, and experiments related to ventilation in different types of pressure reduction stations in the distribution (low-pressure) network in the Netherlands with natural gas and hydrogen	Zenodo (CERN European Organization for Nuclear Research)	2,023	cc-by	55,623	WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Status: definitief Status: definitief Dit project is medegefinancierd door TKI Nieuw Gas \| Topsector Energie uit de PPS-toeslag onder referentienummer TKI2022-HyDelta. WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Document samenvatting Corresponderende auteur Corresponderende auteur Ir. Sander van Woudenberg Michiel van der Laan, MSc. Verbonden aan Kiwa Technology BV E-mail adres sander.van.woudenberg@kiwa.com michiel.van.der.laan@kiwa.com Document historie Versie Datum Auteur Verbonden aan Samenvatting van de wijzigingen 1 14-Mrt- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Eerste versie; concept EAG 2 24-Apr- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Tweede versie; commentaren EAG verwerkt 3 16-Jun- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Derde versie; commentaren sparringsgroep verwerkt 4 27-Jun- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Vierde versie; concept supervisory group 5 19-Jul-2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Vijfde versie; definitieve versie Verspreidingsniveau WP6B – Veiligheid – Gasstations Document historie Versie Datum Auteur Verbonden aan Samenvatting van de wijzigingen 1 14-Mrt- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Eerste versie; concept EAG 2 24-Apr- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Tweede versie; commentaren EAG verwerkt 3 16-Jun- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Derde versie; commentaren sparringsgroep verwerkt 4 27-Jun- 2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Vierde versie; concept supervisory group 5 19-Jul-2023 Sander van Woudenberg/Michiel van der Laan Néstor González Díez/ Aris Twerda Kiwa Technology BV TNO Vijfde versie; definitieve versie Pagina 2/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Document review Partner Naam Kiwa Technology BV Rob van Aerde Westland Rick den Hartog Alliander Peter Verstegen, Rob Nispeling Stedin Ricardo Verhoeve Rendo Johan Jonkman Coteq Jerry Palmers Enexis John Voogt, Raymond van Hooijdonk NBNL, Gasunie, Kiwa, DNV, TNO, NEC HyDelta Supervisory Group WP6B – Veiligheid – Gasstations 1e step: Inventory common types of gas cabinets p y yp g Some 55,000 gas stations are operated by district system operators (DSOs) in the Netherlands. An inventory of gas cabinets used in the Netherlands was carried out. It focused on cabinets installed by DSOs in the last 10 years, because conversion of installations to hydrogen will initially take place with relatively new cabinets and these installations are designed in accordance with NEN1059. This does not mean they are exactly the same, but they are designed with the same minimum functional requirements. Three types of gas cabinets comprise a substantial part of the total population. These are mini-gas cabinets with a volume < 0.5 m3 (these are mostly used for a high-pressure delivery station), ½ m³ gas cabinets and 4 m³ gas cabinet stations. Document review Partner Naam Kiwa Technology BV Rob van Aerde Westland Rick den Hartog Alliander Peter Verstegen, Rob Nispeling Stedin Ricardo Verhoeve Rendo Johan Jonkman Coteq Jerry Palmers Enexis John Voogt, Raymond van Hooijdonk NBNL, Gasunie, Kiwa, DNV, TNO, NEC HyDelta Supervisory Group Document review Pagina 3/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof What is a realistic leakage size? There has been a lot of focus in this study on which leakage is representative in (normal) operation and for which leakage rate ventilation should be effective. Different sources use different assumptions to determine the expected leakage rate. This is not surprising, as leakage rates can differ due to operational pressures, maintenance or environmental factors. This study tested both with the highest leakage rate from a recent field study (40 l/h) of more than 700 gas stations. Also, leakage flow rates were based on leakage openings in other standards (0.025 mm² and 0.25 mm²). This is still a wide range of leakages where, especially for the larger leaks, it is expected to be noticed by the public coincidently in close proximity of the gas cabinet with a leak. The leaks measured the field study of more than 700 stations are considered realistic, with the largest measured leak being 40 liters per hour. Executive summary Executive summary This research report is a follow-up to the gas stations work package from Hydelta 1.0. That study showed that for various types of gas cabinets, hydrogen more often leads to a combustible mixture at the ventilation openings than natural gas, assuming the leakage flow rates chosen in that study. This led to the recommendation to carry out additional research to identify the effects of smaller, more common leakage flow rates. It was also recommended to investigate which types of different gas cabinets are frequently used in the Netherlands. This report further develops these recommendations. This follow-up research is important because ventilation is an important measure in the event of an unintentional gas leak. Ventilation dilutes the gas and minimizes the risk of ignition or explosion. For the transition to hydrogen, it is important to know whether these gas pressure regulating stations with the same types of gas cabinets carry the same risk with the hydrogen of application. The aim of this study is to gain further insight into how hydrogen behaves in existing gas cabinets compared to natural gas. That insight was obtained by looking at the issue from different angles: experiments as well as simulations using finite element method (CFD). This provided interesting insights that will help policymakers determine whether, and if so what, further measures can be taken. 2e step: Measurements An extensive test program was carried out with a ½ m3 cabinet and a 4 m3 cabinet station. With a mini- cabinet, some indicative measurements were performed to get a first impression. This was done by positioning a reference leak in the center of the cabinet during the experiments. From this, gas (hydrogen or natural gas) flows at a known flow rate controlled by a Mass Flow Controller. The gas concentration was then measured at various points in the gas cabinet, directly at the vent openings outside the gas cabinet and at a distance of about 0.5 meter away from the gas cabinet. The smallest leak selected has a flow rate of 40l/h of natural gas (125 l/h for hydrogen). The largest leak is based on a leak opening of 0.25mm2 at a pressure of 8 bar (that is: 1.8 m3n /h natural gas or 5.6 m3n /h hydrogen). Between these extremes, several other leakage flow rates were chosen. Key data from the experiments are shown in the three graphs below for the ½ m3 cabinet, the 4m3 cabinet station and the mini-cabinet where the leakage rate on the x-axis decreases in size. The first dataset on the x-axis represents both 1.8m3n /h natural gas and 5.6m3n /h hydrogen. This reasoning also applies to the other, smaller leakage flow rates. Pagina 4/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figure 1; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a ½ m3 cabinet Figure 2; Gas concentration (vol%) at different leakage rates of natural gas and hydrogen at the 4 m3 cabinet station In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m3n /h natural gas or 0.56 m3n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). het distributie (lagedruk)net in Nederland met aardgas en waterstof Figure 1; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a ½ m3 cabinet Figure 2; Gas concentration (vol%) at different leakage rates of natural gas and hydrogen at the 4 m3 cabinet station In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m3n /h natural gas or 0.56 m3n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). Figure 3; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a mini-cabinet Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 0,18 m³(n)/h Aardgas 0,04 m³(n)/h Waterstof 0,56 m³(n)/h Waterstof 0,125 m³(n)/h Figure 1; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a ½ m3 cabinet Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Figure 1; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a ½ m3 cabinet Figure 2; Gas concentration (vol%) at different leakage rates of natural gas and hydrogen at the 4 m3 cabinet station In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m3n /h natural gas or 0.56 m3n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Figure 2; Gas concentration (vol%) at different leakage rates of natural gas and hydrogen at the 4 m3 cabinet station In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m3n /h natural gas or 0.56 m3n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). 2e step: Measurements Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figure 1; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a ½ m3 cabinet Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figure 3; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a mini cabinet Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 0,18 m³(n)/h Aardgas 0,04 m³(n)/h Waterstof 0,56 m³(n)/h Waterstof 0,125 m³(n)/h Figure 2; Gas concentration (vol%) at different leakage rates of natural gas and hydrogen at the 4 m3 cabinet station Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m3n /h natural gas or 0.56 m3n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). Figure 3; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a mini-cabinet Aardgas 0,18 m³(n)/h Aardgas 0,04 m³(n)/h Waterstof 0,56 m³(n)/h Waterstof 0,125 m³(n)/h Figure 3; Gas concentration (vol%) at different leakage rates natural gas and hydrogen in a mini-cabinet Pagina 5/123 Measurement results mini-cabinet Some measurements were carried out at the mini-cabinet. These aimed to provide input to any follow-up research and are too limited to draw firm conclusions. - Some measurements were carried out at the mini-cabinet. These aimed to provide input to any follow-up research and are too limited to draw firm conclusions. - An actual observation is that no combustible mixture was measured with natural gas (40l/h), while it was measured with hydrogen (125l/h). - An actual observation is that no combustible mixture was measured with natural gas (40l/h), while it was measured with hydrogen (125l/h). Measurement results 4 m3 cabinet station - For both natural gas (40l/h) and hydrogen (125 l/h), the concentration remains well below the lower flammability limit. The measured concentrations with hydrogen are higher, but in all cases below the limit of a combustible mixture. - In measurements with leak openings of 0.25 mm² and 0.025 mm², no combustible mixture was measured with natural gas. In the case of hydrogen, a combustible mixture was measured at leak openings of 0.25 mm². When the measurements are examined more closely, the following can be concluded: Measurement results ½ m3 cabinet - For a 40l/h natural gas leak, the maximum gas concentration in a ½ m3 cabinet is about 1 vol%, a mixture below the lower flammability limit1 . For a similar leak for hydrogen (125l/h), the concentration rises to 2.7 vol%, which is also below the lower flammability limit. Directly at the vent openings, similar concentrations are measured, with 3.2 vol% as a local, time- dependent outlier. Half a meter away from the gas cabinet, the concentration is well below the lower flammability limit in all cases. - Measurements with leakage openings of 0.25 mm² and 0.025 mm² measured flammab mixtures with both natural gas and hydrogen. Measurements with leakage openings of 0.25 mm² and 0.025 mm² measured flammable mixtures with both natural gas and hydrogen. 3e step: verification with CFD calculations p As verification and to make measurements visual, CFD calculations were carried out. The aim of these modelling activities is to understand the factors affecting the flow phenomena of hydrogen gas in gas stations. Measurement data from Hydelta 1.0 were used as a reference for the mathematical models. The quality of the CFD validation is not (yet) so high that the results can be used independently of the field measurements. This is therefore an additional tool. Several reasons can be given, such as the effect of wind (variable wind) or differences in temperature (in different experiments). The main lesson to be drawn from the CFD calculations is the influence of the roof configuration. The ventilation path is important and should facilitate the upward flow of hydrogen driven by the buoyancy. For the modelled gas cabinet with the overhanging roof lid, it can be seen that mainly hydrogen is limited to escape from the gas cabinet by a "siphon" effect. The buoyancy of hydrogen is insufficient to overcome the hydraulic resistance. An alternative CFD geometry where the overhang of the gas cabinet lid is eliminated confirms a significant improvement in ventilation. 1 With definition as added in the glossary 1 With definition as added in the glossary Pagina 6/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Recommendations: Recommendations: The following recommendations emerge from this study: g g y - Consider incorporating concentration measurements into standard operating procedures by recording them for research and monitoring over longer periods. This can provide interesting insights into how populations of gas pressure regulating stations evolve. - Consider incorporating concentration measurements into standard operating procedures by recording them for research and monitoring over longer periods. This can provide interesting insights into how populations of gas pressure regulating stations evolve. - Consider adapting work instructions so that, when working on gas stations, technicians take a gas concentration measurement in the vent opening before opening the cabinet itself. By recording the measured gas concentration, network operators will gain more insight into actual frequency of occurrence of larger leaks with limited extra effort. In addition, leaks found can then be repaired, reducing the population of stations with leaks. p g p p - The application of hydrogen at gas stations seems to require additional precautions to achieve the same level of safety as for natural gas. For the 4 m3 cabinet station, this difference is more evident than for the ½ m3 cabinet. Based on the precautionary principle, for both the ½ m3 and the 4 m3 cabinet station, additional precautionary measures are sensible. There are several possibilities here, such as further increasing the ventilation area of the gas cabinet, modifying the gas cabinet, placing fencing around at least one meter away from the gas cabinet or more intensive monitoring for leaks than is usual for natural gas stations. g g - Further research on HAS cabinets (mini-cabinets) is recommended before HAS cabinets are converted to hydrogen. Until such research is conducted, it is advisable to apply additional precautions when distributing hydrogen. g g - Further research on HAS cabinets (mini-cabinets) is recommended before HAS cabinets are converted to hydrogen. Until such research is conducted, it is advisable to apply additional precautions when distributing hydrogen. - It is important that the standards committee of NEN 1059 makes a statement for which leakage flow rates ventilation should be an effective measure under which specific circumstances. Leakage openings and practical measurements still seem to be far apart. - It is important that the standards committee of NEN 1059 makes a statement for which leakage flow rates ventilation should be an effective measure under which specific circumstances. Leakage openings and practical measurements still seem to be far apart. Insights gained: Insights gained: g g If hydrogen is transported using gas pressure regulating stations and the existing gas cabinets, it is expected that in case of a leakage, the concentrations in the gas cabinet will be higher than with natural gas. This study also looked at possible modifications and their effect. These are modifications that can, in principle, be carried out in the field with the existing gas cabinets. For instance, under- ventilation can be added by lifting the casing a few centimeters from its base, creating a slot at the bottom. In addition, it is possible to raise the roof edge a few millimeters by unscrewing it and replacing it with longer bolts with thicker spacers (o-rings). The effects of these practical, applicable modifications were examined. This revealed the following: Natural gas Hydrogen Distribution of concentrations in case of leakage The medium mixes throughout the cabinet There is a "blanket" effect. Concentrations are higher at the top than at the bottom. Increase top ventilation Lowering concentrations throughout the housing The high concentration at the top remains high, but the blanket becomes thinner. Effect of under-ventilation Little effect The thickness of the blanket decreases. Increase top ventilation (up to 4%) + bottom ventilation (2%) Best effect Best effect, but less than expected. Flammable mixtures are still possible. Adding additional ventilation effect on LEL/LFL limits Lowering effect on all measurements Lowering effect, but some leakage rates exceed flammability limit. 1e stap: Inventarisatie veelvoorkomende typen behuizingen p yp g In Nederland worden er door de regionale netbeheerders zo’n 55.000 gasstations beheerd. Er is een inventarisatie uitgevoerd van de in Nederland toegepaste behuizingen. Daarbij is gefocust op behuizingen die in de laatste 10 jaar door RNB’s zijn geplaatst, omdat omzetting van installaties naar waterstof in eerste instantie zal plaatsvinden met relatief nieuwe behuizingen en deze installaties conform de NEN1059 zijn ontworpen. Dit betekent niet dat ze exact hetzelfde zijn, maar wel met dezelfde minimale functionele eisen zijn ontworpen. Drie types behuizingen omvatten een substantieel deel van de totale populatie. Dat zijn mini-kasten met een inhoud < 0,5 m3 (deze worden veelal gebruikt voor een hogedrukafleverstation), ½ m³ kasten en 4 m³ kaststations. Recommendations: Pagina 7/123 Wat is een reële lekgrootte? Er is in dit onderzoek veel aandacht geweest welke lekkage representatief is bij (normale) bedrijfsvoering en voor welk lekdebiet de ventilatie effectief moet zijn. Verschillende bronnen gebruiken verschillende uitgangspunten om het te verwachten lekdebiet te bepalen. Dit is ook niet vreemd, omdat de lekdebieten kunnen verschillen door operationele druk, onderhoud of omgevingsfactoren. In dit onderzoek is zowel getoetst met het hoogste lekdebiet uit een recent praktijkonderzoek (40 l/u) van meer dan 700 gasstations. Tevens zijn lekdebieten gebaseerd op lekopeningen in andere normen (0,025 mm² en 0,25 mm²). Dit is nog steeds een breed spectrum aan lekkages waarbij, zeker bij de grotere lekken, de verwachting is dat deze door de omgeving zullen worden opgemerkt. De lekkages die zijn gemeten bij praktijkmetingen bij meer dan 700 stations worden als realistisch beschouwd, het grootste gemeten lek hierbij was 40 l/u. Samenvatting Dit onderzoeksrapport is een vervolg op het werkpakket gasstations uit Hydelta 1.0. Uit dat onderzoek bleek dat bij diverse typen behuizingen waterstof vaker leidt tot een brandbaar mengsel bij de ventilatieopeningen dan aardgas, uitgaande van de in dat onderzoek gekozen lekdebieten. Hieruit volgde de aanbeveling om aanvullend onderzoek uit te voeren om de effecten van kleinere, meer in de praktijk voorkomende lekdebieten in kaart te brengen. Daarnaast werd aanbevolen om na te gaan welke types verschillende behuizingen veelvuldig in Nederland worden toegepast. In dit rapport wordt verder invulling gegeven aan deze aanbevelingen. Dit vervolgonderzoek is belangrijk omdat de ventilatie een belangrijke maatregel is bij een onbedoelde gaslekkage. Ventilatie zorgt ervoor dat het gas wordt verdund en de kans op ontbranding of explosie wordt geminimaliseerd. Voor transitie naar waterstof, is het van belang om te weten of deze gasdrukregelstations met dezelfde typen behuizingen ook met toepassing van waterstof hetzelfde risico draagt. Het doel van dit onderzoek is om verder inzicht te verkrijgen hoe waterstof zich gedraagt in bestaande behuizingen t.o.v. aardgas. Dat inzicht is verkregen door het vraagstuk vanuit verschillende invalshoeken te bekijken: praktijkmetingen en simulaties middels eindige elementen methode (CFD). Dit leverde interessante inzichten op die beleidsmakers helpen vast te stellen of, en welke maatregelen er verder genomen kunnen worden. 1e stap: Inventarisatie veelvoorkomende typen behuizingen betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof grootte. De eerste dataset op de x-as is een weergave van zowel 1,8 m3n/h aardgas en 5,6 m3n/h waterstof. Deze redenatie geldt ook voor de andere, kleinere lekdebieten. Figuur 4; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een ½ m3 kast Figuur 5; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in het 4 m3 kaststation In het geval van de mini-kast is middels enkele indicatieve metingen gekeken naar lekkages bij een lekopening van 0,025 mm² bij 8 bar voordruk (dat is: 0,18 m3n/h aardgas of 0,56 m3n/h waterstof) en 1 bar voordruk (dat is: 40l/u aardgas of 125 l/u waterstof). Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen ( g ) g grootte. De eerste dataset op de x-as is een weergave van zowel 1,8 m3n/h aardgas en 5,6 m3n/h waterstof. Deze redenatie geldt ook voor de andere, kleinere lekdebieten. Figuur 4; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een ½ m3 kast Figuur 5; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in het 4 m3 kaststation In het geval van de mini-kast is middels enkele indicatieve metingen gekeken naar lekkages bij een lekopening van 0,025 mm² bij 8 bar voordruk (dat is: 0,18 m3n/h aardgas of 0,56 m3n/h waterstof) en 1 bar voordruk (dat is: 40l/u aardgas of 125 l/u waterstof). 2e stap: Metingen Een uitgebreid testprogramma is uitgevoerd met een ½ m3 kast en een 4 m3 kaststation. Met een mini- kast zijn enkele indicatieve metingen uitgevoerd om een eerste indruk te krijgen. Dit gebeurde door bij de experimenten een referentielek in het midden van de behuizing te positioneren. Hieruit stroomt gas (waterstof of aardgas) met een bekend debiet dat geregeld is door een Mass Flow Controller. Vervolgens is de gasconcentratie gemeten op verschillende plekken in de behuizing, direct bij de ventilatieopeningen buiten de behuizing en op een afstand van een 0,5 meter buiten de behuizing. Het kleinst gekozen lek heeft een debiet van 40l/u aardgas (125 l/u voor waterstof). Het grootste lek is op basis van een lekopening van 0,25mm2 bij een voordruk van 8 bar (dat is: 1,8 m3n/h aardgas of 5,6 m3n/h waterstof). Tussen deze uitersten zijn diverse andere lekdebieten gekozen. De belangrijkste gegevens van de experimenten zijn weergegeven in de drie onderstaande grafieken voor de 1/2m3 kast, het 4m3 kaststation en de mini-kast waarbij het lekdebiet op de x-as afneemt in Pagina 8/123 Pagina 8/123 Figuur 6; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een mini-kast Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Aardgas 0,18 m³(n)/h Aardgas 0,04 m³(n)/h Waterstof 0,56 m³(n)/h Waterstof 0,125 m³(n)/h Figuur 4; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een ½ m3 kast Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen Figuur 5; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in het 4 m3 kaststation Aardgas 1,8 m³(n)/h 0,4 m³(n)/h 0,18 m³(n)/h 0,04 m³(n)/h Waterstof 5,8 m³(n)/h 1,25 m³(n)/h 0,56 m³(n)/h 0,125 m³(n)/h vergelijkbaar met praktijkmetingen In het geval van de mini-kast is middels enkele indicatieve metingen gekeken naar lekkages bij een lekopening van 0,025 mm² bij 8 bar voordruk (dat is: 0,18 m3n/h aardgas of 0,56 m3n/h waterstof) en 1 bar voordruk (dat is: 40l/u aardgas of 125 l/u waterstof). In het geval van de mini-kast is middels enkele indicatieve metingen gekeken naar lekkages bij een lekopening van 0,025 mm² bij 8 bar voordruk (dat is: 0,18 m3n/h aardgas of 0,56 m3n/h waterstof) en 1 bar voordruk (dat is: 40l/u aardgas of 125 l/u waterstof). Figuur 6; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een mini-kast Aardgas 0,18 m³(n)/h Aardgas 0,04 m³(n)/h Waterstof 0,56 m³(n)/h Waterstof 0,125 m³(n)/h Figuur 6; Gasconcentratie (vol%) bij verschillende lekdebieten aardgas en waterstof in een mini-kast Pagina 9/123 Pagina 9/123 Meetresultaten 4 m3 kast - Bij zowel aardgas (40l/u) als waterstof (125 l/u) blijft de concentratie ver onder de onderst brandbaarheidsgrenzen. De gemeten concentraties bij waterstof zijn wel hoger, maar in a gevallen onder de grens van een brandbaar mengsel - Bij zowel aardgas (40l/u) als waterstof (125 l/u) blijft de concentratie ver onder de onderste brandbaarheidsgrenzen. De gemeten concentraties bij waterstof zijn wel hoger, maar in alle gevallen onder de grens van een brandbaar mengsel - Bij metingen met lekopeningen van 0,25 mm² en 0,025 mm² zijn geen brandbare mengsel gemeten bij aardgas. Bij waterstof is een brandbaar mengsel gemeten bij een lekopening van 0,25 mm². - Bij metingen met lekopeningen van 0,25 mm² en 0,025 mm² zijn geen brandbare mengsel gemeten bij aardgas. Bij waterstof is een brandbaar mengsel gemeten bij een lekopening van 0,25 mm². Wanneer de metingen nader worden bestudeerd, kan het volgende worden geconcludeerd: Meetresultaten ½ m3 kast - Bij een lek van 40l/u aardgas is de maximum gasconcentratie in een ½ m3 kast ongeveer 1 vol%, een mengsel onder de onderste brandbaarheidsgrens2. Bij een vergelijkbaar lek voor waterstof (125 l/u), loopt de concentratie op tot 2,7 vol%, ook dat is onder de onderste brandbaarheidsgrens. Direct bij de ventilatieopeningen worden vergelijkbare concentraties gemeten, met 3,2 vol% als lokale, tijdsafhankelijke uitschieter. Op een halve meter afstand van de behuizing is de concentratie in alle gevallen ver onder de brandbaarheidsgrens. - Bij metingen met lekopeningen van 0,25 mm² en 0,025 mm² zijn brandbare mengsels gemeten bij zowel aardgas als waterstof. Meetresultaten mini-kast - Er zijn enkele metingen uitgevoerd bij de mini-kast. Deze hadden als doel om input te gev aan eventueel vervolgonderzoek en zijn te beperkt om harde conclusies aan te verbinden - Er zijn enkele metingen uitgevoerd bij de mini-kast. Deze hadden als doel om input te geven aan eventueel vervolgonderzoek en zijn te beperkt om harde conclusies aan te verbinden. - Een feitelijke waarneming is dat bij aardgas (40l/u) geen brandbare mengsel is gemeten, bij waterstof (125 l/u) wel. g j p - Een feitelijke waarneming is dat bij aardgas (40l/u) geen brandbare mengsel is gemeten, waterstof (125 l/u) wel. - Een feitelijke waarneming is dat bij aardgas (40l/u) geen brandbare mengsel is gemeten, bij waterstof (125 l/u) wel. Aanbevelingen: Aanbevelingen: Aanbevelingen: Uit dit onderzoek komen de volgende aanbevelingen: g Uit dit onderzoek komen de volgende aanbevelingen: g g - Overweeg om concentratiemetingen in de standaard werkwijze op te nemen door deze te registreren voor onderzoek en monitoring over langere periode. Dit kan interessante inzichten geven in hoe populaties gasdrukregelstations zich ontwikkelen. - Overweeg om concentratiemetingen in de standaard werkwijze op te nemen door deze te registreren voor onderzoek en monitoring over langere periode. Dit kan interessante inzichten geven in hoe populaties gasdrukregelstations zich ontwikkelen. - Overweeg om de werkinstructies aan te passen zodat monteurs bij werkzaamheden aan gasstations een gasconcentratiemeting in de ventilatieopening uitvoeren voordat ze de kast openen. Door de gemeten gasconcentratie te registreren krijgen de netbeheerders met een beperkte extra inspanning meer inzicht in daadwerkelijk frequentie van het optreden van grotere lekken. Daarnaast kunnen aangetroffen lekken vervolgens worden gerepareerd, waardoor de populatie van stations met lekkage kleiner wordt. p p g - Voor toepassing van waterstof in gasstations lijken aanvullende voorzorgsmaatregelen nodig te zijn om hetzelfde veiligheidsniveau te bereiken als voor aardgas. Bij het 4 m3 kaststation is dit verschil duidelijker dan bij de ½ m3 kast. Uitgaande van het voorzorgsprincipe, zijn voor zowel de ½ m3 als de 4 m3 kast aanvullende voorzorgsmaatregelen verstandig. Hier zijn verschillende mogelijkheden, zoals het verder vergroten van het ventilatieoppervlakte van de behuizing, het aanpassen van de behuizing, het plaatsen van hekwerk rondom minimaal een meter afstand van de behuizing of intensievere controle op lekken dan gebruikelijk is voor stations op aardgas. p g - Nader onderzoek op het gebied van HAS-kasten is aan te raden voordat HAS-kasten omgezet worden naar waterstof. Zolang dat onderzoek nog niet is uitgevoerd, is het bij distributie van waterstof raadzaam om aanvullende voorzorgsmaatregelen toe te passen. p g - Nader onderzoek op het gebied van HAS-kasten is aan te raden voordat HAS-kasten omgezet worden naar waterstof. Zolang dat onderzoek nog niet is uitgevoerd, is het bij distributie van waterstof raadzaam om aanvullende voorzorgsmaatregelen toe te passen. g g - Het is belangrijk dat de normcommissie van de NEN 1059 een uitspraak doet voor welke lekdebieten de ventilatie een effectieve maatregel moet zijn voor specifieke situaties. Lekopeningen en praktijkmetingen lijken nog factoren van elkaar af te staan. - Het is belangrijk dat de normcommissie van de NEN 1059 een uitspraak doet voor welke lekdebieten de ventilatie een effectieve maatregel moet zijn voor specifieke situaties. 3e stap: verificatie met CFD berekeningen Als verificatie en om alle metingen visueel te maken zijn er CFD berekeningen uitgevoerd. Het doel van deze modelleringsactiviteiten is om te begrijpen welke factoren de verspreiding van waterstofgas in gasstations beïnvloeden. De gegevens van de metingen uit Hydelta 1.0 zijn gebruikt als referentie voor de modellen. De kwaliteit van de CFD-validatie is (nog) niet zo hoog dat de resultaten onafhankelijk van de praktijkmetingen kunnen worden gebruikt. Dit is dus een extra hulpmiddel. Verschillende redenen zijn aan te dragen, zoals het effect van wind (variabele wind) of verschillen in temperatuur (bij verschillende experimenten). De belangrijkste les die te trekken valt uit de CFD berekeningen, is de invloed van de overhang van de dakrand. De ventilatieroute is belangrijk en zou de stijgende stroom van waterstof gedreven door de opwaartse kracht moeten vergemakkelijken. Voor de gemodelleerde behuizing met de overhangende dakrand is te zien dat vooral waterstof wordt gelimiteerd om uit de behuizing te ontsnappen door een “sifon” effect. De opwaartse kracht van waterstof is onvoldoende in staat om de hydraulische weerstand te overbruggen. Een alternatieve CFD geometrie waarbij de overhang van het kastdeksel is geëlimineerd, bevestigt een significante verbetering van de ventilatie. Pagina 10/123 2 Met definitie zoals toegevoegd in begrippenlijst 2 Met definitie zoals toegevoegd in begrippenlijst Pagina 10/123 Pagina 10/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Opgedane inzichten: Opgedane inzichten: pg Als waterstof met gasdrukregelstations en de bestaande behuizingen wordt getransporteerd, dan is de verwachting dat bij een lekkage de concentraties in de behuizing hoger zijn dan met aardgas. In dit onderzoek is ook gekeken naar mogelijke aanpassingen en het effect ervan. Dat zijn aanpassingen die in principe in het veld uitgevoerd kunnen worden met de bestaande kasten. Zo kan onderventilatie worden toegevoegd door de behuizing enkele centimeters van de basis te lichten waardoor een sleuf aan de onderkant ontstaat. Daarnaast is het mogelijk om de dakrand enkele millimeters te verhogen door deze los te schroeven en terug te plaatsen met langere bouten met dikkere afstandhouders (o- ringen). Met deze praktisch, toepasbare modificaties is gekeken naar de effecten van deze ingrepen. Hieruit bleek het volgende: Aardgas Waterstof Verdeling van concentraties bij lekkage Het medium vermengt zich in de hele kast Er is een “deken”-effect. De concentraties zijn bovenin hoger dan beneden. Vergroten bovenventilatie Verlaging van de concentraties in de hele behuizing De hoge concentratie bovenin blijft hoog, maar de deken wordt dunner. Effect onderventilatie Weinig effect De dikte van de deken neemt af. Vergroten bovenventilatie (tot 4%) + onderventilatie (2%) Beste effect Beste effect, maar minder dan verwacht. Er blijven brandbare mengsels ontstaan. Toevoegen extra ventilatie effect op LEL/LFL grenzen Verlagend effect op alle metingen Verlagend effect, maar sommige lekdebieten komen boven de brandbaarheidsgrens uit. Aanbevelingen: Aanbevelingen: Lekopeningen en praktijkmetingen lijken nog factoren van elkaar af te staan. Pagina 11/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Begrippenlijst Behuizing opstellingsruimte voor een gasdrukregel- en/of meetinstallatie Gasdrukregel- en meetstation geheel van gasdrukregel- en/of meetinstallatie, behuizing, locatie en bijbehorende (elektrische) voorzieningen, begrensd door en inclusief de scheidingsafsluiters. Kast behuizing met een inhoud (Ir) kleiner dan of gelijk aan 0,5 m³ Opmerking 1 bij de term: De kast wordt bij definitie als niet-betreedbaar beschouwd. Kaststation behuizing met een inhoud (Ir) groter dan 0,5 m³ en kleiner dan 15 m³ Opmerking 1 bij de term: Een kaststation kan wel- of niet-betreedbaar zijn. Mini-kast behuizing met een inhoud (Ir) kleiner dan of gelijk aan 0,1 m³ Opmerking 1 bij de term: De kast wordt bij definitie als niet-betreedbaar beschouwd. Veiligheidsafstand te hanteren minimumafstand tussen de behuizing van een opstelling of bovengrondse componenten van een open opstelling en andere objecten. Aansluitleiding leiding vanaf de hoofdleiding naar het punt van levering ten behoeve van de eindverbruiker. Districtstation (DS) gasdrukregel- en meetstation voor gasdrukregeling van gas uit hogedrukdistributienet naar een lagedrukdistributienet. Gasdrukregelinstallatie inrichting, bevattend alle componenten, inclusief in- en uitlaatleiding tot en met de scheidingsafsluiters, die samen functioneren als gasdrukregeling en/of drukbeveiliging. Hogedrukafleverstation (HAS) een gasdrukregel- en/of meetstation, uitpandig opgesteld, van de categorie A7, A8 of A9 volgens de NEN-1059 (2019). m³ bij normaal condities (m³n) hoeveelheid gas die in droge toestand en bij een druk van 101 325 Pa en bij een temperatuur van 273,15 K (0 °C) een volume van 1 m³ bezit. Ontvangstation (OS) gasdrukregelstation waar gas in druk wordt verlaagd. Drukregelsysteem systeem dat ervoor zorg draagt dat de druk in de uitlaatleiding wordt gehandhaafd binnen de vereiste grenzen. Netto inhoud (Ir) inhoud van de opstellingsruimte minus de inhoud van de installatie. Bovenventilatie/ onderventilatie/ kruisventilatie Bij het toepassen van bovenventilatie zijn aan de bovenzijde van een behuizing ventilatieopeningen gemaakt waardoor natuurlijke ventilatie kan plaatsvinden. De inhoud van de behuizing kan zo ververst worden. In het geval van onderventilatie zijn aan de onderzijde van de behuizing ventilatieopeningen aangebracht. In het geval van kruisventilatie zijn zowel aan de boven- als onderzijde ventilatieopeningen aangebracht. ATEX De ATEX richtlijn 153 (1999/92/EG en voorheen ATEX 137) beschrijft de veiligheidseisen die werkgevers of eigenaren van ATEX-installaties verplicht Begrippenlijst Behuizing opstellingsruimte voor een gasdrukregel- en/of meetinstallatie Gasdrukregel- en meetstation geheel van gasdrukregel- en/of meetinstallatie, behuizing, locatie en bijbehorende (elektrische) voorzieningen, begrensd door en inclusief de scheidingsafsluiters. Kast behuizing met een inhoud (Ir) kleiner dan of gelijk aan 0,5 m³ Opmerking 1 bij de term: De kast wordt bij definitie als niet-betreedbaar beschouwd. Kaststation behuizing met een inhoud (Ir) groter dan 0,5 m³ en kleiner dan 15 m³ Opmerking 1 bij de term: Een kaststation kan wel- of niet-betreedbaar zijn. Mini-kast behuizing met een inhoud (Ir) kleiner dan of gelijk aan 0,1 m³ Opmerking 1 bij de term: De kast wordt bij definitie als niet-betreedbaar beschouwd. Veiligheidsafstand te hanteren minimumafstand tussen de behuizing van een opstelling of bovengrondse componenten van een open opstelling en andere objecten. Aansluitleiding leiding vanaf de hoofdleiding naar het punt van levering ten behoeve van de eindverbruiker. Districtstation (DS) gasdrukregel- en meetstation voor gasdrukregeling van gas uit hogedrukdistributienet naar een lagedrukdistributienet. Gasdrukregelinstallatie inrichting, bevattend alle componenten, inclusief in- en uitlaatleiding tot en met de scheidingsafsluiters, die samen functioneren als gasdrukregeling en/of drukbeveiliging. Hogedrukafleverstation (HAS) een gasdrukregel- en/of meetstation, uitpandig opgesteld, van de categorie A7, A8 of A9 volgens de NEN-1059 (2019). m³ bij normaal condities (m³n) hoeveelheid gas die in droge toestand en bij een druk van 101 325 Pa en bij een temperatuur van 273,15 K (0 °C) een volume van 1 m³ bezit. Ontvangstation (OS) gasdrukregelstation waar gas in druk wordt verlaagd. Drukregelsysteem systeem dat ervoor zorg draagt dat de druk in de uitlaatleiding wordt gehandhaafd binnen de vereiste grenzen. Netto inhoud (Ir) inhoud van de opstellingsruimte minus de inhoud van de installatie. Bovenventilatie/ onderventilatie/ kruisventilatie Bij het toepassen van bovenventilatie zijn aan de bovenzijde van een behuizing ventilatieopeningen gemaakt waardoor natuurlijke ventilatie kan plaatsvinden. De inhoud van de behuizing kan zo ververst worden. In het geval van onderventilatie zijn aan de onderzijde van de behuizing ventilatieopeningen aangebracht. In het geval van kruisventilatie zijn zowel aan de boven- als onderzijde ventilatieopeningen aangebracht. ATEX De ATEX richtlijn 153 (1999/92/EG en voorheen ATEX 137) beschrijft de veiligheidseisen die werkgevers of eigenaren van ATEX-installaties verplicht Niet gevaarlijk gebied Een opstellingsruimte mag als niet-gevaarlijk gebied (NGG) worden geclassificeerd als een gaslek met een opening van maximaal 1 mm2 bij de onder normale bedrijfsomstandigheden heersende druk, geen gevaarlijke hoeveelheid explosief gas-luchtmengsel kan opleveren volgens de NEN-1059 (2019). Een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden” (zie par 5.5.2 van de NPR7910-1 (2021)). Het aantal malen dat de lucht van een ruimte wordt ververst. Een eenheid waarmee aangegeven wordt wat de verhouding is tussen de verversing van lucht in een ruimte en het volume van deze ruimte bij het beschouwde lekdebiet van de in de ruimte aanwezige gevarenbron. Een kracht per oppervlakte eenheid waarbij in het SI systeem wordt gesproken van Pa of bar. Relatieve druk is altijd een gemeten waarde ten opzichte van een referentiewaarde, meestal ten opzichte van de luchtdruk. De relatieve druk kan hoger of lager zijn dan die referentiewaarde, respectievelijk over- en onderdruk. Om het verschil aan te geven wordt wel de letter g (gauge) of a (absolute) achter de eenheid geplaatst, zodat bijvoorbeeld 2 bara overeenkomt met 1 barg. Met LEL wordt de onderste brandbaarheidsgrens bedoeld. Onder de onderste brandbaarheidsgrens is er onvoldoende brandstof aanwezig om een verbrandingsreactie in stand te houden. Met LEL en LFL wordt dezelfde onderste brandbaarheidsgrens bedoeld. Voor waterstof is de LEL/LFL 4 vol% waterstof in lucht, voor aardgas is de LEL/LFL 5 vol% aardgas in lucht. Met UEL wordt de bovenste brandbaarheidsgrens bedoeld. Boven de bovenste brandbaarheidsgrens is er onvoldoende zuurstof aanwezig om een verbrandingsreactie in stand te houden. Met UEL en UFL wordt dezelfde bovenste brandbaarheidsgrens bedoeld. Voor waterstof is de UEL/UFL 75 vol% waterstof in lucht, voor aardgas is de UEL/UFL 15 vol% aardgas in lucht. De mediaan is de middelste waarde van een groep getallen die gerangschikt wordt volgens grootte. Het is het getal dat exact in het midden ligt zodat 50% van de gerangschikte getallen boven 50% ligt en 50% onder de mediaan. Volledige toelichting van dit begrip wordt gegeven in NEN-EN1127-1:2019. Bij componenten met “enhanced tightness” kan er van worden uitgegaan dat er geen lekkage plaats kan vinden. Deze componenten zijn hier specifiek voor ontworpen en er worden permanent gemonitord op afwijkingen middels gedocumenteerd periodiek onderhoud. Klein type: behuizingen t.b.v. HAS installatie zijn meestal gesitueerd met een inhoud < 0,25 m³. Behuizing opstellingsruimte voor een gasdrukregel- en/of meetinstallatie opstellingsruimte voor een gasdrukregel- en/of meetinstallatie geheel van gasdrukregel- en/of meetinstallatie, behuizing, locatie en bijbehorende (elektrische) voorzieningen, begrensd door en inclusief de scheidingsafsluiters. behuizing met een inhoud (Ir) kleiner dan of gelijk aan 0,5 m³ Opmerking 1 bij de term: De kast wordt bij definitie als niet-betreedbaar beschouwd. De ATEX richtlijn 153 (1999/92/EG en voorheen ATEX 137) beschrijft de veiligheidseisen die werkgevers of eigenaren van ATEX-installaties verplicht Pagina 12/123 moeten treffen zodat medewerkers veilig en gezond kunnen werken in omgevingen met explosiegevaar. Niet gevaarlijk gebied Medium type: behuizingen t.b.v. DS/AS stations vaak gesitueerd met een inhoud < 0,5 m³. Groot type behuizingen t.b.v. grotere DS/OS stations vaak gesitueerd met een inhoud < 4 m³. Pagina 13/123 WP6B – Veiligheid – Gasstations , g p betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Inhoud Document samenvatting ......................................................................................................................... 2 Executive summary ................................................................................................................................. 4 Samenvatting ........................................................................................................................................... 8 Begrippenlijst ......................................................................................................................................... 12 1. Aanleiding ...................................................................................................................................... 17 2. Inventarisatie ................................................................................................................................. 19 2.1 Behuizingen voor gasdrukregelstations in Nederland .......................................................... 20 De behuizing en het gasstation ..................................................................................................... 20 Meest voorkomende typen, naar functie en inlaatdruk ............................................................... 20 Veel voorkomende afmetingen ..................................................................................................... 21 2.2 Belangrijkste typen van behuizingen laatste 10 jaar ............................................................. 22 2.3 Expert opinion voor de hand liggende behuizingen .............................................................. 23 Veel voorkomende maatvoeringen ............................................................................................... 23 Principe van ventilatie ................................................................................................................... 24 Groter dan ½ m3 ............................................................................................................................ 27 2.4 Interpretatie van de inventarisatie ....................................................................................... 29 2.5 Selectie van behuizingen en de mate waarin deze representatief zijn (beantwoording deelvraag 1) ....................................................................................................................................... 30 3. Beschouwing lekopening en relevante normen ............................................................................ 32 3.1 Terugblik op onderzoek HyDelta 1.0 (werkpakket D1B.3A) .................................................. 32 3.2 Beschikbare informatie over lekkages .................................................................................. 33 3.3 Gekozen lekopening .............................................................................................................. 36 3.4 Lekopeningen en eerdere experimenten .............................................................................. 37 3.5 Waarom ATEX? ...................................................................................................................... 39 3.6 ATEX zone bepaling volgens NPR-7910-1 .............................................................................. 40 3.7 ATEX zone bepaling volgens de NEN-EN-IEC60079-10-1 ...................................................... 40 3.8 Deelconclusie toegestane lekopening en relatie tot normen ............................................... 41 4. Plan van aanpak ............................................................................................................................. 42 4.1 Testmethode op hoofdlijnen ................................................................................................. 42 4.2 Meetpunten en meetapparatuur .......................................................................................... 42 4.3 Wind en ventilatie ................................................................................................................. 43 4.4 Foto’s van de opstellingen..................................................................................................... 44 5. Bespreking resultaten van bestaande behuizingen ...................................................................... 46 5.1 Lekdebiet 0,18 m3n/hr - aardgas ............................................................................................ 46 5 2 Lekdebiet 0 56 m3 /hr waterstof 47 Inhoud Pagina 14/123 WP6B – Veiligheid – Gasstations 114 Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 115 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 116 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 117 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 118 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 119 X Meetresultaten – HAS kast .......................................................................................................... 120 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 120 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 121 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 122 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 123 WP6B – Veiligheid – Gasstations betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 5.3 Samenvatting resultaten ....................................................................................................... 48 5.4 Daadwerkelijke ventilatievoud (afgeleid van meetwaardes) ................................................ 51 6. Aanpassingen ventilatie bij ½ m3 kast (oriënterend onderzoek) .................................................. 52 6.1 Aanleiding .............................................................................................................................. 52 6.2 Het effect van additionele onderventilatie ........................................................................... 52 6.3 Het effect van extra boven- en/of onderventilatie ............................................................... 54 6.4 Beschouwing en samenvatting resultaten aanpassingen extra ventilatie ............................ 58 7. Modellering van waterstofverspreiding bij lekkage in gasstations ............................................... 60 7.1 Inleiding ................................................................................................................................. 60 7.2 Literatuurstudie ..................................................................................................................... 61 7.3 Modelleringsaanpak .............................................................................................................. 62 Basis en opstelling van de HyRAM+ calculations .......................................................................... 62 Basis en opstelling van de CFD-berekeningen ............................................................................... 63 7.4 Resultaten van de simulaties met HyRAM+ .......................................................................... 66 7.5 Resultaten van de CFD-simulaties ......................................................................................... 67 Simulatie-01 – Waterstof, 6 m³(n)/h ............................................................................................. 67 Simulatie-02 – Waterstof,  0.6 m³(n)/h ....................................................................................... 70 Simulatie-05 & Simulatie-06 – aardgas, 3 m³(n)/h en 0.45 m³(n)/h ............................................. 72 Simulation-03 – Lekkage in horizontale richting ........................................................................... 74 Simulatie-04 – Verschillend ventilatieontwerp ............................................................................. 75 7.6 Conclusies CFD ....................................................................................................................... 78 7.7 Aanbevelingen CFD ................................................................................................................ 78 8. Conclusies ...................................................................................................................................... 79 9. Aanbevelingen ............................................................................................................................... 82 Referenties ............................................................................................................................................ 83 I Overzicht samenstelling begeleidings- en sparringsgroep ............................................................ 85 II Berekening van ventilatie effecten ............................................................................................... 86 III Zonebepaling ................................................................................................................................. 89 IV Berekening ventilatievoud geteste behuizingen ........................................................................... 94 V Foto’s van de testopstelling .......................................................................................................... 98 VI Gebruikte meetapparatuur ......................................................................................................... 100 VII Overzicht alle metingen........................................................................................................... 101 VIII Meetresultaten – ½ m³ kast .................................................................................................... 104 Lekdebiet 1,8 m³n/hr – aardgas (lekopening 0,25 mm²) – 8 bar ..................................................... 104 Pagina 15/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 107 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 108 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 109 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 110 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 111 IX Meetresultaten – 4 m3 kaststation ............................................................................................. 112 Lekdebiet 1,8 m3n/hr – aardgas (lekopening 0,25 mm²) – 8 bar ..................................................... 112 Lekdebiet 5,6 m3n/hr – waterstof (lekopening 0,25 mm²) – 8 bar .................................................. 113 Lekdebiet 0,4 m3n/hr – aardgas (lekopening 0,25 mm²) – 1 bar ..................................................... het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 107 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 108 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 109 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 110 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 111 Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 107 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 108 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 109 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 110 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 111 IX Meetresultaten – 4 m3 kaststation ............................................................................................. 112 Lekdebiet 1,8 m3n/hr – aardgas (lekopening 0,25 mm²) – 8 bar ..................................................... 112 Lekdebiet 5,6 m3n/hr – waterstof (lekopening 0,25 mm²) – 8 bar .................................................. 113 Lekdebiet 0,4 m3n/hr – aardgas (lekopening 0,25 mm²) – 1 bar ..................................................... 114 Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 115 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 116 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 117 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 118 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 119 X Meetresultaten – HAS kast .......................................................................................................... 120 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 120 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 121 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 122 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 123 IX Meetresultaten – 4 m3 kaststation ............................................................................................. 112 Lekdebiet 1,8 m3n/hr – aardgas (lekopening 0,25 mm²) – 8 bar ..................................................... 112 Lekdebiet 5,6 m3n/hr – waterstof (lekopening 0,25 mm²) – 8 bar .................................................. 113 Lekdebiet 0,4 m3n/hr – aardgas (lekopening 0,25 mm²) – 1 bar ..................................................... 114 Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar ................................................ 115 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 116 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 1. Aanleiding HyDelta is een nationaal samenwerkingsprogramma waarbij barrières die zorgen dat waterstofprojecten worden vertraagd of zelfs stilgezet, worden weggenomen. Een onderdeel van het HyDelta programma is het werkpakket Gasstations. Gasdrukregelstations (voor aardgas) staan in Nederland doorgaans in de openbare ruimte. Dat is toegestaan, omdat buiten de behuizing van een gasdrukregelstation geen ATEX- zone aanwezig is, mits deze voldoet aan de ventilatie-eisen en voorwaarden zoals gesteld in de NEN1059: 2019. In de NEN:1059, 2019 [1] wordt gesteld dat bij een ventilatievoud van meer dan 5 keer per uur, binnen de behuizing een ATEX zone 2 van toepassing is. Buiten de behuizing is er dan geen ATEX zonering. De beschrijving zoals gesteld in de bovenstaande norm roept allerlei vragen op. Wat is de praktische betekenis van ATEX zone 2: hoe vaak mag welke concentratie optreden? Waar is de ventilatievoud van 5 keer per uur op gebaseerd en zien we het verdunnen van een lekkage/ de opbouw van een brandbaar mengsel ook daadwerkelijk terug in metingen en modellen? Wat is de kwantitatieve invulling van “kleine” lekkages en “voldoende” ventilatie? En, misschien wel de belangrijkste vraag binnen WP6B-1 van HyDelta, op welke wijze ventileert waterstof anders dan aardgas uit een behuizing van een station? Al deze vragen hangen met elkaar samen. Daarom is er in samenspraak met de expertgroep voor gekozen om in dit onderzoek niet te werken aan de hand van één enkele algemene hoofdvraag. Wel is er één doelstelling geformuleerd. Het doel van dit onderzoek is: Het verkrijgen van inzicht in de geschiktheid van de ventilatie van bestaande behuizingen voor de distributie van waterstof. Het verkrijgen van inzicht in de geschiktheid van de ventilatie van bestaande behuizingen voor de distributie van waterstof. Om dit inzicht te verkrijgen worden de volgende deelvragen behandeld. Om dit inzicht te verkrijgen worden de volgende deelvragen behandeld. 1. Deelvraag 1: Wat zijn de meest voor de hand liggende typen behuizingen van gasstations voor de toepassing van waterstofdistributie? 2. Deelvraag 2: Aan welke voorwaarden moet een behuizing van een gasstation voldoen voordat deze geschikt is voor de distributie van waterstof? 3 3. Deelvraag 3: In hoeverre voldoen de meest voor hand liggende behuizingen aan de voorwaarden, al dan niet met additionele ventilatie? 4 Het antwoord op deelvraag 1 wordt gegeven in paragraaf 0, op deelvraag 2 in paragraaf 3.8. Deelvraag 3 wordt beantwoord in de conclusie (hoofdstuk 8). 117 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 118 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 119 X Meetresultaten – HAS kast .......................................................................................................... 120 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar ................................................. 120 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar .............................................. 121 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar ................................................. 122 Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar ............................................ 123 Pagina 16/123 Pagina 16/123 3 De eisen voor ventilatie van huidige behuizingen is vastgelegd in de NEN1059: 2019 en de NEN12186: 2014 waarbij onderscheid wordt gemaakt tussen verschillende soorten behuizingen. De gekozen lekopening in de NEN1059: 2019 is vastgesteld op 1 mm² voor het zoneren van opstellingsruimten. Met de resultaten uit HyDelta 1.0 staat deze lekopening ter discussie (zie ook paragraaf 3.2) voor het uitvoeren van de testen binnen dit werkpakket (voor het beproeven van de ventilatie en het vaststellen van de zonering). In HyDelta 1.0 is expliciet een aanbeveling gemaakt dat er verschil zou moeten zijn tussen een incident en een reguliere lekkage/ storing. 4 Het uitgangspunt voor het toepassen van ventilatie zal in dit onderzoeksprogramma allereerst gebaseerd zijn op bestaande behuizingen en geldende normen. Aanpassingen van de behuizingen ter verbetering van de ventilatie zullen, waar nodig, in samenspraak worden bepaald. Dit onderzoek kan het beste worden gezien als een doorontwikkeling van HyDelta 1.0. Daarin wordt aanbevolen om kleinere, meer in de praktijk voorkomende, lekken goed in kaart te brengen. Waar in HyDelta 1.0 (werkpakket D1B.3A) een realistische lekgrootte werd gedefinieerd aan de hand van de (geldende norm) NEN 1059: 2019, namelijk een lekopening met een diameter van 1mm², wordt in dit 3 De eisen voor ventilatie van huidige behuizingen is vastgelegd in de NEN1059: 2019 en de NEN12186: 2014 waarbij onderscheid wordt gemaakt tussen verschillende soorten behuizingen. De gekozen lekopening in de NEN1059: 2019 is vastgesteld op 1 mm² voor het zoneren van opstellingsruimten. Met de resultaten uit HyDelta 1.0 staat deze lekopening ter discussie (zie ook paragraaf 3.2) voor het uitvoeren van de testen binnen dit werkpakket (voor het beproeven van de ventilatie en het vaststellen van de zonering). In HyDelta 1.0 is expliciet een aanbeveling gemaakt dat er verschil zou moeten zijn tussen een incident en een reguliere lekkage/ storing. 4 Het uitgangspunt voor het toepassen van ventilatie zal in dit onderzoeksprogramma allereerst gebaseerd zijn op bestaande behuizingen en geldende normen. Aanpassingen van de behuizingen ter verbetering van de ventilatie zullen, waar nodig, in samenspraak worden bepaald Pagina 17/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof onderzoek onderbouwd dat andere lekopeningen ook en (wellicht zelfs beter) als reëel kunnen worden beschouwd. Met deze lekopeningen worden vervolgens ook experimenten gedaan. Daarnaast wordt in HyDelta 1.0 de aanbeveling gedaan om in kaart te brengen welk type behuizingen veelvuldig in Nederland worden toegepast. In dit onderzoek wordt deze aanbeveling opgevolgd. onderzoek onderbouwd dat andere lekopeningen ook en (wellicht zelfs beter) als reëel kunnen worden beschouwd. Met deze lekopeningen worden vervolgens ook experimenten gedaan. Daarnaast wordt in HyDelta 1.0 de aanbeveling gedaan om in kaart te brengen welk type behuizingen veelvuldig in Nederland worden toegepast. In dit onderzoek wordt deze aanbeveling opgevolgd. onderzoek onderbouwd dat andere lekopeningen ook en (wellicht zelfs beter) als reëel kunnen worden beschouwd. Met deze lekopeningen worden vervolgens ook experimenten gedaan. Daarnaast wordt in HyDelta 1.0 de aanbeveling gedaan om in kaart te brengen welk type behuizingen veelvuldig in Nederland worden toegepast. In dit onderzoek wordt deze aanbeveling opgevolgd. Pagina 18/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Doel van de inventarisatie Om meer inzicht te verkrijgen, is het noodzakelijk om een beeld te hebben van de verschillende behuizingen die veelvuldig voorkomen in Nederland. Daarnaast heeft de inventarisatie als hoofddoel om uitspraken te kunnen doen over de mate waarin de testen en CFD berekeningen in het onderzoek representatief zijn voor de behuizingen die gebruikt (zullen) worden voor waterstofdistributie. Binnen het onderzoek is het mogelijk een beperkt aantal testen en modelberekeningen uit te voeren en deze moeten zo veel mogelijk aansluiten bij de (te verwachten) praktijk. Daarnaast geeft dit deel van het onderzoek een mogelijk waardevol overzicht van de bestaande situatie van de behuizingen voor gasstations in Nederland. In dit hoofdstuk wordt om die reden de volgende deelvraag beantwoord: “Wat zijn de meest voor de hand liggende typen behuizingen van gasstations voor de toepassing van waterstofdistributie?” Met de "voor de hand liggende behuizingen" worden behuizingen bedoeld die nu gebruikelijk zijn voor nieuwe aardgas stations. Deze zijn gestandaardiseerd en veelvuldig gebruikt. In overleg met de expertbegeleidingsgroep (EAG) is nadere invulling aan gegeven aan het begrip “voor de hand liggend”. Behuizingen die het meest waarschijnlijk zullen worden toegepast in een waterstofnetwerk zijn: In overleg met de expertbegeleidingsgroep (EAG) is nadere invulling aan gegeven aan het begrip “voor de hand liggend”. Behuizingen die het meest waarschijnlijk zullen worden toegepast in een waterstofnetwerk zijn: - Zo standaard mogelijk. Dit zijn namelijk de grootste aantallen waardoor ervaringen bij één behuizing kunnen worden toegepast op een groot aantal identieke exemplaren. Bovendien is hierbij ook de meeste ervaring opgedaan met aardgas. Tenslotte zijn dit ook de meest uitontwikkelde behuizingen. - Behuizingen van relatief nieuwe stations. Voor waterstofprojecten zal een netbeheerder (logischerwijs) willen werken met de meest betrouwbare installaties en het is dan onwaarschijnlijk dat een tientallen jaren oude installatie zal worden omgezet van aardgas naar waterstof. Dat is in het bijzonder het geval voor pilot-projecten. In hoeverre later bij grootschalige waterstofdistributie ook uitsluitend gekozen wordt voor nieuwere stations, met nieuwere behuizingen, is nu nog een open vraag. Voor dit moment ligt de focus in elk geval niet op oudere typen behuizingen. - Tenslotte is een open vraag gesteld aan de EAG welk type behuizingen zij, op basis van hun expertise, zouden selecteren voor waterstofprojecten. 2. Inventarisatie Deze aanpak is deels geslaagd. De netbeheerders hebben gegevens over al hun stations kunnen herleiden. De aantallen per type (HHAS, HAS, DS, OS) zijn bekend en ook de voordruk van de stations is bekend. Andere gegevens, zoals dimensies van behuizingen en fabrikanten, bleken in veel gevallen niet beschikbaar voor de gehele populatie. In paragraaf 2.1 is met de beschikbare informatie een overzicht gegeven van de behuizingen in Nederland voor de hele populatie. Van helikopter view naar steeds meer details In eerste instantie is getracht om een volledig beeld te verkrijgen van alle behuizingen van stations van de RNB’s. Met dit volledige beeld kon dan, was de gedachte, worden nagegaan hoeveel verschillende types er als “standaard” in het veld aanwezig waren, met vergelijkbare maatvoering en wijze van ventileren. Daarmee zou kennis worden opgedaan over welk percentage van de gehele populatie als “standaard” kon worden aangeduid. Van een aantal standaard behuizingen zou additionele detailinformatie worden opgevraagd, met name over de exacte lay-out van de ventilatieopeningen. Deze detailinformatie kon dan weer vergeleken worden met de lay-out van de gemodelleerde en geteste behuizingen. Pagina 19/123 Pagina 19/123 EAG expert-opinion Tenslotte is aan de EAG gevraagd welke behuizing zij, op basis van hun expertise, voor de hand vinden liggen voor toepassing in een waterstof distributienetwerk. 2.1 Behuizingen voor gasdrukregelstations in Nederland 2.1 Behuizingen voor gasdrukregelstations in Nederland Focus op de laatste 10 jaar De focus is gelegd op behuizingen die in de laatste 10 jaar zijn aangeschaft door de RNB’s. Van deze behuizingen is meer informatie beschikbaar dan de oudere behuizingen en het is ook waarschijnlijker dat deze worden toegepast voor waterstofdistributie. Van deze behuizingen is opgevraagd en deels ontvangen: materiaal, fabricaat en % van de totale inkoop. In paragraaf 2.2 is een overzicht gegeven van de behuizingen zoals in de laatste 10 jaar ingekocht door de netbeheerders. EAG expert-opinion 2.1 Behuizingen voor gasdrukregelstations in Nederland De behuizing en het gasstation Er zijn in Nederland drie partijen die gasdrukregelinstallaties (kortweg: gasstations) leveren aan de regionale gasnetbeheerders: gAvilar, Raak en van Voskuilen. RENDO maakt zelf zijn eigen gasdrukregelinstallaties. Een gasstation bestaat uit een behuizing, leidingwerk en componenten (het filter, drukregelaars, de drukbeveiliging, manometers en afsluiters). Er zijn een beperkt aantal fabrikanten per component en ook een beperkt aantal fabrikanten van behuizingen. In de praktijk komen vooral vrijstaande stations (inhoud ≥ 15 m3), kaststations (inhoud tussen 0,5 m3 en 15 m3) en kasten (inhoud ≤ 0,5m3) voor. In de laatste categorie vallen ook de HAS installatie, met een inhoud van circa 1/8 m3. De huidige behuizingen zijn gemaakt van RVS. Vanaf eind jaren ’70 tot 2000 was polyester ook gebruikelijk, maar dat bleek na veroudering minder degelijk te zijn dan RVS. Er is sprake van een geleidelijke toename van partijen die de behuizingen levert. Avedko maakt de kasten al geruime tijd, Zador is “enkele jaren geleden” tot de markt toegetreden. Dat was destijds gedreven vanuit de wens van de netbeheerders om minder afhankelijk te zijn van één partij, met kostenbesparing als uiteindelijke doel. In de recente inventarisatie zijn nieuwe namen van fabrikanten naar voren gekomen: Quintall, Idra-tech en van Veen Deventer. Dit zijn metaalbewerkingsbedrijven. Meest voorkomende typen, naar functie en inlaatdruk Meest voorkomende typen, naar functie en inlaatdruk Er is navraag gedaan hoeveel stations de RNB’s beheren per type, per inlaatdruk en per volume-inhoud van de kast. De aantallen per type zijn exact bekend. Voor de aantallen per inlaatdruk en per volume inhoud van de kast zijn de gegevens niet helemaal volledig maar van een voldoende kwaliteit om een betrouwbare inschatting te geven van de verhouding. Dit overzicht is weergegeven in Error! Reference source not found.. Pagina 20/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof T b l 1 G i d RNB’ 2022 i l d k Di b f i l i d i di ij l h WP6B – Veiligheid – Gasstations 2.2 Belangrijkste typen van behuizingen laatste 10 jaar g j yp g j De netbeheerders is gevraagd welke behuizingen ze in de laatste 10 jaar hebben ingekocht. Dat geeft het volgende beeld: De netbeheerders is gevraagd welke behuizingen ze in de laatste 10 jaar hebben ingekocht. Dat geeft het volgende beeld: het volgende beeld: Tabel 2. Inventarisatie type behuizingen zoals in de laatste 10 jaar zijn ingekocht door de netbeheerders RNB Type kast Materiaal behuizing Fabrikant behuizing Inhoud (m³) % van het totaal (per netbeheerder) Opmerkingen Coteq kast RVS Avedko 0,5 57% afmetingen 1x0.5x1 m Coteq kaststation RVS Avedko 5,12 19% afmetingen 1.6x1.6x2 m Coteq HAS RVS Avedko 0,06 8% afmetingen 0.5x0.4x0.3 m Coteq kaststation RVS Avedko 4 6% afmetingen 2x1x2 m Coteq kaststation RVS Avedko 1 5% afmetingen 1x1x1 m Coteq kast RVS Zador 0,5 4% afmetingen 1x0.5x1 m Enexis kast RVS Zador 0,5 60% aantallen is 2021 en 2022 t/m 19-10 Enexis kaststation RVS Zador 4 18% aantallen is 2021 en 2022 t/m 19-10 Enexis kaststation RVS Zador 1 14% aantallen is 2021 en 2022 t/m 19-10 Enexis kaststation RVS Zador 2,25 8% aantallen is 2021 en 2022 t/m 19-10 Liander kast RVS Zador 41% inhoud 0,25 t/m 0,5 m3 Liander kast RVS Adedko 12% inhoud 0,25 t/m 0,5 m3 Liander kaststation RVS Zador 8% inhoud 1 tot 15 m3 Liander kaststations RVS Adedko 6% inhoud 1 tot 15 m3 Liander kaststations RVS Idra-tech 3% inhoud 1 tot 15 m3 Liander kast RVS Voskuilen 3% inhoud 0,25 t/m 0,5 m3 Liander vrijstaand gebouw steen Mavo 3% inhoud > 15m3 Liander kast RVS onbekend 2% inhoud 0,25 t/m 0,5 m3 Liander kast RVS Idra-tech 2% inhoud 0,25 t/m 0,5 m3 Liander vrijstaand gebouw steen onbekend 2% Amsterdam. Inhoud > 15m3 RENDO Mini kast RVS Avedko 0,07 44% ter vervanging polyester kastje. Rendo heeft relatief veel mini kastjes en de laatste 10 jaar zijn relatief veel vervangen. Daarom mogelijk vertekend beeld. hogedruknetwerk. Type station Totaal (afgerond) Netdruk inlaatzijde (percentage van totaal per type station) 1 bar >1 t/m 4 bar 8 bar Afleverstation 10.000 8% 13% 79% Districtstation 10.000 5% 17% 78% Hogedrukafleverstation 34.000 8% 50% 42% Overslagstation 500 0% 2% 98% Hogedrukafleverstations (HAS) zijn de grootste aantallen en zijn vooral aangesloten op het 4 bar of 8 bar deelnet. Afleverstations (AS, een aansluiting voor een bedrijf) zijn tevens een grote groep. De capaciteit van een AS is een veel grotere range dan van een HAS waarmee de behuizingen ook een grotere range zullen hebben. Van de populatie districtstations hebben verreweg de meeste een voordruk van 8 bar. g Om de afmetingen van behuizingen te achterhalen is in eerste instantie begonnen met een indeling per groep. Deze groep zijn: a) ≤ 0,5m3, b) 0,5 – 2 m3, c) 2 t/m 4 m3 en d) groter dan 4 m3. Om de afmetingen van behuizingen te achterhalen is in eerste instantie begonnen met een indeling per groep. Deze groep zijn: a) ≤ 0,5m3, b) 0,5 – 2 m3, c) 2 t/m 4 m3 en d) groter dan 4 m3. In de systemen van de netbeheerders is in veel gevallen geen koppeling gemaakt tussen de afmetingen van een gasstation en de functie van een station. Alhoewel het niet altijd goed geregistreerd is, zijn veelal van de stations 0,5 m3 kasten of kleiner. Grotere installaties waren vroeger gebruikelijk, maar door standaardisatie en ontwikkelingen zijn ze kleiner geworden. Overslagstations zijn qua formaat ongeveer gelijkwaardig verdeeld in alle groepen. Van de districtstations en afleverstations was de dataset niet compleet genoeg om hieraan conclusies te verbinden. In de systemen van de netbeheerders is in veel gevallen geen koppeling gemaakt tussen de afmetingen van een gasstation en de functie van een station. Alhoewel het niet altijd goed geregistreerd is, zijn veelal van de stations 0,5 m3 kasten of kleiner. Grotere installaties waren vroeger gebruikelijk, maar door standaardisatie en ontwikkelingen zijn ze kleiner geworden. Overslagstations zijn qua formaat ongeveer gelijkwaardig verdeeld in alle groepen. Van de districtstations en afleverstations was de dataset niet compleet genoeg om hieraan conclusies te verbinden. Pagina 21/123 WP6B – Veiligheid – Gasstations RENDO kast RVS Avedko 0,5 32% betreft met name vervanging polyester kasten RENDO kaststation RVS Avedko 4 24% betreft met name vervanging polyester kaststations Stedin kast RVS Avedko 0,5 48% Stedin maatwerk RVS Avedko 31% > 0,5 m3 en < 15 m3 Stedin kleine kast RVS Avedko 18% < 0,5 m3 Westland onbekend aluminium Quintall 0,5 36% Westland onbekend aluminium Quintall 0,1 31% afmetingen: 690x500x300 mm Westland onbekend RVS Avedko 0,12 28% afmetingen: 565x650x330 mm Westland onbekend onbekend Weva 0,25 4% Westland onbekend RVS Avedko 0,5 0.7% Westland onbekend RVS Avedko 0,5 0.1% Tabel 2. Inventarisatie type behuizingen zoals in de laatste 10 jaar zijn ingekocht door de netbeheerders Pagina 22/123 Hieruit blijkt dat ook in de laatste 10 jaar, Avedko de belangrijkste partij was om de behuizing van stations te leveren. RVS is het meest gebruikte materiaal. Polyester wordt niet meer toegepast in nieuwe behuizingen, aluminium wordt wel gebruikt. Voordelen van aluminium ten opzicht van RVS zijn een lagere prijs en een lager gewicht. Overigens maakt het gebruikte plaatmateriaal voor de ventilatie van een station geen verschil. Verder valt uit de data op te maken dat het merendeel van de behuizingen die de laatste 10 jaar zijn geplaatst een relatief kleine inhoud hebben, zoals is samengevat in tabel 2. Tabel 3. Percentage van de behuizingen in de laatste tien jaar met een inhoud van ≤ 0,5 m3 Tabel 3. Percentage van de behuizingen in de laatste tien jaar met een inhoud van ≤ 0,5 m3 Tabel 3. Percentage van de behuizingen in de laatste tien jaar met een inhoud van ≤ 0,5 m RNB ≤ 0,5 m3 Coteq 69% Enexis 60% Liander 60% Rendo 76% Stedin 66% Westland 100% De inhoud van de behuizing wisselt tussen de 60 liter en de 108 liter. Aangezien dit al de “meest- voorkomende” types zijn, is niet uit te sluiten dat in het veld nog kleinere of grotere mini-kasten te vinden zijn. Wel kan gesteld worden dat alle netbeheerders streven naar standaardisatie. Zeker wanneer gekeken wordt naar behuizingen van de laatste 10 jaar, kan worden gesteld dat elke netbeheerder een standaard model heeft. Liander heeft de laatste 10 jaar hoofdzakelijk mini-kasten geplaatst met dimensies 0,6 x 0,6 x 0,3 meter. Enexis meldt dat voor mini-kasten (HAS) en afleverstations in de regel in 5 Detail tekening geleverd met maten 0,5/0,4/0,3 m 6 Brochure gAvilar geleverd met maten 0,75/0,45/0,75 m orkomende” types zijn, is niet uit te sluiten dat in het veld nog kleinere of grotere mini-kasten Principe van ventilatie De netbeheerders hebben foto’s en tekeningen aangeleverd van de behuizingen, zie figuur 7. Daarnaast is één mini-kast aangeleverd voor het uitvoeren van testen (hierover meer in hoofdstuk 5). Figuur 7. Veel voorkomende afmetingen van mini-kasten Figuur 7. Veel voorkomende afmetingen van mini-kasten Uit bestudering van deze tekeningen valt op dat de ventilatie in hoofdlijnen op vier verschillende methodes wordt gebaseerd, waarbij optie 3 door één netbeheerder wordt toegepast en optie 4 als curiositeit kan worden beschouwd. Het gaat om de volgende principes: 1. Bovenventilatie7 door middel van een overmaats dak. Dit zien we terug in de tekeningen van Liander, Stedin en Rendo. Het dak bevindt zich op enige afstand (orde grootte 1 cm) boven de rest van de behuizing. Het dak rust op enkele schroeven. Het dak is overmaats, hetgeen betekent dat de dakrand enkele millimeters breder en/of langer is. Ter voorkoming van een volledige open ruimte (met het risico dat men zaken naar binnen kan gooien), hebben de dakranden verticale strips. Uit de tekening van Liander is op te maken dat de dakrand op drie zijden (1x lang, 2x kort) groter is dan de voet. De tekening van Stedin is minder duidelijk, maar hier lijkt de dakrand op alle vier de zijden overmaats te zijn. 2. Bovenventilatie door middel van een sleuf. De mini behuizing die door Enexis is aangeleverd beschikt over één sleuf aan een lange zijde, zie figuur 8. Ook hier is door middel van een opstaande strip voorkomen dat er een volledig open ruimte ontstaat. 2. Bovenventilatie door middel van een sleuf. De mini behuizing die door Enexis is aangeleverd beschikt over één sleuf aan een lange zijde, zie figuur 8. Ook hier is door middel van een opstaande strip voorkomen dat er een volledig open ruimte ontstaat. 3. De behuizingen van Westland beschikken over boven- en onderventilatie. De minikast rust op de gasvoerende buizen en in de bodemplaat zitten twee openingen. Daarnaast beschikken beide de korte zijden van de dakrand over kieren als gevolg van het overmaatse dak. Westland kent ook een polyester (0,25 m³) minikast die aan de onderkant geheel open is en enige centimeters boven maaiveld wordt geplaatst, figuur 10. 4. Tijdens een voortgangsoverleg werd een foto getoond gemaakt van mini-kasten met een rond ventilatierooster. Dit wordt beschouwd als curiositeit en tegelijkertijd als een mogelijke oplossingsrichting wanneer extra ventilatie gerealiseerd moet worden. 4. 7 Bij het toepassen van bovenventilatie zijn aan de bovenzijde van een behuizing ventilatieopeningen gemaakt waardoor natuurlijke ventilatie kan plaatsvinden. De inhoud van de behuizing kan zo ververst worden. In het geval van onderventilatie zijn aan de onderzijde van de behuizing ventilatieopeningen aangebracht. In het geval van kruisventilatie zijn zowel aan de boven- als onderzijde ventilatieopeningen aangebracht. 2.3 Expert opinion voor de hand liggende behuizingen Enexis meldt dat voor mini-kasten (HAS) en afleverstations in de regel in Pagina 23/123 5 Detail tekening geleverd met maten 0,5/0,4/0,3 m 6 Brochure gAvilar geleverd met maten 0,75/0,45/0,75 m 5 Detail tekening geleverd met maten 0,5/0,4/0,3 m 6 Brochure gAvilar geleverd met maten 0,75/0,45/0,75 m Pagina 23/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof “nagenoeg” dezelfde behuizingen worden geleverd. Dit beeld wordt bevestigd door alle betrokken netbeheerders. “nagenoeg” dezelfde behuizingen worden geleverd. Dit beeld wordt bevestigd door alle betrokken netbeheerders. Principe van ventilatie 2.3 Expert opinion voor de hand liggende behuizingen Uit analyse van de belangrijkste gegevens die verkregen zijn bij de uitvraag naar de vijf tot tien meest voorkomende behuizingen bleek dat er veel verschillende soorten “standaard” maten in omloop zijn. Zoals benoemd lag de focus voornamelijk op de meest frequent toegepaste behuizingen in de afgelopen 10 jaar. Hieronder zijn de standaarden en bijzonderheden per type station samengevat. Minikast / HAS / AS Vrijwel alle netbeheerders hebben hun eigen “standaard” maten voor de behuizingen. Voor het kleinste type behuizing, werden de volgende dimensies opgegeven als veelvoorkomende behuizingen: Tabel 4. Veel typische voorkomende afmetingen van kleinste typen behuizingen per netbeheerder. Tabel 4. Veel typische voorkomende afmetingen van kleinste typen behuizingen per netbeheerder. RNB Lengte (m) Breedte (m) Hoogte (m) Totaal (m³) Coteq 0,5 0,4 0,3 0,06 Enexis5 0,415 0,3 0,5 0,062 Liander 0,6 0,3 0,6 0,108 Rendo 0,45 0,25 0,65 0,073 Stedin6 0,8 0,35 0,4 0,14 Westland 0,3 0,5 0,7 0,105 De inhoud van de behuizing wisselt tussen de 60 liter en de 108 liter. Aangezien dit al de “meest- voorkomende” types zijn, is niet uit te sluiten dat in het veld nog kleinere of grotere mini-kasten te vinden zijn. Wel kan gesteld worden dat alle netbeheerders streven naar standaardisatie. Zeker wanneer gekeken wordt naar behuizingen van de laatste 10 jaar, kan worden gesteld dat elke netbeheerder een standaard model heeft. Liander heeft de laatste 10 jaar hoofdzakelijk mini-kasten geplaatst met dimensies 0,6 x 0,6 x 0,3 meter. Enexis meldt dat voor mini-kasten (HAS) en afleverstations in de regel in e inhoud van de behuizing wisselt tussen de 60 liter en de 108 liter. Aangezien dit al de “meest- vinden zijn. Wel kan gesteld worden dat alle netbeheerders streven naar standaardisatie. Zeker wanneer gekeken wordt naar behuizingen van de laatste 10 jaar, kan worden gesteld dat elke netbeheerder een standaard model heeft. Liander heeft de laatste 10 jaar hoofdzakelijk mini-kasten geplaatst met dimensies 0,6 x 0,6 x 0,3 meter. Figuur 8. Mini-kast Enexis met enkele sleuf. Figuur 9. Kruisventilatie in de 0,1 m³ minikast van Westland Figuur 10. De "open" 1/4 m³ polyester kast van Westland. Figuur 8. Mini-kast Enexis met enkele sleuf. Figuur 8. Mini-kast Enexis met enkele sleuf. Figuur 8. Mini-kast Enexis met enkele sleuf. Figuur 9. Kruisventilatie in de 0,1 m³ minikast van Westland Figuur 9. Kruisventilatie in de 0,1 m³ minikast van Westland Figuur 10. De "open" 1/4 m³ polyester kast van Westland. Figuur 10. De "open" 1/4 m³ polyester kast van Westland. Principe van ventilatie Tijdens een voortgangsoverleg werd een foto getoond gemaakt van mini-kasten met een rond ventilatierooster. Dit wordt beschouwd als curiositeit en tegelijkertijd als een mogelijke oplossingsrichting wanneer extra ventilatie gerealiseerd moet worden. 7 Bij het toepassen van bovenventilatie zijn aan de bovenzijde van een behuizing ventilatieopeningen gemaakt waardoor natuurlijke ventilatie kan plaatsvinden. De inhoud van de behuizing kan zo ververst worden. In het geval van onderventilatie zijn aan de onderzijde van de behuizing ventilatieopeningen aangebracht. In het geval van kruisventilatie zijn zowel aan de boven- als onderzijde ventilatieopeningen aangebracht. Pagina 24/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten m betrekking tot ventilatie in verschillende typen gasdrukregelstatio distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 8. Mini-kast Enexis met enkele sleuf. Figuur 9. Kruisventilatie in de 0,1 m³ minikast van Westland Figuur 10. De "open" 1/4 m³ polyester kast van Westland. ½ ³ k ½ m³ kast Door alle netbeheerders is aangegeven dat de “½ m³ kast” veel wordt toegepast. Een veelgebruikt ½ m³ kast van gAvilar (type DS1) heeft als afmetingen 1,0 x 0,5 x 1,0 meter (l x b x h). De ½m³ kast van Zador (zoals onder andere gebruikt door Liander), heeft als afmetingen 1,05 x 0,56 x 1 meter. Omdat bij de bepaling van de inhoud van een station wordt gerekend exclusief de ruimte die door de installatie wordt Pagina 25/123 Pagina 25/123 Figuur 12. Tekening 1/2 m3 kast Liander (Zador) Figuur 12. Tekening 1/2 m3 kast Liander (Zador) Groter dan ½ m3 Een andere veel-voorkomende maat zijn kaststations van 2 x 1 x 2 meter. Verder zijn allerhande maten in omloop. Van Liander is een tekening verkregen van de 4m3 kast zoals deze door Zador wordt gefabriceerd, (figuur 13). Ook hier is duidelijk de overmaatse dakrand als ventilatieprincipe terug te zien. Figuur 12. Tekening 1/2 m3 kast Liander (Zador) Figuur 12. Tekening 1/2 m3 kast Liander (Zador) Groter dan ½ m3 ingenomen, wordt hier nog steeds gesproken over een ½ m3 kast. Beide worden veel toegepast. Alle ½ m³ kasten waarvan wij de tekeningen hebben ingezien, maken gebruik van het overmaatse dak en bovenventilatie. Het dak van deze behuizing bevindt zich enkele millimeters boven de bovenkant van de voet. Het dak rust op O-ringen rondom de bevestigingsbout, zoals is op te maken uit deze foto (figuur 11) die vanuit de binnenzijde van de kast is genomen. Figuur 11. Dak (boven) van 1/2 m3 kast, bevestigd met bout op de voet. De spleet is in verbinding met de buitenlucht. Figuur 11. Dak (boven) van 1/2 m3 kast, bevestigd met bout op de voet. De spleet is in verbinding met de buitenluc Het standaard principe van een ½ m3 kast is dat van een kantelkast of klapkast, zoals weergegeven is in figuur 12. Het kantelende deel van de kast rust op een rubberen strip, alle ventilatieopeningen bevinden zich dus ook bij een klapkast rondom het dak. Het is bekend dat niet alle ½ m3 kasten ook klapkasten zijn, maar aangezien dit geen invloed lijkt te hebben op de ventilatie is geen nader onderzoek gedaan naar de verdeling klapkast/vaste kast bij de ½ m3 kasten. Het standaard principe van een ½ m3 kast is dat van een kantelkast of klapkast, zoals weergegeven is in figuur 12. Het kantelende deel van de kast rust op een rubberen strip, alle ventilatieopeningen bevinden zich dus ook bij een klapkast rondom het dak. Het is bekend dat niet alle ½ m3 kasten ook klapkasten zijn, maar aangezien dit geen invloed lijkt te hebben op de ventilatie is geen nader onderzoek gedaan naar de verdeling klapkast/vaste kast bij de ½ m3 kasten. Alle ½ m3 kasten die in de laatste 10 jaar zijn geleverd beschikken over een plat dak. ½ m3 kasten met een puntdak kunnen voorkomen, dit zijn oudere modellen. Inzet van deze oudere en afwijkende kasten voor waterstofprojecten ligt niet voor de hand en deze worden daarom verder in dit onderzoek buiten beschouwing gelaten. Pagina 26/123 Groter dan ½ m3 Een andere veel-voorkomende maat zijn kaststations van 2 x 1 x 2 meter. Verder zijn allerhande maten in omloop. Van Liander is een tekening verkregen van de 4m3 kast zoals deze door Zador wordt gefabriceerd, (figuur 13). Ook hier is duidelijk de overmaatse dakrand als ventilatieprincipe terug te zien. Een andere veel-voorkomende maat zijn kaststations van 2 x 1 x 2 meter. Verder zijn allerhande maten in omloop. Van Liander is een tekening verkregen van de 4m3 kast zoals deze door Zador wordt gefabriceerd, (figuur 13). Ook hier is duidelijk de overmaatse dakrand als ventilatieprincipe terug te zien. Figuur 13. Uitsnede constructietekening 4 m3 kaststation Liander (Zador) W t d lt i d t k i i d t d b h i i l t d k h ft (t t k Figuur 13. Uitsnede constructietekening 4 m3 kaststation Liander (Zador) Figuur 13. Uitsnede constructietekening 4 m3 kaststation Liander (Zador) Wat verder opvalt in deze tekening is dat deze behuizing geen plat dak heeft, maar (tot op zekere hoogte) een puntdak. Dat type dak zien we terug op meerdere plekken. Het is in gebruik door Rendo (figuur 14), Westland Infra (figuur 15) en is tevens aanwezig in de behuizing die wordt gebruikt in dit onderzoek en is aangeleverd door Enexis (figuur 16). Wat verder opvalt in deze tekening is dat deze behuizing geen plat dak heeft, maar (tot op zekere hoogte) een puntdak. Dat type dak zien we terug op meerdere plekken. Het is in gebruik door Rendo (figuur 14), Westland Infra (figuur 15) en is tevens aanwezig in de behuizing die wordt gebruikt in dit onderzoek en is aangeleverd door Enexis (figuur 16). Pagina 27/123 Pagina 27/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 14. Een 4 m3 kaststation van Rendo Figuur 15. Een 4 m3 kaststation van Westland Infra WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 14. Een 4 m3 kaststation van Rendo 2.4 Interpretatie van de inventarisatie 2.4 Interpretatie van de inventarisatie Uit de inventarisatie kunnen een aantal conclusies worden verbonden. Algemeen: 2.4 Interpretatie van de inventarisatie Uit de inventarisatie kunnen een aantal conclusies worden verbonden. Algemeen: Uit de inventarisatie kunnen een aantal conclusies worden verbonden. - Behuizingen worden hoofdzakelijk gemaakt van RVS en hoofdzakelijk door Avedko en Zador, al - Behuizingen worden hoofdzakelijk gemaakt van RVS en hoofdzakelijk door Avedko en Zador, al leveren andere metaalbewerkingsbedrijven ook behuizingen. Minikasten: Minikasten: Minikasten: - Elke netbeheerder heeft zijn eigen standaard minikast. - Als ventilatieprincipe worden diverse methoden gekozen, met het overmaatse dak als meest voorkomende oplossing. voorkomende oplossing. Figuur 16. Het 4 m3 kaststation zoals wordt gebruikt voor de testen (aangeleverd door Enexis) De reden dat dit “puntdak” hier wordt benoemd, is dat het vermoeden bestaat dat dit invloed heeft op de ventilatie. Hierover wordt in paragraaf 6.4 verder aandacht besteed. De reden dat dit “puntdak” hier wordt benoemd, is dat het vermoeden bestaat dat dit invloed heeft op de ventilatie. Hierover wordt in paragraaf 6.4 verder aandacht besteed. De reden dat dit “puntdak” hier wordt benoemd, is dat het vermoeden bestaat dat dit invloed heeft op de ventilatie. Hierover wordt in paragraaf 6.4 verder aandacht besteed. Figuur 14. Een 4 m3 kaststation van Rendo Figuur 14. Een 4 m3 kaststation van Rendo Figuur 15. Een 4 m3 kaststation van Westland Infra Figuur 15. Een 4 m3 kaststation van Westland Infra Figuur 15. Een 4 m3 kaststation van Westland Infra Figuur 15. Een 4 m3 kaststation van Westland Infra Pagina 28/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 16. Het 4 m3 kaststation zoals wordt gebruikt voor de testen (aangeleverd door Enexis) De reden dat dit “puntdak” hier wordt benoemd, is dat het vermoeden bestaat dat dit invloed heeft op de ventilatie. Hierover wordt in paragraaf 6.4 verder aandacht besteed. In hoofdstuk 1 is de volgende deelvraag gesteld: In hoofdstuk 1 is de volgende deelvraag gesteld: Wat zijn de meest voor de hand liggende typen behuizingen van gasstations voor de toepassing van waterstofdistributie? In de laatste 10 jaar zijn vooral kleinere behuizingen geplaatst, deze behuizingen zijn tamelijk uniform. In het bijzonder zijn kasten met een inhoud van 0,5 m³ en kaststations met een inhoud van 4 m³ in grote mate vergelijkbaar. Het is hiermee aannemelijk dat de resultaten uit dit onderzoek een goede indicatie zijn voor alle stations met deze inhoudsmaat, zolang deze gelijk zijn met betrekking tot de wijze van ventileren en het ventilatieoppervlakte. De mini kasten (inhoud < 0,5 m³) zijn in mindere mate uniform. Elke netbeheerder heeft zijn eigen standaard die soms veel en soms weinig afwijkt van de standaard van andere netbeheerders. Deze drie types behuizingen (de mini-kast, de ½ m³ kast en de 4 m³ kast) kunnen gebruikt worden als indicatie voor een substantieel deel van de hele populatie. Deze drie types worden beschouwd als het meest voor de hand liggend om te onderzoeken voor toepassing met waterstof vanwege het feit dat deze veel voorkomen. De mini-kast wordt toegepast in een hogedrukafleverstation (HAS). In absolute aantallen is dit het meest voorkomende gasstation. Daarnaast zal deze kast, gezien het beperkte volume, bij een relatief klein lek snel hoge gasconcentraties bereiken. Binnen het project moesten keuzes worden gemaakt welke behuizingen wel en welke niet getest konden worden. Gedurende een lange tijd in het project was het uitgangspunt om te kiezen voor uitwerking van de 1/2m3 kast en de 4m3 kast, waarbij de mini-kast buiten beschouwing gelaten zou worden. Redenen om deze buiten te beschouwen te houden waren: - Er zijn bij de meeste netbeheerders geen waterstof(pilot)projecten gaande of in de planning waarbij mini-kasten worden gebruikt. - Er zijn bij de meeste netbeheerders geen waterstof(pilot)projecten gaande of in de planning waarbij mini-kasten worden gebruikt. - Testen en modellen zouden waarschijnlijk weinig inzichten opleveren. Extrapolatie van gegevens van testen met de ½ m3 kast uit HyDelta 1.0 en de eerste testen in dit HyDelta 2.0 onderzoek gaven een indicatie dat de ventilatie met de huidige lay-out van de kleinere mini-kasten volstrekt onvoldoende zou kunnen zijn. - Gezien de grotere diversiteit van de mini-kasten zou het onderzoek tamelijk uitvoerig moeten zijn. Daarmee zouden te veel andere aandachtspunten van het onderzoek geschrapt moeten worden. 2.5 Selectie van behuizingen en de mate waarin deze representatief zijn (beantwoording deelvraag 1) 0,5 m3 en 4 m3 kasten - Behuizingen zijn tot in grote mate gestandaardiseerd. Wel zijn er verschillende tekeningen in omloop en worden de kasten geleverd door diverse leveranciers. - Uit de verkregen foto’s en tekeningen blijkt telkens één principe van ventilatie: het overmaatse dak. Andere principes van ventilatie (bijvoorbeeld met roosters of kruisventilatie) worden nauwelijks toegepast. Pagina 29/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 2.5 Selectie van behuizingen en de mate waarin deze representatief zijn (beantwoording deelvraag 1) In hoofdstuk 1 is de volgende deelvraag gesteld: Tegen het eind van het project is alsnog gekozen om enkele testen uit te voeren met de mini-kast. Dit had als doel om alvast enige kennis op te doen van het ventilatiegedrag van mini-kasten zodat een eventueel vervolgonderzoek gericht op de mini-kast zo effectief mogelijk kan worden vormgegeven. De onderzochte mini kast (HAS) heeft artikelnummer 74025, type A en is geproduceerd door de firma RAAK. De behuizing is gemaakt door Van Veen metal Productie Deventer . Deze heeft als afmetingen 40,7 x 30,8 x 51,4 cm (lbh). Aan één lange zijde bevindt zich een ventilatiesleuf, direct onder de dakrand. De ventilatiesleuf is 13,5 mm bij 404 mm, daarmee bedraagt de ventilatie 4,6% van het vloeroppervlakte. De ½ m³ kast is door alle netbeheerders genoemd als interessant type. Dit model is het meest gestandaardiseerd van alle maatvoeringen. Bovendien is in HyDelta 1.0 ook getest met een ½ m3 kast, Pagina 30/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 3. Beschouwing lekopening en relevante normen waardoor resultaten uit HyDelta 1.0 gekoppeld konden worden aan 2.0. De geteste ½ m3 kast is een klapkast, aangeleverd door Enexis. Afmetingen van deze kast zijn 106 x 55,5 x 106 cm (lxbxh). De omlaagstaande lip van de dakrand heeft een hoogte van 5 cm. Aan drie zijden is bovenventilatie toegevoegd, aan de twee lange zijden en één korte zijde. De spleet onder het dak is 4mm. Het totale oppervlakte van de ventilatie is hiermee 2% van het vloeroppervlakte. De 4 m³ kast is ook door de meeste netbeheerders genoemd als interessant type. Van de kaststations “groter dan ½ m³” is de 4 m3-kast het meest gestandaardiseerd. Mogelijk kunnen metingen en berekeningen aan de 4m³ kast dienen als indicatie voor alle “grotere” kaststations. Ook hier geldt dat deze ook in HyDelta 1.0 is getest, waardoor resultaten uit HyDelta 1.0 gekoppeld konden worden aan 2.0. De afmetingen van dit kaststation zijn 200x100x200 cm. De maten komen overeen met Avedko artikel nummer 014.30.01. Rondom de gehele omtrek van het station bevindt zich een ventilatiesleuf van 16mm. Het totale ventilatieoppervlakte is daarmee 4,8% van het vloeroppervlakte. Pagina 31/123 In HyDelta 1.0 is in de werkgroep uitgebreid gesproken over lekkages in gasdrukregelstations. In dit onderzoek is gekozen om lekkages te testen die goed overeenkomen met heersende drukken uit de praktijk. Het toepassingsgebied van gasdrukregelstations weergegeven in tabel 1. De heersende druk vormt samen met de gekozen lekopening de input voor de berekening van de volumetrische uitstroom van aardgas of waterstof. In HyDelta 1.0 is in de werkgroep uitgebreid gesproken over lekkages in gasdrukregelstations. In dit onderzoek is gekozen om lekkages te testen die goed overeenkomen met heersende drukken uit de praktijk. Het toepassingsgebied van gasdrukregelstations weergegeven in tabel 1. De heersende druk vormt samen met de gekozen lekopening de input voor de berekening van de volumetrische uitstroom van aardgas of waterstof. Tabel 1 - type stations en toepassingen Type station inlaatdruk uitlaatdruk ≤ 200 mbar 1 bar 4 bar 8 bar ≤ 200 mbar 1 bar 4 bar 8 bar Afleverstation x x x x afhankelijk van wensen v/d afnemer Districtstation x x x x Hogedrukafleverstation x x x x Overslagstation x x x x Tabel 1 - type stations en toepassingen De vraag “Wat is het maximale lekdebiet in een gasdrukregelstation waar rekening mee moet worden gehouden?” kan dus alleen goed beantwoord worden wanneer de maximale druk (8 bar) en de maximale lekopening gedefinieerd zijn. Deze uitgangspunten zijn, net als de link met de praktijk, noodzakelijk om te beantwoorden om tot een goed uitgangspunt te komen. De grootte van een lekopening wordt in diverse bronnen genoemd. De vraag “Wat is het maximale lekdebiet in een gasdrukregelstation waar rekening mee moet worden gehouden?” kan dus alleen goed beantwoord worden wanneer de maximale druk (8 bar) en de maximale lekopening gedefinieerd zijn. Deze uitgangspunten zijn, net als de link met de praktijk, noodzakelijk om te beantwoorden om tot een goed uitgangspunt te komen. De grootte van een lekopening wordt in diverse bronnen genoemd. 3.1 Terugblik op onderzoek HyDelta 1.0 (werkpakket D1B.3A) Tijdens de start van het HyDelta onderzoek in 2020 is als onderdeel van het werkpakket 1B “gasstations” onder andere gekeken naar de ventilatie van behuizingen voor gasdrukregelstations [2]. Wanneer gasdrukregelstations gebruikt gaan worden voor de distributie van waterstof, moeten de nu geldende normen en eisen opnieuw tegen het licht worden gehouden. De vraag die destijds voorlag, was tweeledig en als volgt geformuleerd; 1. Zijn er aanpassingen aan de behuizing nodig voor een veilig gebruik met waterstof en zo j welke? 2. Kan er onder normale omstandigheden buiten de kast een brandbaar of explosief aardgas of waterstof mengsel optreden? eze hoofdvragen konden beantwoord worden wanneer de volgende deelvragen werden beantw 1. Wat is het maximale lekdebiet in een station waar rekening mee moet worden gehouden in het veld tijdens normale bedrijfsomstandigheden? 2. Voldoet de binnenzijde van het station inderdaad aan ATEX zonering 2, bij de huidig toegepaste ventilatie en het maximale lekdebiet, zowel voor aardgas als voor waterstof? 3. Is er sprake van een brandbaar mengsel buiten de behuizing bij de huidige toegepaste ventilatie en het maximale lekdebiet, zowel voor aardgas als voor waterstof? Uit de resultaten is naar voren gekomen dat de gekozen lekopening van 1 mm² leidt tot aanzienlijke lekdebieten waarbij zowel de concentraties aardgas als waterstof opliepen tot ruim in het ontsteekbare gebied. De vraag hierbij was wel in hoeverre de geselecteerde lekopening en de gekozen lekdebieten realistisch worden geacht voor een gaslekkage onder normale bedrijfsomstandigheden. De 1 mm² zou daarbij moeten worden beschouwd als een incident en niet als een reguliere lekkage/ storing. Hieruit is in HyDelta 1.0 onder andere de aanbeveling gedaan om opnieuw te kijken naar de resultaten van dat onderzoek. Voornaamste doel zou moeten zijn wat een goede inschatting is voor realistische lekdebieten voor reguliere lekkages. Een opsomming van realistische lekdebieten staat beschreven in paragraaf 3.2, waarna in paragraaf 3.3 een keuze wordt gemaakt voor meest realistische lekdebieten die vervolgens ook zijn toegepast in de experimenten. Pagina 32/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekgrootte uit de 1059 staat ter discussie In paragraaf 7.3.11 van de norm NEN 1059:2019 staat een lekopening genoemd van 1 mm2. Daarin staat “Een opstellingsruimte mag als niet-gevaarlijk gebied (NGG) worden geclassificeerd als een gaslek met een opening van maximaal 1 mm2 bij de onder normale bedrijfsomstandigheden heersende druk, geen gevaarlijke hoeveelheid explosief gas-luchtmengsel kan opleveren. Zie bijlage E.” In bijlage E wordt genoemd dat opstellingsruimten voor meteropstellingen met een maximale inlaatdruk van 100mbar als NGG kunnen worden geclassificeerd, indien aan de voorwaarden wordt voldaan. Zoals in de bovenstaande paragraaf reeds is toegelicht, is deze lekgatgrootte te herleiden is tot de opening bij het springen van een bourdonveer in een manometer. Een tweede argument dat wordt gebruikt in de normcommissie van de NEN 1059 is dat 1 mm2 de ademopening is van een regelaar. Op het moment dat een membraan is een regelaar scheurt, ontstaat er een uitstroomopening ter grootte van deze ademopening van circa 1 mm2. Binnen de normcommissie van de NEN 1059 staat de lekopening van 1 mm2 al enige tijd ter discussie. Verreweg de meeste gasstations van de netbeheerder hebben een inlaatdruk van 8 bar, 4 bar of 1 bar. Echter, bijlage E van de 1059 benoemd specifiek gasmeteropstellingen met een inlaatdruk van 100 mbar. De NEN 1059 geeft in feite geen direct antwoord op de vraag welke lekopening aangenomen moet worden bij het bepalen van de benodigde ventilatie voor stations bij een druk groter dan 100 mbar. Bij gebrek aan dit duidelijke antwoord, zijn binnen Hydelta 1.0 metingen gedaan bij een lekopening van deze 1 mm2, maar dan wel bij een in de praktijk optredende bedrijfsdruk, namelijk 8 bar. Het lekdebiet dat ontstaat bij een voordruk van 8 bar en een lekopening van 1 mm2 is veel groter dan wat door de experts binnen de normcommissie als realistisch wordt beschouwd voor een regulier optredend lek. Een dergelijk groot gaslek is zowel hoorbaar als ruikbaar op enige meters van het gasstation en zal beschouwd worden als een incident en niet als een reguliere lekkage. In de sector en bij de leden van de NEN 1059 commissie is er geen enkel incident bekend dat er door een vonk van buitenaf, onder normale omstandigheden, een kast vlam heeft gevat [8]. Wel worden met enige regelmaat gaslucht gemeld bij gasstations, waarop gereageerd wordt door de storingsdiensten van de netbeheerder. In de NEN-EN-IEC-60079-10-1 (2021) [6], “Explosive atmospheres – part 10-1: classification of areas – explosive gas atmospheres” worden de richtlijnen uitgezet voor Europese ATEX regulering. In deze norm die toepasbaar is op explosieve atmosferen in het algemeen komen dezelfde waarden terug voor de lekkages van pakkingen. In het HyDelta rapport “ventilatie” is de aanbeveling verstrekt om metingen te verrichten in de bandbreedte tussen 0,025 mm² en 0,25 mm² waar in annex B van deze norm (Table B.1 – “Suggested hole cross sections for secondary grade of releases”) wordt gerefereerd aan “typical values for the conditions at which the release opening will not expand”. Deze norm laat meer ruimte voor interpretatie dan de IGEM/SR/25 omdat hier een bandbreedte wordt gegeven van 0,025 mm² en 0,25 mm² waarbij geen relatie wordt gelegd met de heersende druk. In het onderzoeksrapport van de Britse Health and Safety Executive (RR630 - “Area classification for secondary releases from low pressure natural gas systems”) [7] worden soortgelijke waarden gedefinieerd voor secundaire gevarenbronnen waarbij de inhoud van dit onderzoeksrapport is ingebracht in de IGEM/SR/25 Edition 2. Informatie uit normen Pagina 33/123 Pagina 33/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Informatie uit normen In Nederland worden gasdrukregelstations ontworpen aan de hand van de NEN1059; 2019 [3], “gasvoorzieningsystemen - Gasdrukregel- en meetstations voor transport en distributie”. In deze norm wordt een lekopening van 1 mm² gedefinieerd voor het bepalen van de zonering van een opstellingsruimte. Van deze waarde is de referentiebron niet eenduidig te achterhalen maar het vermoeden bestaat dat het hier gaat om de ademopening van een falende regelaar of het falen van een bourdonveer van een manometer. In het artikel “de hinderwet en gasdrukregel- en meetstations” in het tijdschrift GAS van februari 1968 wordt beschreven dat de menselijke factor een grote rol speelt bij lekkages en dat behoorlijke ventilatie een noodzaak is. Uit vroege richtlijnen blijkt dat het niet uitgesloten kan worden dat gas uitstroomt door een opening van circa 1 mm² door het springen van een bourdonveer in een manometer [4]. In de hinderwet (1968) zelf wordt in hoofdstuk IX beschreven dat “in een omsloten opstelling moeten openingen, groter dan 1 mm, waar – bijvoorbeeld als gevolg van het bezwijken van een membraan, balg, bourdonbuis of het in werking treden van een veiligheidsklep – gas in gevaarlijke hoeveelheden kan uitstromen, zijn voorzien van een afvoerleiding, waardoor dit gas op een veilige plaats in de buitenlucht wordt afgevoerd”. Hier wordt dus gesproken over 1 mm in plaats van 1 mm². (nb. een cirkelvormig lek met een doorsnee van 1 mm heeft een oppervlakte van ¼ * π * 12 = 0,79 mm²). Deze 1 mm2 staat al enige tijd ter discussie, zoals zal worden toegelicht in de volgende paragraaf. Naast de NEN1059:2019 kan ook de IGEM/SR/25 Edition 2 (2010) [5], “Hazardous area classification of Natural Gas installations” geraadpleegd worden. In deze norm van IGEM, the institution of gas engineers and managers is het advies om voor de beschouwing van aardgas lekkages in gasdrukregelstations een lekopening te kiezen van 0,25 mm² voor secundaire gevarenbronnen bij een ingaande druk van meer dan 100 mbar. Voor installaties met een ingaande druk tot 100 mbar wordt in de UK aangeraden om voor secundaire gevarenbronnen te rekenen met een lekopening van 0,025 mm² (paragraaf 4.4.1 van IGEM /SR/25). de ventilatieopeningen bewogen. Vervolgens is nog 1 minuut gemeten op de plek waar de male concentratie tijdens de rondgang is gemeten. Er is tijdens het uitvoeren van deze metingen in geen enkel geval een brandbaar mengsel gevonden in de ventilatieopening. Bij tweederde van de metingen is de methaanconcentratie lager dan 11 ppm. Bij 99,4% van alle metingen is de gasconcentratie onder de 10% LEL. Twee metingen hadden een meetwaarde van 10% LEL en 2 metingen zitten daarboven, met als hoogst gemeten waarde 40% LEL. Er is geen enkel brandbaar mengsel gemeten in de ventilatieopeningen van alle 713 stations die bemeten zijn. Dit project heeft daarmee aangetoond dat het zeer onwaarschijnlijk is dat een station tijdens regulier bedrijf vanuit de ventilatieopeningen tot ontbranding kan worden gebracht [8]. Lekdebietmetingen vanuit het onderwerp methaanemissies Het project waarbij de ventilatieopeningen van 713 stations zijn doorgemeten, heeft in 2022 ook een spin-off project gekregen vanuit het oogpunt van kwantificeren van methaanemissies. Informatie over gasconcentraties bij de ventilatieopeningen zijn buitengewoon nuttig vanuit het oogpunt van veiligheid. Vanuit het oogpunt van kwantificeren van methaanemissies, is het ook noodzakelijk om een uitspraak te kunnen doen over de omvang van een lekdebiet. De monteur die de concentratiemeting in de ventilatieopeningen doet, heeft bij een meetwaarde van meer dan 500 ppm contact opgenomen met Kiwa. Bij 13 districtstations heeft Kiwa vervolgens door middel van een “Hi Flow Sampling” meting het lekdebiet bepaald. Het grootste gemeten lek was 0,3 liter/minuut (18 liter/uur). Een gemiddelde lekgrootte is 2 liter uur [9]. Ook in Duitsland is in 2022 onderzoek gedaan naar lekkages in gasdrukregelstations. Uit DVGW onderzoek [10] is naar voren gekomen dat bij reguliere lekkages aan 159 gasdrukregelstations een gemiddeld lek van 1,8 liter per uur is gemeten. De maximale gevonden lekkage bedroeg 31 liter per uur. Tevens is vastgesteld dat tweederde van alle metingen kleiner zijn dan 2 liter per uur met een mediaan van 1 liter per uur. Dit gemiddelde lek (of: deze emissiefactor) vervangt daarmee de eerder gebruikte emissiefactoren van 25,7 liter per uur voor stations op lagedruk en 105,6 liter per uur voor stations op middendruk op die worden aangehaald in het omvangrijke onderzoek Methane Emission Estimation Method for the Gas Distribution Grid (MEEM [11]. De originele bron van deze waarden is een onderzoek van Ruhrgas AG uit 2000 en is op basis van het geringe aantal van vijf metingen. Praktijkmetingen aan lekken bij stations j g j Om een beter beeld te verkrijgen welke gasconcentraties zich daadwerkelijke voordoen bij ventilatieopeningen in districtregelstations, is in 2022 onderzoek uitgevoerd door Kiwa Technology, in opdracht van de normcommissie NEN 1059 en Netbeheer Nederland. Bij een totaal van 713 districtstations is met een gasconcentratiemeter secuur (en volgens gestandaardiseerd protocol) gemeten langs de ventilatieopeningen. De aanzuigopening van de gasconcentratiemeter is zo dicht als mogelijk bij of net in de opening gehouden. In deze positie is met een snelheid van 2 meter per minuut Pagina 34/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Gasdrukregelstations worden in de regel toegepast als transitie tussen hoge druk en la Hierbij worden aan de hogedrukzijde drukken toegepast van 1 tot 8 barg. Daarom is te naar dezelfde lekopeningen bij inlaatdruk van 1 barg. Daarmee wordt de volgende set v lekopeningen” gedefinieerd: Bron en toelichting Lekdebiet aardgas Equivalente waterstof b lekopen Voordruk 8 bar bij Lekopening 0,25 mm² 1,80 m3n/h 5,6 m Voordruk 8 bar bij Lekopening 0,025 mm² 0,18 m3n/h 0,56 Voordruk 1 bar bij Lekopening 0,25 mm² 0,4 m3n/h 1,25 Voordruk 1 bar bij Lekopening 0,025 mm² Voor aardgas tevens in lijn met onderzoek naar lekgroottes methaanemissies (maximaal 40 liter/uur) 0,04 m3n/h 0,125 ) afrondingen van getallen voorbehouden WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 3.3 Zoals reeds in paragraaf 3.2 omschreven, zijn verschillende bronnen geraadpleegd om een eenduidig beeld te kunnen schetsen van lekkages. Voor reguliere lekkages (zogenaamde secundaire gevarenbronnen 8) is bij hogere drukken dan 100 mbar te vinden dat een typische lekopening van 0,25 mm² zouden moeten worden geselecteerd (IGEM/HSE). Een kleinere lekopening van bijvoorbeeld 0,025 mm² kan wellicht ook interessant zijn om te beschouwen met het oog op de inhoud van de Europese norm voor explosieve gas atmosferen. Een lekopeningen van 0,025 mm2 geeft bij een voordruk van 1 bar een lekgrootte van 40 liter/ uur. Lekgrootte metingen met hi flow samplers aan gasstations geven vergelijkbare waardes met een maximum van 18 liter/uur (Nederlandse metingen) en 31 liter/uur (Duitse metingen). Een lekgrootte van 40 liter / uur is daarmee een realistische lekgrootte. In een gasdrukregelstation heerst een maximale druk van 8 barg waardoor bij een lekopening in de bandbreedte tussen 0,025 mm² en 0,25 mm² leidt tot lekdebieten van 0,18 m3n/h en 1,80 m3n/h aardgas. In eerder onderzoek is gekeken naar de ratio die toegepast zou moeten worden tussen aardgas en waterstof [2] [13]. Afhankelijk van de druk, stroomopwaarts van de lekopening, is de stroming laminair of turbulent. Tevens speelt de geometrie van een lekopening hierbij een rol. Door de band genomen kan worden verondersteld dat in dit onderzoek een factor 3 tussen aardgas en waterstof een logische keuze is op basis van de resultaten uit de eerder aangehaalde onderzoeken. Wanneer dit wordt gedaan voor bovenstaande lekdebieten voor aardgas, volgt dat de lekdebieten voor waterstof tussen de 0,56 m3n/h en 5,6 m3n/h liggen. Gasdrukregelstations worden in de regel toegepast als transitie tussen hoge druk en lage(re) druk. Hierbij worden aan de hogedrukzijde drukken toegepast van 1 tot 8 barg. Daarom is tevens gekeken naar dezelfde lekopeningen bij inlaatdruk van 1 barg. Daarmee wordt de volgende set van “relevante lekopeningen” gedefinieerd: Bron en toelichting Lekdebiet aardgas Equivalente lekgrootte waterstof bij dezelfde lekopeningen Voordruk 8 bar bij Lekopening 0,25 mm² 1,80 m3n/h 5,6 m3n/h Voordruk 8 bar bij Lekopening 0,025 mm² 0,18 m3n/h 0,56 m3n/h Voordruk 1 bar bij Lekopening 0,25 mm² 0,4 m3n/h 1,25 m3n/h Voordruk 1 bar bij Lekopening 0,025 mm² Voor aardgas tevens in lijn met onderzoek naar lekgroottes methaanemissies (maximaal 40 liter/uur) 0,04 m3n/h 0,125 m3n/h ) afrondingen van getallen voorbehouden 8 een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden” (zie par 5.5.2 van de NPR7910-1 (2021)) Alle Nederlandse gasnetbeheerder zijn partner van de Oil & Gas Methane Partnership 2.0. Dit is een initiatief van UNEP (United Nations Environment Program) en heeft als doel om de accuraatheid en transparantie van het rapporteren van methaanemissies te vergroten in de olie en gassector. Binnen OGMP 2.0 wordt onderscheidt gemaakt uit drie bronnen van methaanemissies: fugitives (kleine, “normaal optredende” lekken), incidenten en emissies bij werkzaamheden. Incidenten hebben in de regel een groter lekdebiet dan “regulier optredende lekken”, maar worden in de regel gemeld en snel veilig gesteld. Bij de vaststelling van een maximale lekgrootte is het daarom van belang om vast te stellen of hierbij ook gekeken moet worden naar incidenten of uitsluitend naar lekgroottes die meer permanent kunnen optreden. Pagina 35/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en expe betrekking tot ventilatie in verschillende typen gasdruk distributie (lagedruk)net in Nederland met aardgas en w 3.3 Gekozen lekopening Zoals reeds in paragraaf 3.2 omschreven, zijn verschillende bronnen geraadpleegd om beeld te kunnen schetsen van lekkages. Voor reguliere lekkages (zogenaamde secunda gevarenbronnen 8) is bij hogere drukken dan 100 mbar te vinden dat een typische leko mm² zouden moeten worden geselecteerd (IGEM/HSE). Een kleinere lekopening van bi mm² kan wellicht ook interessant zijn om te beschouwen met het oog op de inhoud va norm voor explosieve gas atmosferen. Een lekopeningen van 0,025 mm2 geeft bij een v bar een lekgrootte van 40 liter/ uur. Lekgrootte metingen met hi flow samplers aan gas vergelijkbare waardes met een maximum van 18 liter/uur (Nederlandse metingen) en 3 (Duitse metingen). Een lekgrootte van 40 liter / uur is daarmee een realistische lekgroo In een gasdrukregelstation heerst een maximale druk van 8 barg waardoor bij een leko bandbreedte tussen 0,025 mm² en 0,25 mm² leidt tot lekdebieten van 0,18 m3n/h en 1, In eerder onderzoek is gekeken naar de ratio die toegepast zou moeten worden tussen waterstof [2] [13]. Afhankelijk van de druk, stroomopwaarts van de lekopening, is de st of turbulent. Tevens speelt de geometrie van een lekopening hierbij een rol. Door de b worden verondersteld dat in dit onderzoek een factor 3 tussen aardgas en waterstof ee is op basis van de resultaten uit de eerder aangehaalde onderzoeken. Wanneer dit wordt gedaan voor bovenstaande lekdebieten voor aardgas, volgt dat de waterstof tussen de 0,56 m3n/h en 5,6 m3n/h liggen. Lekopeningen en eerdere experimenten Wanneer bovenstaande getallen worden vertaald naar een aandachtsgebied voor lekkages, kunnen zowel voor aardgas als voor waterstof overzichten gemaakt worden. Wanneer hierbij tevens de informatie uit HyDelta 1.0 wordt toegevoegd, ontstaat een goed beeld van alle testen. Hierbij valt op te merken dat gekozen lekdebieten uit HyDelta 1.0 deels een overlap hebben met lekkages uit HyDelta 2.0. Met het oog op de praktische haalbaarheid van het instellen van het lekdebiet met behulp van een MFC zijn afrondingen en significante cijfers in deze tabellen voorbehouden. Wanneer de druk wordt uitgezet tegen het lekdebiet (flow), kunnen verschillende diagrammen opgesteld worden waarbij het “operating window” wordt gevisualiseerd. Op deze manier is voor zowel aardgas als waterstof duidelijk te zien welk operationele gebied wordt afgedekt voor de veronderstelde lekkages. Figuur 17. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor aardgas Figuur 18. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor waterstof Figuur 17. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor aardgas Figuur 18. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor waterstof Op basis van de gepresenteerde gegevens uit de verschillende onderzoeken rondom het gebied van methaanemissie en het eerder genoemde DVGW onderzoek valt op te maken dat de gevonden lekkages in gasdrukregelstations bij benadering overeenkomen met de kleinste lekkage (0,025 mm²) bij de laagste voorgestelde druk (1 barg). Daarmee kan veilig worden gesteld dat een reguliere lekkage zoals die voorkomt bij normale bedrijfsomstandigheden is meegenomen in dit testprogramma. Figuur 17. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor aardgas Figuur 17. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor aardgas Figuur 18. Gekozen lekdebieten HyDelta 1.0 versus HyDelta 2.0 voor waterstof Op basis van de gepresenteerde gegevens uit de verschillende onderzoeken rondom het gebied van methaanemissie en het eerder genoemde DVGW onderzoek valt op te maken dat de gevonden lekkages in gasdrukregelstations bij benadering overeenkomen met de kleinste lekkage (0,025 mm²) bij de laagste voorgestelde druk (1 barg). Daarmee kan veilig worden gesteld dat een reguliere lekkage zoals die voorkomt bij normale bedrijfsomstandigheden is meegenomen in dit testprogramma. Op basis van de gepresenteerde gegevens uit de verschillende onderzoeken rondom het gebied van methaanemissie en het eerder genoemde DVGW onderzoek valt op te maken dat de gevonden lekkages in gasdrukregelstations bij benadering overeenkomen met de kleinste lekkage (0,025 mm²) bij de laagste voorgestelde druk (1 barg). Bron en toelichting Pagina 36/123 8 een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden” (zie par 5.5.2 van de NPR7910-1 (2021)) 8 een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden” (zie par 5 5 2 van de NPR7910 1 (2021)) Pagina 36/123 Daarmee kan veilig worden gesteld dat een reguliere lekkage zoals die voorkomt bij normale bedrijfsomstandigheden is meegenomen in dit testprogramma. Op basis van de gepresenteerde gegevens uit de verschillende onderzoeken rondom het gebied van methaanemissie en het eerder genoemde DVGW onderzoek valt op te maken dat de gevonden lekkages in gasdrukregelstations bij benadering overeenkomen met de kleinste lekkage (0,025 mm²) bij de laagste voorgestelde druk (1 barg). Daarmee kan veilig worden gesteld dat een reguliere lekkage zoals die voorkomt bij normale bedrijfsomstandigheden is meegenomen in dit testprogramma. Pagina 37/123 WP6B – Veiligheid – Gasstations WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof ) afrondingen van getallen en significante cijfers voorbehouden ) afrondingen van getallen en significante cijfers voorbehouden 3.5 Waarom ATEX? Een behuizing moet voldoen aan normeringen die voorschrijven welke lekkages nog veilig kunnen worden gemitigeerd en welke ventilatie daarvoor minimaal vereist is. Hierbij wordt onderscheid gemaakt tussen continue, primaire en secundaire gevarenbronnen hetgeen een beeld geeft met betrekking tot de oorsprong van een lekkage. Vanuit de Europese richtlijn ATEX 153 zijn werkgevers wettelijk verplicht om een ATEX zonering aan te geven in gebieden waar explosieve stoffen aanwezig zijn. Het risico wordt vastgesteld op basis van een risico inventarisatie- en evaluatie. Hiermee wordt de link gelegd met de ATEX 114 richtlijn. De ATEX 114 goedgekeurde apparatuur is onderverdeeld in categorieën, die aangeeft in welke zones deze mag worden toegepast, zodat deze apparatuur een explosieve atmosfeer niet kan ontsteken. Bij de distributie van waterstof is de insteek dat de veiligheid vergelijkbaar is met aardgas Daarmee moet (de behuizing van) een gasstation voor waterstof ook minstens net zo veilig zijn als (de behuizing van) een gasstation voor aardgas. Om dit te bereiken moet worden zorggedragen dat het gebied dat vrij toegankelijk is voor het algemene publiek kan worden aangeduid als “Niet Gevaarlijk Gebied (NGG)”. Bij aardgas geldt dat binnen de behuizing van een gasstation sprake is van een ATEX zone 2 en dat direct buiten de behuizing het “Niet Gevaarlijk Gebied” begint. Om die reden wordt er normaal gesproken geen hekwerk geplaatst rondom een station voor de distributie van aardgas. Datzelfde zou moeten gelden voor stations met waterstof. Een behuizing (of: de behuizing inclusief een afgeschermd stuk grondoppervlakte) moet geclassificeerd kunnen worden als een ATEX zone 2. Dan is direct naast deze zone geen ATEX zonering noodzakelijk en mag dat gebied toegankelijk zijn voor het algemene publiek. In de NEN:1059, 2019 wordt gesteld dat bij een ventilatievoud van meer dan 5 keer per uur, binnen de behuizing ATEX zone 2 van toepassing is. Dan is er buiten de behuizing geen zonering van toepassing. In het geval van een kleine lekkage zou voldoende ventilatie aanwezig moeten zijn om het ontsnapte gas goed weg te kunnen ventileren. De basis van deze eis is reeds lang geleden opgesteld en gepresenteerd in een GAS artikel [3]. Hier is via een set van differentiaalvergelijkingen bepaald dat een onderlinge relatie tussen het volume van de behuizing, de ventilatievoud en het lekdebiet kan worden vastgesteld (zie appendix II). Bovenstaande informatie is vertaald naar een testmatrix die wordt gebruikt voor dit onderzoek met daarin de twee lekopeningen van 0,25 en 0,025 mm². Met twee drukken (1 bar en 8 bar) en aardgas/ waterstof levert dit 23 (= 8) mogelijke combinaties op. Deze zijn in onderstaande tabel samengevat. Tabel 5. Lekdebieten voor aardgas en waterstof. Lekdebiet 1 bij 8 bar voordruk, lekdebiet 2 bij 1 bar voordruk. Tabel 5. Lekdebieten voor aardgas en waterstof. Lekdebiet 1 bij 8 bar voordruk, lekdebiet 2 bij 1 bar voordruk. Lekdebiet 1 - 0,25 mm² Lekdebiet 2 - 0,25 mm² Lekdebiet 1 - 0,025 mm² Lekdebiet 2 - 0,025 mm² Aardgas (m3n/h) 1,8 0,40 0,18 0,04 Aardgas (g/s) * 0,42 0,09 0,04 0,01 Waterstof (m3n/h) * 5,6 1,25 0,56 0,125 Waterstof (g/s) * 0,14 0,03 0,014 0,003 ) afrondingen van getallen en significante cijfers voorbehouden ) afrondingen van getallen en significante cijfers voorbehouden Pagina 38/123 Deze Nederlandse praktijkrichtlijn geeft informatieve aanwijzingen bij het opstellen van een gevarenzone-indeling met betrekking tot gasexplosiegevaar. Deze praktijkrichtlijn dient als praktisch hulpmiddel ter nadere uitvoering van NEN-EN-IEC60079-10-1 (2021) en behoort in samenhang met deze norm gelezen te worden. Door een samenloop van omstandigheden kan een lekkage ontstaan in een gastechnische installatie. De lekkage kan heel klein zijn waardoor deze niet eens opgemerkt wordt. In de Nederlandse normering is hiervoor een stappenplan gemaakt die classificeert hoe hiermee om te gaan. Allereerst kan bij een zeer kleine lekkage (in combinatie met goede ventilatie) gekeken worden of een indeling als niet gevaarlijk gebied (NGG) mogelijk is. Als dit vanwege de uitstromende hoeveelheid gas in combinatie met de beschikbare ventilatie niet mogelijk is, moet een ATEX zonering worden toegepast. In bijlage III wordt aan de hand van de NPR-7910 een bepaling gedaan van de ATEX zonering voor een gasdrukregelstation. WP6B – D6B.1A betrekk distribu 3.6 ATEX zone bepaling volg Deze Nederlandse praktijkrichtlijn geeft gevarenzone-indeling met betrekking to hulpmiddel ter nadere uitvoering van N norm gelezen te worden. Door een samenloop van omstandighed lekkage kan heel klein zijn waardoor de hiervoor een stappenplan gemaakt die kleine lekkage (in combinatie met goed gebied (NGG) mogelijk is. Als dit vanwe beschikbare ventilatie niet mogelijk is, m In bijlage III wordt aan de hand van de N gasdrukregelstation. 3.7 ATEX zone bepaling volg Zoals eerder benoemd is de NPR-7910- EN-IEC-60079-10-1 (2021)(Explosive atm atmospheres) die in detail beschrijft we uitvoeren van een analyse inzake explo waar mogelijk, gecomplementeerd wor beschikbaar is. In appendix C en D van deze internation analyse moet worden uitgevoerd. Een t ✓ Bepaal het type van de lekbron ✓ Bepaal het lekdebiet van de lek ✓ Bepaal de mate van verdunning ✓ Bepaal de bijbehorende ATEX z beschikbaarheid van ventilatie Toch heeft deze werkwijze ook limiterin Het gaat in dit onderzoek om zeer klein gebruikte grafieken voor het bepalen va influence” geëxtrapoleerd moeten wor In bijlage III wordt aan de hand van de N voor een gasdrukregelstation. ✓ Bepaal de mate van verdunning aan de hand van ✓ Bepaal de bijbehorende ATEX zonering aan de ha beschikbaarheid van ventilatie (betrouwbaarheid Toch heeft deze werkwijze ook limiteringen, zeker voor d Het gaat in dit onderzoek om zeer kleine lekkages vanuit gebruikte grafieken voor het bepalen van de mate van ve influence” geëxtrapoleerd moeten worden WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering betrekking tot ventilatie in verschillende typen g distributie (lagedruk)net in Nederland met aard 3.6 ATEX zone bepaling volgens NPR-7910-1 Deze Nederlandse praktijkrichtlijn geeft informatieve aanwijzingen bij het opste gevarenzone-indeling met betrekking tot gasexplosiegevaar. Deze praktijkrichtl hulpmiddel ter nadere uitvoering van NEN-EN-IEC60079-10-1 (2021) en behoor norm gelezen te worden. Door een samenloop van omstandigheden kan een lekkage ontstaan in een gas lekkage kan heel klein zijn waardoor deze niet eens opgemerkt wordt. In de Ned hiervoor een stappenplan gemaakt die classificeert hoe hiermee om te gaan. Al kleine lekkage (in combinatie met goede ventilatie) gekeken worden of een ind gebied (NGG) mogelijk is. Als dit vanwege de uitstromende hoeveelheid gas in c beschikbare ventilatie niet mogelijk is, moet een ATEX zonering worden toegep In bijlage III wordt aan de hand van de NPR-7910 een bepaling gedaan van de A gasdrukregelstation. 3.7 ATEX zone bepaling volgens de NEN-EN-IEC60079-10-1 Zoals eerder benoemd is de NPR-7910-1 (2021) een praktisch hulpmiddel voor EN-IEC-60079-10-1 (2021)(Explosive atmospheres - Part 10-1: Classification of a atmospheres) die in detail beschrijft welke specifieke berekeningen gemaakt ku uitvoeren van een analyse inzake explosieveiligheid. Hierbij worden berekening waar mogelijk, gecomplementeerd worden met praktische ervaring (of meetwa beschikbaar is. In appendix C en D van deze internationale standaard staat in stappen beschrev analyse moet worden uitgevoerd. Een typische analyse bestaat uit de volgende ✓ Bepaal het type van de lekbron (continue/ primair/ secundair). ✓ Bepaal het lekdebiet van de lekbron. ✓ Bepaal de mate van verdunning aan de hand van lekdebiet en ventilatie ✓ Bepaal de bijbehorende ATEX zonering aan de hand de mate van de ma beschikbaarheid van ventilatie (betrouwbaarheid) en het type van de le Toch heeft deze werkwijze ook limiteringen, zeker voor de situatie die beschouw Het gaat in dit onderzoek om zeer kleine lekkages vanuit secundaire gevarenbro gebruikte grafieken voor het bepalen van de mate van verdunning en het bepal influence” geëxtrapoleerd moeten worden. In bijlage III wordt aan de hand van de NEN-EN-IEC60079 een bepaling gedaan v voor een gasdrukregelstation. De basis van bovenstaande normering ligt in het technisch dicht verklaren van de installatie volgens de volgens de NEN-EN 1127 Annex B (B.3 – “Enhanced tightness”) of de NPR-7910-1 (2021). Echter, omdat hele kleine lekkages niet meteen opgemerkt worden, is het raadzaam om het binnenste van de kast als ATEX zonering 2 in te delen waarbij de zone niet buiten de kast komt. Zo hoeft er buiten de behuizing niet gezoneerd te worden. Uit onderzoek weten we van reguliere lekkages dat ze klein zijn en door de omgeving niet altijd opgemerkt worden. Een realistische lekopening voor een reguliere lekkage wordt echter in referenties niet expliciet benoemd. In paragraaf 3.2 wordt beschreven dat op basis van recent onderzoek is vastgesteld dat zo’n lekkage zelden tot nooit groter is dan 40 liter per uur [9]. Op basis van paragraaf 3.3 is richting gegeven aan de grootte van een dergelijke lekkage en deze kan dienen als startpunt voor het maken van een ATEX berekening en het vaststellen van de daarbij behorende ATEX zonering. Pagina 39/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in he distributie (lagedruk)net in Nederland met aardgas en waterstof 3.6 ATEX zone bepaling volgens NPR-7910-1 Deze Nederlandse praktijkrichtlijn geeft informatieve aanwijzingen bij het opstellen van een gevarenzone-indeling met betrekking tot gasexplosiegevaar. Deze praktijkrichtlijn dient als praktisch hulpmiddel ter nadere uitvoering van NEN-EN-IEC60079-10-1 (2021) en behoort in samenhang met deze norm gelezen te worden. Door een samenloop van omstandigheden kan een lekkage ontstaan in een gastechnische installatie. De lekkage kan heel klein zijn waardoor deze niet eens opgemerkt wordt. In de Nederlandse normering is hiervoor een stappenplan gemaakt die classificeert hoe hiermee om te gaan. Allereerst kan bij een zeer kleine lekkage (in combinatie met goede ventilatie) gekeken worden of een indeling als niet gevaarlijk gebied (NGG) mogelijk is. Als dit vanwege de uitstromende hoeveelheid gas in combinatie met de beschikbare ventilatie niet mogelijk is, moet een ATEX zonering worden toegepast. In bijlage III wordt aan de hand van de NPR-7910 een bepaling gedaan van de ATEX zonering voor een gasdrukregelstation. 3.7 ATEX zone bepaling volgens de NEN-EN-IEC60079-10-1 Zoals eerder benoemd is de NPR-7910-1 (2021) een praktisch hulpmiddel voor het gebruik van de NEN- EN-IEC-60079-10-1 (2021)(Explosive atmospheres - Part 10-1: Classification of areas - explosive gas atmospheres) die in detail beschrijft welke specifieke berekeningen gemaakt kunnen worden bij het uitvoeren van een analyse inzake explosieveiligheid. Hierbij worden berekeningen gepresenteerd die, waar mogelijk, gecomplementeerd worden met praktische ervaring (of meetwaarden) als deze beschikbaar is. In appendix C en D van deze internationale standaard staat in stappen beschreven hoe een dergelijke analyse moet worden uitgevoerd. Een typische analyse bestaat uit de volgende stappen; ✓ Bepaal het type van de lekbron (continue/ primair/ secundair). ✓ Bepaal het lekdebiet van de lekbron. ✓ Bepaal de mate van verdunning aan de hand van lekdebiet en ventilatie. ✓ Bepaal de bijbehorende ATEX zonering aan de hand de mate van de mate van verdunning, de beschikbaarheid van ventilatie (betrouwbaarheid) en het type van de lekbron. Toch heeft deze werkwijze ook limiteringen, zeker voor de situatie die beschouwd wordt in dit rapport. Het gaat in dit onderzoek om zeer kleine lekkages vanuit secundaire gevarenbronnen waardoor de gebruikte grafieken voor het bepalen van de mate van verdunning en het bepalen van de “circle of WP6B – Veiligheid – Gass D6B.1A & D6B.1B – Inven betrekking tot ventilatie distributie (lagedruk)net 3.6 ATEX zone bepaling volgens NPR-7910 Deze Nederlandse praktijkrichtlijn geeft informatieve aan gevarenzone-indeling met betrekking tot gasexplosiegev hulpmiddel ter nadere uitvoering van NEN-EN-IEC60079- norm gelezen te worden. Door een samenloop van omstandigheden kan een lekka lekkage kan heel klein zijn waardoor deze niet eens opge hiervoor een stappenplan gemaakt die classificeert hoe h kleine lekkage (in combinatie met goede ventilatie) geke gebied (NGG) mogelijk is. Als dit vanwege de uitstromen beschikbare ventilatie niet mogelijk is, moet een ATEX zo In bijlage III wordt aan de hand van de NPR-7910 een be gasdrukregelstation. 3.7 ATEX zone bepaling volgens de NEN-E Zoals eerder benoemd is de NPR-7910-1 (2021) een prak EN-IEC-60079-10-1 (2021)(Explosive atmospheres - Part atmospheres) die in detail beschrijft welke specifieke be uitvoeren van een analyse inzake explosieveiligheid. Hier waar mogelijk, gecomplementeerd worden met praktisc beschikbaar is. In appendix C en D van deze internationale standaard sta analyse moet worden uitgevoerd. Een typische analyse b ✓ Bepaal het type van de lekbron (continue/ prima ✓ Bepaal het lekdebiet van de lekbron. In hoofdstuk 1 is de volgende deelvraag gesteld: Deelvraag 2: Aan welke voorwaarden moet een behuizing van een gasstation voldoen voordat deze geschikt is voor de distributie van waterstof? De distributie van waterstof moet minstens net zo veilig zijn als de distributie van aardgas. Daarmee moet (de behuizing van) een gasstation voor waterstof ook minstens net zo veilig zijn als (de behuizing van) een gasstation voor aardgas. Uiteindelijk moet aan deze voorwaarde worden voldaan. Om dat te bereiken, moet aangetoond kunnen worden dat de behuizing (of: de behuizing inclusief een afgeschermd stuk grondoppervlakte) geclassificeerd kan worden als een ATEX zone 2. Alleen dan is direct naast deze zone geen ATEX zonering noodzakelijk en mag dat gebied toegankelijk zijn voor het algemene publiek. Om een gebied te mogen classificeren als ATEX zone 2, moet aangetoond kunnen worden dat sprake is van een secundaire gevarenbron en dat er voldoende ventilatie aanwezig is. Om een gebied te mogen classificeren als ATEX zone 2, moet aangetoond kunnen worden dat sprake is van een secundaire gevarenbron en dat er voldoende ventilatie aanwezig is. Een secundaire gevarenbron betekent dat er in minder dan 0,1% van de tijd sprake mag zijn van een ontbrandbaar mengsel. Om dat aan te kunnen tonen, is kennis nodig van de in de praktijk optredende lekken. Er mogen lekken voor komen in gasstations die leiden tot een concentratie boven de 100% LEL, zolang deze maar zelden (< 0,1% van de tijd) voor komen. Door een gedegen onderhoudsregime moet door de eigenaar immers zorg worden gedragen voor een technisch dichte installatie. De NPR7910-1 stelt: “In het geval van een kleine lekkage zou voldoende ventilatie aanwezig moeten zijn om het ontsnapte gas goed weg te kunnen ventileren”. Een praktische interpretatie hiervan is dat de concentratie onder de LEL/LFL moet blijven. Tevens dient opgemerkt te worden dat wanneer aan de hand van de NPR 7910-1 een analyse wordt gedaan voor een behuizing, hierbij enkel een eis wordt neergelegd waaraan de behuizing zou moeten voldoen. Daarmee doet de NPR7910-1 geen uitspraak of de gebouwde geometrie van een behuizing daadwerkelijk geschikt is om adequaat te kunnen ventileren. De geldende normen geven diverse aanknopingspunten voor een definitie van het begrip “kleine lekkage”. De NEN 1059: 2019 noemt een lekopening van 1 mm². De IGEM/SR/25 noemt een lekopening van 0,25 mm² (P > 100mbar) en 0,025 mm² (P < 100 mbar). De NEN-EN-IEC-60079-10-1 (2021) noemt een lekopening in de range van 0,025 mm² t/m 0,25 mm². 3.7 ATEX zone bepaling volgens de NEN-EN-IEC60079-10-1 Zoals eerder benoemd is de NPR-7910-1 (2021) een praktisch hulpmiddel voor het gebruik van de NEN- EN-IEC-60079-10-1 (2021)(Explosive atmospheres - Part 10-1: Classification of areas - explosive gas atmospheres) die in detail beschrijft welke specifieke berekeningen gemaakt kunnen worden bij het uitvoeren van een analyse inzake explosieveiligheid. Hierbij worden berekeningen gepresenteerd die, waar mogelijk, gecomplementeerd worden met praktische ervaring (of meetwaarden) als deze beschikbaar is. In appendix C en D van deze internationale standaard staat in stappen beschreven hoe een dergelijke analyse moet worden uitgevoerd. Een typische analyse bestaat uit de volgende stappen; ✓ Bepaal het type van de lekbron (continue/ primair/ secundair). ✓ Bepaal het lekdebiet van de lekbron. ✓ Bepaal de mate van verdunning aan de hand van lekdebiet en ventilatie. ✓ Bepaal de bijbehorende ATEX zonering aan de hand de mate van de mate van verdunning, de beschikbaarheid van ventilatie (betrouwbaarheid) en het type van de lekbron. Toch heeft deze werkwijze ook limiteringen, zeker voor de situatie die beschouwd wordt in dit rapport. Het gaat in dit onderzoek om zeer kleine lekkages vanuit secundaire gevarenbronnen waardoor de gebruikte grafieken voor het bepalen van de mate van verdunning en het bepalen van de “circle of influence” geëxtrapoleerd moeten worden. In bijlage III wordt aan de hand van de NEN-EN-IEC60079 een bepaling gedaan van de ATEX zonering voor een gasdrukregelstation. Pagina 40/123 In hoofdstuk 1 is de volgende deelvraag gesteld: In hoofdstuk 1 is de volgende deelvraag gesteld: Op het moment dat berekeningen worden uitgevoerd aan lekken van deze omvang, wordt duidelijk dat ATEX zone 2 niet in alle gevallen gehaald wordt, zoals blijkt uit paragraaf 3.7 en appendix III/IV. Er zit echt een verschil tussen de lekgroottes zoals die in normen worden genoemd en de lekgroottes die daadwerkelijk in praktijk zijn gemeten. Praktijkmetingen tonen aan dat bij de ventilatieopening van districtstations geen gasconcentraties boven 25% LEL/LFL voorkomen. Uit deze metingen blijkt ook dat lekkages groter dan 40 l/u niet zijn waargenomen bij gasstations op aardgas. Wanneer dit debiet (aan de veilige kant) wordt omgerekend naar waterstof, is een reëel maximumwaarde voor een lekdebiet 125 l/u waterstof. Om zo veel mogelijk inzicht te verkrijgen zijn experimenten uitgevoerd bij zowel lekgroottes zoals die in de normen zijn aangedragen als lekgroottes zoals die in de praktijk optreden. Deze staan samengevat in tabel 5 van paragraaf 3.4. Pagina 41/123 4.1 Testmethode op hoofdlijnen 4.1 Testmethode op hoofdlijnen Zoals al eerder benoemd, kan dit onderzoek het beste gezien worden als een doorontwikkeling van HyDelta 1.0. Daarin wordt aanbevolen om de gevolgen van kleinere, meer in de praktijk voorkomende, lekken goed in kaart te brengen. Hiervoor is in de basis dezelfde testopstelling gebruikt, waardoor de beschrijving ook is overgenomen uit de rapportage van HyDelta 1.0. De testopstelling bestaat uit: • Gasflessen met waterstof, aardgas en stikstof • Een gasdrukregelaar en een mass flow controller (MFC). • Een standaard kast (1/2m3) met alleen bovenventilatie (~ 2%), geleverd door Enexis. E d d k i (4 3) ll b il i ( 4%) l d d E • Een standaard Hogedrukafleverstation (HAS)(< 0,1 m3) met alleen bovenventilatie (~ 4%), geleverd door Enexis. • Een leiding zodat een lek met een uitstroomopening van respectievelijk 0,25 en 0,025 mm² in de behuizing wordt aangebracht. Het lek wordt gepositioneerd in het midden van de kast. De uitstroomopening heeft tevens de mogelijkheid tot het aansluiten van een naaldafsluiter. • Een leiding zodat een lek met een uitstroomopening van respectievelijk 0,25 en 0,025 mm² in de behuizing wordt aangebracht. Het lek wordt gepositioneerd in het midden van de kast. De uitstroomopening heeft tevens de mogelijkheid tot het aansluiten van een naaldafsluiter. Figuur 19. Schematische weergave testopstelling Gasfles H2 MFC Gasstation Figuur 19. Schematische weergave testopstelling 4. Plan van aanpak 4. Plan van aanpak • Direct buiten de kast op 4 punten (punten 5, 6, 7 en 8), alle zijden op 2 cm afstand van de ventilatieopeningen in het midden. Hiervoor worden de Riken Keiki sensoren gebruikt, 0 – 100 vol% waterstof. Bij de experimenten met aardgas worden de MultiRae sensoren gebruikt. (0-100% LEL/LFL en 0-100vol%). • Direct buiten de kast op 4 punten (punten 5, 6, 7 en 8), alle zijden op 2 cm afstand van de ventilatieopeningen in het midden. Hiervoor worden de Riken Keiki sensoren gebruikt, 0 – 100 vol% waterstof. Bij de experimenten met aardgas worden de MultiRae sensoren gebruikt. (0-100% LEL/LFL en 0-100vol%). • Tevens zal de aardgas of waterstof concentratie op 0,5 meter van de behuizing gemeten worden (punten 9 en 10). De sensoren (allen MultiRae’s) zijn op 1 meter boven grondniveau geplaatst. Deze meters bevatten zowel een ppm-sensor als een LEL/LFL-sensor voor waterstof. • Tevens zal de aardgas of waterstof concentratie op 0,5 meter van de behuizing gemeten worden (punten 9 en 10). De sensoren (allen MultiRae’s) zijn op 1 meter boven grondniveau geplaatst. Deze meters bevatten zowel een ppm-sensor als een LEL/LFL-sensor voor waterstof. 2 Meetpunten en meetapparatuur De aardgas of waterstof concentratie wordt op de volgende punten gemeten: g p g p g • In de kast op 4 punten (onder, midden en 2x boven). Het meetpunt “midden” (2), zit 10 cm verwijderd van de uitstroomopening. De meetpunten “onder” (1) en “boven” (3) zitten recht onder en boven het meetpunt “midden”, op 5 cm afstand van maaiveld dan wel de bovenkant van de kast (voor de 1/2m3 kast). Het meetpunt “boven” (4) zit in het kaststation ter hoogte van de daklijn, hetgeen niet het hoogste punt betreft. Voor andere geteste behuizingen, is eenzelfde werkwijze gehanteerd. Schematisch ziet de opstelling van de meetpunten er als volgt uit; • In de kast op 4 punten (onder, midden en 2x boven). Het meetpunt “midden” (2), zit 10 cm verwijderd van de uitstroomopening. De meetpunten “onder” (1) en “boven” (3) zitten recht onder en boven het meetpunt “midden”, op 5 cm afstand van maaiveld dan wel de bovenkant van de kast (voor de 1/2m3 kast). Het meetpunt “boven” (4) zit in het kaststation ter hoogte van de daklijn, hetgeen niet het hoogste punt betreft. Voor andere geteste behuizingen, is eenzelfde werkwijze gehanteerd. Schematisch ziet de opstelling van de meetpunten er als volgt uit; Figuur 20. Setup van de meetpunten in de testopstelling (1/2m3 kast), zijaanzicht (links) en bovenaanzicht (rechts) Figuur 20. Setup van de meetpunten in de testopstelling (1/2m3 kast), zijaanzicht (links) en bovenaanzicht (rechts) Pagina 42/123 Pagina 42/123 Aan de hand van windrozen valt op te maken dat windkracht 1 tot 5 regelmatig voorkomt. Als voorbeeld zijn voor weerstation “de Bilt” een aantal windrozen toegevoegd. Figuur 21. Windrozen gedurende de seizoenen in de bilt op basis van langjarig gemiddelden Tijdens alle experimenten is de windsnelheid bij de behuizing gemeten om te verifiëren. In de basis is geprobeerd om alle metingen uit te voeren op windluwe dagen. Wanneer het vermoeden bestaat dat de experimenten zijn beïnvloed door de wind, is dit benoemd in de rapportage. Figuur 21. Windrozen gedurende de seizoenen in de bilt op basis van langjarig gemiddelden Tijdens alle experimenten is de windsnelheid bij de behuizing gemeten om te verifiëren. In de basis is geprobeerd om alle metingen uit te voeren op windluwe dagen. Wanneer het vermoeden bestaat dat de experimenten zijn beïnvloed door de wind, is dit benoemd in de rapportage. Tijdens alle experimenten is de windsnelheid bij de behuizing gemeten om te verifiëren. In de basis is geprobeerd om alle metingen uit te voeren op windluwe dagen. Wanneer het vermoeden bestaat dat de experimenten zijn beïnvloed door de wind, is dit benoemd in de rapportage. Tijdens alle experimenten is de windsnelheid bij de behuizing gemeten om te verifiëren. In de basis is geprobeerd om alle metingen uit te voeren op windluwe dagen. Wanneer het vermoeden bestaat dat de experimenten zijn beïnvloed door de wind, is dit benoemd in de rapportage. 4.3 Wind en ventilatie De wind kan een grote invloed hebben op de experimenten. Tijdens HyDelta 1.0 zijn daarom twee situaties gecreëerd, namelijk een binnensituatie (windstille situatie) en een buitensituatie. Om een windstille situatie te simuleren, wordt het kaststation in een grote tent geplaatst. Daarnaast worden dezelfde serie metingen uitgevoerd op een dag die, naar verwachting, een constante windkracht (van windkracht 2 of 4) heeft om zo de invloed ten gevolge van natuurlijke trek te beperken. Tijdens HyDelta 2.0 is gekozen om alle experimenten zoveel mogelijk uit te voeren in een windluwe situatie waarbij een grote tent is gebruikt. Wanneer door de grootte van de behuizing plaatsing in een tent niet mogelijk was, is met behulp van blinderende dranghekken een windluwe situatie gecreëerd. Ter controle of er daadwerkelijke sprake was van een windluwe (dus niet per definitie: windstille) situatie, is tijdens de metingen gebruik gemaakt van een indicatieve windmeter. Dit betreft een anemometer van merk en Type Skywatch OELE, met een minimaal meetbereik van 0,6 m/s. In de NEN1059:2019 is een ventilatievoud van 5 of groter gesteld voor behuizingen van gasdrukregelstations. Echter is de ventilatievoud afhankelijk van de windsnelheid. In de NPR7910-1 (2021) en in de NEN-EN-60079-10-1 (2021) worden windsnelheden gesteld voor het berekenen van ventilatie. In de NPR7910-1 (2021) wordt gesteld dat de luchtverversing waarbij zonder mechanische hulpmiddelen de luchtsnelheid meestal hoger is dan 2 m/s en zelden lager dan 0,5 m/s. Tevens staat in de NPR7910-1 (2021) beschreven dat voor het berekenen van de ventilatieopeningen een windsnelheid van 0,5 m/s als richtlijn kan worden aangenomen. In de NEN-EN-60079-10-1 (2021) is een tabel opgenomen waarin windsnelheden voor verschillende situaties benoemd zijn. Ook deze waarden kunnen gebruikt worden voor een analyse van de ventilatievoud en komen overeen met de NPR7910-1 (2021). Behuizingen voor gasdrukregelstations staan overal in Nederland. Het KNMI heeft richtlijnen opgesteld voor de gemiddelde windsnelheid. Aan de hand van langjarige gemiddelden (1991-2020) kan inzichtelijk gemaakt worden hoe de wind gedurende de maanden van het jaar varieert. Deze metingen zijn echter genormaliseerd voor een hoogte van 10 meter boven het maaiveld om te corrigeren voor globale terreinruwheid. De bruikbaarheid van deze gegevens voor de hoogte van de ventilatie van een gasstation (0,5 meter tot 2 meter boven maaiveld) is zeer beperkt. Voor deze hoogte zijn helaas geen betrouwbare datasets beschikbaar. Pagina 43/123 4.4 Foto’s van de opstellingen De stations zijn dus zowel in een buitensituatie als in een binnen-/ windluwe situatie getest. Stations staan in de praktijk altijd buiten. Echter, het is de verwachting dat natuurlijke ventilatie op windstille en/ of windluwe momenten minder goed zal functioneren dan wanneer er sprake is van natuurlijk ventilatie door wind. Om windstille en/ of windluwe situaties na te bootsen, is zowel de 1/2m3 kast als het 4m3 kaststation in een tent geplaatst. Figuur 22. De ½ m³ kast in een windluwe situatie Figuur 23. Het 4m³ kaststation in een windluwe situatie Figuur 23. Het 4m³ kaststation in een windluwe situatie Figuur 22. De ½ m³ kast in een windluwe situatie Figuur 23. Het 4m³ kaststation in een windluwe situatie Pagina 44/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 5.1 Lekdebiet 0,18 m3n/hr - aardgas Bij de eerste test is de kleinste lekopening conform de IGEM/SR/25 Edition 2 gecreëerd bij een voordruk van 8 bar in de ½ m3 behuizing waarbij de concentratie oploopt tot maximaal 6,0 vol%. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 24. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatieopeningen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Figuur 24. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Figuur 24. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Figuur 24. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatieopeningen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatieopeningen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatieopeningen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 6,0vol% 6,0vol% 2,0vol% 0,1vol% Figuur 25. Concentratie bij de ventilatie openingen (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd incidenteel 100 ppm bereikt. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. guur 25. Concentratie bij de ventilatie openingen (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. 5. Bespreking resultaten van bestaande behuizingen In de twee onderstaande paragrafen wordt één meting voor aardgas en één meting voor waterstof gepresenteerd. Beide testen zijn uitgevoerd in de 1/2m3 kast bij de hoogste voordruk in een windluwe situatie (met de kleinste lekopening). Deze paragrafen dienen als voorbeeld hoe alle resultaten uit tabel 5 verwerkt zijn. Deze resultaten zijn toegevoegd in bijlage VII tot en met X van dit rapport. Vanuit een gasfles is respectievelijk aardgas (L-gas) en waterstof met een mass flow controller (MFC) gecontroleerd toegevoegd in de behuizing. Dit gas stroomt uit in het midden van de kast vanuit een lekopening met een oppervlakte van respectievelijk 0,25 en 0,025 mm2. In de kast zijn meetpunten aangebracht, één op 5 cm van de bodem, één in het midden, één op 5 cm van de daklijn en één in de nabijheid van een ventilatieopening. Daarnaast zijn op alle vier de zijden van de kast sensoren geplaatst die buiten de kast vlak onder de ventilatieopening meten. Tenslotte zijn sensoren geplaatst die op 0,5 meter afstand meten. Deze sensoren zijn op 1 meter boven de grond opgehangen zoals te zien op bovenstaande foto’s. Op die manier wordt het verspreidingsgedrag van het gaslek in kaart gebracht. Additionele informatie over de testopstelling kan gevonden worden in appendix IV. Pagina 45/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,56 m3n/hr - waterstof Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 6,9 vol% in de behuizing. Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 6,9 vol% in de behuizing. De concentratie op alle meetpunten in de behuizing stabiliseert na ongeveer 10 minuten en blijft daarna nagenoeg gelijk tijdens de hele test. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Figuur 26. Concentratie (vol % waterstof) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 27. Concentratie bij de ventilatieopeningen (vol% waterstof) in de ½ m3 kast bij een lek van 0,025mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Figuur 26. Concentratie (vol % waterstof) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Figuur 26. Concentratie (vol % waterstof) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 27. Concentratie bij de ventilatieopeningen (vol% waterstof) in de ½ m3 kast bij een lek van 0,025mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. 5.1 Lekdebiet 0,18 m3n/hr - aardgas Deze hebben een meetbereik tot 1000 ppm en werd incidenteel 100 ppm bereikt. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Pagina 46/123 Pagina 46/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 5.2 Lekdebiet 0,56 m3n/hr - waterstof Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 6,9 vol% in de behuizing. De concentratie op alle meetpunten in de behuizing stabiliseert na ongeveer 10 minuten en blijft daarna WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 5.3 De resultaten van het testprogramma voor de verschillende behuizingen (de 1/2m³ kast, het 4m³ kaststation en de HAS kast) zijn verzameld in bijlage VII tot en met X. De belangrijkste gegevens van al deze testen zijn kort samengevat in onderstaande tabellen. De waarden in rood geven aan wanneer de maximale concentratiemeting binnen de brandbaarheidsgrenzen van het gas-luchtmengsel valt. Wanneer een veld in onderstaande tabellen niet is ingevuld, betekent dit dat er geen waarde is gemeten. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 27. Concentratie bij de ventilatieopeningen (vol% waterstof) in de ½ m3 kast bij een lek van 0,025mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. guur 27. Concentratie bij de ventilatieopeningen (vol% waterstof) in de ½ m3 kast bij een lek van 0,025mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Pagina 47/123 Pagina 47/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 1/2m³ kast Tabel 2 – 1/2m3 kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. ) lekkage levert aardgas concentraties op boven de UEL/ UFL maar omdat altijd ergens een overgang in concentratie zal zijn die binnen de brandbaarheidsgrenzen van het gas valt, zijn deze waarden rood weergegeven. Aardgas/ natural gas 1/2m3 kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,179 8 6,0% 6,0% 6,0% 2,0% 0,1% - 100 0,04 1 1,0% 1,0% 0,5% 0,1% 0,4% - - Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 1,8 8 26,6% 21,5% 25,1% 2,7% 23,1% 7,0% >1000 0,4 1 11,9% 11,3% 11,6% 4,0% 10,8% - - Waterstof/ hydrogen 1/2m3 kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,558 8 6,9% 6,3% 6,4% 4,5% 6,4% - 580 0,125 1 2,7% 2,9% 2,6% 1,4% 3,2% - 260 Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 5,8 8 24,0% 20,0% 20,0% 18,0% 21,0% 18,0% >1000 1,25 1 11,0% 10,0% 11,0% 6,5% 11,0% - 580 Tabel 2 – 1/2m3 kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Tabel 2 – 1/2m3 kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. ) lekkage levert aardgas concentraties op boven de UEL/ UFL maar omdat altijd ergens een overgang in concentratie zal zijn die binnen de brandbaarheidsgrenzen van het gas valt, zijn deze waarden rood weergegeven. Pagina 48/123 Pagina 48/123 WP6B – Veiligheid – Gasstations 4m³ kaststation Tabel 3 – 4m3 kaststation resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Tabel 3 – 4m3 kaststation resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Aardgas/ natural gas 4m3 kaststation Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,179 8 0,8% 0,2% 0,7% 0,4% 0,4% - - 0,04 1 0,1% 0,0% 0,0% 0,0% 0,1% - - Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 1,8 8 4,3% 2,6% 1,0% 3,7% 3,7% - - 0,4 1 1,1% 0,9% 0,3% 0,5% 1,8% - - Waterstof/ hydrogen 4m3 kaststation Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,558 8 2,2% 0,9% 1,6% 1,1% 1,3% - 70 0,125 1 1,0% 0,3% 0,4% 0,4% 0,4% - 10 Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 5,8 8 12,0% 9,1% 4,0% 9,3% 6,0% - 570 1,25 1 5,5% 4,1% 4,0% 3,6% 4,1% - 120 Pagina 49/123 Pagina 49/123 WP6B – Veiligheid – Gasstations In paragraaf 3.4 wordt beschreven dat de afleiding van de ventilatievoud tot stand is gekomen via een set van differentiaalvergelijkingen die een onderlinge relatie beschrijven tussen het volume van de behuizing, de ventilatievoud en het lekdebiet [3]. Het resultaat van deze berekening is de gasconcentratie die behaald kan worden in een behuizing met deze specifieke randvoorwaarden. Omdat de oorsprong van deze informatie erg belangrijk wordt geacht voor het beschouwen van het onderwerp ventilatie, is een deel van het artikel toegevoegd in appendix II. Hiermee is namelijk de ventilatievoud van meer dan 5 (zoals gebruikt in de NPR-7910-1 (2021)) vastgesteld. Deze redenatie kan ook worden omgedraaid. Aan de hand van het volume van de behuizing, het lekdebiet en de gestabiliseerde eindconcentratie van een experiment kan worden vastgesteld wat de ventilatievoud van een specifieke geometrie is. In onderstaande grafieken is respectievelijk voor 1 mm², 0,25 mm² en voor 0,025 mm² de gemeten concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Hierbij zijn de verschillende meetpunten kast laag (KL), kast midden (KM), kast hoog (KH) en kast hoog ventilatie (KHV) uitgezet. Daarnaast is met behulp van de theorie een curve berekend waarbij de gemiddelde gasconcentratie wordt berekend. De gemiddelde gasconcentratie van alle metingen (kast laag (KL), kast midden (KM), kast hoog (KH) en kast hoog ventilatie (KHV)) is tevens uitgezet. Figuur 28. Aardgasconcentratie als functie van tijd (½ m3 kast) bij een lek van 1 mm², 0,25 mm² en 0,025mm2 bij 8 bar guur 28. Aardgasconcentratie als functie van tijd (½ m3 kast) bij een lek van 1 mm², 0,25 mm² en 0,025mm2 bij 8 bar Figuur 28. De inhoud van de HAS kast bedraagt circa 0,06 m³. De inhoud van de HAS kast bedraagt circa 0,06 m³. Tabel 4 – HAS kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Tabel 4 – HAS kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Aardgas/ natural gas HAS kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% Links Rechts %LFL ppm 0,179 8 10,3% 9,1% 7,9% 0,0% 0 0,04 1 3,3% 3,9% 2,7% 0,0% 0 Waterstof/ hydrogen HAS kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% Links Rechts %LFL ppm 0,558 8 12,0% 12,0% 12,0% 0,0% 410 0,125 1 4,8% 5,6% 4,1% 0,0% 20 * experiment voortijdig gestopt (ivm veiligheid) door snelle opbouw van waterstof concentratie. Pagina 50/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 5 4 D d k lijk til ti d ( f l id t d ) Aardgasconcentratie als functie van tijd (½ m3 kast) bij een lek van 1 mm², 0,25 mm² en 0 anneer deze drie metingen worden beschouwd voor aardgas bij 8 bar (bij lekopeningen van spectievelijk 1 mm² 0 25 mm² en 0 025 mm² ) in een 1/2m3 behuizing dan kunnen we op bas Wanneer deze drie metingen worden beschouwd voor aardgas bij 8 bar (bij lekopeningen van respectievelijk 1 mm², 0,25 mm² en 0,025 mm² ) in een 1/2m3 behuizing, dan kunnen we op basis van de metingen de volgende gegevens afleiden; Wanneer deze drie metingen worden beschouwd voor aardgas bij 8 bar (bij lekopeningen van respectievelijk 1 mm², 0,25 mm² en 0,025 mm² ) in een 1/2m3 behuizing, dan kunnen we op basis van de metingen de volgende gegevens afleiden; Tabel 5 – Lekopening, lekdebiet en maximale aardgasconcentratie Tabel 5 Lekopening, lekdebiet en maximale aardgasconcentratie Lekopening [mm²] Lekdebiet (m3n/h) Maximale concentratie (vol%)* Ventilatievoud [-] 1 7,5 45,9 25 0,25 1,8 26,6 14 0,025 0,18 6,0 12 ) maximale aardgasconcentratie gemeten in de behuizing De ventilatievoud is dus afhankelijk van het lekdebiet, de geometrie van de behuizing, het volume van de behuizing en de omstandigheden (wind/temperatuur) van de omgeving. Eenzelfde soort conclusie is vastgesteld aan de hand van eerdere literatuur [3]. Voor deze specifieke geometrie geldt dat de vastgestelde ventilatievoud altijd groter is dan de vereiste ventilatievoud van 5 (en dus op papier beschikt over voldoende capaciteit) waarmee kan worden gesteld dat op basis van de werkwijze in de NPR7910-1 (2021), tabel 7 een gevarenzone 2 van toepassing mag zijn. Echter kan ook gesteld worden dat bij een ventilatievoud van meer dan 5 nog steeds een aanzienlijke gasconcentratie (van meer dan 25% LEL/LFL) kan ontstaan in de behuizing. Het stellen van de gewenste ventilatievoud is hiermee geen dekkend middel voor het garanderen van voldoende ventilatie bij deze specifieke behuizing. Pagina 51/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 6.1 Aanleiding In de hoofdstuk 5 is zijn de resultaten besproken van testen met verschillende lekkage aan standaard behuizingen met onaangepaste ventilatie. Voor de 1/2m³ behuizing is enkel bovenventilatie toegepast. Hetzelfde geldt voor het 4m³ kaststation en de HAS kast. In de NEN1059:2019 wordt in tabel 3 uiteen gezet wat de totale minimale vrije doorlaat van ventilatieopeningen en/of -kokers als percentage van de vloeroppervlakte moet zijn. Hierbij wordt aangegeven dat voor niet betreedbare kaststations en kasten zowel bovenventilatie als boven- en onderventilatie mag worden toegepast. Hierbij wordt het percentage van het vloeroppervlak gespecificeerd waarbij deze percentages verschillen per situatie. Zo is bijvoorbeeld gespecificeerd dat een kast minimaal 2% bovenventilatie moet hebben en dat in het geval van boven- en onderventilatie minimaal 1% per ventilatiepositie moet worden aangehouden. Andere onderzoeken [14] [15] hebben aangetoond dat gecombineerde boven- en onderventilatie kan leiden tot verbeterde ventilatiecondities en dientengevolge ook lagere gasconcentraties bij een lekkage. Om meer inzicht te verkrijgen is onderzocht wat het effect is van het aanpassen van de ventilatie van de ½ m3 kast. Dat is gedaan in twee stappen. In eerste instantie is onderventilatie toegevoegd (2% van het vloeroppervlakte). Vervolgens is verder gegaan met het uitbouwen van de ventilatie voor zover dit mogelijk was binnen het ontwerpprincipe van de behuizing. De gestreepte lijnen zijn de meetwaardes van de standaard kast, de vaste lijnen van de aangepaste behuizing met kruisventilatie. De kleuren van de lijnen en streeplijnen zijn gelijk en houden verband met de locatie van het meetpunt. Figuur 30. Concentratie (vol % aardgas links en vol % waterstof rechts) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Figuur 30. Concentratie (vol % aardgas links en vol % waterstof rechts) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar In bovenstaande grafieken is te zien dat de gemeten gasconcentratie op de meetposities “midden”, “hoog” en “hoog-ventilatie” beduidend lager ligt voor de gecombineerde boven/ onderventilatie in vergelijking met alleen bovenventilatie. Dit geldt zowel voor aardgas als voor waterstof. Wat tevens opvalt, is dat aardgas met gecombineerde boven/ onderventilatie net onder LEL/LFL niveau stabiliseert voor deze lekcondities. Voor waterstof met kruisventilatie is wel een verbetering waar te nemen ten opzichte van de uitgangssituatie, echter zijn de concentraties waterstof op de meetposities op of net boven de LEL/LFL. In bovenstaande grafieken is te zien dat de gemeten gasconcentratie op de meetposities “midden”, “hoog” en “hoog-ventilatie” beduidend lager ligt voor de gecombineerde boven/ onderventilatie in vergelijking met alleen bovenventilatie. Dit geldt zowel voor aardgas als voor waterstof. Wat tevens opvalt, is dat aardgas met gecombineerde boven/ onderventilatie net onder LEL/LFL niveau stabiliseert voor deze lekcondities. Voor waterstof met kruisventilatie is wel een verbetering waar te nemen ten opzichte van de uitgangssituatie, echter zijn de concentraties waterstof op de meetposities op of net boven de LEL/LFL. Tevens zijn voor waterstof lekkages additionele experimenten uitgevoerd met/ zonder wind bij een lekopening van 0,025 mm² en 8 bar om zo de effecten van aangepaste ventilatie en wind in beeld te kunnen brengen. Voor het overzichtelijk vergelijken van experimenten zijn de gemiddelde waarden van deze metingen weergegeven in staafdiagrammen. In de onderstaande grafiek is de gemiddelde concentratie waterstof (in vol%) weergegeven voor verschillende meetposities in de behuizing. De gemiddelde concentratie is bepaald in de tijdsspanne tussen 5 en 30 minuten. Figuur 31. Concentratie (vol % waterstof) voor verschillende meetposities in de ½ m3 kast Figuur 31. Concentratie (vol % waterstof) voor verschillende meetposities in de ½ m3 kast De gemeten gemiddelde gasconcentratie ligt op de meetposities “midden”, “hoog” en “hoog-ventilatie” beduidend lager voor de kruisventilatie in vergelijk met alleen bovenventilatie (oranje versus blauw). 6.2 Het effect van additionele onderventilatie Deze kast heeft alleen “standaard” 2% bovenventilatie, zoals vereist in de NEN1059:2019. Vanuit de EAG was de wens om de behuizing minimaal te beschadigen bij het aanbrengen van extra ventilatie. Door de basis van deze kast op metalen strips te plaatsen, is de ventilatie aangepast van bovenventilatie naar boven- en onderventilatie. Hierbij is 2% onderventilatie toegevoegd waardoor kruisventilatie ontstaat. De kast heeft dan dus 2% bovenventilatie en 2% onderventilatie. Figuur 29. Aangepaste ventilatie ½ m3 kast ten behoeve van kruisventilatie met behulp van metalen strips Figuur 29. Aangepaste ventilatie ½ m3 kast ten behoeve van kruisventilatie met behulp van metalen strips Achtereenvolgens zijn een aantal experimenten uitgevoerd met aardgas en waterstof met de kleinste lekkage bij 8 bar (0,56 m3n/h). In eerdere experimenten voor de 1/2m3 behuizing is waargenomen dat de maximale aardgas- of waterstofconcentratie in de behuizing net boven de LEL/LFL komt. Er is onderzocht of de additionele ventilatie kan leiden tot lagere concentraties. In de onderstaande linker grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Hetzelfde experiment is uitgevoerd met waterstof en weergegeven in de rechter grafiek. Pagina 52/123 Pagina 52/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Het effect van wind is duidelijk zichtbaar bij zowel bovenventilatie (oranje versus geel) maar minder duidelijk bij kruisventilatie (blauw versus grijs). De meeste gemeten gemiddelde waarden in de behuizing zijn boven of net op de brandbaarheidsgrens. Pagina 53/123 6.3 Het effect van extra boven- en/of onderventilatie Het effect van extra boven- en/of onderventilatie Omdat het effect van het toevoegen van 2% extra onderventilatie beperkt bleek te zijn, is verder gegaan met het vergroten van het ventilatieoppervlakte. Dat is gedaan op een wijze die paste binnen het ontwerpprincipe van deze behuizing. Door het toevoegen van additionele ringen aan de bovenzijde van de kast kan het dak verder gelift worden en kan de bovenventilatie vergroot worden. Aan de onderzijde kan met behulp van de metalen strip (zie Figuur 29) onderventilatie worden toegevoegd. De bovenventilatie varieert tussen de 2% en 6% van het vloeroppervlak. Verdere vergroting boven de 6% was niet mogelijk omdat dan de kier tussen de behuizing en de omlaagslaande rand ook vergroot zou moeten worden, en dat zou een aanpassing betekenen die indruist tegen het ontwerpprincipe van deze kast. De bovenventilatie van de behuizing kan met zeer geringe inspanning worden vergroot tot 6%, een groter percentage zou extra inspanningen vereisen waarbij de kast bijvoorbeeld van extra roosters moet worden voorzien. De onderventilatie varieert tussen de 0% (dicht) en 2%. Figuur 32. Aanpassing van ½ m3 kast met extra boven- en/of onderventilatie In deze stap is achtereenvolgens een aantal experimenten uitgevoerd met aardgas en waterstof met de kleinste lekkage bij 8 bar (0,56 m3n/h) als uitgangspositie. In eerdere experimenten voor de 1/2m³ behuizing is waargenomen dat de maximale aardgas- of waterstofconcentratie in de behuizing net boven de LEL/LFL komt. Figuur 32. Aanpassing van ½ m3 kast met extra boven- en/of onderventilatie Figuur 32. Aanpassing van ½ m3 kast met extra boven- en/of onderventilatie Figuur 32. Aanpassing van ½ m3 kast met extra boven- en/of onderventilatie In deze stap is achtereenvolgens een aantal experimenten uitgevoerd met aardgas en waterstof met de kleinste lekkage bij 8 bar (0,56 m3n/h) als uitgangspositie. In eerdere experimenten voor de 1/2m³ behuizing is waargenomen dat de maximale aardgas- of waterstofconcentratie in de behuizing net boven de LEL/LFL komt. Pagina 54/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Deze aanpak is uitgewerkt in een dynamische testmatrix die besproken is met de “expert assessment group” voor de aanvang van de uitvoer van alle experimenten. Deze aanpak is uitgewerkt in een dynamische testmatrix die besproken is met de “expert assessment group” voor de aanvang van de uitvoer van alle experimenten. Figuur 33. Opzet van dynamische testmatrix. Figuur 33. Opzet van dynamische testmatrix. Na het uitvoeren van de eerste set experimenten (stap 1, aangegeven met een blauwe pijl), wordt gekozen tussen het verder verbeteren van de ventilatie (wanneer de gemeten aardgas- of waterstofconcentraties boven de LEL/LFL uitkomen) en het vergroten van de lekopening (wanneer de gemeten aardgas- of waterstofconcentraties onder de LEL/LFL uitkomen). Achtereenvolgens wordt als vervolg een deel van de testmatrix uitgevoerd in stap 2 en stap 3 zoals aangegeven met rode kaders. Die geven aan welke route is gevolgd en er is tevens benoemd welk percentage ventilatie in elke stap is toegepast om het ventilatiegedrag maximaal te beïnvloeden. Figuur 34. Stappen in de dynamische testmatrix. Figuur 34. Stappen in de dynamische testmatrix. De gekozen route geeft al wat details prijs met betrekking tot de resultaten van de experimenten die zijn uitgevoerd. Hierbij valt op dat de route voor het verbeteren van de ventilatie wordt gevolgd en niet het vergroten van de gekozen lekopening. Hieruit valt indirect op te maken dat verbetering van ventilatie onvoldoende is om de gemeten gasconcentratie in de behuizing (voor zowel aardgas als waterstof) adequaat te verlagen. Pagina 55/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Stap 1; verhoogde bovenventilatie (4%) bij een lekopening van 0,025 mm² en 8 bar. In vergelijk met de uitgangspositie (alleen 2% bovenventilatie) was het criterium gesteld dat de gemiddelde gemeten concentraties aardgas en waterstof onder de LEL/LFL zouden uitkomen. Figuur 35. Vergroten van de bovenventilatie van 2% naar 4% bij ½ m3 kast Figuur 35. Vergroten van de bov ventilatie van 2% naar 4% bij ½ m3 kast Figuur 35. Vergroten van de bovenventilatie van 2% na Wanneer de 1/2m³ kast in ongewijzigde vorm (met alleen 2% bovenventilatie) wordt vergeleken met dezelfde kast en verbeterde bovenventilatie, valt op dat de gasconcentraties voor aardgas over de hele linie licht afnemen. Hier heeft het vergroten van de bovenventilatie een positief effect op de gasconcentratie in de behuizing. De gemiddeld gemeten waarden bevinden zich onder de LEL/LFL, de hoogst gemeten waarden bevinden zich boven de LEL/LFL. Voor waterstof valt op dat de concentratie afneemt in het midden van de kast (kast midden) maar hoog in de behuizing nagenoeg gelijk blijft bij verbeterde bovenventilatie. De gemiddeld gemeten waarden en de hoogst gemeten waarden bevinden zich boven de LEL/LFL. De foutenbalken zijn gebruikt om inzicht te geven in de spreiding van de metingen, waarbij deze balken in het staafdiagram de gemiddelde meetwaarde representeren. Stap 2; onderventilatie (2%) en verhoogde bovenventilatie (4%) bij een lekopening van 0,025 mm² en 8 bar. In vergelijk met de uitgangspositie (alleen 2% bovenventilatie) was het criterium gesteld dat de gemiddelde gemeten concentraties aardgas en waterstof onder de LEL/LFL zouden uitkomen. Wanneer dit niet het geval is, wordt overgaan tot het verder vergroten van de bovenventilatie. Figuur 36. Vergroten van de bovenventilatie van 2% naar 4% en toepassen kruisventilatie bij ½ m3 kast 2% naar 4% en toepassen kruisventilatie bij ½ m3 kast Figuur 36. Vergroten van de bovenventilatie van 2% naar 4% en toepassen kruisventilatie bij ½ m3 kast Wanneer de 1/2m³ kast in ongewijzigde vorm (met alleen 2% bovenventilatie) wordt vergeleken met dezelfde kast en verbeterde boven- en onderventilatie, valt op dat de gasconcentraties voor aardgas over de hele linie licht afnemen. Hier heeft het verder vergroten van de bovenventilatie effect op de gasconcentratie in de behuizing, echter is de situatie ten opzichte van stap 1 niet significant. Wanneer de 1/2m³ kast in ongewijzigde vorm (met alleen 2% bovenventilatie) wordt vergeleken met dezelfde kast en verbeterde boven- en onderventilatie, valt op dat de gasconcentraties voor aardgas over de hele linie licht afnemen. Hier heeft het verder vergroten van de bovenventilatie effect op de gasconcentratie in de behuizing, echter is de situatie ten opzichte van stap 1 niet significant. Voor waterstof valt op dat de concentratie in het midden van de kast sterk afneemt maar hoog in de behuizing nagenoeg gelijk blijft bij verbeterde boven- en onderventilatie. Er vormt zich een deken van uniforme waterstof concentratie in de behuizing die niet weggeventileerd wordt. Zowel de gemiddeld gemeten waarden als de hoogst gemeten waarden bevinden zich boven de LEL/LFL. Pagina 56/123 Pagina 56/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Stap 3; onderventilatie (2%) en verhoogde bovenventilatie (6%) bij een lekopening van 0,025 mm² en 8 bar. In vergelijk met de uitgangspositie (alleen 2% bovenventilatie) was het criterium gesteld dat de gemiddelde gemeten concentraties aardgas en waterstof onder de LEL/LFL zouden moeten uitkomen. Figuur 37. Vergroten van de bovenventilatie van 2% naar 6% en toepassen kruisventilatie bij ½ m3 kast 2% 6% t k i til ti bij ½ 3 k t an 2% naar 6% en toepassen kruisventilatie bij ½ m3 kast Figuur 37. Vergroten van de bovenventilatie van 2% naar 6% en toepassen kruisventilatie bij ½ m3 k Wanneer de 1/2m³ kast in ongewijzigde vorm (met alleen 2% bovenventilatie) wordt vergeleken met dezelfde kast en verbeterde boven- en onderventilatie, valt op dat de gasconcentraties voor aardgas over de hele linie licht afnemen. Hier heeft het verder vergroten van de bovenventilatie weinig effect op de gasconcentratie in de behuizing. Wanneer de 1/2m³ kast in ongewijzigde vorm (met alleen 2% bovenventilatie) wordt vergeleken met dezelfde kast en verbeterde boven- en onderventilatie, valt op dat de gasconcentraties voor aardgas over de hele linie licht afnemen. Hier heeft het verder vergroten van de bovenventilatie weinig effect op de gasconcentratie in de behuizing. Voor waterstof valt op dat de concentratie in het midden van de kast sterk afneemt maar hoog in de behuizing nagenoeg gelijk blijft bij verbeterde boven- en onderventilatie. Ook hier vormt zich een deken van uniforme waterstof concentratie in de behuizing die niet weggeventileerd wordt. Zowel de gemiddeld gemeten waarden als de hoogst gemeten waarden bevinden zich net boven de LEL/LFL. Om een overzichtelijk vergelijk te maken, zijn alle metingen opgenomen in één staafdiagram voor aardgas en één voor waterstof. Hierbij valt voor aardgas de verbetering op bij het toevoegen van bovenventilatie terwijl bij waterstof vooral de verschuiving van gasconcentratie (van midden naar hoog in de behuizing opvalt). Figuur 38. De effecten van het aanpassen van de ventilatie bij een ½ m3 kast Resumerend valt voor deze specifieke behuizing op basis van bovenstaande stappen te concluderen dat; d valt voor deze specifieke behuizing op basis van bovenstaande stappen te concluderen dat; • Het vergroten van de bovenventilatie leidt tot verlaging van de gemeten aardgas concentraties; • Het vergroten van de bovenventilatie leidt tot verplaatsing van de gemeten waterstof concentraties. In paragraaf 5.3 worden alle resultaten uit de testmatrix samengevat voor de ½ m³ kast, het 4 m³ kaststation en de HAS kast. Alle gerapporteerde waarden zijn de maximaal gemeten concentraties aardgas of waterstof in en om de behuizing. Deze concentraties zijn uit de metingen gehaald die een looptijd hadden van 30 minuten tenzij de gemeten concentratie leidde tot een verhoogd risico met betrekking tot de brandbaarheidsgrenzen. In die specifieke gevallen is de meting korter geweest en wordt dit vermeld bij de beschrijving van de meting in de bijlage. Ook is de meting gestopt wanneer de gemeten concentratie niet meer noemenswaardig toenam. Aan de hand van dit specifieke experiment is tevens een vergelijk gemaakt tussen de werkwijze uit de NPR en de metingen. Hierbij is de mogelijkheid tot ventileren (geometrie specifiek) vergeleken met de experimenten (zie appendix IV). De berekening aan de hand van de NPR laat zien of, afhankelijk van de effectieve windsnelheid die de geometrie binnentreedt, voldoende is om de concentratie in de behuizing tot respectievelijk 25% LEL/LFL of 10% LEL/LFL te doen dalen. Door het veronderstellen van een effectieve windsnelheid van 0,1 m/s (windstil) als intredende windsnelheid kan op basis van deze analyse geconcludeerd worden dat de berekening volgens de NPR redelijk goed overeenkomt met de resultaten uit de experimenten voor waterstof bij een lekopening van 0,025 mm². In het geval van aardgas bij dezelfde lekopening oordeelt de NPR positiever in vergelijking met de experimenten. Bij het opstellen van de behuizing in een normale buitensituatie (niet windstil) zijn geen meetresultaten beschikbaar, echter kan wel een NPR berekening gedaan worden. Wanneer wordt verondersteld dat de effectieve windsnelheid 0,5 m/s bedraagt, is het oordeel van de NPR voor de ventilatie positief met uitzondering van de lekkage waterstof bij 8 bar. Wat tevens opgemerkt dient te worden, is het vergelijk tussen de NPR methode en het bijbehorend experiment bij 40l/u aardgas. In een windstille situatie is de gemeten gasconcentratie in de behuizing kleiner dan 25% LEL/LFL, hetgeen bevestigd wordt door de NPR berekening. Voor het experiment met waterstof is de gemeten gasconcentratie in de behuizing groter dan 25% LEL/LFL, waarbij de NPR berekening net wel positief uitpakt en aantoont dat de behuizing zou voldoen. Bij grotere lekopeningen (en lekdebieten) komen de experimenten en de NPR redelijk goed overeen. Hierbij wordt een deken van uniforme concentratie gevormd die onvoldoende weggeventileerd kan worden (alleen nog hoge concentratie op “kast hoog” en “kast ventilatie”); • Het toevoegen van onderventilatie lijkt vooral een positief effect te hebben voor de gemeten • Het vergroten van de bovenventilatie leidt tot verlaging van de gemeten aardgas concentraties; • Het vergroten van de bovenventilatie leidt tot verplaatsing van de gemeten waterstof concentraties. Hierbij wordt een deken van uniforme concentratie gevormd die onvoldoende weggeventileerd kan worden (alleen nog hoge concentratie op “kast hoog” en “kast ventilatie”); • Het toevoegen van onderventilatie lijkt vooral een positief effect te hebben voor de gemeten • Het vergroten van de bovenventilatie leidt tot verplaatsing van de gemeten waterstof concentraties. Hierbij wordt een deken van uniforme concentratie gevormd die onvoldoende weggeventileerd kan worden (alleen nog hoge concentratie op “kast hoog” en “kast ventilatie”); • Het toevoegen van onderventilatie lijkt vooral een positief effect te hebben voor de gemeten waterstof concentraties. De dikte van de “deken” van uniforme concentratie neemt af. Pagina 57/123 Op basis van een oriënterende stap is inzichtelijk gemaakt dat het toevoegen van extra ventilatie kan leiden tot een verbetering. Door verschillende scenario’s te vergelijken, is een lagere concentratie gemeten in de behuizing. Deze stap is gebruikt als opmaat naar een systematische aanpak van de analyse ter verbetering van de ventilatie met behulp van een dynamische testmatrix. Door het vergroten van de ventilatiecapaciteit middels het toevoegen van boven- en/of onderventilatie is aannemelijk gemaakt dat een verbetering van ventilatie mogelijk is. Er vindt minder ophoping van gas plaats in de behuizing. Hierbij dient wel opgemerkt te worden dat sommige meetwaarden bij grotere lekkages de onderste brandbaarheidsgrens overschrijden. Op basis van de waarnemingen lijkt de geometrie van het dak van invloed te zijn op de efficiëntie van ventilatie van deze 1/2m³ kast. Dit is al eerder geconstateerd tijdens HyDelta 1.0 voor het 4m³ kaststation waarbij de concentratie op de ventilatieopeningen tijdelijk gelijk of zelfs hoger is dan de metingen in de behuizing. Door toevoegen van het meetpunt “top kast” boven de uitgang van de ventilatie (in de dakpunt) is tijdens HyDelta 2.0 aan de hand van metingen inzichtelijk gemaakt dat vooral bij de grotere lekdebieten de gemeten gasconcentratie in de dakpunt hoger kan worden dan de gemeten concentratie bij de ventilatieopeningen. Dit duidt erop dat ophoping van gas in de dakpunt kan plaatsvinden. Door de gekozen dak geometrie (ten behoeve van hemelwaterafvoer) hebben deze behuizingen een hydraulische weerstand die door het ontsnappende gas overwonnen dient te worden. Door de lage dichtheid van waterstof is meer drijvende kracht nodig, hetgeen leidt tot een dikkere deken van uniforme gasconcentratie. Daarbij speelt het dichtheidsverschil tussen lucht en aardgas of lucht en waterstof tevens een belangrijke rol. Voor het ontwerp van een behuizing die geschikt is voor waterstof (H2 ready ontwerp), valt o het ontwerp van een behuizing die geschikt is voor waterstof (H2 ready ontwerp), valt onder re te denken aan de volgende aandachtspunten met betrekking tot het ontwerp: 1. Het percentage ventilatie ten opzichte van het vloeroppervlak zou afhankelijk moeten zij de inhoud van de behuizing. Dit punt dient nader onderzocht te worden. 2. Het vergroten van het ventilatie oppervlak dient nader onderzocht te worden met als vertrekpunt de bevindingen uit deze rapportage en de huidige normering. 3. Het toepassen van kruisventilatie lijkt bij te dragen tot een verbeterde mogelijkheid tot ventilatie in vergelijking met alleen bovenventilatie. Hierbij tonen zowel de NPR methode als de experimenten aan dat het ventilatiegedrag niet kan voorkomen dat de gasconcentratie oploopt tot meer dan 25% LEL/LFL. In paragraaf 6.2 en 6.3 is de rol van ventilatie verder onderzocht door het uitvoeren van metingen met de ½ m³ kast. Om hierbij de rol van externe factoren te minimaliseren, is ervoor gekozen om de staafdiagrammen te rapporteren met gemiddelde waarden. Deze waarden zijn een representatie van de instantane metingen over tijdsspanne tussen 4 minuten en 30 minuten (tenzij situatie als mogelijk risicovol werd geacht, dan is de tijdsduur van de meting ingekort). Vervolgens is in de staafdiagrammen gebruik gemaakt van foutbalken voor het inzichtelijk maken van de minimale en de maximale gemeten concentratie. De lengte van de foutenbalk geeft per meetpunt aan wat de bandbreedte van de gemeten concentraties tijdens het experiment. Met deze spreiding kan de lezer zien wat het effect is van de externe invloeden (zoals wind/ natuurlijke trek). Op deze manier kunnen metingen in één oogopslag met elkaar vergeleken worden. Pagina 58/123 7.1 Inleiding Waterstofgas is een gas dat ernstige risico's kan opleveren als het in een afgesloten ruimte wordt vrijgegeven. De mogelijke gevaren die gepaard gaan met de verspreiding van waterstofgas in afgesloten ruimten zijn onder meer brand en explosie. Het is belangrijk om het gedrag van waterstofgas in afgesloten ruimten te begrijpen om effectieve strategieën te ontwikkelen voor het verminderen van de risico's en het waarborgen van de veiligheid van mensen en apparatuur. Het doel van de modelleringsactiviteiten is om te begrijpen welke factoren de verspreiding van waterstofgas in gasstations beïnvloeden, in termen van de concentratie en verdeling van waterstofgas in de kasten van het gasstation. Deze kennis kan worden gebruikt om de gasstations te ontwerpen en te bedienen op een manier die de risico’s op waterstof gerelateerde incidenten minimaliseren en om de effectiviteit van verschillende risicoverminderingsstrategieën te evalueren. Modellering kan worden uitgevoerd om het effect van gasstation behuizingen van verschillende maten en vormen te bestuderen. Factoren zoals de eigenschappen van het waterstofgas, de ventilatiesnelheid, de aanwezigheid van obstakels of menging in het gasstation en de initiële concentratie en verdeling van het gas kunnen worden meegenomen. De stappen en factoren die uiteindelijk de waterstofconcentratieprofielen in een gasstation bepalen, zijn als volgt. De stappen en factoren die uiteindelijk de waterstofconcentratieprofielen in een gasstation bepalen, zijn als volgt. 1. Lekdebiet: de hoeveelheid waterstof of aardgas die via een lekopening (in het defecte onderdeel) uit de leiding ontsnapt. Bij voldoende hoge bedrijfsdrukken, wat het geval is aan de stroom inlaatzijde van het gasstation (~8 barg), kunnen explosieve mengsels worden vastgesteld. Voor een bepaalde gas, zijn de enige parameters die de lekkagesnelheid bepalen, de druk en temperatuur in de leiding, de lekopening en een discharge coëfficiënt die zorgt voor correctie van de effectieve doorlaatbaarheid. 2. Verspreiding: de snelheid waarmee waterstof of aardgas zich in het gasstation verspreidt, wordt bepaald door verschillende fysische processen. Het eerste mechanisme is convectie, vanwege de hoofdbeweging van het gas door de uitstootstraal en de opwaartse kracht (dichtheidsverschillen) tussen het vrijgegeven gas en de lucht in het gasstation. Ten tweede drijft diffusie moleculen van een gebied met een hoge concentratie naar een gebied met een lage concentratie als gevolg van willekeurige moleculaire beweging. Een derde mechanisme is de ventilatie, die natuurlijk kan worden uitgevoerd door middel van openingen of gedwongen in geval van gebruik van een mechanisch apparaat. In een gasstation wordt alleen natuurlijke ventilatie toegepast. De praktische positionering van onderventilatie (in relatie tot plantgroei en sneeuw) dient hierbij in overweging genomen te worden. 4. De ventilatieopeningen dienen op enige afstand van de bovenzijde van de kast geplaatst te worden. Ook wordt in literatuur [3] gesteld dat de kwaliteit van ventilatie maximaal rendeert wanneer de ventilatie openingen 5-10% van de kasthoogte onder het dak worden aangebracht. Plaatst men de ventilatie openingen te hoog, dan strijkt de dwarswind teveel over het gasmengsel in de kast en dient men het oppervlak van de ventilatieopeningen met een factor 2 of 3 te vermenigvuldigen om hetzelfde effect te krijgen. Dit effect dient verder onderzocht te worden. 5. Bovenmaatse daken met als doel het voorkomen van inwateren dienen onderzocht te worden met het oog op hydraulische weerstand en ventileren. Een dergelijke constructie lijkt de ventilatiecapaciteit te limiteren. 5. Bovenmaatse daken met als doel het voorkomen van inwateren dienen onderzocht te worden met het oog op hydraulische weerstand en ventileren. Een dergelijke constructie lijkt de ventilatiecapaciteit te limiteren. Voor een dergelijke stap kunnen experimenten gebruikt worden als startpunt, waarbij CFD berekeningen kunnen worden gebruikt om meer inzicht te krijgen in het gedrag van gas tijdens een lekkage. Voor een dergelijke stap kunnen experimenten gebruikt worden als startpunt, waarbij CFD berekeningen kunnen worden gebruikt om meer inzicht te krijgen in het gedrag van gas tijdens een lekkage. Pagina 59/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 7. Modellering van gasverspreiding bij lekkages in gasstations 7.2 Literatuurstudie De literatuurstudie heeft zich gericht op de technieken die gebruikt worden bij het simuleren van de verspreiding van gassen in afgesloten ruimtes. Publicaties richten zich doorgaans op scenario’s die plaatsvinden in huishoudelijke, mobiliteit of industriële omgevingen. Zoals vermeld in de vorige sectie kan het probleem van een waterstoflek in een gasstation meer algemeen worden beschreven als dat van een onontstoken gasuitstoot in een afgesloten ruimte met passieve ventilatie. Omdat het gebruik van CFD-simulaties voor onderzoek naar ongevallen met waterstof momenteel veel aandacht krijgt, is de literatuurstudie die in dit rapport wordt samengevat beperkt tot publicaties waarvan het onderzoeksgebied veel overeenkomsten vertoont met het onderzoeksgebied van dit rapport. Ref. [16] presenteert een blinde CFD-benchmark van waterstofuitstoot en verspreiding in een afgesloten ruimte met passieve ventilatie. Dit is echter niet de eerste referentie die probeert CFD te benchmarken voor waterstofophoping in afgesloten ruimtes: het artikel herinnert aan twee andere vergelijkbare CFD- benchmarks die werden uitgevoerd met vergelijkbare doelstellingen, maar voor veel grotere afgesloten ruimtes. Het volume dat in ref. [16] wordt overwogen, is van dezelfde orde van grootte (~1 m³) als dat van het gasstation dat is geselecteerd voor CFD-simulaties (0,5 m³). In de experimenten werd helium gebruikt in plaats van waterstof vanwege veiligheidsredenen en eigenschappen van het gas. In de experimenten werd ook het effect van windomstandigheden [17] onderzocht, hoewel de simulaties werden uitgevoerd onder omstandigheden zonder wind. Er werden verschillende openingen gebruikt. Drie verschillende partijen werden uitgenodigd om CFD-simulaties uit te voeren om de concentraties van de verontreinigende stof (helium) op verschillende posities in de afgesloten ruimte te bepalen. Elke partij gebruikte een andere CFD-solver en turbulentie model. Alle partijen ontwikkelden een model waarin de buitenomgeving wordt weergegeven door een voldoende groot volume. Als het gekozen volume te klein was, zouden er onnauwkeurigheden ontstaan in de berekend stromingsprofiel rond het ventilatieopening. In alle gevallen werd drijfvermogen impliciet vastgelegd door te berekenen voor de massaverhouding van helium over het hele model. De belangrijkste verschillen tussen de CFD- benaderingen waren het turbulentie model (URANS, LES), de numerieke discretisatiemethoden en het gebruik van grensvoorwaarden. Over het algemeen leverden alle CFD-formuleringen bevredigende resultaten op voor alle overwogen openingen. Er werd echter vastgesteld dat numerieke methoden, vooral op basis van URANS, de concentratie van de verontreinigende stof op de bodem van de afgesloten ruimte grotendeels overschatten, hypothetisch vanwege een kunstmatige overschatting van turbulente diffusie. 7.1 Inleiding Deze mechanismen worden weergegeven in Figuur 39. Pagina 60/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 39. Mechanismen die het transport van een gasvervuilende stof (in dit geval waterstof) binnen het gasstation (groene doos) aandrijven nadat er een lekkage heeft plaatsgevonden. Ventilatiespleten zijn beschikbaar aan de bovenkant van de kast. WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 39. Mechanismen die het transport van een gasvervuilende stof (in dit geval waterstof) binnen het gasstation (groene doos) aandrijven nadat er een lekkage heeft plaatsgevonden. Ventilatiespleten zijn beschikbaar aan de bovenkant van de kast. Figuur 39. Mechanismen die het transport van een gasvervuilende stof (in dit geval waterstof) binnen het gasstation (groene doos) aandrijven nadat er een lekkage heeft plaatsgevonden. Ventilatiespleten zijn beschikbaar aan de bovenkant van de kast. De resultaten van de benchmark-inspanningen zijn later door dezelfde auteurs verwerkt tot een richtlijn voor de beste werkwijze en een bredere beoordeling van hoe CFD-simulaties uit te voeren voor waterstofverspreidingsproblemen in afgesloten ruimtes [18] [19]. Meer recentelijk voerden Lee et al. [20] CFD-onderzoeken uit voor een waterstofdrukregelstation dat qua kenmerken vergelijkbaar is met de gasstations die in HyDelta zijn onderzocht. Er werd een goede overeenkomst gevonden tussen CFD en experimenten, waardoor de onderzoekers het effect van alternatieve ventilatie konden onderzoeken. Een combinatie van de Bousinessq-hypothese, het species transport model en de k- SST-turbulentieformulering presteerde het beste. Op basis van deze referenties en de hierboven besproken referenties worden in Tabel 6 een aantal aanbevelingen gegeven voor de CFD-simulaties die nuttig zijn voor het onderzoek naar gasstations. gegeven in referenties [18], [19]. Onderwerp Aanbevelingen Domein Breid de simulatie uit verder dan alleen de positie van de ventilatie om eventuele randeffecten van de fysieke lucht-waterstofuitwisseling bij de ventilatie te verwijderen. Rooster Pas verhoogde resolutie toe in regio's met sterke gradiënten, zoals rond de waterstofinjectienozzle, jet-wall impingement-punt en ventilatie. Lek injectie Ofwel als een volumebronterm of als een grensbron gekarakteriseerd door zijn snelheid op een geselecteerd gebied/oppervlak. De laatste is de voorkeur. Gas eigenschappen Volg een mengbenadering op basis van massafractionering van de verschillende soorten en volumegebaseerd voor de dichtheid. Turbulentie model Geen duidelijke aanbeveling tussen op RANS gebaseerde of LES-gebaseerde benaderingen. RANS heeft de voorkeur vanwege de lagere kosten, maar met extra bouyancy-termen. Randvoorwaarden Als het domeinadvies wordt gevolgd, zijn vaste drukgrenzen, nulgradiënt of zelfs symmetriegrenzen mogelijk. Tijdstap Over het algemeen wordt aangenomen dat de grootte van de tijdstap klein moet zijn, met CFL-waarden (Courant getal) rond de 10 (5-20) in geval van k- turbulentiemodellering. Voor LES is dit lager dan 1. Numeriek schema Er zijn geen specifieke aanbevelingen, maar in alle benchmarkstudies werden ten minste tweede orde schema's gebruikt voor de convectieve termen. 7.2 Literatuurstudie De benchmark concludeerde dat CFD een geschikte benadering vertegenwoordigt om ventilatie-eigenschappen van afgesloten ruimtes te onderzoeken die vergelijkbaar zijn met de ruimte die in de validatie-experimenten is gebruikt en met vergelijkbare openingprincipes Pagina 61/123 7.3 Modelleringsaanpak 7.3 Er kunnen verschillende modelleringsmethoden worden gebruikt om de concentratieprofielen van waterstof in een (geventileerde) ruimte te voorspellen. Het doel van de modellering is om een beter begrip te geven van de experimentele resultaten. Er worden twee modelleringsniveaus nagestreefd. Het eerste niveau is gebaseerd op een combinatie van empirisch gevalideerde modellen die zijn opgenomen in de HyRAM+ toolkit, die bijna onmiddellijke resultaten biedt. Het tweede niveau is gebaseerd op CFD, wat lang kan duren om antwoorden te geven voor een bepaald geval. Basis en opstelling van de HyRAM+ calculations Basis en opstelling van de HyRAM+ calculations HyRAM+ [21] is gekozen om een eerste voorspelling en perspectief te geven op welke concentratieniveaus kunnen worden bereikt bij een waterstof- of aardgaslek in een gasstation. De resultaten van HyRAM+ kunnen worden vergeleken met de experimenten en met de CFD- modelleringsresultaten. Pagina 62/123 Pagina 62/123 De modelleringsaanpak die in de HyRAM+ tool wordt gebruikt, volgt ref. [2]. De basis van het model is een massabalans binnen een veronderstelde homogene laag die boven in de ruimte ontstaat wanneer waterstof wordt vrijgegeven. De vrijgave wordt gemodelleerd met een jet model. De gevormde pluim voert waterstof naar de laag. Ventilatie vermindert de ophoping van waterstof in de veronderstelde laag. Ventilatie kan van nature worden aangedreven door opwaartse kracht, standaard aan de bovenkant, maar kan ook worden additioneel worden ondersteund door wind of een mechanische ventilator. Voor de berekeningen uitgevoerd met HyRAM+, worden de volgende aannames of randvoorwaarden gebruikt: • Waterstof wordt vrijgegeven uit een tank bij een druk van 8 barg, wat de stroom opwaartse zijde van een gasstation vertegenwoordigt. • De lekdiameter is geselecteerd zodat het waterstoflekdebiet overeenkomt met de experimentele waarden. Het lek bevindt zich in het midden en 30 cm boven de grond, en naar boven gericht. • Ventilatie: alleen plafondventilatie zonder aanwezigheid van wind of mechanische ondersteuning. Het gebied van de plafondventilatie is gebaseerd op een 15 mm spleet die zich uitstrekt over 3 van de 4 zijden van het deksel van het gasstation behuizing (de twee lange zijden en een korte zijde - de andere korte zijde heeft een plaat die grotendeels de ventilatiestroom blokkeert). • De temperatuur van zowel de vrijgegeven waterstof als de omgeving wordt verondersteld gelijk te zijn op 15 graden Celsius. Basis en opstelling van de CFD-berekeningen Basis en opstelling van de CFD-berekeningen De modelleringsopstelling moest eerst worden geverifieerd aan de hand van een selectie van de experimenten die zijn uitgevoerd in de voormalige HyDelta 1-fase (gerapporteerd in [2]) en de huidige HyDelta 2-fase. Omdat de HyDelta 1-gegevens al beschikbaar waren aan het begin van HyDelta 2, werd deze set eerst geselecteerd voor validatiedoeleinden. De HyDelta 1-experimenten hielden rekening met een aantal onafhankelijke omstandigheden bij het onderzoeken van de ventilatie van de gasstations: • Het ontwerp van het gasstation kabinet: een 1/2 m³ of een 4 m³ station. • Het gas dat voor de lekkage wordt gebruikt: aardgas (G-gas) of waterstof. • Bescherming tegen de wind of niet • Bescherming tegen de wind of niet • De daadwerkelijke opwaartse druk / spuitmond diameter die uiteindelijk het lekpercentage bepaalt. Een overzicht van de experimentele matrix van HyDelta 1 wordt getoond in Figuur 40. Vanwege de kosten per CFD-berekening is slechts een subset geselecteerd voor validatie van de CFD-simulaties. De 1/2 m³ kast is geselecteerd omdat hogere concentratieniveaus werden gevonden in de experimenten. De gevallen met windbescherming door een tent zijn ook interessant maar niet gekozen, omdat dit een moeilijk te bepalen randvoorwaarde is voor de CFD vanuit een dynamisch oogpunt (gemiddelde windsnelheidsvariaties en windstoten). Tevens is een geen-windconditie ook een conservatiever scenario, omdat wind de ventilatie zal verbeteren. Over het algemeen betekent dit dat het 1/2 m³-kast wordt gesimuleerd voor lekken van G-gas en H2 zonder rekening te houden met wind Pagina 63/123 g p betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 42. Lokale roosterverfijning is toegepast bij de lekopeningkage en bij de wanden van de kast, aan beide zijden. Het totale aantal cellen bedraagt ongeveer 3,7 miljoen. Figuur 42. Lokale roosterverfijning is toegepast bij de lekopeningkage en bij de wanden van de kast, aan beide zijden. Het totale aantal cellen bedraagt ongeveer 3,7 miljoen. Het model werd opgezet volgens ref. [16], i.e. met RANS-turbulentie modellering met de realizable k-ε turbulentie model en uitgebreide wandfunctie beschrijving. Volledige beschrijving van opwaartse effecten werden meegenomen wat betekent dat opwaartse krachten invloed hebben op de productie en afbraak van turbulentie. Specietransport werd gemodelleerd met behulp van een gegroepeerde specie voor "lucht" en een andere voor waterstof. De energievergelijking werd niet opgelost. De dichtheid op elk punt waar een mengsel van lucht en waterstof aanwezig is, werd berekend met behulp van de volumegewogen gemiddelde. De eigenschappen van de individuele soorten werden constant gehouden en zijn te vinden in Tabel 7. Hoewel de viscositeit van waterstof van dezelfde orde van grootte is als die van lucht, is de dichtheid een orde van grootte kleiner. Dit betekent dat waterstof sterk opwaarts beweegt wanneer het in de behuizing wordt vrijgegeven. Met betrekking tot de randvoorwaarden, werden alle oppervlakken die de grond of bodem van de kast vertegenwoordigen, gemodelleerd als wanden. Ook werden de vier zijoppervlakken van het volume dat de buitenomgeving voorstelt, gemodelleerd als wanden. Het bovenoppervlak werd gemodelleerd met een constante drukrandvoorwaarde. De lekkage werd gemodelleerd als een instroom met een oppervlakte van 2 mm2, wat aanzienlijk groter is dan de eigenlijke opening die in geen van de experimenten werd gebruikt. Dit werd gedaan om zeer kleine roosterafstanden te voorkomen die de rekentijd aanzienlijk zouden verhogen. De massastroom werd echter gehandhaafd op de beoogde waarde om te kunnen vergelijken met de experimenten. Dit heeft invloed op de omgeving dicht bij opening, maar niet op de algehele verdeling van waterstof in de kast. Dit komt feitelijk overeen met het plaatsen van de "virtuele" instroom op een punt waar de straal al is uitgezet. De belangrijkste resultaten die worden gezocht in de CFD-simulaties zijn de concentraties waterstof in lucht. De virtuele meetpunten zijn de punten P1-P2-P3-P4, die overeenkomen met meetlocaties 1-2-3-4 (Figuur 20). Tabel 7. Eigenschappen van de gassen die zijn gebruikt in de CFD-simulaties. Lokale roosterverfijning is toegepast bij de lekopeningkage en bij de wanden van de kast, aan beide zijden. Het totale aantal cellen bedraagt ongeveer 3,7 miljoen. Een voorbeeld van het rooster is te zien in Pagina 64/123 Pagina 64/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en exp betrekking tot ventilatie in verschillende typen gasdru distributie (lagedruk)net in Nederland met aardgas en Figuur 42. Lokale roosterverfijning is toegepast bij de lekopeningkage en bij de wande beide zijden. Het totale aantal cellen bedraagt ongeveer 3,7 miljoen. Het model werd opgezet volgens ref. [16], i.e. met RANS-turbulentie modellering met turbulentie model en uitgebreide wandfunctie beschrijving. Volledige beschrijving van effecten werden meegenomen wat betekent dat opwaartse krachten invloed hebben en afbraak van turbulentie. Specietransport werd gemodelleerd met behulp van een g specie voor "lucht" en een andere voor waterstof. De energievergelijking werd niet op dichtheid op elk punt waar een mengsel van lucht en waterstof aanwezig is, werd ber van de volumegewogen gemiddelde. De eigenschappen van de individuele soorten w gehouden en zijn te vinden in Tabel 7. Hoewel de viscositeit van waterstof van dezelfd grootte is als die van lucht, is de dichtheid een orde van grootte kleiner. Dit betekent opwaarts beweegt wanneer het in de behuizing wordt vrijgegeven. Met betrekking tot de randvoorwaarden, werden alle oppervlakken die de grond of b vertegenwoordigen, gemodelleerd als wanden. Ook werden de vier zijoppervlakken v de buitenomgeving voorstelt, gemodelleerd als wanden. Het bovenoppervlak werd ge een constante drukrandvoorwaarde. De lekkage werd gemodelleerd als een instroom oppervlakte van 2 mm2, wat aanzienlijk groter is dan de eigenlijke opening die in geen experimenten werd gebruikt. Dit werd gedaan om zeer kleine roosterafstanden te voo rekentijd aanzienlijk zouden verhogen. De massastroom werd echter gehandhaafd op waarde om te kunnen vergelijken met de experimenten. Dit heeft invloed op de omge opening, maar niet op de algehele verdeling van waterstof in de kast. Dit komt feitelij plaatsen van de "virtuele" instroom op een punt waar de straal al is uitgezet. De belangrijkste resultaten die worden gezocht in de CFD-simulaties zijn de concentra lucht. De virtuele meetpunten zijn de punten P1-P2-P3-P4, die overeenkomen met me (Figuur 20). Tabel 7. Eigenschappen van de gassen die zijn gebruikt in de CFD-simulaties. Eigenschap gassen Lucht Waterstof Aardgas Di hth id [k / 3] 1 225 0 082 0 833 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 42. D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 40. Overzicht van experimenten uitgevoerd in HyDelta 1.0 [23]. Het pad dat groen is gemarkeerd, geeft de gevallen aan die zijn geselecteerd voor validatie van de CFD. Figuur 40. Overzicht van experimenten uitgevoerd in HyDelta 1.0 [23]. Het pad dat groen is gemarkeerd, geeft de gevallen aan die zijn geselecteerd voor validatie van de CFD. Figuur 41. Het initiële model setup toonde het gasstation met apparatuur (links) en een beperkte dekking van het model voor de buitenomgeving. Er werd ontdekt dat er sprake was van verbeterde menging als gevolg van het “botsen” van de straal op een flens (rechts), die niet kon worden gevalideerd aan de hand van de experimenten. De definitieve model setup wordt weergegeven in Figuur 42. Figuur 41. Het initiële model setup toonde het gasstation met apparatuur (links) en een beperkte dekking van het model voor de buitenomgeving. Er werd ontdekt dat er sprake was van verbeterde menging als gevolg van het “botsen” van de straal op een flens (rechts), die niet kon worden gevalideerd aan de hand van de experimenten. De definitieve model setup wordt weergegeven in Figuur 42. De simulaties werden uitgevoerd op de ½ m³ kast. Enkele vereenvoudigingen werden toegepast op de geometrie van de kast om het model klaar te maken voor het genereren van het net. Apparatuur in het gasstation (bijv. leidingwerk, kleppen, enz.) werd uiteindelijk niet gemodelleerd (Figuur 42), omdat dit de rooster kwaliteit verminderde en de mengsnelheid verhoogde omdat de straalstroom tegen de flens botst (Figuur 41). Er werd een domein van 4 m × 3 m × 3 m gebouwd rond de kast om de omgeving van het gasstation te modelleren. Deze uiteindelijke grootte van het buitenste domein is het resultaat van vroege pogingen met een kleiner domein (1,5 m × 2 m × 2 m), dat in combinatie met drukuitlaat randvoorwaarde onfysische neveneffecten naar het interessegebied bracht. Het maken van het rekenrooster werd uitgevoerd in Pointwise resulterend in hybride rekenroosters die voornamelijk tetraëdrische cellen bevatte en enkele hexaëdrische cellen. Na het importeren van het rekenrooster in ANSYS Fluent werden de tetraëdrische cellen getransformeerd in polyhedrale cellen om het aantal cellen te verminderen en de roosterkwaliteit te verbeteren. Tabel 7. Eigenschappen van de gassen die zijn gebruikt in de CFD-simulaties. Figuur 42. Polyhedrale rekenrooster gebruikt in de CFD-simulaties en lokale rooster verfijning rondom de injectie nozzle (rechtsboven). Figuur 42. Polyhedrale rekenrooster gebruikt in de CFD-simulaties en lokale rooster verfijning rondom de injectie nozzle (rechtsboven). (rechtsboven). Eigenschap gassen Lucht Waterstof Aardgas Dichtheid [kg/m3] 1.225 0.082 0.833 Viscositeit [Pa s] 17. 9 8.4 11.4 Tabel 7. Eigenschappen van de gassen die zijn gebruikt in de CFD-simulaties. Pagina 65/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 43. Links: vergelijking van de resultaten van HyRAM+ met die van de experimenten uitgevoerd in HyDelta 1.0 en HyDelta 2.0. Als alle punten in deze grafiek samenvielen met de gestippelde lijn, zou dat betekenen dat er een perfecte overeenstemming is tussen de modelresultaten en de experimentele resultaten. Het model onderschat de gemeten concentraties in de kast. Rechts: oppervlak van de ventilatie dat in HyRAM+ -model moet worden geïntroduceerd om de gemeten concentraties overeen te laten komen, onderscheid makend tussen de gevallen waarin de kast blootgesteld wordt aan (wisselende sterkte van) wind en de gevallen waarin deze wordt beschermd door een tent. 0 5 10 15 0 5 0 5 0 5 50 0 10 0 0 0 50 R .1 as is, 0.0 5 entilation easured a . on entration in gas station abinet 0 0.00 0.00 0.00 0.00 0.01 0.01 0 10 0 0 0 50 ent area to at h R .1 easured a . on entration in gas station abinet ind tent 0 0.00 0.00 0.00 0.00 0.01 0.01 0 10 0 0 0 50 ent area to at h R .1 easured a . on entration in gas station abinet ind tent 0 5 10 15 0 5 0 5 0 5 50 0 10 0 0 0 50 R .1 as is, 0.0 5 entilation easured a . on entration in gas station abinet Figuur 43. Links: vergelijking van de resultaten van HyRAM+ met die van de experimenten uitgevoerd in HyDelta 1.0 en HyDelta 2.0. Als alle punten in deze grafiek samenvielen met de gestippelde lijn, zou dat betekenen dat er een perfecte overeenstemming is tussen de modelresultaten en de experimentele resultaten. Het model onderschat de gemeten concentraties in de kast. Rechts: oppervlak van de ventilatie dat in HyRAM+ -model moet worden geïntroduceerd om de gemeten concentraties overeen te laten komen, onderscheid makend tussen de gevallen waarin de kast blootgesteld wordt aan (wisselende sterkte van) wind en de gevallen waarin deze wordt beschermd door een tent. 7.5 Resultaten van de CFD-simulaties 7.5 Er zijn veel simulaties gedaan om de gevoeligheid van modelleringsopties te bestuderen totdat een definitieve instelling was bepaald (zoals beschreven in sectie 7.3). Hier worden alleen de meest relevante simulatieresultaten beschreven met betrekking tot de meetcampagnes in HyDelta 1.0 [2] en Hydelta 2.0. Het overzicht van de uitgevoerde simulaties wordt gegeven in Tabel 8, die de basis vormt van de volgende subsecties. Wanneer de resultaten worden vergeleken met experimenten, worden alleen de gevallen in de windluwe situatie (binnen de tent) in overweging genomen. Tabel 8. Samenvatting van uitgevoerde simulaties met de beschreven opstelling. Simulation ID Leaked fluid Lek debiet [Nm³/h] Comments 01 Waterstof 6.0 HyDelta 1, 8 barg, 0.25 mm² 02 Waterstof 0.6 HyDelta 2, 8 barg, 0.025 mm² 03 Waterstof 0.6 02 met gevoeligheid voor jet richting 04 Waterstof 0.6 02 met gevoeligheid voor ontwerp van ventilatieopening 05 Aardgas 3.0 HyDelta 1, maximale debiet met tentbescherming 06 Aardgas 0.45 HyDelta 1, lager debiet dan 05 Tabel 8. Samenvatting van uitgevoerde simulaties met de beschreven opstelling. Resultaten van de simulaties met HyRAM+ Simulaties met de HyRAM+ tool zijn berekend voor alle experimentele gevallen die zijn uitgevoerd gedurende HyDelta 1.0 [2] en de 1/2 m³ kast van HyDelta 2.0 (zoals gepresenteerd in dit rapport), inclusief die voor waterstof en aardgas. Onder de aannames beschreven in sectie 7.3, voorspelt het model consequent concentraties die ongeveer een factor 3 lager zijn dan de gemeten concentraties tijdens de experimenten. Dit wordt getoond in Figuur 43. De belangrijkste reden hiervoor is waarschijnlijk dat de geometrie van het ventilatiegebied niet overeenkomt met het archetype dat wordt aangenomen in het HyRAM+ model, aangezien waterstof een kronkelend pad moet volgen om uit de kast te komen. Dit pad omvat zelfs een neerwaarts segment (tegen de opwaartse kracht van de buoyancy) dat alleen kan worden doorlopen door middel van een drijvende overdruk. In het HyRAM+ model bevindt het ventilatiegebied zich in hetzelfde vlak als de verticale wand van de kast, die de binnenkant van de behuizing scheidt van de buitenomgeving. Om de gemeten concentraties te evenaren, zou een veel kleiner ventilatiegebied moeten worden geïntroduceerd, dat ongeveer 10 keer kleiner is dan de geometrisch geselecteerde waarde. Als alternatief zou de uitstroomcoëfficiënt moeten worden aangepast, hoewel het effect gelijkwaardig is. De waarde fluctueert, waarbij gevallen blootgesteld aan wind een kleinere oppervlaktecorrectie vereisen dan de gevallen waarin de tent aanwezig was. Dit is redelijk, aangezien wind extra ventilatie biedt, wat kan worden vertaald in een effectief groter ventilatieraam. Windeffecten kunnen potentieel worden vastgelegd in HyRAM+, maar alleen door van invoer van de gebruiker, en niet als gevolg van het fysische model zelf. Over het algemeen is de belangrijkste conclusie dat de geometrie van de ventilatiespleten van een typische Nederlandse gasstation kast het gebruik van moderne, standaardmodellen zoals die in HyRAM+ moeilijk maakt om toe te passen. Een correctiefactor op het ventilatiegebied van ongeveer 10 zou redelijke resultaten opleveren in vergelijking met de experimenten die in dit programma zijn uitgevoerd. Pagina 66/123 Simulatie-01 – Waterstof, 6 m³(n)/h Simulatie-01 – Waterstof, 6 m³(n)/h Simulatie 01 Waterstof, 6 m (n)/h De eerste gesimuleerde case betrof de uitstoot van 6 m3(n)/h H2. Deze case simuleert een lekopening van 0,25 mm2 met een stuur-druk van 8 barg vanuit het netwerk. Dit komt overeen met een massastroom van 0,15 g/s die werd toegepast in de CFD. Na 8 minuten wordt de uitstoot van waterstof gestopt. Een contourplot van de snelheidsmagnitude wordt getoond in Figuur 44. Hierbij is te zien dat de straal de deksel van de kast raakt en vervolgens langs de bovenkant naar de zijkanten van de kast stroomt. De impuls van de straal drijft circulatie naar beneden en langs de muren van de kast. In Figuur 46 worden contourplots van de H2 (%v) gepresenteerd. Vanwege de hoge snelheid van de straal mengt de waterstof zich redelijk goed in de kast en ontsnapt ook wat H2 via de ventilatieroosters naar buiten, waar het vanwege de opwaartse drijfkracht omhoog beweegt. Het gedrag lijkt niet zo op de theoretische archetypen gebaseerd op een homogene laag. Dit komt door de grote impuls van de straal. De waterstofconcentratie bereikt na ongeveer 6 minuten een stabiele toestand, zoals te zien is in Figuur Pagina 67/123 Pagina 67/123 45 waar de tijdsevolutie is geplot voor verschillende locaties in de kast. Ook zijn de gemeten waarden in deze figuur geplot, waarbij 'L' staat voor de lagere locatie en 'H' voor de hogere locatie. Figuur 44. Contourplot van de snelheidsmagnitude in een vlak door het lek. Figuur 44. Contourplot van de snelheidsmagnitude in een vlak door het lek. Figuur 45. Concentratie van H2 op drie locaties (P1-P2-P3) vergeleken met de experimentele metingen (EXP L en EXP H) voor debiet 6 m3(n)/h H2. Figuur 45. Concentratie van H2 op drie locaties (P1-P2-P3) vergeleken met de experimentele metingen (EXP L en EXP H) voor debiet 6 m3(n)/h H2. Pagina 68/123 Pagina 68/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Waterstof, 6 m³(n)/h t = 1 s t = 5 s t = 10 s t = 15 s t = 20 s t = 25 s t = 30 s t = 40 s Figuur 46. Contourplots van de concentratie H2 (volume%) op verschillende tijdstippen. t=1,5,10,15,20,25,30 en 40 s. De contour is afgekapt op 4,4%, de onderste brandbaarheidsgrens van H2. t = 1 s Figuur 46. Contourplots van de concentratie H2 (volume%) op verschillende tijdstippen. t=1,5,10,15,20,25,30 en 40 s. De contour is afgekapt op 4,4%, de onderste brandbaarheidsgrens van H2. Pagina 69/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Enkele observaties die kunnen worden gemaakt vanuit Figuur 45 zijn: Simulatie-02 – Waterstof,  0.6 m³(n)/h , ( )/ Vervolgens wordt een geval uit HyDelta 2.0 berekend voor een lekkage van ongeveer 0,6 m3(n)/h, wat overeenkomt met een massastroom van 0,015 g/s. Dit geval simuleert een lekopening van 0,025 mm² met een inlaatdruk van 8 barg. Kijkend naar de contourplots in Figuur 47 zien we dat de stroming minder turbulent is en dat waterstof langzaam een laag vormt aan de bovenkant van de kast maar niet gemakkelijk door de ventilatieopening stroomt. Het gedrag is zeer vergelijkbaar met theoretische kaders die uitgaan van een homogene laag van groeiende dikte totdat een evenwicht tussen bron en ventilatie is bereikt. Het gedrag is ook zichtbaar in Figuur 48, waar de concentraties op verschillende meetpunten zijn geplot. Ter vergelijk wordt hier ook de CFD-gegevens van het vorige geval gegeven. Zoals verwacht, is de opbouw trager, maar bereikt het geen plateau zoals in het vorige geval. Ook zijn de concentratiewaarden verkregen in de simulatie veel hoger dan die verkregen in de experimenten, wat aangeeft dat het model een minder effectieve ventilatie vastlegt dan tijdens de experimenten. Bovendien zijn de verkregen concentraties hoger dan die bij een hogere stroomsnelheid, wat aangeeft dat de kast zonder aandrijvende impuls van een straal niet ventileert. In het geval van de hogere stroomsnelheid (simulatie-01) is de impuls die de waterstof nog bevat bij het raken van de deksel voldoende om een deel ervan naar buiten te drijven. Bij een kleinere stroomsnelheid is dat niet het geval en is de kast in staat om meer waterstof te verzamelen. Na 18 minuten simulatietijd wordt de simulatie gestopt. Vervolgens wordt een geval uit HyDelta 2.0 berekend voor een lekkage van ongeveer 0,6 m3(n)/h, wat overeenkomt met een massastroom van 0,015 g/s. Dit geval simuleert een lekopening van 0,025 mm² met een inlaatdruk van 8 barg. Kijkend naar de contourplots in Figuur 47 zien we dat de stroming minder turbulent is en dat waterstof langzaam een laag vormt aan de bovenkant van de kast maar niet gemakkelijk door de ventilatieopening stroomt. Het gedrag is zeer vergelijkbaar met theoretische kaders die uitgaan van een homogene laag van groeiende dikte totdat een evenwicht tussen bron en ventilatie is bereikt. Het gedrag is ook zichtbaar in Figuur 48, waar de concentraties op verschillende meetpunten zijn geplot. Ter vergelijk wordt hier ook de CFD-gegevens van het vorige geval gegeven. Enkele observaties die kunnen worden gemaakt vanuit Figuur 45 zijn: • De initiële opbouw van waterstof aan de bovenkant van de kast komt redelijk overeen met de experimenten. De uiteindelijke evenwichtsconcentraties aan de bovenkant van de kast (P2, P3) zijn ook zeer vergelijkbaar en komen overeen met de gemeten waarden. • De initiële opbouw van waterstof aan de bovenkant van de kast komt redelijk overeen met de experimenten. De uiteindelijke evenwichtsconcentraties aan de bovenkant van de kast (P2, P3) zijn ook zeer vergelijkbaar en komen overeen met de gemeten waarden. De tijd waarop evenwicht wordt bereikt is ook redelijk vergelijkbaar, tussen 5 en 6 minuten. • De tijd waarop evenwicht wordt bereikt is ook redelijk vergelijkbaar, tussen 5 en 6 mi • Voor P1, het laagste punt, overschat de simulatie de evenwichtsconcentratie tot het punt dat er weinig tot geen stratificatie binnen de kast te zien is. Dit betekent dat de circulatie gecreëerd door de jet een dominante invloed heeft in vergelijking met opwaartse krachten, maar mogelijk ook dat de ventilatie onvoldoende is. Er is weinig concurrentie tussen zwaardere verse lucht en opwaartse waterstof om de ruimte bij de bodem van de kast te bezetten. • Na 8 minuten, wanneer de lekkage in de experimenten stopt, veranderen de concentraties niet in de simulatie, terwijl de waarden in het experiment snel dalen. Dit geeft aan dat de ventilatie volgens de simulatie slecht is en dat er geen verse lucht van buitenaf de kast binnenkomt (of omgekeerd dat de waterstof vastzit in de kast). De diffusie naar buiten is onvoldoende om de concentratie in de kast te laten dalen. Simulatie-02 – Waterstof,  0.6 m³(n)/h Zoals verwacht, is de opbouw trager, maar bereikt het geen plateau zoals in het vorige geval. Ook zijn de concentratiewaarden verkregen in de simulatie veel hoger dan die verkregen in de experimenten, wat aangeeft dat het model een minder effectieve ventilatie vastlegt dan tijdens de experimenten. Bovendien zijn de verkregen concentraties hoger dan die bij een hogere stroomsnelheid, wat aangeeft dat de kast zonder aandrijvende impuls van een straal niet ventileert. In het geval van de hogere stroomsnelheid (simulatie-01) is de impuls die de waterstof nog bevat bij het raken van de deksel voldoende om een deel ervan naar buiten te drijven. Bij een kleinere stroomsnelheid is dat niet het geval en is de kast in staat om meer waterstof te verzamelen. Na 18 minuten simulatietijd wordt de simulatie gestopt. Pagina 70/123 g yp g g distributie (lagedruk)net in Nederland met aardgas en waterstof Waterstof, 0,6 m³(n)/h t = 1 s t = 5 s t = 15 s t = 25 s t = 30 s t = 40 s t = 50 s t = 60 s Figuur 47. Contourplots van de concentratie H2 (volume%) op verschillende tijdstippen. t=1,5,15,25,30,40,50 en 60 s. De contour is afgekapt op 4,4%, de onderste brandbaarheidsgrens van H2 t = 1 s Figuur 47. Contourplots van de concentratie H2 (volume%) op verschillende tijdstippen. t=1,5,15,25,30,40,50 en 60 s. De contour is afgekapt op 4,4%, de onderste brandbaarheidsgrens van H2 Pagina 71/123 Pagina 71/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 48. Concentratie van H2 op drie locaties vergeleken met de experimentele metingen van debiet 0,588 m3(n)/h van H2. Ter vergelijking worden ook de simulatieresultaten van de 6 m3(n)/h case gegeven. Figuur 48. Concentratie van H2 op drie locaties vergeleken met de experimentele metingen van debiet 0,588 m3(n)/h van H2. Ter vergelijking worden ook de simulatieresultaten van de 6 m3(n)/h case gegeven. Simulatie-05 & Simulatie-06 – aardgas, 3 m³(n)/h en 0.45 m³(n)/h Twee gevallen met een lekkage van aardgas (NG) worden uitgevoerd. Het verschil zit alleen in de lekstroom. De hoge stroom met 3,0 Nm3/h en een lagere stroom van 0,45 m3(n)/h. Resultaten zijn te zien in Figuur 49 en Figuur 50. In Figuur 49 zijn de contouren van NG geplot voor beide gevallen. Het geval met de hoge stroomsnelheid mengt veel sneller maar, vanwege het momentum van de straal, kan het aardgas sterk naar buiten het kabinet worden uitgestoten. Dit trekt ook meer verse lucht naar binnen. Bij het geval van de lage lekstroom wordt er een opbouw van aardgas gevormd vanaf het bovenste deksel van het kabinet, en de uitstoot van aardgas naar buiten is nauwelijks zichtbaar in de contourplots. Door naar de waarden op de meetpunten in Figuur 51 te kijken, kan worden gezien dat de evenwichtsconcentratie bijna hetzelfde is voor beide gevallen. Dit is een onverwacht resultaat, omdat de experimenten een hogere waarde laten zien voor de hogere stroomsnelheid. Het is echter consistent met het gedrag dat ook werd waargenomen in het waterstofgeval bij het vergelijken van simulaties 01 en 02, waarbij een lagere lekstroom resulteert in een hogere concentratie binnenin het kabinet. Pagina 72/123 Pagin Aardgas. Links: 0,45 m3(n)/h. Rechts: 3,0 m3(n)/h t = 1 s t = 1 s t = 10 s t = 10 s t = 100 s t = 100 s t = 400 s t = 400 s t = 1000 s t = 1000 s Figuur 49. Contourplots van de concentratie aardgas voor beide-gevallen. Links: 0,45 m3(n)/h. Rechts: 3,0 m3(n)/h. Tijdswaarden (van boven naar beneden) zijn respectievelijk 1,0 s, 10 s, 100 s, 400 s en 1000 s. Aardgas. Links: 0,45 m3(n)/h. Rechts: 3,0 m3(n)/h Aardgas. Links: 0,45 m3(n)/h. Rechts: 3,0 m3(n)/h t = 1 s t = 1 s t = 10 s t = 10 s t = 10 s t = 10 s t = 100 s t = 100 s t = 100 s t = 400 s t = 400 s t = 1000 s t = 1000 s t = 1000 s t = 1000 s Figuur 49. Contourplots van de concentratie aardgas voor beide-gevallen. Links: 0,45 m3(n)/h. R Tijdswaarden (van boven naar beneden) zijn respectievelijk 1,0 s, 10 s, 100 s, 400 s en 1000 s. Pagina 73/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B Veiligheid Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten me betrekking tot ventilatie in verschillende typen gasdrukregelstations distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 50. Concentratie van NG op drie locaties voor zowel de uitstroom van 3 m3(n)/h als 0,45 Nm3 (n)/h aardgas. Ter vergelijking worden ook de experimentele resultaten van de 3 m3(n)/h debiet gegeven. Figuur 50. Concentratie van NG op drie locaties voor zowel de uitstroom van 3 m3(n)/h als 0,45 Nm3 (n)/h aardgas. Ter vergelijking worden ook de experimentele resultaten van de 3 m3(n)/h debiet gegeven. Simulation-03 – Lekkage in horizontale richting Er wordt een sensitiviteit scenario gesimuleerd met een horizontale lekkage van waterstof. Het stromingsveld is afgebeeld in Figuur 51. Hier is te zien dat de stroom tegen de wand botst en voornamelijk omhoog stroomt. Een deel stroomt ook naar beneden. Wanneer we kijken naar de concentraties op de meetpunten die worden getoond in Figuur 52, is er weinig verschil tussen de verticale en horizontale gevallen. De waarden van P2 en P3 liggen dichter bij elkaar voor de horizontale lekkage, wat betekent dat er betere menging plaatsvindt binnenin de kast in het geval van horizontale lekkage. 7Figuur 51. Contourplot van het snelheidsveld van simulatie 03, waarbij de lekkage is vrijgegeven in horizontale richting. Deze configuratie is niet met experimenten 7Figuur 51. Contourplot van het snelheidsveld van simulatie 03, waarbij de lekkage is vrijgegeven in horizontale richting. Deze configuratie is niet met experimenten Pagina 74/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations i distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 527. Concentratie van H2 op drie locaties voor het geval van een horizontale lekkage-uitstroom, vergeleken met de experimentele metingen van de ~0.6 m3(n)/h uitstroom van H2. Figuur 527. Concentratie van H2 op drie locaties voor het geval van een horizontale lekkage-uitstroom, vergeleken met de experimentele metingen van de ~0.6 m3(n)/h uitstroom van H2. Simulatie-04 – Verschillend ventilatieontwerp Simulatie-04 – Verschillend ventilatieontwerp In voorgaande secties is besproken dat de geometrie van de ventilatiekanalen van het gasstation weerstand biedt tegen de uitstroom van waterstof in geval van een lek. Een belangrijke barrière is de overhang van het deksel. Waterstof die bovenaan arriveert, moet eerst naar beneden stromen (tegen zijn natuurlijke opwaartse neiging in) om de kast van het gasstation te verlaten. Dat kan alleen gebeuren als er voldoende tegenovergestelde druk is. Alleen wanneer de gasstroom sterk genoeg is, is er voldoende momentum in de stroming (of stagnatiedruk) om de onderkant van de overhang te bereiken en naar buiten te stromen. Het was interessant om te begrijpen in hoeverre de overhang inderdaad een effectieve barrière vormt voor de uitstroom van waterstof. Daarom is een simulatie uitgevoerd waarbij de overhang is verwijderd. Dit wordt getoond in Figuur 53. De gekozen lekdebiet is ~ 0,6 m3(n)/h. Praktische gevolgen van het verwijderen van de overhang, zoals bescherming tegen waterindringing of kwaadwillige handelingen, worden momenteel genegeerd. De resultaten van de simulatie worden getoond in Figuur 54 en Figuur 55. Het verwijderen van de overhang heeft een duidelijk effect op het verminderen van de concentratie van waterstof binnen de gasstation behuizing. Zo'n ontwerp zorgt voor een betere ventilatie. Het is daarom aan te bevelen om alternatieve ventilatieontwerpen te onderzoeken die waterstof in staat stellen zijn natuurlijke opwaartse beweging te volgen, terwijl er voldoende bescherming is tegen externe factoren. Pagina 75/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 537. Detail van de ontbrekende overhang die wordt verwijderd in simulatie 04. De wandoppervlakken van het deksel worden weergegeven in groen, wat waterstof dwingt om de gele pijlen te volgen om de kast van het gasstation te verlaten. In simulatie 04 worden de laterale oppervlakken (overhang) verwijderd, waardoor waterstof de behuizing van het gasstation kan verlaten. WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 537. Detail van de ontbrekende overhang die wordt verwijderd in simulatie 04. De wandoppervlakken van het deksel worden weergegeven in groen, wat waterstof dwingt om de gele pijlen te volgen om de kast van het gasstation te verlaten. Simulatie-04 – Verschillend ventilatieontwerp In simulatie 04 worden de laterale oppervlakken (overhang) verwijderd, waardoor waterstof de behuizing van het gasstation kan verlaten. 7Figuur 54. Concentratie van H2 op drie locaties voor het geval zonder overhang, vergeleken met de experimentele metingen van de ~0,6 m3(n)/h release van H2. Ter vergelijking wordt ook het geval met overhang getoond (geldt ook voor experimentele resultaten). 7Figuur 54. Concentratie van H2 op drie locaties voor het geval zonder overhang, vergeleken met de experimentele metingen van de ~0,6 m3(n)/h release van H2. Ter vergelijking wordt ook het geval met overhang getoond (geldt ook voor experimentele resultaten). Pagina 76/123 7.6 Conclusies CFD 7.6 Conclusies CFD De verschillende modelleringsactiviteiten werden geïntroduceerd om het resultaat van de metingen beter te begrijpen. Om dat te kunnen doen, is een goede overeenstemming tussen de modellen en de experimenten een vereiste. De volgende conclusies kunnen worden getrokken: • De modellering gebaseerd op moderne, standaard engineering tools zoals HyRAM+ of de CFD brengen het ontwerp van de ventilatie van behuizing in beeld. Met beide technieken worden voor de gekozen lekdebieten de opbouw van gasconcentratie in kaart gebracht waarbij qua orde grootte dezelfde conclusies getrokken kunnen worden. • De modellering gebaseerd op moderne, standaard engineering tools zoals HyRAM+ of de CFD brengen het ontwerp van de ventilatie van behuizing in beeld. Met beide technieken worden voor de gekozen lekdebieten de opbouw van gasconcentratie in kaart gebracht waarbij qua orde grootte dezelfde conclusies getrokken kunnen worden. • Modellering bevestigt dat de gasconcentraties voor grote lekkages van waterstof-luchtmengsels binnen de kast boven de LFL liggen, zoals gevonden in de experimenten. Ventilatie voor deze scenario’s zijn gelimiteerd in het huidige ontwerp. • De ventilatieroute is belangrijk en zou de stijgende stroom van waterstof gedreven door de opwaartse kracht moeten vergemakkelijken. Een alternatieve geometrie die de overhang van het kastdeksel verwijdert, bevestigt een significante verbetering van de ventilatie. • Of de lekkage nu in verticale of horizontale richting plaatsvindt, heeft geen sterke invloed op de concentraties binnen de kast. • De kwaliteit van de CFD-validatie in termen van overeenkomst tussen de gemeten evenwichtsconcentraties en de CFD-modelleringsresultaten wordt niet als bevredigend genoeg beschouwd. Het model is nuttig om te visualiseren welke fenomenen de verspreiding van waterstof binnen de behuizing aandrijven, maar in vergelijking met de experimenten ontbreken nauwkeurige voorspellende capaciteiten. Verschillende redenen zijn mogelijk, zoals het effect van wind (windstoten) of verschillen in temperatuur. Waterstof, 0,6 m³(n)/h Waterstof, 0,6 m³(n)/h 7Figuur 55. Contourplot van de concentratie van H2 (volume%) op verschillende tijdstippen. t=1, 10, 60, 100, 200, 300, 600 en 900 s. De contour is afgekapt bij 4,4%, de onderste brandbaarheidsgrens van H2. 7Figuur 55. Contourplot van de concentratie van H2 (volume%) op verschillende tijdstippen. t=1, 10, 60, 100, 200, 300, 600 en 900 s. De contour is afgekapt bij 4,4%, de onderste brandbaarheidsgrens van H2. Pagina 77/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 8. Conclusies De doelstelling van dit onderzoek was om inzicht te verkrijgen in de geschiktheid van de ventilatie van de bestaande behuizingen van gasstations voor de distributie van waterstof. Om dat inzicht te verkrijgen zijn de volgende deelvragen gesteld: 1. Deelvraag 1: Wat zijn de meest voor de hand liggende typen behuizingen van gasstations voor de toepassing van waterstofdistributie? 1. Deelvraag 1: Wat zijn de meest voor de hand liggende typen behuizingen van gasstations voor de toepassing van waterstofdistributie? 2. Deelvraag 2: Aan welke voorwaarden moet een behuizing van een gasstation voldoen voordat deze geschikt is voor de distributie van waterstof? 9 3. Deelvraag 3: In hoeverre voldoen de meest voor hand ligge voorwaarden, al dan niet met additionele ventilatie? 10 3. Deelvraag 3: In hoeverre voldoen de meest voor hand liggende behuizingen aan de voorwaarden, al dan niet met additionele ventilatie? 10 De eerste deelvraag kan worden beantwoord. Er zijn drie typen behuizingen die een afspiegeling zijn van een substantieel deel van hele populatie: de HAS kast, de ½ m3 kast en de 4 m3 kast. Deze zijn geselecteerd na een inventarisatie bij de netbeheerders. Hierbij zijn de behuizingen die in de laatste tien jaar zijn geplaatst, meegenomen en is een inschatting gemaakt welke type behuizing in eerste instantie gebruikt zal gaan worden in een waterstofnetwerk. Uit de gegevens van de laatste 10 jaar blijkt dat recent vooral kleinere behuizingen zijn geplaatst, deze behuizingen zijn tamelijk uniform. In het bijzonder zijn kasten met een inhoud van 0,5 m³ en kaststations met een inhoud van 4 m³ in grote mate vergelijkbaar. Het is hiermee aannemelijk dat de resultaten uit dit onderzoek een goede indicatie zijn voor alle stations met deze inhoudsmaat, zolang deze gelijk zijn met betrekking tot de wijze van ventileren en het ventilatieoppervlakte. De HAS kasten (inhoud < 0,5 m³) zijn in mindere mate uniform. Elke netbeheerder heeft zijn eigen standaard die soms veel en soms weinig afwijkt van de standaard van andere netbeheerders. Ook de tweede deelvraag kan worden beantwoord, al is dit antwoord wel complexer. De belangrijkste voorwaarde is dat (de behuizing van) een gasstation voor waterstof minstens net zo veilig is als (de behuizing van) een gasstation voor aardgas. Om dit te bereiken moet worden zorggedragen dat het gebied dat vrij toegankelijk is voor het algemene publiek kan worden aangeduid als Niet Gevaarlijk Gebied (NGG). 9 De eisen voor ventilatie van huidige behuizingen is vastgelegd in de NEN1059: 2019 en de NEN12186: 2014 waarbij onderscheid wordt gemaakt tussen verschillende soorten behuizingen. De gekozen lekopening in de NEN1059: 2019 is vastgesteld op 1 mm² voor het zoneren van opstellingsruimten. Met de resultaten uit HyDelta 1.0 staat deze lekopening ter discussie voor het uitvoeren van de testen binnen dit werkpakket (voor het beproeven van de ventilatie en het vaststellen van de zonering). In HyDelta 1.0 is expliciet een aanbeveling gemaakt dat er verschil zou moeten zijn tussen een incident en een reguliere lekkage/ storing. 10 Het uitgangspunt voor het toepassen van ventilatie zal in dit onderzoeksprogramma allereerst gebaseerd zijn op bestaande behuizingen en geldende normen. Aanpassingen van de behuizingen ter verbetering van de ventilatie zullen, waar nodig, in samenspraak worden bepaald. 7.7 Aanbevelingen CFD De volgende aanbevelingen voor de modellering worden gegeven met als uiteindelijk doel de ontwerpvereisten van de gasstation kast te bepalen om waterstofconcentraties in de lucht onder de LFL te houden. • Een breder validatieonderzoek gebaseerd op academische gevallen is essentieel om de oorzaak van de afwijkingen tussen modellering en experimenten tot op dit punt te vinden. In andere woorden, experimenten die uitgevoerd zijn in een zeer gecontroleerde omgeving (vooral zonder wind), zouden gereproduceerd moeten worden met de CFD modelleringstechniek om modelleringsgebreken te onderscheiden van fouten die te wijten zijn aan de complexiteit van deze specifieke zaak. • Itereer verschillende ontwerpen van ventilatieroosters, zodat de uitstroom van waterstof kan worden vergemakkelijkt. De ontwerpen moeten compatibel zijn met voldoende bescherming van de apparatuur binnenin • Itereer verschillende ontwerpen van ventilatieroosters, zodat de uitstroom van waterstof kan worden vergemakkelijkt. De ontwerpen moeten compatibel zijn met voldoende bescherming van de apparatuur binnenin Pagina 78/123 Dat betekent dat lekken die 0,1% van de tijd of vaker voor kunnen komen, niet mogen leiden ontbrandbaar mengsel binnen de fysieke grenzen van de ATEX zone 2. Hier bovenop stelt de Dat betekent dat lekken die 0,1% van de tijd of vaker voor kunnen komen, niet mogen leiden tot een ontbrandbaar mengsel binnen de fysieke grenzen van de ATEX zone 2. Hier bovenop stelt de NPR7910-1 (2021) dat de ventilatievoud groter moet zijn dan 5 keer per uur. Tevens stelt deze norm: “In het geval van een kleine lekkage zou voldoende ventilatie aanwezig moeten zijn om het ontsnapte gas goed weg te kunnen ventileren” Een praktische interpretatie hiervan is dat de concentratie onder de LEL/LFL moet blijven. ontbrandbaar mengsel binnen de fysieke grenzen van de ATEX zone 2. Hier bovenop stelt de NPR7910 1 (2021) dat de ventilatievoud groter moet zijn dan 5 keer per uur. Tevens stelt deze norm: “In het geval van een kleine lekkage zou voldoende ventilatie aanwezig moeten zijn om het ontsnapte gas goed weg te kunnen ventileren” Een praktische interpretatie hiervan is dat de concentratie onder de LEL/LFL moet blijven. Het is van belang om een getalswaarde te noemen en te onderbouwen voor deze “kleine lekkage”. Bovenstaande leidt tot de vraag hoe groot een lek kan zijn dat wél 0,1% van de tijd of vaker voorkomt. In hoofdstuk 3 van dit rapport wordt een nadere beschouwing gegeven over lekopeningen en daaruit blijkt dat er meerdere definities mogelijk zijn van een “kleine lekkage”: daaruit blijkt dat er meerdere definities mogelijk zijn van een “kleine lekkage”: - De NEN 1059: 2019 noemt een lekopening van 1 mm² - De IGEM/SR/25 noemt een lekopening van 0,25 mm² (P > 100mbar) en 0,025 mm² (P < 100 mbar) - De NEN-EN-IEC-60079-10-1 (2021) noemt een lekopening in de range van 0,025 mm² t/m 0,25 mm² - Praktijkmetingen tonen aan dat bij de ventilatieopening van districtstations geen gasconcentraties van 100% LEL of hoger voorkomen. Uit aanvullende metingen met een high flow sampler, blijkt ook dat lekkages groter dan 40 l/u niet of nauwelijks voorkomen bij stations op aardgas. Wanneer we dit debiet (aan de veilige kant) omrekenen naar waterstof komen we op 125 l/u. Binnen dit onderzoek zijn metingen uitgevoerd aan lekkages met openingen van 0,25 mm² , 0,025 mm² en bij een lekkage van 40 l/u aardgas en 125 l/u waterstof. 4 m3 kast, vergelijking aardgas en waterstof 4 m3 kast, vergelijking aardgas en waterstof Bij zowel aardgas (40l/u) als waterstof (125 l/u) blijft de concentratie ver onder de onderste brandbaarheidsgrenzen. De gemeten concentraties bij waterstof zijn wel hoger, maar in beide gevallen veilig. Bij metingen met lekopeningen van 0,25 mm² en 0,025 mm² zijn geen brandbare mengsel gemeten bij aardgas. Bij waterstof is een brandbaar mengsel gemeten bij een lekopening van 0,25 mm². Om een vergelijk te maken tussen het veiligheidsniveau van stations op aardgas en waterstof moet vooral gekeken worden naar de laatstgenoemde lekopeningen, omdat deze in de praktijk voorkomen. Dit alles leidt tot de beantwoording van deelvraag 3. Voldoen de behuizingen aan de voorwaarden? Of, met andere woorden, is de situatie met waterstof net zo veilig of met aardgas? 8. Conclusies Bij aardgas geldt dat binnen de behuizing van een gasstation sprake is van een ATEX zone 2 en dat direct buiten de behuizing het Niet Gevaarlijk Gebied begint. Om die reden wordt er normaal gesproken geen hekwerk geplaatst rondom een station voor de distributie van aardgas. Datzelfde moet ook gelden voor stations met waterstof. Een behuizing (of: de behuizing inclusief een afgeschermd stuk grondoppervlakte) moet geclassificeerd kunnen worden als een ATEX zone 2. Dan is direct naast deze zone geen ATEX zonering meer noodzakelijk en mag dat gebied toegankelijk zijn voor het algemene publiek. Om een gebied te mogen classificeren als ATEX zone 2, moet aangetoond kunnen worden dat sprake is van een secundaire gevarenbron en dat voldoende ventilatie aanwezig is. Dat betekent dat er in minder dan 0,1% van de tijd sprake mag zijn van een brandbaar mengsel. Om dat aan te kunnen tonen, is kennis nodig van de in de praktijk optredende lekken. Er mogen lekken voorkomen in gasstations die leiden tot een concentratie boven de 100% LEL/LFL, zolang deze maar zelden (< 0,1% van de tijd) voorkomen. Door een gedegen onderhoudsregime moet door de eigenaar immers zorg worden gedragen voor een technisch dichte installatie. Pagina 79/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof ½ m3 kast, vergelijking aardgas en waterstof Bij een lek van 40l/u aardgas is de maximum gasconcentratie in een ½ m3 kast ongeveer 1 vol%, een mengsel onder de onderste brandbaarheidsgrens. Bij een vergelijkbaar lek voor waterstof (125 l/u), loopt de concentratie op tot 2,7 vol%. Direct bij de ventilatieopeningen worden vergelijkbare concentraties gemeten, met 3,2 vol% als lokale, tijdsafhankelijke uitschieter. Op een halve meter afstand van de behuizing is de concentratie in alle gevallen ver onder de ontbrandbaarheidsgrens. Bij metingen met lekopeningen van 0,25 mm² en 0,025 mm² zijn brandbare mengsels gemeten bij zowel aardgas als waterstof. Minikast, vergelijking aardgas en waterstof Er zijn enkele metingen uitgevoerd bij de mini-kast. Deze hadden als doel om input te geven aan eventueel vervolgonderzoek en zijn te beperkt om harde conclusies aan te verbinden. Een feitelijke waarneming is dat bij aardgas (40l/u) geen ontbrandbare mengsel zijn gemeten, bij waterstof (125 l/u) wel. Effecten toevoegen extra ventilatie Pagina 80/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Vervolgens is onderzocht in hoeverre deze specifieke ½ m3 kast aangepast kan worden zodat er een groter ventilatieoppervlakte ontstaat. Hierbij is gemeten met lekopening van 0,25 mm² en 0,025 mm². Het toevoegen van extra ventilatieoppervlakte heeft zeker effect, maar niet in die mate dat daarmee alle concentratie boven de 100% LEL/LFL vermeden kan worden. Ook bij een verdubbeling van de bovenventilatie (tot 4% van het vloeroppervlakte) en aanvullende toevoeging van onderventilatie (2% van het vloeroppervlakte) blijven brandbare mengsels waterstof ontstaan wanneer lekdebieten conform kentallen uit de normen worden toegepast. Er zijn geen metingen gedaan bij een lekdebiet van 125 l/u waterstof omdat de concentratie hier al onder 100% LEL/LFL bleven bij standaard ventilatie. Een ander interessant inzicht uit de proeven met additionele ventilatie is invloed op het verspreidingsgedrag van het gasmengsel. Bij aardgas wordt met name de gasconcentratie bovenin de behuizing lager bij het toevoegen van extra ventilatie, bij waterstof wordt juist de concentratie op lagere meetpunten lager. De verklaring hiervoor is dat bij een waterstoflek een “deken” ontstaat van een waterstof/luchtmengsel met een vergelijkbare concentratie. Dat beeld wordt bevestigd door resultaten van de CFD berekeningen. Het toevoegen van extra ventilatie zorgt er wel voor dat die deken dunner wordt maar niet dat de concentratie in die deken omlaag gaat. Het effect hiervan is dat extra ventilatie er wel voor zorgt dat er minder gas in de behuizing aanwezig is. Vervolgens is onderzocht in hoeverre deze specifieke ½ m3 kast aangepast kan worden zodat er een groter ventilatieoppervlakte ontstaat. Hierbij is gemeten met lekopening van 0,25 mm² en 0,025 mm². Het toevoegen van extra ventilatieoppervlakte heeft zeker effect, maar niet in die mate dat daarmee alle concentratie boven de 100% LEL/LFL vermeden kan worden. Ook bij een verdubbeling van de bovenventilatie (tot 4% van het vloeroppervlakte) en aanvullende toevoeging van onderventilatie (2% van het vloeroppervlakte) blijven brandbare mengsels waterstof ontstaan wanneer lekdebieten conform kentallen uit de normen worden toegepast. Er zijn geen metingen gedaan bij een lekdebiet van 125 l/u waterstof omdat de concentratie hier al onder 100% LEL/LFL bleven bij standaard ventilatie. kentallen uit de normen worden toegepast. Er zijn geen metingen gedaan bij een lekdebiet van 125 l/u waterstof omdat de concentratie hier al onder 100% LEL/LFL bleven bij standaard ventilatie. Een ander interessant inzicht uit de proeven met additionele ventilatie is invloed op het verspreidingsgedrag van het gasmengsel. Bij aardgas wordt met name de gasconcentratie bovenin de behuizing lager bij het toevoegen van extra ventilatie, bij waterstof wordt juist de concentratie op lagere meetpunten lager. De verklaring hiervoor is dat bij een waterstoflek een “deken” ontstaat van een waterstof/luchtmengsel met een vergelijkbare concentratie. Dat beeld wordt bevestigd door resultaten van de CFD berekeningen. Het toevoegen van extra ventilatie zorgt er wel voor dat die deken dunner wordt maar niet dat de concentratie in die deken omlaag gaat. Het effect hiervan is dat extra ventilatie er wel voor zorgt dat er minder gas in de behuizing aanwezig is. Er bestaat een sterk vermoeden dat de configuratie van het dak van de behuizing (de rand van het dak) leidt tot een hydraulische weerstand bij het ventileren, dat is ook één van de inzichten uit de CFD berekeningen. Hierdoor heeft zowel aardgas als waterstof onvoldoende mogelijkheid om uit de behuizing te ontsnappen. Pagina 81/123 9. Aanbevelingen Uit dit onderzoek volgen de volgende aanbevelingen: Uit dit onderzoek volgen de volgende aanbevelingen: - Overweeg om concentratiemetingen in de standaard werkwijze op te nemen door deze te registreren voor onderzoek en monitoring over langere periode. Dit kan interessante inzichten geven in hoe populaties gasdrukregelstations zich ontwikkelen. - Overweeg om concentratiemetingen in de standaard werkwijze op te nemen door deze te registreren voor onderzoek en monitoring over langere periode. Dit kan interessante inzichten geven in hoe populaties gasdrukregelstations zich ontwikkelen. - Overweeg om de werkinstructies aan te passen zodat monteurs bij werkzaamheden aan gasstations een gasconcentratiemeting in de ventilatieopening uitvoeren voordat ze de kast openen. Door de gemeten gasconcentratie te registreren krijgen de netbeheerders met een beperkte extra inspanning meer inzicht in daadwerkelijk frequentie van het optreden van grotere lekken. - Voor toepassing van waterstof in gasstations lijken aanvullende voorzorgsmaatregelen nodig te zijn om hetzelfde veiligheidsniveau te bereiken als voor aardgas. Bij het 4 m3 kaststation is dit verschil duidelijker dan bij de ½m3 kast. Uitgaande van het voorzorgsprincipe, zijn voor zowel de ½ m3 als de 4 m3 kast aanvullende voorzorgsmaatregelen verstandig. Hier zijn verschillende mogelijkheden, zoals het verder vergroten van het ventilatieoppervlakte van de behuizing, het aanpassen van de behuizing, het plaatsen van hekwerk rondom minimaal een meter afstand van de behuizing of intensievere controle op lekken dan gebruikelijk is voor stations op aardgas. - Nader onderzoek op het gebied van HAS-kasten is aan te raden voordat HAS-kasten omgezet worden naar waterstof. Zolang dat onderzoek nog niet is uitgevoerd, is het raadzaam om aanvullende voorzorgsmaatregelen toe te passen wanneer HAS-kasten voor waterstof worden toegepast. Dat onderzoek kan op twee verschillende manieren worden benaderd. Een mogelijkheid is om uit te gaan van de bestaande situatie, waarbij elke netbeheerder zijn eigen standaard behuizing heeft. Elke behuizing zal dan onderzocht worden d.m.v. CFD berekeningen en/of metingen. Een tweede aanpak is om te werken naar een nieuw ontwerp voor behuizingen van mini-kasten. Dan zullen een aantal concepten worden bedacht (in samenspraak met de constructeurs) en de effectiviteit van deze nieuwe concepten worden gecontroleerd met CFD betekeningen en/of metingen. - Het is belangrijk dat de normcommissie van de NEN 1059 een uitspraak doet voor welke lekdebieten (voor welke specifieke situatie) de ventilatie een effectieve maatregel moet zijn. Lekopeningen en praktijkmetingen lijken nog factoren van elkaar af te staan. 9. Aanbevelingen - Het is belangrijk dat de normcommissie van de NEN 1059 een uitspraak doet voor welke lekdebieten (voor welke specifieke situatie) de ventilatie een effectieve maatregel moet zijn. Lekopeningen en praktijkmetingen lijken nog factoren van elkaar af te staan. Pagina 82/123 Referenties Molkov, “CFD benchmark on hydrogen release and dispersion in a ventilated enclosure: passive ventilation and Pagina 83/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof the role of an external wind,” International Journal of Hydrogen Energy, vol. 40, pp. 6465-6477, 2015. Referenties [1] NEN1059, Gasvoorzieningssystemen - Gasdrukregel- en meetstations voor transport en distributie, Nederlandse Norm, 2019. [1] NEN1059, Gasvoorzieningssystemen - Gasdrukregel- en meetstations voor transport en distributie, Nederlandse Norm, 2019. [2] van Woudenberg, “HyDelta 1.0 - WP1B Gasstations - Ventilatie,” New Energy Coalition - TKI2020 - HyDelta, 2022. [3] G. Gaikhorst, O van Uchelen, “Ventilatie in kaststations - VEG gasinstituut,” GAS, pp. 433 - 441, November 1971. [4] Ir. O. van Uchelen, “de Hinderwet em gasdrukregel- en meetstations,” GAS, pp. 49-92, Februari 1968. [5] 2, IGEM/SR/25 Edition, “HAZARDOUS AREA CLASSIFICATION OF NATURAL GAS,” Institute of gas engineers and managers, 2010. [6] NEN-EN-IEC60079-10-1, “Explosive atmospheres - Part 10-1: Classification of areas - explosive gas atmospheres (IEC-60079-10-1:2020, IDT),” Nederlandse norm, 2021. [7] Dr M J Ivings, Mr S Clarke, Dr S E Gant,, “Area classification for secondary releases from low pressure natural gas systems,” HSE, 2008. [8] N. Vermeltfoort, M. van der Laan, “GT-220257 Concentratiemetingen in ventilatieopeningen van districtregelstations,” Kiwa Technologie BV, 2022. [9] N. Vermeltfoort, M.van der Laan, “GT-220349 HI FLOW Sampler metingen bij gasstations,” Kiwa Technology, 2022. [10] Charlotte Große, Melanie Eyßer, Stefanie Lehmann, Jenny Sammüller, Marco Behnke, Klaus Peters, “Research Project Methane Emmision DSO Has Been Completed,” Energy, Wasser-praxis, Vols. 5 - 2022, pp. 1 - 8, 2022. [11] DBI, “Methane emission estimation methode for the gas distribution grid (MEEM),” 2018. [12] Nestor Gonzalez Diez, “Impact of high speed hydrogen flow on system integrity and noise,” TNO, 2021. [13] K. Pulles, M. van der Laan, “Zonering van gasstations, fase 2,” Kiwa Technology, 2017. [14] Jaewon Lee, Sunghyun Cho, Hyungtae Cho, Seungsik Cho, Il Moon, Jungkwam Kim, “CFD modeling on natural and forced ventilation during hydrogen leaks in a pressure regulator process of a residential area,” Process Safety and Environmental Protection, vol. 161, pp. 436-446, 2022. [15] S. Giannissi, V. Shentsov, D. Melideo, B. Cariteau, D. Baraldi, A. Venetsanos and V. Molkov, “CFD benchmark on hydrogen release and dispersion in confined, naturally ventilated space with one vent,” Int. J. Hydrogen ENergy, vol. 40, p. 2415–2429, 2015. [16] S. Giannissi, J. Hoyes, B. Chernyavskiy, P. Hooker, J. Hall, A. Venetsanos and V. [17] I. Tolias, S. Giannissi, A. Venetsanos, J. Keenan, V. Shentsov, D. Makarov, S. Coldrick, A. Kotchourko, K. Ren, O. Jedicke and D. Melideo, “Best practice guidelines in numerical simulations and CFD benchmarking for hydrogen safety applications,” International Journal of Hydrogen Energy, vol. 44, no. 17, pp. 9050-9062, 2019. [18] S. Giannissi, I. Tolias, D. Melideo, D. Baraldi, V. Shentsov, D. Makarov, V. Molkov and A. Venetsanos, “On the CFD modelling of hydrogen dispersion at low-Reynolds number release in closed facility,” International Journal of Hydrogen Energy, vol. 46, no. 57, pp. 29745-29761, 2021. [19] J. Lee, S. Cho, H. Cho, S. Cho, I. Lee, I. Moon and J. Kim, “CFD modeling on natural and forced ventilation during hydrogen leaks in a pressure regulator process of a residential area,” Process Safety and Environmental Protection, vol. 161, pp. 436-446, 2022. [20] C. S. E. S. H. B. B. S. K. M. G. J. T. R. a. G. W. W. Brian D. Ehrhart, HyRAM+ (Hydrogen Plus Other Alternative Fuels Risk Assessment Models), Version 4.1., Sandia National Laboratories, April 2022. [21] S. van Woudenberg, “WP 1B Gas Stations D1B.3a Ventilation,” HyDelta, 2022. [22] NPR7910-1, Gevarenzone-indeling met betrekking tot explosiegevaar - Deel 1: gasexplosiegevaar, gebaseerd op NEN-EN-IEC-60079-10-1 (2015), Nederlandse Praktijkrichtlijn, 2021. [23] T. Sniffers, “High Flow Sampler report HAS-kasten / Kiwa 2020. Projectnr. 200600A,” 2018. [24] G. H. C. S. a. M. S. B. Lowesmith, “Gas build-up in a domestic property following releases of methane/hydrogen mixtures,” International Journal of Hydrogen Energy, vol. 34, pp. 5932-5939, 2009. Pagina 84/123 WP6B – Veiligheid – Gasstations Tabel 9 – Samenstelling begeleidingsgroep/sparringsgroep Tabel 9 – Samenstelling begeleidingsgroep/sparringsgroep Tabel 9 – Samenstelling begeleidingsgroep/sparringsgroep Naam Werkgever Begeleidingsgroep Sparringsgroep R. den Hartog Westland V F. Verweij Westland V J. Jonkman Rendo V R. Scholten Rendo V P. Verstegen Alliander V R. Nispeling Alliander V R. Verhoeve Stedin V J. Palmers Coteq V J. Voogt Enexis V W. Koppenol Enexis V R. van Hooijdonk Enexis V M. van der Laan Kiwa Technology V S. van Woudenberg Kiwa Technology V De sparringsgroep is betrokken bij het beoordelen van de concept-rapportages. Pagina 85/123 Pagina 86/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Pagina 87/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof III Zonebepaling III Pagina 88/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof NPR-7910-1 (2021) In de richtlijn wordt een reguliere lekkage in een gasstation geclassificeerd als secundaire gevarenbron met als definitie “een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden” (zie par 5.5.2 van de NPR7910-1 (2021)). Daarbij is te denken aan minder dan 0,1% van de bedrijfsduur van een installatie of van de duur van een activiteit, waarbij de langste duur bepalend is zoals benoemd in de ATEX153 richtlijn. Tevens wordt in de IGEM/SR/25 Edition 2 (2010) [5] in appendix 7 aangehaald wat richtinggevende getallen zijn voor het falen van componenten in gasdrukregelstations. Deze genoemde waarden zijn vele malen lager dan 0,1% van de bedrijfsduur en mogen in die context als kort gezien worden. Van de Nederlandse situatie zijn dergelijke gegevens op het moment van schrijven niet voor alle netbeheerders beschikbaar. Tevens dient opgemerkt te worden dat de basis voor bovenstaande situatie altijd een installatie in bedrijf betreft (en dus niet tijdens werkzaamheden). Tijdens in bedrijf wordt de volgende definitie genoemd in de NEN1059 (2019), paragraaf 7.3.11 een gasdrukregelinstallatie een secundaire gevarenbron is. De installatie moet bij in bedrijfname en tijdens bedrijf technisch gasdicht zijn. De opstellingsruimte van de gasdrukregelinstallatie wordt geclassificeerd als gevarenzone 1, of als gevarenzone 2 indien kan worden aangetoond dat het ventilatievoud van de opstellingsruimte groter is dan 5 h–1 (classificatie volgens NEN-EN-IEC 60079-10-1). Naast de bepaling van de gevarenbron, speelt ook de ventilatie een belangrijke rol. Wanneer voor een behuizing wordt gekeken naar ventilatie, kan dit gedaan worden met behulp van de NPR7910-1 (2021)(Nederlandse Praktijkrichtlijn – gevarenzone-indeling met betrekking tot explosiegevaar) [24]. In paragraaf 8 (ventilatie in de omgeving van een gevarenbron) wordt volgens praktijkrichtlijn wordt geëvalueerd welke ventilatiecapaciteit een behuizing zou moeten hebben. Hierbij wordt onderscheid gemaakt tussen gematigde capaciteit (k = 0,25) en voldoende capaciteit (k = 0,1) waarbij de relatie tot het percentage LEL/LFL een maat is voor de beoordeling van deze capaciteit zoals beschreven in paragraaf 8.4.2.2 van de NPR. Tevens staat hier beschreven dat “bij voldoende capaciteit er bij de gevarenzone-indelingen geen rekening wordt gehouden met de kans dat het gas blijft hangen, aangezien de omstandigheden die daartoe kunnen leiden, slechts zelden voorkomen en dan gedurende korte tijd optreden”. De berekening in de NPR7910-1 (2021) bestaat uit een beknopt stappenplan. Invloed van wind en temperatuur Natuurlijke ventilatie is gedreven door een verschil in druk (binnen en buiten de behuizing) of een verschil in concentratie. Men spreekt hier van convectie of diffusie. Beide processen verlopen relatief langzaam en worden eerder gedomineerd door externe factoren zoals wind en/of temperatuur. Ventilatie kan bijvoorbeeld plaatsvinden door wind die tegen de behuizing blaast en via de ventilatieopeningen de lucht in de behuizing ververst. De effectieve windsnelheid heeft dus invloed op de verversing van lucht door natuurlijke trek. In het eerder aangehaalde artikel uit GAS [3] is eerder een conclusie gepresenteerd dat de ventilatie van een kaststation grotendeels door windinvloeden ontstaat waarbij vooral het totale oppervlak van de ventilatieopeningen en de mogelijkheid tot kruisventilatie van belang zijn. Ook wordt gesteld dat de kwaliteit van ventilatie maximaal rendeert wanneer de ventilatie openingen 5-10% van de kasthoogte onder het dak worden aangebracht. Plaatst men de ventilatie openingen te hoog, dan strijkt de dwarswind teveel over het gasmengsel in de kast en dient men het oppervlak van de ventilatieopeningen met een factor 2 of 3 te vermenigvuldigen om hetzelfde effect te krijgen. Tevens kan een behuizing ventileren door een gering verschil in temperatuur. Op warme dagen is de temperatuur in de behuizing hoger dan in de buitenlucht. Dit leidt tot een drijvende kracht die vele malen kleiner is dan middels natuurlijke trek. Met behulp van de effectieve windsnelheid kan dus de verversing van lucht in de behuizing berekend worden. Door de vereiste ventilatievoud (uit bovenstaande berekening) te vergelijken met de praktische ventilatievoud kan bepaald worden of de behuizing voldoende zal ventileren bij een bepaalde windsnelheid. In de NEN-EN-60079-10-1 (2021) wordt als stelregel 0,5 m/s aangehouden voor de windsnelheid in zowel een belemmerde als onbelemmerde gebied op een hoogte van 2 meter of minder boven het maaiveld (voor gas- en dampuitstromingen lichter dan lucht). Binnen in een behuizing zal de effectieve windsnelheid in de behuizing echter aanzienlijk lager liggen. Hiervoor kan met behulp van de NEN-EN-60079-10-1 (2021), formule C.2 een inschatting worden gemaakt die rond de 0,1 tot 0,2 m/s ligt wanneer de windsnelheid buiten de behuizing 0,5 m/s bedraagt. In de NEN1059;2019 wordt als stelregel aangehouden dat het beschikbare ventilatieoppervlak voor kasten/kaststations minimaal 2% van het vloeroppervlak moet bedragen. Vanuit andere standaarden zoals de EN12186;2014 wordt een referentiewaarde van slechts 1% van het vloeroppervlak genoemd. In stap 2 wordt de benodigde ventilatiecapaciteit gekoppeld aan het volume van de behuizing. Deze berekening behoort een ventilatievoud van de ruimte op te leveren die groter of gelijk is aan 5. De ventilatievoud, het aantal malen dat de lucht per uur wordt ververst (VV), wordt bepaald met de onderstaande formule; Is de ventilatievoud VV groter dan 1 en kleiner dan 5, dan wordt de ventilatiecapaciteit beschouwd als een gematigde capaciteit (verdunning tot 25% LEL/LFL mogelijk). Wanneer de ventilatievoud VV groter of gelijk is aan 5, dan wordt de ventilatiecapaciteit beschouwd als voldoende capaciteit (verdunning tot 10% LEL/LFL mogelijk). Deze classificatie zegt vooral iets over de benodigde mate van verdunning bij een lekkage, de lokaal heersende concentratie in en rond de behuizing volgt niet uit deze analyse. Ook volgt niet uit deze analyse of de gebouwde geometrie van een behuizing geschikt is om adequaat te kunnen ventileren. Het legt enkel een eis neer waaraan de behuizing zou moeten voldoen. NPR-7910-1 (2021) In stap 1 wordt de benodigde ventilatiecapaciteit van de behuizing berekend (zie 8.4). Met; Met; Voor gematigde capaciteit geldt dat k gelijk is aan 25% en voor voldoende capaciteit geldt dat k gelijk is aan 10%. Op deze manier zit in stap 1 een beoordeling van de LEL/LFL waarde (25% LEL of 10% LEL) voor de eis met betrekking tot de ventilatie, welke ventilatiecapaciteit is nodig om een lekdebiet terug te ventileren naar respectievelijk 25% LEL of 10% LEL. Daar dient wel opgemerkt te worden dat er vanuit wordt gegaan dat gas uniform verdund wordt, hetgeen beschreven wordt in paragraaf 8.4.2 van de NPR7910. Pagina 89/123 Pagina 89/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof In stap 2 wordt de benodigde ventilatiecapaciteit gekoppeld aan het volume van de behuizing. Deze berekening behoort een ventilatievoud van de ruimte op te leveren die groter of gelijk is aan 5. De ventilatievoud, het aantal malen dat de lucht per uur wordt ververst (VV), wordt bepaald met de onderstaande formule; Deze berekening is gemaakt voor de 1/2m³ kast in appendix IV voor verschillende lekdebieten bij aardgas en waterstof. Het resultaat van deze berekening is een oordeel over de geometrie op basis van ventilatiecapaciteit (wat is minimaal nodig voor het verlagen van de concentratie onder 25% LEL/LFL of 10% LEL/LFL) bij een specifieke lekkage. De kleurcode groen geeft aan dat de behuizing theoretisch voldoende kan ventileren bij de gestelde lekkages, de kleur oranje geeft aan dat de ventilatie bij die omstandigheden onvoldoende kan ventileren. Deze waarden kunnen met de experimenten vergeleken worden om te kijken of dit bij benadering overeenkomt. Ook kan met de bovenstaande gegevens (het type lekkage en de kwaliteit van de ventilatie) op basis van paragraaf 9.4 uit de NPR7910 (2021), tabel 7 worden vastgesteld wat de zonering bij de beschouwde situatie hoort. In het geval van een lekkage vanuit een secundaire bron in een gesloten gebouw met beperkte ventilatie volgens paragraaf 8.4.2 en een gematigde capaciteit (1 ≤ VV ≤ 5) leidt deze situatie tot een ATEX zonering 2. Ditzelfde geldt wanneer voldoende capaciteit (VV ≥ 5) beschikbaar is waarbij de afmeting van de zone in alle gevallen de gehele binnenzijde van een gasstation zal beslaan11. Hierin verschilt de aanpak van aardgas ten opzichte van waterstof slechts in het feit dat de brandbaarheidsgrenzen anders zijn en dat de uitstroom van gas in het geval van een lekkage groter zal zijn voor waterstof in vergelijking met aardgas waardoor de afmetingen van de zone ook anders zullen uitpakken. Hiervoor is op moment van schrijven nog geen onderzoek beschikbaar. Het veranderen van de zonering zal enkel plaats moeten vinden wanneer de classificering van de bron verandert van secundair naar primair. Ook kan gekeken worden naar het toepassen van een zone 2 NE (negligible extent, verwaarloosbare mate) waarbij de classificering wordt afgezwakt tot een verwaarloosbaar risico. Echter gaat men er dan vanuit dat een gasstation altijd technisch dicht is, waarbij uit eerder onderzoek [8] is gebleken dat dit in de praktijk niet haalbaar blijkt. 11 De genoemde afstanden in tabel 7 zijn van toepassing bij lekdebieten tussen de 1 gram en 10 gram per seconde. Deze waarden refereren vooral naar de huidige situatie met aardgas en zullen in de toekomst met waterstof heroverwogen moeten worden. Invloed van wind en temperatuur Op basis van de windsnelheid en het beschikbare ventilatie oppervlak kan bepaald worden hoe vaak een inhoud van de behuizing per tijdseenheid ververst kan worden. Dit is de combinatie van ventilatiecapaciteit (wat is nodig) en effectieve windsnelheid (wat is beschikbaar). Pagina 90/123 Pagina 90/123 NEN-EN-IEC-60079-10-1 (2021) De NEN-EN-IEC-60079-10-1 (2021)(Explosive atmospheres - Part 10-1: Classification of areas - explosive gas atmospheres) bevat een gedetailleerde beschrijving voor de zone bepaling van een technische installatie. Hierbij worden berekeningen gepresenteerd die, waar mogelijk, gecomplementeerd worden met praktische ervaring (of meetwaarden) als deze beschikbaar is. In appendix C en D van deze internationale standaard staat in stappen beschreven hoe een dergelijke analyse moet worden uitgevoerd. Een typische analyse bestaat uit de volgende stappen; ✓ Bepaal het type van de lekbron (continue/ primair/ secundair). ✓ Bepaal het lekdebiet van de lekbron. ✓ Bepaal de bijbehorende ATEX zonering aan de hand de mate van de mate van verdunning, de beschikbaarheid van ventilatie (betrouwbaarheid) en het type van de lekbron. ✓ Bepaal de bijbehorende ATEX zonering aan de hand de mate van de mate van verdunning, de beschikbaarheid van ventilatie (betrouwbaarheid) en het type van de lekbron. Aan de hand van deze stappen is de zone bepaling van een gasdrukregelstation uitgevoerd. De stappen zijn genummerd om een duidelijke link te maken met bovenstaande lijst die zo als checklist gebruikt kan worden . Stap 1; voor aardgas en waterstoflekkages wordt uitgegaan van een secundaire lekbron (“een secundaire gevarenbron is een gevarenbron van waaruit het vrijkomen van een brandbare stof onder normaal bedrijf niet waarschijnlijk is, en indien dit wel gebeurt, niet frequent en slechts gedurende korte perioden”). Stap 2; voor aardgas en waterstof kiezen we even een kleine lekkage, dus bijvoorbeeld een lekbron van 0,025 mm² bij een voordruk van 8 bar. Dit leidt tot een lekdebiet van 0,18 m3n/h aardgas en 0,56 m3n/h waterstof. Normaliter zou in stap 3 aan de hand van figuur C.1 uit de NEN-EN-IEC-60079-10-1 (2021) een inschatting gemaakt worden van de mate van verdunning in de behuizing. Op basis van experimenten zullen we leren dat bij een lekkage aardgas (of waterstof) bij 0,025 mm² bij een voordruk van 8 bar de concentraties in de 1/2m³ behuizing oplopen tot respectievelijk 6.0 vol% (voor aardgas) en 6.9 vol% (voor waterstof). In paragraaf 3.6.2 van de norm wordt gesteld dat de mate van verdunning moet als laag worden beschouwd als de achtergrondconcentratie hoger is dan 25 % van de LEL/LFL. De volgende stap in het proces is het vellen van een oordeel over de beschikbaarheid van ventilatie. De kwaliteit van de ventilatie wordt beïnvloed door de geometrie van de behuizing. Daarvoor wordt een kental voor de ventilatie-inefficiëntie toegevoegd (de f-waarde). Relatie veiligheidsafstanden en ATEX zonering In paragraaf 6.2 van de NEN 1059-2019 zijn veiligheidsafstanden bepaald voor alle categorieën gasstations ten opzichte van kwetsbare en beperkt kwetsbare objecten. De vraag of en hoe deze veiligheidsafstanden aangepast moeten worden voor stations op waterstof ligt buiten de scope van dit onderzoek. Pagina 91/123 11 De genoemde afstanden in tabel 7 zijn van toepassing bij lekdebieten tussen de 1 gram en 10 gram per seconde. Deze waarden refereren vooral naar de huidige situatie met aardgas en zullen in de toekomst met waterstof heroverwogen moeten worden. 11 De genoemde afstanden in tabel 7 zijn van toepassing bij lekdebieten tussen de 1 gram en 10 gram per seconde. Deze waarden refereren vooral naar de huidige situatie met aardgas en zullen in de toekomst met waterstof heroverwogen moeten worden. Pagina 91/123 NEN-EN-IEC-60079-10-1 (2021) De ventilatie-inefficiëntie is de mate waarin de lucht in de behuizing buiten de invloedssfeer van de lekopening goed gemengd wordt. f=1: de achtergrondconcentratie is uniform en de lekopening is ver verwijderd van de ventilatie openingen, zodat de concentratie bij de ventilatie openingen de gemiddelde achtergrond weergeeft concentratie. f>1: er is een gradiënt van achtergrondconcentratie in de behuizing als gevolg van inefficiënte menging en de ventilatie opening is “ver” verwijderd van de lekkage zelf, zodat de concentratie bij de uitlaat kleiner is dan de gemiddelde achtergrondconcentratie. De waarde van f kan tussen 1,5 liggen voor licht inefficiënt mengen en 5 voor zeer inefficiënt mengen. Deze factor zorgt voor correctie van niet Pagina 92/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof uniformiteit in luchtstromingspatronen in een behuizing met obstakels waar ventilatieopeningen niet optimaal gepositioneerd zijn voor maximale ventilatie effect. 1/2m³ kast – aardgas bij een lekopening van 0,025 mm² Afmetingen kast Inputs Lengte 1000 [mm] k1 (NPR) 10 % Breedte 500 [mm] k2 (NPR) 25 % Hoogte 1000 [mm] LEL (aardgas) 5,0 vol% M (aardgas) 18,636 [g/mol] Inhoud - bruto 0,5000 [m3] Windsnelheid 0,5 [m/s] Af 0,0750 [m3] Gehinderde windsnelheid, 20% effectief 0,1 [m/s] Inhoud - netto 0,4250 [m3] Vloeroppervlak 0,5 [m2] Aard van lekkage volgens NPR7910 secundair Ventilatie 2,0% ATEX zonering volgens tabel 7, NPR7910 zone 2 Ventilatie oppervlak 0,01 [m2] Berekeningen 0,025 [mm2] bij 8 [bar] 0,025 [mm2] bij 1 [bar] WINDLUW Lekopening 0,025 [mm2] Lekopening 0,025 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,04139 [g/s] Lekdebiet (a_m) 0,00917 [g/s] Lekdebiet (a) 0,17889 [m3/h(n)] Lekdebiet (a) 0,03996 [m3/h(n)] VC (25%) - minimaal nodig 14,31 (gematigd) VC (25%) - minimaal nodig 3,1968 (gematigd) VC (10%) - minimaal nodig 35,78 (goed) VC (10%) - minimaal nodig 7,992 (goed) Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatievoud (Vv) 8,5 Ventilatievoud (Vv) 8,5 Wat is minimaal nodig als VC (voor 25% LEL) 14,3 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 3,2 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 35,8 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 8,0 (goed) Welke Vv is beschikbaar obv windsnelheid 8,5 Welke Vv is beschikbaar obv windsnelheid 8,5 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Ja Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Ja Concentratie in behuizing (experiment) 6,0% (max) Concentratie in behuizing (experiment) 1,0% (max) WIND bij 0,5m/s Lekopening 0,025 [mm2] Lekopening 0,025 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? D6B.1A & D6B.1B Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof uniformiteit in luchtstromingspatronen in een behuizing met obstakels waar ventilatieopeningen niet optimaal gepositioneerd zijn voor maximale ventilatie effect. Uit eerdere experimenten in HyDelta is gebleken dat de aanwezige gasconcentratie in een behuizing verschilt. Er is een concentratiegradiënt aanwezig die oploopt met de verticale positie in de behuizing voor veel van de uitgevoerde metingen. De ventilatie-inefficiëntie zal dus als hoog moeten worden ingeschat voor de 1/2m³ behuizing. Wanneer tabel D.1 “Zones for grade of release and effectiveness of ventilation” uit de NEN-EN-IEC- 60079-10-1 (2021) wordt geraadpleegd, blijkt dat met het strikt volgen van de werkwijze en de resultaten uit de experimenten geconcludeerd zou moeten worden dat de mate van verdunning laag is (de achtergrondconcentratie hoger is dan 25 % van de LEL/LFL) waarbij een ATEX zone 1 moet worden toegekend. De sleutel tot succes ligt hier in het verbeteren van de ventilatie en het opschuiven naar links in de onderstaande tabel. Deze hele redenatie staat of valt met het kiezen van het lekdebiet en het definiëren van een realistische lekkage. Onderzoek naar lekkages bij gasstations toont aan dat lekken in de praktijk kleiner zijn dan in normen genoemde lekkages [10] [9]. Voor deze specifieke lekdebieten zou de conclusie dus anders kunnen uitpakken. Dit is een belangrijk onderdeel van dit onderzoek, welke keuze leidt tot welke gevolgen en wat weten we op dit moment vanuit de praktijk. Pagina 93/123 ( g ) g 1/2m³ kast – aardgas bij een lekopening van 0,25 mm² Afmetingen kast Inputs Lengte 1000 [mm] k1 (NPR) 10 % Breedte 500 [mm] k2 (NPR) 25 % Hoogte 1000 [mm] LEL (aardgas) 5,0 vol% M (aardgas) 18,636 [g/mol] Inhoud - bruto 0,5000 [m3] Windsnelheid 0,5 [m/s] Af 0,0750 [m3] Gehinderde windsnelheid, 20% effectief 0,1 [m/s] Inhoud - netto 0,4250 [m3] Vloeroppervlak 0,5 [m2] Aard van lekkage volgens NPR7910 secundair Ventilatie 2,0% ATEX zonering volgens tabel 7, NPR7910 zone 2 Ventilatie oppervlak 0,01 [m2] Berekeningen 0,25 [mm2] bij 8 [bar] 0,25 [mm2] bij 1 [bar] WINDLUW Lekopening 0,25 [mm2] Lekopening 0,25 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,41389 [g/s] Lekdebiet (a_m) 0,09167 [g/s] Lekdebiet (a) 1,78890 [m3/h(n)] Lekdebiet (a) 0,3996 [m3/h(n)] VC (25%) - minimaal nodig 143,11 (gematigd) VC (25%) - minimaal nodig 31,97 (gematigd) VC (10%) - minimaal nodig 357,78 (goed) VC (10%) - minimaal nodig 79,92 (goed) Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatievoud (Vv) 8,5 Ventilatievoud (Vv) 8,5 Wat is minimaal nodig als VC (voor 25% LEL) 143,1 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 32,0 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 357,8 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 79,9 (goed) Welke Vv is beschikbaar obv windsnelheid 8,5 Welke Vv is beschikbaar obv windsnelheid 8,5 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Concentratie in behuizing (experiment) 26,6% (max) Concentratie in behuizing (experiment) 11,9% (max) WIND bij 0,5m/s Lekopening 0,25 [mm2] Lekopening 0,25 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,04139 [g/s] Lekdebiet (a_m) 0,00917 [g/s] Lekdebiet (a) 0,17889 [m3/h(n)] Lekdebiet (a) 0,03996 [m3/h(n)] VC (25%) - minimaal nodig 14,31 (gematigd) VC (25%) - minimaal nodig 3,1968 (gematigd) VC (10%) - minimaal nodig 35,78 (goed) VC (10%) - minimaal nodig 7,992 (goed) Windsnelheid 0,50 [m/s] Windsnelheid 0,50 [m/s] Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatievoud (Vv) 42,4 Ventilatievoud (Vv) 42,4 Wat is minimaal nodig als VC (voor 25% LEL) 14,3 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 3,2 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 35,8 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 8,0 (goed) Welke Vv is beschikbaar obv windsnelheid 42,4 Welke Vv is beschikbaar obv windsnelheid 42,4 Is de Vv voldoende voor minder dan 25% LEL ? Ja Is de Vv voldoende voor minder dan 25% LEL ? Ja Is de Vv voldoende voor minder dan 10% LEL ? Ja Is de Vv voldoende voor minder dan 10% LEL ? Ja 1/2m³ kast – aardgas bij een lekopening van 0,025 mm² 1/2m³ kast – aardgas bij een lekopening van 0,025 mm² Pagina 94/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof WP6B – Veiligheid – Gasstations VALID VALID Lekdebiet (a_m) 0,04139 [g/s] Lekdebiet (a_m) 0,00917 [g/s] Lekdebiet (a) 1,78890 [m3/h(n)] Lekdebiet (a) 0,3996 [m3/h(n)] VC (25%) - minimaal nodig 143,11 (gematigd) VC (25%) - minimaal nodig 31,97 (gematigd) VC (10%) - minimaal nodig 357,78 (goed) VC (10%) - minimaal nodig 79,92 (goed) Windsnelheid 0,50 [m/s] Windsnelheid 0,50 [m/s] Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatievoud (Vv) 42,4 Ventilatievoud (Vv) 42,4 Wat is minimaal nodig als VC (voor 25% LEL) 143,1 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 32,0 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 357,8 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 79,9 (goed) Welke Vv is beschikbaar obv windsnelheid 42,4 Welke Vv is beschikbaar obv windsnelheid 42,4 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Ja Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee 1/2m³ kast – aardgas bij een lekopening van 0,25 mm² Pagina 95/123 WP6B – Veiligheid – Gasstations het distributie (lagedruk)net in Nederland met aardgas en waterstof 1/2m³ kast – waterstof bij een lekopening van 0,025 mm² Afmetingen kast Inputs Lengte 1000 [mm] k1 (NPR) 10 % Breedte 500 [mm] k2 (NPR) 25 % Hoogte 1000 [mm] LEL (waterstof) 4,0 vol% M (waterstof) 2,016 [g/mol] Inhoud - bruto 0,5000 [m3] Windsnelheid 0,5 [m/s] Af 0,0750 [m3] Gehinderde windsnelheid, 20% effectief 0,1 [m/s] Inhoud - netto 0,4250 [m3] Vloeroppervlak 0,5 [m2] Aard van lekkage volgens NPR7910 secundair Ventilatie 2,0% ATEX zonering volgens tabel 7, NPR7910 zone 2 Ventilatie oppervlak 0,01 [m2] Berekeningen 0,025 [mm2] bij 8 [bar] 0,025 [mm2] bij 1 [bar] WINDLUW Lekopening 0,025 [mm2] Lekopening 0,025 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,01389 [g/s] Lekdebiet (a_m) 0,00306 [g/s] Lekdebiet (a) 0,55800 [m3/h(n)] Lekdebiet (a) 0,12500 [m3/h(n)] VC (25%) - minimaal nodig 55,80 (gematigd) VC (25%) - minimaal nodig 12,5 (gematigd) VC (10%) - minimaal nodig 139,50 (goed) VC (10%) - minimaal nodig 31,25 (goed) Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatievoud (Vv) 8,5 Ventilatievoud (Vv) 8,5 Wat is minimaal nodig als VC (voor 25% LEL) 55,8 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 12,5 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 139,5 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 31,3 (goed) Welke Vv is beschikbaar obv windsnelheid 8,5 Welke Vv is beschikbaar obv windsnelheid 8,5 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee WIND bij 0,5m/s Lekopening 0,025 [mm2] Lekopening 0,025 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,01389 [g/s] Lekdebiet (a_m) 0,00306 [g/s] Lekdebiet (a) 0,55800 [m3/h(n)] Lekdebiet (a) 0,12500 [m3/h(n)] VC (25%) - minimaal nodig 55,80 (gematigd) VC (25%) - minimaal nodig 12,5 (gematigd) VC (10%) - minimaal nodig 139,50 (goed) VC (10%) - minimaal nodig 31,25 (goed) Windsnelheid 0,50 [m/s] Windsnelheid 0,50 [m/s] Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatievoud (Vv) 42,4 Ventilatievoud (Vv) 42,4 Wat is minimaal nodig als VC (voor 25% LEL) 55,8 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 12,5 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 139,5 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 31,3 (goed) Welke Vv is beschikbaar obv windsnelheid 42,4 Welke Vv is beschikbaar obv windsnelheid 42,4 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Ja Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Ja 1/2m³ kast – waterstof bij een lekopening van 0,025 mm² Pagina 96/123 WP6B – Veiligheid – Gasstations betrekking tot ventilatie in verschillende typen gasdrukregelstations i het distributie (lagedruk)net in Nederland met aardgas en waterstof 1/2m³ kast – waterstof bij een lekopening van 0,25 mm² Afmetingen kast Inputs Lengte 1000 [mm] k1 (NPR) 10 % Breedte 500 [mm] k2 (NPR) 25 % Hoogte 1000 [mm] LEL (waterstof) 4,0 vol% M (waterstof) 2,016 [g/mol] Inhoud - bruto 0,5000 [m3] Windsnelheid 0,5 [m/s] Af 0,0750 [m3] Gehinderde windsnelheid, 20% effectief 0,1 [m/s] Inhoud - netto 0,4250 [m3] Vloeroppervlak 0,5 [m2] Aard van lekkage volgens NPR7910 secundair Ventilatie 2,0% ATEX zonering volgens tabel 7, NPR7910 zone 2 Ventilatie oppervlak 0,01 [m2] Berekeningen 0,25 [mm2] bij 8 [bar] 0,25 [mm2] bij 1 [bar] WINDLUW Lekopening 0,25 [mm2] Lekopening 0,25 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,13944 [g/s] Lekdebiet (a_m) 0,03111 [g/s] Lekdebiet (a) 5,58000 [m3/h(n)] Lekdebiet (a) 1,25000 [m3/h(n)] VC (25%) - minimaal nodig 558,00 (gematigd) VC (25%) - minimaal nodig 125 (gematigd) VC (10%) - minimaal nodig 1395,00 (goed) VC (10%) - minimaal nodig 312,5 (goed) Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Gehinderde windsnelheid, 20% effectief 0,10 [m/s] Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatiecapaciteit obv windsnelheid 3,6 Ventilatievoud (Vv) 8,5 Ventilatievoud (Vv) 8,5 Wat is minimaal nodig als VC (voor 25% LEL) 558,0 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 125,0 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 1395,0 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 312,5 (goed) Welke Vv is beschikbaar obv windsnelheid 8,5 Welke Vv is beschikbaar obv windsnelheid 8,5 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Concentratie in behuizing (experiment) 24,0% (max) Concentratie in behuizing (experiment) 11,0% (max) WIND bij 0,5m/s Lekopening 0,25 [mm2] Lekopening 0,25 [mm2] Druk 8 [bar] Druk 1 [bar] Flow Sonic Flow Sonic Formula for flow correct ? VALID VALID Lekdebiet (a_m) 0,13944 [g/s] Lekdebiet (a_m) 0,03111 [g/s] Lekdebiet (a) 5,58000 [m3/h(n)] Lekdebiet (a) 1,25000 [m3/h(n)] VC (25%) - minimaal nodig 558,00 (gematigd) VC (25%) - minimaal nodig 125 (gematigd) VC (10%) - minimaal nodig 1395,00 (goed) VC (10%) - minimaal nodig 312,5 (goed) Windsnelheid 0,50 [m/s] Windsnelheid 0,50 [m/s] Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatiecapaciteit obv windsnelheid 18,0 Ventilatievoud (Vv) 42,4 Ventilatievoud (Vv) 42,4 Wat is minimaal nodig als VC (voor 25% LEL) 558,0 (gematigd) Wat is minimaal nodig als VC (voor 25% LEL) 125,0 (gematigd) Wat is minimaal nodig als VC (voor 10% LEL) 1395,0 (goed) Wat is minimaal nodig als VC (voor 10% LEL) 312,5 (goed) Welke Vv is beschikbaar obv windsnelheid 42,4 Welke Vv is beschikbaar obv windsnelheid 42,4 Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 25% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee Is de Vv voldoende voor minder dan 10% LEL ? Nee 1/2m³ kast – waterstof bij een lekopening van 0,25 mm² Pagina 97/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Foto’s van de testopstelling Foto’s van de testopstelling Figuur 58. Het 4m³ kaststation in een buitensituatie Figuur 59. Het 4m³ kaststation met daarin het gasdrukregelstation Figuur 60. De HAS kast in een buitensituatie Figuur 61. De HAS kast met daarin de hogedrukafleverstation Figuur 56. De ½ m3 kast in een buitensituatie Figuur 57. De ½ m3 kast met daarin het gasregelstation Figuur 57. De ½ m3 kast met daarin het gasregelstation Figuur 57. De ½ m3 kast met daarin het gasregelstation Figuur 56. De ½ m3 kast in een buitensituatie Figuur 56. De ½ m3 kast in een buitensituatie Figuur 59. Het 4m³ kaststation met daarin het gasdrukregelstation Figuur 59. Het 4m³ kaststation met daarin het gasdrukregelstation Figuur 59. Het 4m³ kaststation met daarin het gasdrukregelstation Figuur 58. Het 4m³ kaststation in een buitensituatie Figuur 60. De HAS kast in een buitensituatie Figuur 61. De HAS kast met daarin de hogedrukafleverstation Figuur 61. De HAS kast met daarin de hogedrukafleverstation Figuur 60. De HAS kast in een buitensituatie Pagina 98/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Figuur 62. Set nozzles met verschillende diameters Figuur 63. Nozzle met lekopening van 0,25 mm² in de 1/2m³ kast Figuur 63. Nozzle met lekopening van 0,25 mm² in de 1/2m³ kast Figuur 62. Set nozzles met verschillende diameters Figuur 65. Riken Keiki NP 1000 Waterstofdetector Figuur 64. MultiRAE Lite IR Aardgasdetector Figuur 65. Riken Keiki NP 1000 Waterstofdetector Figuur 64. MultiRAE Lite IR Aardgasdetector Pagina 99/123 WP6B – Veiligheid – Gasstations VI Gebruikte meetapparatuur VI Gebruikte meetapparatuur Tabel 10 – Gebruikte meetapparatuur details Tabel 10 – Gebruikte meetapparatuur details Omschrijving Fabrikaat en type Kiwa-nr Aardgasdetector MultiRAE - Lite IR 114033 Aardgasdetector MultiRAE Lite IR 114034 Aardgasdetector MultiRAE Lite IR 114036 Aardgasdetector MultiRAE Lite IR 114037 Aardgasdetector MultiRAE Lite IR 114038 Aardgasdetector MultiRAE Lite IR 114039 Aardgasdetector MultiRAE Lite IR 114040 Aardgasdetector MultiRAE Lite IR 114041 Aardgasdetector MultiRAE Lite IR 114043 Waterstofdetector Riken Keiki NP 1000 - MFC (aardgas/ waterstof) Bronkhorst, type F-201CV-RAD-22-V - MFC (aardgas/ waterstof) Bronkhorst, type F-203AV-M50-RBD-FF-V - Aardgas Aardgas - De Multirae hebben een meetbereik van 0 - 100% LEL/LFL en 0-100vol% aardgas). / g ) - De vier Multirae sensoren die zijn gebruikt als meetpunt op grotere afstand (op 0,5 m en 1 m benedenwinds) van het gasdrukregelstation hebben een meetbereik van 0 - 100% LEL/LFL en 0- 100vol% aardgas. Waterstof Waterstof - De Riken Keiki sensoren hebben een meetbereik van 0 – 100 vol% waterstof. - De vier Multirae sensoren die zijn gebruikt als meetpunt op grotere afstand (op 0,5 m en 1 m benedenwinds) van het gasdrukregelstation hebben een meetbereik van 0-1000 ppm waterstof. Tevens zijn deze specifieke Multirae sensoren uitgerust met een 0 – 100% LEL/LFL sensor voor waterstof. Pagina 100/123 VII Overzicht alle metingen VII Overzicht alle metingen 1/2m³ kast Tabel 6 – 1/2m³ kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Tabel 6 – 1/2m³ kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. ) lekkage levert aardgas concentraties op boven de UEL maar omdat altijd ergens een overgang in concentratie zal zijn die binnen de brandbaarheidsgrenzen van het gas valt, zijn deze waarden rood weergegeven. Aardgas/ natural gas 1/2m3 kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,179 8 6,0% 6,0% 6,0% 2,0% 0,1% - 100 0,04 1 1,0% 1,0% 0,5% 0,1% 0,4% - - Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 1,8 * 8 26,6% 21,5% 25,1% 2,7% 23,1% 7,0% >1000 0,4 1 11,9% 11,3% 11,6% 4,0% 10,8% - - Waterstof/ hydrogen 1/2m3 kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,558 8 6,9% 6,3% 6,4% 4,5% 6,4% - 580 0,125 1 2,7% 2,9% 2,6% 1,4% 3,2% - 260 Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 5,8 8 24,0% 20,0% 20,0% 18,0% 21,0% 18,0% >1000 1,25 1 11,0% 10,0% 11,0% 6,5% 11,0% - 580 ) lekkage levert aardgas concentraties op boven de UEL maar omdat altijd ergens een overgang in concentratie zal zijn die binnen de brandbaarheidsgrenzen van het gas valt, zijn deze waarden rood weergegeven. ekkage levert aardgas concentraties op boven de UEL maar omdat altijd ergens een overgang in concentratie zal zijn die binnen de andbaarheidsgrenzen van het gas valt, zijn deze waarden rood weergegeven. Pagina 101/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof 4m³ kaststation Tabel 7 – 4m³ kaststation resultaten - windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Tabel 7 – 4m³ kaststation resultaten - windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Aardgas/ natural gas 4m3 kaststation Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,179 8 0,8% 0,2% 0,7% 0,4% 0,4% - - 0,04 1 0,1% 0,0% 0,0% 0,0% 0,1% - - Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 1,8 8 4,3% 2,6% 1,0% 3,7% 3,7% - - 0,4 1 1,1% 0,9% 0,3% 0,5% 1,8% - - Waterstof/ hydrogen 4m3 kaststation Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 0,558 8 2,2% 0,9% 1,6% 1,1% 1,3% - 70 0,125 1 1,0% 0,3% 0,4% 0,4% 0,4% - 10 Lekopening 0,25 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% N O Z W %LFL ppm 5,8 8 12,0% 9,1% 4,0% 9,3% 6,0% - 570 1,25 1 5,5% 4,1% 4,0% 3,6% 4,1% - 120 Pagina 102/123 VIII Meetresultaten – ½ m³ kast VIII Meetresultaten – ½ m³ kast HAS kast De inhoud van de HAS kast bedraagt circa 0,06 m³. Alleen gemeten met 0,025 mm², geen data beschikbaar voor de lekopening van 0,25 mm². Tabel 8 – HAS kast resultaten – windluwe situatie voor aardgas (boven) en waterstof (onder). LEL/LFL aardgas ≥ 5 vol% en LEL/LFL waterstof ≥ 4 vol%. Buiten de behuizing is gemeten op 0,5 meter afstand. Aardgas/ natural gas HAS kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% Links Rechts %LFL ppm 0,179 8 10,3% 9,1% 7,9% 0,0% 0 0,04 1 3,3% 3,9% 2,7% 0,0% 0 Waterstof/ hydrogen HAS kast Lekopening 0,025 mm^2 Lekdebiet Druk Max concentratie behuizing Max concentratie ventilatieopening op 0,5 m nm3/hr bar vol% vol% Links Rechts %LFL ppm 0,558 8 12,0% 12,0% 12,0% 0,0% 410 0,125 1 4,8% 5,6% 4,1% 0,0% 20 * experiment voortijdig gestopt (ivm veiligheid) door snelle opbouw van waterstof concentratie. Pagina 103/123 Lekdebiet 1,8 m³n/hr – aardgas (lekopening 0,25 mm²) – 8 bar Bij een lek van 1,8 m3n/h loopt de aardgas concentratie in de ½ m3 behuizing op tot een maximum concentratie van 13 vol%. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 66. Concentratie (vol % CH4) in de ½ m³ kast bij een lekdebiet van 1,8 m³(n)/h Figuur 66. Concentratie (vol % CH4) in de ½ m³ kast bij een lekdebiet van 1,8 m³(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 21,5vol% 25,1vol% 2,7vol% 23,1vol% Figuur 67. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 1,8 m3(n)/h Gedurende het experiment loopt de concentratie aardgas buiten het gasdrukregelstation incidenteel op tot maximaal 7% LEL/LFL. Deze concentratie wordt gemeten op 0,5 meter (west) afstand van de behuizing op 1 meter hoogte vanaf de grond. Figuur 67. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 1,8 m3(n)/h Gedurende het experiment loopt de concentratie aardgas buiten het gasdrukregelstation incidenteel op tot maximaal 7% LEL/LFL. Deze concentratie wordt gemeten op 0,5 meter (west) afstand van de behuizing op 1 meter hoogte vanaf de grond. Pagina 104/123 Pagina 104/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 5,6 m3n/hr – waterstof (lekopening 0,25 mm²) – 8 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 5,6 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 22 vol% in de behuizing. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Door de hoge concentratie is de experiment na korte tijd gestopt. grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Door de hoge concentratie is de experiment na korte tijd gestopt. Figuur 68. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 21,0 vol% 21,0 vol% 18,0 vol% 22,0 vol% Figuur 69. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Gedurende het experiment loopt de concentratie waterstof buiten het gasdrukregelstation incidenteel op tot maximaal 18% LEL/LFL. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en die werd ook bereikt. Figuur 68. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Figuur 68. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Figuur 68. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof 68. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 21,0 vol% 21,0 vol% 18,0 vol% 22,0 vol% Figuur 69. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Noord Oost Zuid West Concentratie waterstof 21,0 vol% 21,0 vol% 18,0 vol% 22,0 vol% Figuur 69. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof 69. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Gedurende het experiment loopt de concentratie waterstof buiten het gasdrukregelstation incidenteel op tot maximaal 18% LEL/LFL. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en die werd ook bereikt. Gedurende het experiment loopt de concentratie waterstof buiten het gasdrukregelstation incidenteel op tot maximaal 18% LEL/LFL. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en die werd ook bereikt. Pagina 105/123 Pagina 105/123 Lekdebiet 0,40 m3n/hr – aardgas (lekopening 0,25 mm²) – 1 bar Bij een lek van 0,40 m3n/h loopt de aardgas concentratie in de ½ m3 behuizing op tot een maximum concentratie van 11,9 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen invloed van wind waargenomen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 70. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,40 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Figuur 70. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,40 m3(n)/h Figuur 70. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,40 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 11,3vol% 11,6vol% 4,0vol% 10,8vol% Figuur 71. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 0,40 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 71. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 0,40 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 106/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar Noord Oost Zuid West Concentratie waterstof 10,0 vol% 11,0 vol% 6,5 vol% 11,0 vol% Figuur 72. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Figuur 72. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Figuur 72. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie w (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 10,0 vol% 11,0 vol% 6,5 vol% 11,0 vol% Figuur 73. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 580 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test Noord Oost Zuid West Concentratie waterstof 10,0 vol% 11,0 vol% 6,5 vol% 11,0 vol% Figuur 73. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 580 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. 73. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 1,25 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 11 vol% in de behuizing. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Door de hoge concentratie is de experiment na korte tijd gestopt. circa 1,25 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 11 vol% in de behuizing. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Door de hoge concentratie is de experiment na korte tijd gestopt. Figuur 72. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 10,0 vol% 11,0 vol% 6,5 vol% 11,0 vol% Figuur 73. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 580 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Figuur 72. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Deze hebben een meetbereik tot 1000 ppm en werd 580 ppm bereikt. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 580 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 107/123 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar Bij de eerste test is de kleinste lekopening conform de IGEM/SR/25 Edition 2 gecreëerd bij een voordruk van 8 bar in de ½ m3 behuizing waarbij de concentratie oploopt tot maximaal 6,0 vol%. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Bij de eerste test is de kleinste lekopening conform de IGEM/SR/25 Edition 2 gecreëerd bij een voordruk van 8 bar in de ½ m3 behuizing waarbij de concentratie oploopt tot maximaal 6,0 vol%. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 74. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Figuur 74. Concentratie (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm² bij 8 bar Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatieopeningen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 6,0vol% 6,0vol% 2,0vol% 0,1vol% Figuur 75. Concentratie bij de ventilatie openingen (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd incidenteel 100 ppm bereikt. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. guur 75. Concentratie bij de ventilatie openingen (vol% aardgas) in de ½ m3 kast bij een lek van 0,025 mm2 bij 8 bar Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd incidenteel 100 ppm bereikt. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Pagina 108/123 Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 6,9 vol% in de behuizing. behuizing. De concentratie op alle meetpunten in de behuizing stabiliseert na ongeveer 10 minuten en blijft daarna nagenoeg gelijk tijdens de hele test. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Figuur 76. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 77. Concentratie (vol% H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. De concentratie op alle meetpunten in de behuizing stabiliseert na ongeveer 10 minuten en blijft daarna nagenoeg gelijk tijdens de hele test. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Figuur 76. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof Figuur 76. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof * Figuur 76. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. g g j Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 77. Concentratie (vol% H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Noord Oost Zuid West Concentratie waterstof 6,3vol% 6,4vol% 4,5vol% 6,4vol% Figuur 77. Concentratie (vol% H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Figuur 77. Concentratie (vol% H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet 0,56 m3(n)/h waters Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werden waarden tot 580 ppm bereikt aan de westzijde van de behuizing. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. g ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 109/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar Bij een lek van 0,040 m3n/h loopt de aardgas concentratie in de ½ m3 behuizing op tot een maximum concentratie van 1 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is minimale invloed van wind waargenomen waardoor lichte preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 78. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,040 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in Figuur 78. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,040 m3(n)/h Figuur 78. Concentratie (vol % CH4) in de ½ m3 kast bij een lekdebiet van 0,040 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 1,0vol% 0,5vol% 0,1vol% 0,4vol% Figuur 79. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 0,040 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 79. Concentratie (vol% CH4) bij de ventilatieopeningen van een ½ m3 kast bij een lekdebiet van 0,040 m3(n)/ Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 110/123 het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 2,7 vol% in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Figuur 80. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 2,9 vol% 2,6 vol% 1,4 vol% 3,1 vol% Figuur 81. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 260 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 2,7 vol% in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Figuur 80. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof * Figuur 80. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Figuur 80. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Concentratie (vol % H2) in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 2,9 vol% 2,6 vol% 1,4 vol% 3,1 vol% Figuur 81. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 260 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Noord Oost Zuid West Concentratie waterstof 2,9 vol% 2,6 vol% 1,4 vol% 3,1 vol% Figuur 81. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 260 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. 81. Concentratie (vol % H2) bij de ventilatie openingen in de ½ m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 260 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 260 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 111/123 Meetresultaten – 4 m3 kaststation Lekdebiet 1,8 m3n/hr – aardgas (lekopening 0,25 mm²) – 8 bar Bij een lek van 1,8 m3n/h loopt de aardgas concentratie in de 4 m3 behuizing op tot een maximum concentratie van 4,3 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 82. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 1,8 m3(n)/h Figuur 82. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 1,8 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 2,6vol% 1,0vol% 3,7vol% 3,7vol% Figuur 83. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4 m3 kast bij een lekdebiet van 1,8 m3(n)/h Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. Deze h bb b k d d l b k Figuur 83. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4 m3 kast bij een lekdebiet van 1,8 m3(n)/h Op een afstand van 0,5 meter van de behuizing is ook gemeten met sensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd incidenteel 100 ppm bereikt. Op een afstand van 0,5 meter werden geen LEL/LFL waarden gedetecteerd. Pagina 112/123 Pagina 112/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 5,6 m3n/hr – waterstof (lekopening 0,25 mm²) – 8 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 5,6 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 12 vol% in de behuizing. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is slechts een geringe beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). seconden). Figuur 84. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 9,1 vol% 4,0 vol% 9,3 vol% 6,0 vol% Figuur 85. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 570 ppm bereikt. Op een afstand van 0,5 Figuur 84. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof * Figuur 84. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof * Figuur 84. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 9,1 vol% 4,0 vol% 9,3 vol% 6,0 vol% Figuur 85. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof O f d 0 5 d b h i i i k f 1 Figuur 84. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 9,1 vol% 4,0 vol% 9,3 vol% 6,0 vol% Figuur 85. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Noord Oost Zuid West Concentratie waterstof 9,1 vol% 4,0 vol% 9,3 vol% 6,0 vol% Figuur 85. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof 85. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 5,6 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 570 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 113/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,4 m3n/hr – aardgas (lekopening 0,25 mm²) – 1 bar Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 1,25 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 5,5 vol% in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een geringe beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Door de hoge concentratie is de experiment na korte tijd gestopt. Figuur 88. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 4,1 vol% 4,0 vol% 3,6 vol% 4,1 vol% Figuur 89. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 120 ppm bereikt. Figuur 88. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Figuur 88. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 4,1 vol% 4,0 vol% 3,6 vol% 4,1 vol% Figuur 89. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 120 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Noord Oost Zuid West Concentratie waterstof 4,1 vol% 4,0 vol% 3,6 vol% 4,1 vol% Figuur 89. Bij een lek van 0,4 m3n/h loopt de aardgas concentratie in de 4 m3 behuizing op tot een maximum concentratie van 3,3 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 86. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,4 m3(n)/h Figuur 86. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,4 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 0,9vol% 0,3vol% 0,5vol% 2,8vol% Figuur 87. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4 m3 kast bij een lekdebiet van 0,4 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 87. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4 m3 kast bij een lekdebiet van 0,4 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 114/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 1,25 m3n/hr – waterstof (lekopening 0,25 mm²) – 1 bar Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 120 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. 89. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 1,25 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 120 ppm bereikt. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 120 ppm bereikt. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 115/123 Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar Bij een lek van 0,18 m3n/h loopt de aardgas concentratie in de 4 m3 behuizing op tot een maximum concentratie van 0,8 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 90. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,18 m3(n)/h Figuur 90. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,18 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 0,2vol% 0,7vol% 0,4vol% 0,4vol% Figuur 91. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4m3 kast bij een lekdebiet van 0,18 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 91. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4m3 kast bij een lekdebiet van 0,18 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 116/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 2,6 vol% in de behuizing. De concentratie op alle meetpunten in de behuizing stabiliseert na ongeveer 10 minuten en blijft daarna nagenoeg gelijk tijdens de hele test. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. weergegeven als functie van de tijd in minuten. Figuur 92. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 0,9vol% 1,6vol% 1,1vol% 1,3vol% Figuur 93. Concentratie (vol% H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 70 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Figuur 92. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof Figuur 92. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,56 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie waterstof 0,9vol% 1,6vol% 1,1vol% 1,3vol% Figuur 93. Concentratie (vol% H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Noord Oost Zuid West Concentratie waterstof 0,9vol% 1,6vol% 1,1vol% 1,3vol% Figuur 93. Concentratie (vol% H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof guur 93. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Concentratie (vol% H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 70 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 70 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 117/123 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar Bij een lek van 0,04 m3n/h loopt de aardgas concentratie in de 4 m3 behuizing op tot een maximum concentratie van 0,2 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 94. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,04 m3(n)/h Figuur 94. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,04 m3(n)/h Figuur 94. Concentratie (vol % CH4) in de 4 m3 kast bij een lekdebiet van 0,04 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Noord Oost Zuid West Concentratie aardgas 0,0vol% 0,0vol% 0,0vol% 0,1vol% Figuur 95. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4m3 kast bij een lekdebiet van 0,04 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 95. Concentratie (vol% CH4) bij de ventilatieopeningen van een 4m3 kast bij een lekdebiet van 0,04 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 118/123 het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 1,0 vol% in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Figuur 96. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 0,3 vol% 0,4 vol% 0,4 vol% 0,4 vol% Figuur 97. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 10 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. ) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Wanneer met dezelfde voordruk een lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie van maximaal 1,0 vol% in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is geen beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Figuur 96. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Figuur 96. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof * Figuur 96. Concentratie (vol % H2) in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof * Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten (en seconden). Noord Oost Zuid West Concentratie waterstof 0,3 vol% 0,4 vol% 0,4 vol% 0,4 vol% Figuur 97. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof O f t d 0 5 t d b h i i i k t t t t f 1 t Noord Oost Zuid West Concentratie waterstof 0,3 vol% 0,4 vol% 0,4 vol% 0,4 vol% Figuur 97. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof 97. Concentratie (vol % H2) bij de ventilatie openingen in de 4 m3 kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 10 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 10 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. *) de sensor laag in de behuizing heeft alleen nulwaarden geregistreerd gedurende deze test. Pagina 119/123 Meetresultaten – HAS kast Meetresultaten – HAS kast Meetresultaten – HAS kast Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar Lekdebiet 0,18 m3n/hr – aardgas (lekopening 0,025 mm²) – 8 bar Bij een lek van 0,18 m3n/h loopt de aardgas concentratie in de HAS kast op tot een maximum concentratie van 10,9 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 98. Concentratie (vol % CH4) in de HAS kast bij een lekdebiet van 0,18 m3(n)/h Figuur 98. Concentratie (vol % CH4) in de HAS kast bij een lekdebiet van 0,18 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Links voor Rechts voor Concentratie aardgas 9,1vol% 7,9vol% Figuur 99. Concentratie (vol% CH4) bij de ventilatieopeningen van een HAS kast bij een lekdebiet van 0,18 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 99. Concentratie (vol% CH4) bij de ventilatieopeningen van een HAS kast bij een lekdebiet van 0,18 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 120/123 Pagina 120/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar Lekdebiet 0,56 m3n/hr – waterstof (lekopening 0,025 mm²) – 8 bar Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,56 m3n/h. Dit lek leidde tot een waterstof concentratie in de HAS kast van maximaal 12 vol%. De test is korter dan andere testen in verband met de aanzienlijke gasconcentratie die in korte tijd opbouwt in de behuizing. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een geringe beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten. Figuur 100. Concentratie (vol % waterstof) in de HAS kast bij een lekdebiet van 0,56 m3(n)/h waterstof Figuur 100. Concentratie (vol % waterstof) in de HAS kast bij een lekdebiet van 0,56 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Links voor Rechts voor Concentratie waterstof 12,0vol% 12,0vol% Figuur 101. Concentratie (vol % H2) bij de ventilatie openingen in de HAS kast bij een lekdebiet van 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 410 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. 101. Concentratie (vol % H2) bij de ventilatie openingen in de HAS kast bij een lekdebiet van 0,56 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd 410 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Pagina 121/123 Lekdebiet 0,04 m3n/hr – aardgas (lekopening 0,025 mm²) – 1 bar Bij een lek van 0,04 m3n/h loopt de aardgas concentratie in de HAS kast op tot een maximum concentratie van 3,3 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een kleine invloed van wind waargenomen waardoor preferentie ontstaat op de ventilatieopeningen. In de onderstaande grafiek is de concentratie aardgas (in vol%) weergegeven als functie van de tijd in minuten. Figuur 102. Concentratie (vol % CH4) in de HAS kast bij een lekdebiet van 0,04 m3(n)/h Figuur 102. Concentratie (vol % CH4) in de HAS kast bij een lekdebiet van 0,04 m3(n)/h Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie aardgas (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie aardgas weergegeven als functie van de tijd in minuten. Links voor Rechts voor Concentratie aardgas 3,9vol% 2,7vol% Figuur 103. Concentratie (vol% CH4) bij de ventilatieopeningen van een HAS kast bij een lekdebiet van 0,04 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Figuur 103. Concentratie (vol% CH4) bij de ventilatieopeningen van een HAS kast bij een lekdebiet van 0,04 m3(n)/h Op een afstand van 0,5 meter van de kast, is op geen enkel moment een aardgas concentratie gemeten, deze is daarmee onder de 0,1% LEL/LFL gebleven. Pagina 122/123 WP6B – Veiligheid – Gasstations D6B.1A & D6B.1B – Inventarisatie, modellering en experimenten met betrekking tot ventilatie in verschillende typen gasdrukregelstations in het distributie (lagedruk)net in Nederland met aardgas en waterstof Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar Lekdebiet 0,125 m3n/hr – waterstof (lekopening 0,025 mm²) – 1 bar Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie in de HAS kast van maximaal 5,8 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een geringe beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten en seconden. Figuur 104. Concentratie (vol % waterstof) in de HAS kast bij een lekdebiet van 0,125 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Links voor Rechts voor Concentratie waterstof 5,6vol% 4,1vol% Figuur 105. Concentratie (vol % H2) bij de ventilatie openingen in de HAS kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 20 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Wanneer met dezelfde voordruk het gedefinieerde lek met waterstof wordt gecreëerd, is het gemeten lekdebiet circa 0,125 m3n/h. Dit lek leidde tot een waterstof concentratie in de HAS kast van maximaal 5,8 vol%. De invloed van wind is enkel bekeken als deze experimenten in de windluwe situatie verstoorde. Gedurende deze meting is een geringe beïnvloeding waargenomen. In onderstaande grafiek is de concentratie waterstof weergegeven als functie van de tijd in minuten en seconden. Figuur 104. Concentratie (vol % waterstof) in de HAS kast bij een lekdebiet van 0,125 m3(n)/h waterstof Figuur 104. Concentratie (vol % waterstof) in de HAS kast bij een lekdebiet van 0,125 m3(n)/h waterstof Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Onderstaande grafiek en tabel laat het tijdsafhankelijke beeld van de gemeten concentratie waterstof (in vol%) aan de ventilatie openingen met de maximale concentratie in rood. Wederom is de concentratie waterstof weergegeven als functie van de tijd in minuten. Links voor Rechts voor Concentratie waterstof 5,6vol% 4,1vol% Figuur 105. Concentratie (vol % H2) bij de ventilatie openingen in de HAS kast bij een lekdebiet van 0,125 m3(n)/h waterstof Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 20 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. 105. Concentratie (vol % H2) bij de ventilatie openingen in de HAS kast bij een lekdebiet van 0,125 m3(n)/h watersto Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 20 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Op een afstand van 0,5 meter van de behuizing is ook gemeten met waterstofsensoren op 1 meter hoogte. Deze hebben een meetbereik tot 1000 ppm en werd slechts 20 ppm bereikt. Op een afstand van 0,5 meter werd geen LEL/LFL waarden gedetecteerd. Pagina 123/123
https://openalex.org/W3116032052	https://www.researchsquare.com/article/pex-946/v1.pdf?c=1631848963000	English	null	Zebrafish retinal mRNA RNA-seq data processing	Research Square (Research Square)	2,020	cc-by	1,810	Zebrafish retinal mRNA RNA-seq data processing Nicholas Owen Nicholas Owen UCL Institute of Ophthalmology, London, UK https://orcid.org/0000-0001-5598-6274 Mariya Moosajee (  m.moosajee@ucl.ac.uk ) UCL Institute of Ophthalmology, London, UK https://orcid.org/0000-0003-1688-5360 Nicholas Owen UCL Institute of Ophthalmology, London, UK https://orcid.org/0000-0001-5598-6274 Mariya Moosajee (  m.moosajee@ucl.ac.uk ) ute of Ophthalmology, London, UK https://orcid.org/0000-0003-1688-5360 UCL Institute of Ophthalmology, London, UK https://orcid.org/0000-0003-1688-5360 Method Article Keywords: RNA-seq, zebrafish, spatio-temporal transcriptome, embryo development, retina, optic fissur Posted Date: May 19th, 2020 DOI: https://doi.org/10.21203/rs.3.pex-946/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Keywords: RNA-seq, zebrafish, spatio-temporal transcriptome, embryo development, retina, optic fissure Posted Date: May 19th, 2020 DOI: https://doi.org/10.21203/rs.3.pex-946/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/5 Abstract This protocol details the step-by-step procedures followed to process zebrafish retinal mRNA sequencing data generated by the SMARTSeq2 library preparation protocols in the manuscript Richardson et al 2019 1. Equipment Access to a high-performance computer cluster is recommended. Alternatively, a desktop workstation running either Linux or Mac OS with 32 GB RAM and high-end CPU(s). Cloud computing platforms such as Amazon Web Services (AWS) or Elastic Cloud (EC1) are also appropriate. Requisite software and packages include R 2/R Studio 3 (version 3.5 or higher), FASTQC 4, multiQC 5, Trim Galore 6, STAR 7, HTSEQ 8, DESeq2 9, GOseq 10. Introduction This protocol describes the bioinformatic data processing pipeline for the associated manuscript. Procedure 1. Raw sequencing data from mRNA SMARTSeq2 libraries (100bp paired-end [PE]) was converted from bcl to FASTQ format in BaseSpace per sample. 1. Raw sequencing data from mRNA SMARTSeq2 libraries (100bp paired-end [PE]) was converted from bcl to FASTQ format in BaseSpace per sample. 2. All FASTQ reads were assessed for quality control using FASTQC and FastQ Screen. FastQ Screen is highly recommended as it assesses the library for sequence origin, ensuring the data match expectations. 2. All FASTQ reads were assessed for quality control using FASTQC and FastQ Screen. FastQ Screen is highly recommended as it assesses the library for sequence origin, ensuring the data match expectations. 3. Sequencing adapters (Illumina) and low quality read bases were trimmed using Trim Galore, removing reads with a quality Phred score under 6. 3. Sequencing adapters (Illumina) and low quality read bases were trimmed using Trim Galore, removing reads with a quality Phred score under 6. 4. To align the read sequences to the zebrafish genome, the complete genomic sequence (build GCRz10) file (FASTA) and GTF (general transfer format) / GFF (general feature format) (version 95) were obtained from Ensembl (downloaded: FA : ftp://ftp.ensembl.org/pub/release-95/fasta/danio_rerio/dna/ GTF: ftp://ftp.ensembl.org/pub/release-95/gtf/danio_rerio/) 4. To align the read sequences to the zebrafish genome, the complete genomic sequence (build GCRz10) file (FASTA) and GTF (general transfer format) / GFF (general feature format) (version 95) were obtained from Ensembl (downloaded: FA : ftp://ftp.ensembl.org/pub/release-95/fasta/danio_rerio/dna/ GTF: ftp://ftp.ensembl.org/pub/release-95/gtf/danio_rerio/) Page 2/5 5. Appropriate index files were created using STAR (v2.7.1a): STAR --runMode genomeGenerate --runThreadN 4 -genomeSAindexNbases 8 --genomeChrBinNbits 14 -- genomeDir ./STARIndex/ --genomeFastaFiles ./GRCz11/Danio_rerio.GRCz11.dna.toplevel.fa --sjdbGTFfile ./GRCz11/Danio_rerio.GRCz11.95.gtf --sjdbOverhang 99 STAR --runMode genomeGenerate --runThreadN 4 -genomeSAindexNbases 8 --genomeChrBinNbits 14 -- genomeDir ./STARIndex/ --genomeFastaFiles ./GRCz11/Danio_rerio.GRCz11.dna.toplevel.fa --sjdbGTFfile ./GRCz11/Danio_rerio.GRCz11.95.gtf --sjdbOverhang 99 6. PE FASTQ reads were aligned using STAR (v2.7.1a) in two-pass mode and guided by the GTF/GFF annotation: STAR --readFilesIn ${fq1} ${fq2} --readFilesCommand zcat --genomeDir ./STARIndex/ --runThreadN 4 -- twopassMode Basic --outFileNamePrefix ${sample_id} --outSAMtype BAM SortedByCoordinate -- outSAMunmapped Within --outSAMheaderHD @HD VN:1.4 --outSAMattrRGline ID:${sample_id} CN:UCL DS:RNAseq LB:Truseq PL:${library.platform} SM:${sample_id} 7. Resulting SAM files were sorted by coordinate, converted to BAM format, and indexed using samtools (v1.9): 7. Resulting SAM files were sorted by coordinate, converted to BAM format, and indexed using samtools (v1.9): samtools sort ${input.sam} -o ${output.sorted.bam} -O bam / samtools index ${output.sorted.bam} samtools sort ${input.sam} -o ${output.sorted.bam} -O bam / samtools index ${output.sorted.bam} 8. Procedure After mapping, read duplicates were marked using the Picard (v2.20.4) MarkDuplicates command and gene (or transcript) level counts calculated with a module of HTSeq (v0.11.3), htseq-count keeping duplicate reads: htseq-count --order=pos --stranded=no - ./GRCz11/Danio_rerio.GRCz11.95.gtf > ./counts/${BAM_file_name}_htseq_counts_keepdups.tsv 9. Quality of mapping was carried out, summarizing the number of mapped, multimapped, unmapped, as well as mapped to exons, intergenic, or intragenic regions using RNA-SeQC/picard metrics/Qualimap. 9. Quality of mapping was carried out, summarizing the number of mapped, multimapped, unmapped, as well as mapped to exons, intergenic, or intragenic regions using RNA-SeQC/picard metrics/Qualimap. 10. Summary reports of all metrics were created using MultiQC (v1.8) 5. 11. The R computing environment (>3.5) and Bioconductor packages, DESeq2 (v1.28.0) was used for statistical modelling of the count data, carrying out pairwise comparisons of optic fissure (OF) tissue and dorsal retina (DR) tissue at each specific developmental time point using WALD testing to generate p values. Metadata of the samples should include; sample id, condition, timepoint. Per sample count data was imported using tximport package, creating a single count dataframe. Low counts were filtered and removed (basemean <10) before modelling. Pairwise comparisons were carried out using specific contrasts called in the DESeq2 command. This ensured the complete dataset was preprocessed and consistent for all comparisons. For the time course analysis, DESeq2 was used with a likelihood-ratio test (LRT) to generate significance values on the complete data set using the interaction of time with tissue origin factors (time:tissue) . 11. The R computing environment (>3.5) and Bioconductor packages, DESeq2 (v1.28.0) was used for statistical modelling of the count data, carrying out pairwise comparisons of optic fissure (OF) tissue and dorsal retina (DR) tissue at each specific developmental time point using WALD testing to generate p values. Metadata of the samples should include; sample id, condition, timepoint. Per sample count data was imported using tximport package, creating a single count dataframe. Low counts were filtered and removed (basemean <10) before modelling. Pairwise comparisons were carried out using specific contrasts called in the DESeq2 command. This ensured the complete dataset was preprocessed and consistent for all comparisons. For the time course analysis, DESeq2 was used with a likelihood-ratio test (LRT) to generate significance values on the complete data set using the interaction of time with tissue origin factors (time:tissue) . Page 3/5 Page 3/5 12. Procedure The data was explored using several graphical plots using ggplot2 in R 11, including the MA plot – this shows the log2 fold change, per feature, plotted against the mean of normalized counts, for all the samples. An overview of the level of similarity was created using a sample-to-sample distance heatmap, using hierarchical clustering. Principle component analysis (PCA plot) was used to show the samples in a 2D space, separated by the first two principle components, key to identification of outliers and batch effects. All plots can be saved as resolution independent SVG images. 12. The data was explored using several graphical plots using ggplot2 in R 11, including the MA plot – this shows the log2 fold change, per feature, plotted against the mean of normalized counts, for all the samples. An overview of the level of similarity was created using a sample-to-sample distance heatmap, using hierarchical clustering. Principle component analysis (PCA plot) was used to show the samples in a 2D space, separated by the first two principle components, key to identification of outliers and batch effects. All plots can be saved as resolution independent SVG images. 13. Differentially expressed genes (DEGs) were defined in this dataset as those having an absolute log2 fold change >= 1 and an FDR <= 0.05. 14. (optional) Genes can be further annotated using the biomaRt package 12; providing common gene name, chromosomal location, biotype and strand specificity as well as associated protein identifiers and homologues in a number of species. 15. Gene ontology over representation analysis was carried out using GOseq, which has the benefit of considering any length bias in the data. Alternative tools include GOrilla, DAVID and Enrichr (accessible through web interfaces and APIs). 16. Hierarchical clustering of the identified DEGs can be used to identify clades/groups of differentially expressed genes, either up or down regulated in samples represented together colour coded in the heatmap output. The rows of the heatmap (representing an individual gene/transcript) are reordered according to the clustering result, putting similar observations close to one another. Clustering used the rlog transformed assay data from the DESeq2 object. 16. Hierarchical clustering of the identified DEGs can be used to identify clades/groups of differentially expressed genes, either up or down regulated in samples represented together colour coded in the heatmap output. Procedure The rows of the heatmap (representing an individual gene/transcript) are reordered according to the clustering result, putting similar observations close to one another. Clustering used the rlog transformed assay data from the DESeq2 object. 17. For inter sample comparisons of specific gene(s) expression, transcripts per million (TPM) values were calculated 13 and appropriate plots created. 17. For inter sample comparisons of specific gene(s) expression, transcripts per million (TPM) values were calculated 13 and appropriate plots created. Anticipated Results Please see associated Publication References 1 Richardson, R. et al. Transcriptome profiling of zebrafish optic fissure fusion. Sci Rep 9, 1541, doi:10.1038/s41598-018-38379-5 (2019). Page 4/5 6 Krueger, F. Trim Galore!, <https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/> (2012). 7 Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15-21, doi:10.1093/bioinformatics/bts635 (2013). 8 Anders, S., Pyl, P. T. & Huber, W. HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166-169, doi:10.1093/bioinformatics/btu638 (2015). 9 Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA- seq data with DESeq2. Genome Biol 15, 550, doi:10.1186/s13059-014-0550-8 (2014). 10 Young, M. D., Wakefield, M. J., Smyth, G. K. & Oshlack, A. Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biol 11, R14, doi:10.1186/gb-2010-11-2-r14 (2010). 11 Wickham, H. ggplot2: Elegant Graphics for Data Analysis. (Springer-Verlag New York, 2016). 12 Durinck, S., Spellman, P. T., Birney, E. & Huber, W. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc 4, 1184-1191, doi:10.1038/nprot.2009.97 (2009). 13 Wagner, G. P., Kin, K. & Lynch, V. J. Measurement of mRNA abundance using RNA-seq data: RPKM measure is inconsistent among samples. Theory Biosci 131, 281-285, doi:10.1007/s12064-012-0162-3 (2012). Page 4/5 2 R Core Team. (Vienna, Austria, 2017). 2 R Core Team. (Vienna, Austria, 2017). 3 RStudio Team. (Boston, MA, 2015). 4 Andrews, S. FASTQC - A Quality Control tool for High Throughput Sequence Data, <http://www.bioinformatics.babraham.ac.uk/projects/fastqc/> (2014). 5 Ewels, P., Magnusson, M., Lundin, S. & Kaller, M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047-3048, doi:10.1093/bioinformatics/btw354 (2016). 5 Ewels, P., Magnusson, M., Lundin, S. & Kaller, M. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047-3048, doi:10.1093/bioinformatics/btw354 (2016). 6 Krueger, F. Trim Galore!, <https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/> (2012). Acknowledgements This research was funded by the Wellcome Trust (205174/Z/16/Z). Page 5/5 Page 5/5
https://openalex.org/W1506862630	https://journals.plos.org/plosntds/article/file?id=10.1371/journal.pntd.0003542&type=printable	English	null	Managing the Fight against Onchocerciasis in Africa: APOC Experience	PLoS neglected tropical diseases	2,015	cc-by	4,932	HISTORICAL PROFILES AND PERSPECTIVES Managing the Fight against Onchocerciasis in Africa: APOC Experience Grace Fobi1, Laurent Yameogo1, Mounkaila Noma1, Yaovi Aholou1, Joseph B. Koroma1, Honorat M. Zouré1, Tony Ukety2, Paul-Samson Lusamba-Dikassa1, Chris Mwikisa1, Daniel A. Boakye1, Jean-Baptist Roungou1 1 World Health Organization/African Programme for Onchocerciasis Control (WHO/APOC), Ouagadougou, Burkina Faso, 2 World Health Organization (WHO), Geneva, Switzerland DBoakye@noguchi.ug.edu.gh * DBoakye@noguchi.ug.edu.gh OPEN ACCESS OPEN ACCESS Citation: Fobi G, Yameogo L, Noma M, Aholou Y, Koroma JB, Zouré HM, et al. (2015) Managing the Fight against Onchocerciasis in Africa: APOC Experience. PLoS Negl Trop Dis 9(5): e0003542. doi:10.1371/journal.pntd.0003542 Editor: Jeremiah M. Ngondi, University of Cambridge, UNITED KINGDOM Published: May 14, 2015 Copyright: © 2015 Fobi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Citation: Fobi G, Yameogo L, Noma M, Aholou Y, Koroma JB, Zouré HM, et al. (2015) Managing the Fight against Onchocerciasis in Africa: APOC Experience. PLoS Negl Trop Dis 9(5): e0003542. doi:10.1371/journal.pntd.0003542 doi:10.1371/journal.pntd.0003542 Editor: Jeremiah M. Ngondi, University of Cambridge, UNITED KINGDOM Published: May 14, 2015 Copyright: © 2015 Fobi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The African Programme for Onchocerciasis Control (APOC) was launched in December 1995. In order to reach its objective of onchocerciasis control in all endemic countries in sub- Saharan Africa, the Programme used Rapid Epidemiological Mapping of Onchocerciasis (REMO) [3] to delineate areas of mesoendemicity and hyperendemicity and to estimate the population at high risk of contracting onchocerciasis. Countries included in the APOC pro- gram were: Angola, Burundi, Cameroon, Central African Republic, Chad, Congo, Democratic Republic of Congo, Equatorial Guinea, Ethiopia, Gabon, Kenya, Liberia, Malawi, Mozambique, Nigeria, Rwanda, South Sudan, Sudan, Uganda, and Tanzania. The exercise revealed that 102 million people in the Programme area were at risk and needed ivermectin treatment, while an estimated 37 million people were already infected with the disease [4]. In 1997, APOC adopted community-directed treatment with ivermectin (CDTi) as its core strategy [5–7]. Following CDTi introduction and implementation, coverage and compliance with ivermectin steadily Copyright: © 2015 Fobi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: None of the authors received financial contribution for the development of the manuscript nor for submission of the manuscript for publication. Competing Interests: The authors have declared that no competing interests exist. Introduction Due to the socioeconomic impact of human onchocerciasis (commonly referred to as river blindness) in West Africa, the Onchocerciasis Control Programme in the Volta River Basin (OCP) was instituted [1]. This initial programme started in 1975 and covered seven West Afri- can countries: Benin, Burkina Faso, Cote d’Ivoire, Ghana, Mali, Niger, and Togo. However, later evidence indicated that endemic areas outside the initial area posed a threat to the achievement of the OCP and, hence, the Programme was extended southward and westward to include four additional countries, bringing the total number of countries covered by OCP to eleven. The formal name was then changed to the Onchocerciasis Control Programme in West Africa, retaining the acronym OCP. OCP used aerial larviciding as its principle strategy to control the vectors of human oncho- cerciasis, members of the Simulium damnosum complex, in the absence of a safe drug for mass treatment against the parasites [2]. Efforts to control onchocerciasis evolved in 1987 when iver- mectin was donated to kill the juvenile worms that cause the various symptoms associated with the disease. As a result of the donation, OCP instituted a new strategy of chemotherapy in com- bination with vector control. In the 11 countries covered by OCP, this two-prong approach led to the virtual elimination of onchocerciasis as a public health problem and as an obstacle to so- cioeconomic development. The availability of a donated drug effective against the parasite and safe for mass drug administration, coupled with evidence that other pathological effects of on- chocerciasis were equally important socioeconomic threats, led to the decision that onchocerci- asis should be controlled in all endemic countries in Africa (Fig 1). PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 averted from 2011–2015 [8]. Through co-implementation activities, APOC has also averted an additional 1 million DALYs for other targeted diseases such as ascariasis, trichuriasis, hook- worm, lymphatic filariasis, strongyloidiasis, and epidermal parasitic skin diseases over the du- ration of the Programme [9]. Research now shows that ivermectin treatment can not only control, but in many areas (Mali, Senegal, Uganda, and Nigeria), eliminate river blindness infection and interrupt trans- mission [10–12]. In 2009, taking into account the feasibility of the elimination of onchocercia- sis infection and interruption of its transmission with ivermectin mass treatment alone [10], the Joint Action Forum (JAF), the governing body of APOC (described below), directed the Programme to shift from control to elimination of onchocerciasis. In 2010, the third midterm evaluation of APOC advised the JAF that it would be premature to close the Programme in 2015 given the perspective of onchocerciasis elimination. Thus, in 2011, JAF reaffirmed its en- dorsement for the Programme to pursue the elimination of onchocerciasis in Africa as well as co-implementation of preventive chemotherapy interventions for other selected neglected tropical diseases (NTDs) in the context of increased support to community-level health sys- tems strengthening. The other preventable NTDs susceptible to mass drug administration in- clude lymphatic filariasis (elephantiasis), trachoma, schistosomiasis (bilharzia), and soil- transmitted helminths, which include roundworm (ascariasis), whipworm (trichuris), and hookworm. Participative Governance The Programme is a unique partnership between the affected communities, governments, bi- lateral and multilateral agencies, foundations, non-governmental development organizations (NGDOs), the scientific community, and the private sector. The partnership is built on a legal agreement called the “Memorandum of the African Programme for Onchocerciasis Control” [13]. The APOC Secretariat is responsible for initiating the budget process, taking the lead in preparing a multi-year plan of action for APOC and using this to develop an indicative budget to implement the multi-year plan. APOC is governed by the JAF, consisting of representatives of (a) the participating countries; (b) the contributing development partners; (c) the sponsoring agencies; (d) members of the NGDO Coordination Group; (e) Merck & Co., Inc. representing the private sector as the donor of ivermectin; (f) intergovernmental regional or sub-regional or- ganizations; and (g) other invited entities. The JAF decides on the overall policy and strategy of APOC, assesses progress review, approves the APOC Plan of Action and Budget, and assesses global financing requirements of the Programme. The JAF meets annually and is usually hosted alternatively by a participating country and a donor country. The Committee of Sponsoring Agencies (CSA), comprising representatives from sponsoring agencies, NGDO Coordination Group, Merck & Co., Inc., and the Mectizan Donation Program (MDP), works closely with the APOC Secretariat. CSA makes interim decisions on behalf of JAF when required, follows closely the financial situation of APOC, and scrutinizes documen- tation for the JAF. WHO is the executing agency within this partnership and the World Bank is the fiscal agent of the Programme [13]. The World Bank mobilizes donor contributions into the APOC Trust Fund and provides funds for APOC’s operations. Although APOC provides some contribution to the NGDO Coordination group through the Trust Fund, most of the funds from the NGDO group to countries are raised by the individual members of the group. OPEN ACCESS 1 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 Fig 1. Onchocerciasis-endemic countries in Africa, showing countries covered by the OCP and initially by APOC. Map from 2010. Note that South Sudan gained independence in 2011, becoming the 20th APOC country. Fig 1. Onchocerciasis-endemic countries in Africa, showing countries covered by the OCP and initially by APOC. Map from 2010. Note that South Sudan gained independence in 2011, becoming the 20th APOC country. ntries in Africa, showing countries covered by the OCP and initially by APOC. Map from 2010. Note that South becoming the 20th APOC country. Fig 1. Onchocerciasis-endemic countries in Africa, showing countries covered by the OCP and initially by A Sudan gained independence in 2011, becoming the 20th APOC country. doi:10.1371/journal.pntd.0003542.g001 doi:10.1371/journal.pntd.0003542.g001 improved—the number of persons benefiting from ivermectin treatment increased from 1.5 million in 1997 to 75.8 million in 2010 and to 100.79 million in 2013. CDTi ensured a sustain- able method to deliver ivermectin and also strengthened health systems. Long-term impact assessments of APOC operations revealed a decrease in the number of persons infected from 37.9 million in 1995 to 15.1 million in 2011. An estimated 9.5 million cases of severe itching were prevented, 400,000 persons were protected from low vision, and 200,000 persons were protected from blindness. In most advanced APOC projects, the preva- lence of infection is already close to zero. The operations of APOC prevented 8.9 million dis- ability-adjusted life years (DALYs) from 2005–2010, with another estimated 10.1 million 2 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 Sustaining Partnership APOC’s broad partnership includes the poor in programmatic decision-making. This partner- ship involves over 146,000 local communities, African endemic countries, donor countries and institutions, over 16 NGDOs, Merck & Co. Inc., research institutions, and programs such as the WHO Special Programme for Research and Training in Tropical Diseases (TDR), as well as research institutions within onchocerciasis-endemic countries. which is headed by the APOC Director. WHO Headquarters provides administrative and tech- nical as well as operational research support. APOC maintains close collaboration with WHO offices of all participating countries and with National Onchocerciasis Task Forces (described below) in the implementation and monitoring of CDTi projects. The Secretariat of the Programme is based in Ouagadougou, Burkina Faso, and runs two technical units: Sustainable Drug Distribution Unit and Epidemiology and Vector Elimination Unit. Administrative support is provided through the Director’s Office Coordination Unit. APOC currently supports onchocerciasis control and elimination activities in 31 African coun- tries, including the 19 original signatories of the Memorandum, South Sudan, and the 11 ex- OCP participating countries. These countries included: Angola, Benin, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Congo, Côte d’Ivoire, Democratic Republic of Congo, Equatorial Guinea, Ethiopia, Gabon, Ghana, Guinea, Guinea Bissau, Kenya, Liberia, Malawi, Mali, Mozambique, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, South Sudan, Sudan, Tanzania, Togo, and Uganda. The main activities implemented by the Secretariat in- clude the design, execution, monitoring, and evaluation of community-directed ivermectin dis- tribution projects, as well as mapping of the disease, capacity building for countries, and provision of technical guidance in efforts to control and eliminate river blindness in Africa. Programme Management APOC is one of the few programs that the WHO Regional Office for Africa implements direct- ly. The WHO Regional Director for Africa ensures the overall guidance of APOC Secretariat, 3 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 Engaging Governments and Other Stakeholders All of the APOC participating countries have established a National Onchocerciasis Task Force (NOTF) composed of Ministry of Health (MoH) staff from relevant divisions, representatives of implementing NGDO partners, research institutions and representatives from other related Ministries (e.g., Ministry of Education). The National Onchocerciasis Control Programme (NOCP) serves as the Secretariat of the NOTF. The NOTF has the responsibility of overseeing implementation of the onchocerciasis con- trol efforts at the national level. The MoH has the critical role to create a favorable environment for all partners and enabling policies for community-directed interventions for the control and elimination of onchocerciasis and other NTDs targeted by preventive chemotherapy (PC-NTDs); ensuring entry of ivermectin and other NTD medicines in the country without imposing duty, tax, or other charges; as well as chairing and expanding the NOTF to include coordination of control and elimination of onchocerciasis and other PC-NTDs. The MoH also advocates for and mobilizes national financial contributions. Between 2010 and 2011, govern- ments of 15 countries disbursed US$16,937,214 for CDTi-related equipment and salaries of health personnel of various CDTi implementation units. Those countries included: Angola, Burundi, Cameroun, Central African Republic, Chad, Congo, Democratic Republic of the Congo, Equatorial Guinea, Ethiopia, Liberia, Malawi, Sudan, Nigeria, Uganda and United Re- public of Tanzania. These governments also disbursed US$3,012,750 to core CDTi activities alone [18]. However, in order to ensure elimination of onchocerciasis, the governments of par- ticipating countries need to increase financial and human resources for the following core CDTi activities: health education, sensitization advocacy and mobilization, training, distribu- tion of ivermectin, supervision, monitoring, and reporting. intervention implementation and service absorption within the village and to reach expected results. Using trained community-directed distributors (CDDs) selected by the community, APOC has been able to scale up treatment with ivermectin from 1.5 million in 1995 to over 75 million people in 2010. Engaging communities has yielded multiple health gains to remote communities, providing additional health interventions and commodities such as medicines for the control of other preventable NTDs, insecticide-treated bed nets for malaria control, and vitamin A supplementation. Ensuring Medicine Availability In 1987, Merck committed to donate Mectizan (ivermectin, MSD) for the treatment of oncho- cerciasis to all countries that need it for as long as necessary. In 1998, the donation was expand- ed to the treatment of lymphatic filariasis in the African countries where onchocerciasis and lymphatic filariasis are co-endemic. Medicine donation programs are critical as they cover a major technical and financial component of the control and elimination of onchocerciasis and other PC-NTDs. Sustaining the action of such programs is a key determinant for success in the fight against onchocerciasis and other NTDs. Between 1997 and 2011, the Mectizan Donation Programme (MDP) supplied 2,168,732,700 tablets to APOC participant countries. In 2011 alone, 352,594,500 (77%) tablets were distributed to APOC participant countries and 107,939,500 (23%) tablets were distributed to ex-OCP member countries for onchocerciasis or for integrated lymphatic filariasis and onchocerciasis control [19]. Engaging Communities The CDTi strategy is built on a participation paradigm in which communities play an impor- tant role for planning, leading, and managing interventions that benefit their own health. In the CDTi context, a community is the lowest autonomous unit whose members are linked to one traditional or political head and share resources in common. CDTi focuses on empowering communities to take responsibility for ivermectin delivery by deciding how, when, and by whom the ivermectin treatment should be administered. This strategy seeks to empower the people who are the most affected to take specific roles, responsibilities, and critical decisions for interventions that address their needs. CDTi, using a bottom-up approach, is a well-tested and highly cost-effective strategy that has extended the reach of essential interventions to those at the end of the road at a reasonably low cost. The success of CDTi, especially in remote areas of countries affected by conflict and war, has opened the doors to a number of other health care interventions that lend themselves to the community-directed intervention approach (CDI) [4,14]. There is evidence that at least four to five interventions could effectively be implemented through CDI with a boost in iver- mectin coverage by 10% [15,16]. Empowered communities have contributed to improving the prevailing therapeutic cover- age with ivermectin from 62.2% to at least 65% and geographical coverage to 100% [17]. CDTi/ CDI is undertaken at the community level under the direction of the community itself. The health services and NGDOs introduce the concept of community ownership and role, whereby the community takes charge of the process. Without community engagement in planning, designing, implementing, and monitoring, it is difficult for an external agent to identify the various social factors that will influence 4 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 The NGDO Coordination Group for Onchocerciasis Control provides managerial, technical, and financial support to more than 80% of CDTi projects in participating countries. NGDOs also get involved in opera- tional research activities. During the last few years, intra-country collaboration among mem- bers of the Group has allowed addressing the financial constraints faced by some of their members in order to sustain support to CDTi projects. CDTi Projects for Field Operations APOC has delineated CDTi project areas using data obtained through REMO. Each CDTi project covers a delineated geographic area of an endemic country. The CDTi project approach allows for a phased introduction of CDTi in a country and focuses support on the early phases of CDTi development with the view of applying the lessons learned to the rest of the country. In 2010, treatment activities in 16 APOC participating countries covered 138,448 out of 144,837 (96%) endemic communities [19]. In total, 75.8 million people were treated with an av- erage therapeutic coverage of 79.0% in countries with a stable security situation and 71.4% in post-conflict countries [19]. APOC presently supports 122 CDTi projects in 20 APOC partici- pant countries and four ex-OCP member countries, Cote d’Ivoire, Ghana, Guinea Bissau, and Sierra Leone. became known as the NGDO Coordination Group for Onchocerciasis Control at the launch of APOC in 1995. The membership of the Group increased to 12 NGDOs in order to address the need of expansion of the Programme in Cameroon, Nigeria, and Uganda. Since 2005, addition- al NGDOs have joined the Group as the result of increased momentum towards controlling NTDs. To date, the Group comprises 14 full members and three associate members. The full members include: Charitable Society for Social Welfare (CSSW), Christoffel-Blindenmission (CBM), Helen Keller International (HKI), IMA World Health, Lions Club International Foun- dation (SightFirst Program), Malaria Consortium, Mectizan Donation Program (MDP), Mis- sion to Save the Helpless (MITOSATH), Organisation pour la Prévention de la Cécité (OPC), Schistosomiasis Control Initiative (SCI), SightSavers International, The Carter Center, United Front Against River Blindness (UFAR), and US Fund for UNICEF. The three associate mem- bers include: International Agency for the Prevention of Blindness (IAPB), Liverpool Centre for Neglected Tropical Diseases (CNTD), and Merck & Co, Inc. The NGDO Coordination Group for Onchocerciasis Control provides managerial, technical, and financial support to more than 80% of CDTi projects in participating countries. NGDOs also get involved in opera- tional research activities. During the last few years, intra-country collaboration among mem- bers of the Group has allowed addressing the financial constraints faced by some of their members in order to sustain support to CDTi projects. became known as the NGDO Coordination Group for Onchocerciasis Control at the launch of APOC in 1995. The membership of the Group increased to 12 NGDOs in order to address the need of expansion of the Programme in Cameroon, Nigeria, and Uganda. Since 2005, addition- al NGDOs have joined the Group as the result of increased momentum towards controlling NTDs. To date, the Group comprises 14 full members and three associate members. The full members include: Charitable Society for Social Welfare (CSSW), Christoffel-Blindenmission (CBM), Helen Keller International (HKI), IMA World Health, Lions Club International Foun- dation (SightFirst Program), Malaria Consortium, Mectizan Donation Program (MDP), Mis- sion to Save the Helpless (MITOSATH), Organisation pour la Prévention de la Cécité (OPC), Schistosomiasis Control Initiative (SCI), SightSavers International, The Carter Center, United Front Against River Blindness (UFAR), and US Fund for UNICEF. The three associate mem- bers include: International Agency for the Prevention of Blindness (IAPB), Liverpool Centre for Neglected Tropical Diseases (CNTD), and Merck & Co, Inc. Coordinating NGDO Actions The WHO Programme for the Prevention of Blindness created the NGDO Coordination Group for Ivermectin Distribution in partnership with seven NGDOs in 1991, which later 5 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 Technical Oversight and Evaluation The APOC Technical Consultative Committee (TCC) ensures the technical oversight of APOC operations. Its main function is to review new CDTi projects plans and budget, annual techni- cal reports from CDTi projects and operational research proposals. TCC thus contributes to es- tablishing a research agenda for APOC. Its recommendations are addressed to the Programme Director or, if required, to the CSA. The TCC members meet twice a year. The TCC comprises 12 members which are selected through various mechanisms. One of the 12 TCC members is a representative from MDP and is appointed by Merck & Co. Inc. Eleven members are scientists/ experts appointed by the WHO Director-General based upon the recommendation of the CSA. Among those, three members are proposed by the NGDO Coordination Group for the consid- eration of the CSA. The other eight members are suggested by APOC management to the CSA for their consideration. TCC members appointed by the Executing Agency hold membership for three years renewable for a maximum of another three years, on a staggered basis. However, since MDP oversees the donation of the drug ivermectin, the MDP representative has perma- nent tenure on the TCC. The review function of TCC has been partly devolved to some countries, including Camer- oon, Nigeria, and Uganda, where Technical Review Committees (TRC) have been established 6 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 Fig 2. Possible evolution of APOC structures and mechanisms. doi:10 1371/journal pntd 0003542 g002 Fig 2. Possible evolution of APOC structures and mechanisms. Fig 2. Possible evolution of APOC structures and mechanisms. doi:10.1371/journal.pntd.0003542.g002 doi:10.1371/journal.pntd.0003542.g002 doi:10.1371/journal.pntd.0003542.g002 to review annual technical reports of mature projects (defined as distributing ivermectin to en- demic communities for seven or more years) on behalf of the TCC, and to support in-country operational research agenda in relation with CDTi implementation. An external evaluation sys- tem was established since the inception of the Programme. Evaluations are undertaken every five years, supported financially by the donor community and organized by the CSA. The eval- uation teams are composed of scientists with a relevant background in public health. Three mid-term external evaluations have been undertaken in 2000, 2005, and 2010. They have made recommendations that allowed the JAF to make decisions on the mandate and future of APOC [20–22]. The next evaluation is planned for 2015. PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 References 1. WHO/OCP. (1994) Twenty years of onchocerciasis control in West Africa: review of the work of the On- chocerciasis Control Programme in West Africa from 1974–1994: Introduction. Geneva: WHO pp175. 2. Davies JB, Le Berre R, Walsh JF, Cliff B. (1978) Onchocerciasis and simulium control in the Volta River Basin. Mosquito News; 38 (4): 466–472. 3. Noma M, Nwoke BEB, Nutall I, Tambala PA, Enyong P, et al. (2002) Rapid epidemiological mapping of onchocerciasis (REMO): its application by the African Programme for Onchocerciasis Control (APOC). Ann Trop Med Parasitol; 96 Suppl 1: S29–S39. PMID: 12081248 4. Amazigo U, Okeibunor J, Matovu V, Zoure H, Bump J, et al. (2007) Performance of predictors: evaluat- ing sustainability in community-directed treatment projects of the African programme for onchocerciasis control. Soc Sci Med; 64(2007) 2070–2082. PMID: 17383061 5. Amazigo U. (2008) The African Programme for Onchocerciasis Control (APOC). Ann Trop Med Parasi- tol; 102 Suppl 1:19–22. doi: 10.1179/136485908X337436 PMID: 18718149 6. Dadzie KY. (1997) Onchocerciasis control: the APOC strategy. Afr Health; 19: 13–15. PMID: 12292398 7. Homeida M, Braide E, Elhassan E, Amazigo UV, Liese B, et al. (2002) APOC's strategy of community- directed treatment with ivermectin (CDTi) and its potential for providing additional health services to the poorest populations. Ann Trop Med Parasitol; 96 Suppl 1: S93–104. PMID: 12081254 8. Coffeng LE, Stolk WA, Zouré HGM, Veerman JL, Agblewonu KB, et al. (2014) African Programme for Onchocerciasis Control 1995–2015: updated health impact estimates based on new disability weights. PLOS Negl Trop Dis 8(6): e2759. doi: 10.1371/journal.pntd.0002759 PMID: 24901642 9. de Vlas SJ. (2011) Health Impact Assessment of APOC—Update 2011. Department of Public Health, Erasmus MC, University Medical Center Rotterdam, The Netherlands. PMID: 19301709 10. Diawara L, Traoré MO, Badji A, Bissan Y, Doumbia K, et al. (2009) Feasibility of onchocerciasis elimi- nation with ivermectin treatment in endemic foci in Africa: first evidence from studies in Mali and Sene- gal. PLoS Negl Trop Dis 3(7): e497. doi: 10.1371/journal.pntd.0000497 PMID: 19621091 11. Katabarwa MN, Walsh F, Habomugisha P, Lakwo TL, Agunyo S, Oguttu DW, et al. (2012). Transmis- sion of Onchocerciasis in Wadelai Focus of Northwestern Uganda has been interrupted and the dis- ease eliminated. J of Parasitol Res. 2012, 1–7. 12. Tekle A.H., Elhassan E., Isiyaku S., Amazigo U.V., Bush S., Noma M., et al. (2012). into a new regional entity, provisionally named Programme for the Elimination of Neglected Diseases in Africa (PENDA) [23,24]. This new entity will have a mandate for “the coordination of the implementation of the elimination of onchocerciasis and lymphatic filariasis, and sup- port interventions for other PC-NTDs in Africa” [24]. The global momentum and commitment for the elimination of targeted NTDs [25,26] re- quires putting in place adequate collaboration mechanisms and structures at all levels to ensure effectiveness, efficiency, and synergy of interventions. In this environment, APOC structures and management framework may evolve towards complementarity with other NTDs programs in PENDA. When APOC mechanisms have a competitive advantage, they should be extended to serve other NTDs in PENDA, for example, with the APOC Trust Fund and the JAF. At the same time, PENDA should also benefit from the added value of proven approaches for particu- lar situations such as conflict, post conflict, urban settings, and problematic areas. With respect to global and regional governance and management structures, due attention should be given to the structure and management of NTD control and elimination programs at the national level, in accordance with country NTD master plans. Thus, the NOTF should be extended to cover other NTDs. Fig 2 depicts a possible evolution of APOC structures and mechanisms. Such an evolution would require the revision of the Memorandum of APOC [9]. The APOC Such an evolution would require the revision of the Memorandum of APOC [9]. The APOC Secretariat will support the transition of APOC from a single disease entity to a regional NTD elimination program and help ensure that future generations in Africa live free from the threat of debilitating and preventable diseases. The Way Forward The structure and management framework of APOC was determined based on the OCP expe- rience. This mechanism has demonstrated efficacy in achieving APOC’s initial goal of estab- lishing sustainable community-directed systems for ivermectin distribution that effectively controls onchocerciasis as a public health problem. In addition, the CTDi and CDI strategies have significantly contributed to scaling up other health interventions, such as control of lym- phatic filariasis, distribution of insecticide-treated bed nets, and vitamin A supplementation, among others [20]. The success of APOC has prompted the JAF to extend the Programme be- yond 2015, support countries in achieving elimination of onchocerciasis, use acquired expertise to benefit other targeted NTDs amenable to the PC strategy, and strengthen health systems at the community level across Africa. The evolution of APOC after 2015 is captured in the devel- opment and adoption of a concept note document and indicative budget to transform APOC 7 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 References Impact of long-term treatment of onchocerciasis with ivermectin in Kaduna State, Nigeria: first evidence of the potential for elimination in the operational area of the African Programme for Onchocerciasis Control. Parasites & Vectors, 5:28 8 / 9 PLOS Neglected Tropical Diseases \| DOI:10.1371/journal.pntd.0003542 May 14, 2015 13. APOC. (2011) Seventeenth session of the Joint Action Forum final communiqué. http://www.who.int/ apoc/about/structure/jaf/JAF17_Final_Communique_030112Clean.pdf. Accessed: 4 August 2014. 14. Braide EI, Obono OM, Bassey SE. (1990) Community participation in the control of onchocerciasis in Cross River, Nigeria. Acta Leidensia, 59: 427–432. PMID: 2378223 15. UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Dis- eases-TDR. (2008) Community-directed interventions for major health problems in Africa: a multi-coun- try study: final report. Geneva: WHO. 125 p. 16. The CDI Study Group. (2010) Community-directed interventions for priority health problems in Africa: results of a multi-country study. Bull World Health Organ; 88: 509–518. doi: 10.2471/BLT.09.069203 PMID: 20616970 17. UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Dis- eases-TDR. (1996) Community-directed treatment with ivermectin: report of a multi-country study. Ge- neva: World Health Organization. 4 p. 18. APOC/WHO. (2011) Year 2011 progress report of WHO/APOC. Ouagadougou: World Health Organi- zation. http://www.who.int/apoc/publications/JZF17_OK_EN_APOC_ProgressReport2011.pdf?ua=1. Accessed 1 August 2014. 19. WHO. (2011) Weekly epidemiological record. 48(86): 541–556. Available: http://www.who.int/wer/ 2011/wer8648.pdf?ua=1. Accessed: August 1, 2014. PMID: 22128386 20. Ransome-Kuti O, Hodgkin C, McFarland DA, Muller AS, Philippon BA, et al. (2000) African Programme for Onchocerciasis Control (APOC) report: external mid-term evaluation. Ouagadougou: African Pro- gramme for Onchocerciasis Control. 21. WHO/APOC. (2005) Report of the external evaluation. JAF 11.10. http://sireresources.worldbank.org/ EXTGLOREGPARPROG/Resources/APOC_indep_eval.pdf. Accessed: 1 August 2014. 22. WHO/APOC. (2010) Report of the external mid-term evaluation of the African Programme for Oncho- cerciasis Control. JAF 16.8. http://who.int/apoc/MidtermEvaluation_29Oct2010_final_printed.pdf. Ac- cessed 1 August 2014. 23. WHO/APOC. (2013) Programme for the Elimination of Neglected Diseases in Africa (PENDA): strategic plan of action and indicative budget 2016–2025. JAF19.8. http://www.who.int/entity/apoc/en_apoc_ strategic_plan_2013_ok.pdf?ua=1. Accessed: 1 August 2014. 24. WHO/APOC. (2013) Concept Note: Transforming APOC into a new regional entity for Oncho and LF elimination and support to other PC/NTD. JAF19.7. http://www.who.int/apoc/en_concept_note_ok.pdf? ua=1. Accessed: 1 August 2014. 25. WHO (2012). Accelerating work to overcome the global impact of neglected tropical diseases: a road- map for implementation. http://www.who.int/neglected_diseases/NTD_RoadMap_2012_Fullversion. pdf. Accessed: 1 August 2014. 26. Uniting to Combat NTDs. (2012) London Declaration on Neglected Tropical Diseases. http:// unitingtocombatntds.org/resource/london-declaration. Accessed: 1 August 2014. 9 / 9
https://openalex.org/W4297826773	https://hal.science/hal-03836151/file/Dang_acp-22-9389-2022-includes-suppl..pdf	English	null	Sensitivity analysis of an aerosol-aware microphysics scheme in Weather Research and Forecasting (WRF) during case studies of fog in Namibia	Atmospheric chemistry and physics	2,022	cc-by	19,741	Sensitivity analysis of an aerosol-aware microphysics scheme in Weather Research and Forecasting (WRF) during case studies of fog in Namibia Caroline Dang, Michal Segal-Rozenhaimer, Haochi Che, Lu Zhang, Paola Formenti, Jonathan Taylor, Amie Dobracki, Sara Purdue, Pui-Shan Wong, Athanasios Nenes, et al. To cite this version: Caroline Dang, Michal Segal-Rozenhaimer, Haochi Che, Lu Zhang, Paola Formenti, et al.. Sensitivity analysis of an aerosol-aware microphysics scheme in Weather Research and Forecasting (WRF) during case studies of fog in Namibia. Atmospheric Chemistry and Physics, 2022, 22 (15), pp.10221 - 10245. 10.5194/acp-22-10221-2022. hal-03836151 Sensitivity analysis of an aerosol-aware microphysics scheme in Weather Research and Forecasting (WRF) during case studies of fog in Namibia HAL Id: hal-03836151 https://hal.science/hal-03836151v1 Submitted on 1 Nov 2022 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Research article Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 © Author(s) 2022. This work is distributed under the Creative Commons Attribution 4.0 License. Published by Copernicus Publications on behalf of the European Geosciences Union. Biomass burning and marine aerosol processing over the southeast Atlantic Ocean: a TEM single-particle analysis Caroline Dang1,2, Michal Segal-Rozenhaimer3,4, Haochi Che3, Lu Zhang3, Paola Formenti5, Jonathan Taylor6, Amie Dobracki7, Sara Purdue7, Pui-Shan Wong8, Athanasios Nenes9,10, Arthur Sedlacek III11, Hugh Coe6, Jens Redemann12, Paquita Zuidema7, Steven Howell13, and James Haywood14,15 Caroline Dang1,2, Michal Segal-Rozenhaimer3,4, Haochi Che3, Lu Zhang3, Paola Formenti5, 6 7 7 8 9 10 1NASA Ames Research Center, Moffett Field, CA 94035, USA 2Oak Ridge Associated Universities, Oak Ridge, TN 37831, USA 3Department of Geophysics, Porter School, Tel Aviv University, Tel Aviv, 69978, Israel 4Bay Area Environmental Research Institute, NASA Ames Research Center, Moffett Field, CA, USA 5Université de Paris Cité and Université Paris-Est Creteil, CNRS, LISA, 75013 Paris, France 6Department of Earth and Environmental Sciences, University of Manchester, Manchester, UK 7Rosenstiel School, University of Miami, Miami, FL, USA 8Mount Allison University, Sackville, New Brunswick, CA, USA 9Laboratory of Atmospheric Processes and their Impacts, School of Architecture, Civil & Environmental Engineering, École Polytechnique Fédérale de Lausanne, Lausanne 1015, Switzerland 10Center for Studies of Air Quality and Climate Change, Institute of Chemical Engineering Sciences, Foundation for Research and Technology Hellas, Patras 26504, Greece 11Brookhaven National Laboratory, Brookhaven, NY, USA 12School of Meteorology, University of Oklahoma, Norman, OK, USA 13Department of Oceanography, University of Hawai‘i at M¯anoa, Honolulu, HI, USA 14College of Engineering, Mathematics and Physical Science, University of Exeter, Exeter, UK 15Met Ofﬁce, Exeter, EX1 3PB, UK 1NASA Ames Research Center, Moffett Field, CA 94035, USA 2Oak Ridge Associated Universities, Oak Ridge, TN 37831, USA 3Department of Geophysics, Porter School, Tel Aviv University, Tel Aviv, 69978, Israel 4Bay Area Environmental Research Institute, NASA Ames Research Center, Moffett Field, CA, USA 5Université de Paris Cité and Université Paris-Est Creteil, CNRS, LISA, 75013 Paris, France 6Department of Earth and Environmental Sciences, University of Manchester, Manchester, UK 7Rosenstiel School, University of Miami, Miami, FL, USA 8Mount Allison University, Sackville, New Brunswick, CA, USA 9Laboratory of Atmospheric Processes and their Impacts, School of Architecture, Civil & Environmental Engineering, École Polytechnique Fédérale de Lausanne, Lausanne 1015, Switzerland 10Center for Studies of Air Quality and Climate Change, Institute of Chemical Engineering Sciences, Foundation for Research and Technology Hellas, Patras 26504, Greece 11Brookhaven National Laboratory, Brookhaven, NY, USA 12School of Meteorology, University of Oklahoma, Norman, OK, USA 13Department of Oceanography, University of Hawai‘i at M¯anoa, Honolulu, HI, USA 14College of Engineering, Mathematics and Physical Science, University of Exeter, Exeter, UK 15Met Ofﬁce, Exeter, EX1 3PB, UK 1 Introduction sion electron microscopy (TEM) coupled with energy dis- persive X-ray (EDX) is suited to understanding physical and chemical properties of individual particles including shape, elemental composition, mixing state, volatility and viscosity, and it is particularly useful for complex aerosol which have been processed (Signorell and Reid, 2011; Reid et al., 2018; Li et al., 2003). Therefore TEM-EDX is a useful method for understanding processes affecting aged BB aerosol as well as marine salts which are pervasive over the ocean. With Africa producing almost a third of the Earth’s biomass burning aerosol (BBA) (Roberts et al., 2009), two air- craft campaigns, ObseRvations of Aerosols above CLouds and their intEractionS (ORACLES) and CLoud–Aerosol– Radiation Interaction and Forcing: Year 2017 (CLARIFY- 2017), were focused on understanding African biomass burn- ing aerosol interaction with clouds and radiation in the south- east Atlantic (Haywood et al., 2021; Redemann et al., 2021). The CLARIFY campaign was based on Ascension Island in 2017 and sampled primarily in that vicinity, and ORACLES was based in São Tomé in 2017 and 2018 and generally sam- pled closer to Africa than CLARIFY. CLARIFY ﬁndings de- tail a complex vertical structure in aerosol with a temperature structure inhibiting mixing between layers (Haywood et al., 2021). Over Africa, mixing is inhibited by stable layers at the top of the continental boundary layer (CBL) (Garstang et al., 1996), and over the southeast Atlantic the BBA in the resid- ual CBL moves over the marine boundary layer (MBL) as the air is transported west (Haywood et al., 2021). However, BBA aerosol is more often affected by the MBL than pre- viously accounted for, reaching the MBL through pathways that are not fully articulated (Zuidema et al., 2018), with en- trainment processes through the clouds potentially altering aerosol properties further. An example of this is the low sin- gle scattering albedo in the boundary layer compared to the free troposphere (Zuidema et al., 2018; Pistone et al., 2019). Both campaigns report that a more detailed aerosol process- level understanding including the properties of black carbon, organic carbon and inorganic compounds and how they vary as a function of mixing state and altitude is needed, as is knowledge of properties of the aerosols as they age from emission to deposition and the degree of mixing of BBA into the MBL (Haywood et al., 2021; Redemann et al., 2021). 1 Introduction Previous work of African BBA from the Southern African Regional Science Initiative (SAFARI-2000) showed that the aerosols were primarily composed of black carbon, potas- sium salts and organic / sulfur material (Liu et al., 2000; Pós- fai et al., 2003; Li et al., 2003). The SAFARI campaign was mostly focused on BBA that were less aged than particles in CLARIFY and ORACLES. SAFARI results showed that KCl particles occur in young smoke more often, while K2SO4 and KNO3 particles occur more in aged biomass burning aerosol (Li et al., 2003). This is due to gas-phase oxida- tion of NOx and SO2 and the displacement of HCl by the stronger acids HNO3 and H2SO4 during plume transport. The authors theorized that aging caused sulfate to accumu- late on organic and soot particles due to the large amount of internally mixed soot / sulfate and organic / sulfate parti- cles in haze (Pósfai et al., 2003). Based on the location and composition of the particles, Pósfai et al. (2003) concluded that organic and soot particles were the main cloud conden- sation nuclei (CCN) constituents of BBA. They determined that organic particles with inorganic inclusions likely con- tribute to the high cloud-nucleating capability of biomass burning particles, and Semeniuk et al. (2007), using envi- ronmental TEM, found that the inorganic phases of SAFARI particles took up water, while soot and tar balls did not; there- fore they determined that the inorganic content of mixed or- ganic / inorganic particles determined the hygroscopic prop- erties of BBA. SAFARI-2000 samples were taken in stratus clouds that capped the boundary layer, distinct from the BB haze layer in the free troposphere (FT), and were dominated by sea salt particles (Pósfai et al., 2003). CLARIFY and OR- ACLES online observations also show an aerosol population dominated by coated black carbon (BC), organics and sul- fates, consistent with the SAFARI TEM ﬁndings of BBA. CLARIFY noted a thick inorganic or organic coating around BC (Taylor et al., 2020), while ORACLES noted a less thick coating around BC, as well as a decreasing amount of coating with plume age (Sedlacek III et al., 2022). ORACLES AMS data also noted a decrease in organic aerosol with plume age (Dobracki et al., 2022). Correspondence: Caroline Dang (carolinevandang@gmail.com) and Michal Segal-Rozenhaimer (msegalro@tauex.tau.ac.il) Received: 26 August 2021 – Discussion started: 21 September 2021 nce: Caroline Dang (carolinevandang@gmail.com) and Michal Segal-Rozenhaimer (msegalro@tauex.tau.ac.il) Correspondence: Caroline Dang (carolinevandang@gmail.com) and Michal Segal-Rozenhaime (msegalro@tauex.tau.ac.il) Abstract. This study characterizes single-particle aerosol composition from ﬁlters collected during the Ob- seRvations of Aerosols above CLouds and their intEractionS (ORACLES) and CLoud–Aerosol–Radiation In- teraction and Forcing: Year 2017 (CLARIFY-2017) campaigns. In particular the study describes aged biomass burning aerosol (BBA), its interaction with the marine boundary layer and the inﬂuence of biomass burning (BB) air on marine aerosol. The study ﬁnds evidence of BBA inﬂuenced by marine boundary layer processing as well as sea salt inﬂuenced by BB air. Secondary chloride aerosols were observed in clean marine air as well as in BB- inﬂuenced air in the free troposphere. Higher-volatility organic aerosol appears to be associated with increased age of biomass burning plumes, and photolysis or oxidation may be a mechanism for the apparent increased volatility. Aqueous processing and interaction with the marine boundary layer air may be a mechanism for the presence of sodium on many aged potassium salts. By number, biomass burning potassium salts and modiﬁed sea salts are the most observed particles on ﬁlter samples. The most commonly observed BC coatings are inorganic salts. These results suggest that atmospheric processes such as photolysis, oxidation and cloud processing are key drivers in the elemental composition and morphology of aged BBA. Fresh BBA inorganic salt content, as it has an important role in the particles’ ability to uptake water, may be a key driver in how aqueous processing and atmospheric aging proceed. C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9390 https://doi.org/10.5194/acp-22-9389-2022 1 Introduction (1995) performed TEM analysis of ma- rine aerosol as part of the Atlantic Stratocumulus Transition Experiment/Marine Aerosol and Gas Exchange (ASTEX/- MAGE) campaign and found that polluted continental air af- fected sea salt aerosol processing, heterogeneity and mixing with sulfates and nitrates. surement (FAAM). Lacey carbon TEM grids (Ted Pella, Inc, #01881) were attached to 400 nm hole size polycarbonate nuclepore (WhatmanTM WHA10417112) ﬁlters. The Bae- 146 has been used for ﬁlter analysis for single-particle anal- ysis (Chou et al., 2008), as well as bulk analysis (Sanchez- Marroquin et al., 2019; Hand et al., 2010; Andreae et al., 2000). The AFS was composed of a ﬁlter holder manifold with ﬁve separate ﬁlters, connected to the aerosol in situ suite inlet during ORACLES 2017 and 2018. A vacuum pump connected to a ﬂow meter to maintain ﬂow of 30 L min−1 was used for sampling, with ﬁve manually controlled valves that were used to switch the sampling to ﬁlter holders. The ﬁlter manifold was preloaded before each ﬂight with ﬁlters. Samples for both campaigns were deposited on TEM grids at the locations, sampling times and total ﬂow volumes listed in Table S1 in the Supplement. After sampling, the ORACLES 2017 and CLARIFY 2017 ﬁlters were sealed in polycarbonate ﬁlter holders and wrapped in Paraﬁlm® and aluminum foil and transported to- gether with ice packs in a cooler and placed in a designated freezer immediately at the University of Manchester. The ORACLES 2018 ﬁlters were sealed in the same manner and transported with ice packs and stored in a designated freezer at Tel Aviv University. A preliminary set of TEM analysis was conducted on the ORACLES and CLARIFY 2017 ﬁlters at the University of Manchester, and then they were sealed and transported together with ice packs in a cooler for analy- sis at Tel Aviv University. Care was taken to maintain similar handling, storage, transport and analysis of all ﬁlters in both campaigns. All data included in this study are from the Tel Aviv University analysis. With both biomass burning salts and marine salts being major contributors to aerosol in the southeast Atlantic re- gion, a technique that can detect salts is important to accu- rately represent the aerosol in the region. Further, the plumes sampled during the CLARIFY and ORACLES campaigns are aged up to 15 and 7 d, respectively, according to back trajectories initialized at ﬁlter sampling times and locations. 1 Introduction As the single-particle soot photome- ter, used to detect coatings on BC, does not differentiate be- tween organic and inorganic material, TEM can help eluci- date the type and source of coating on BC. While in situ instruments provide data over large temporal and spatial scales, the instruments which analyzed chemical composition in the ORACLES and CLARIFY campaigns an- alyzed bulk aerosol; detailed ofﬂine single-particle analysis can offer valuable information to complement these online measurements. The principal in situ instrument used in these campaigns to determine aerosol chemical composition is the aerosol mass spectrometer (AMS). The AMS can detect or- ganic and non-refractory inorganic mass at high time resolu- tion. There are limitations on the size range of aerosols de- tected depending on the inlet system employed, with no de- tection above 1 µm and a decreasing efﬁciency above 700 nm. Salts do not vaporize easily and tend to recombine with op- positely charged ions and make quantiﬁcation of salts in the mass spectra difﬁcult, if not impossible (Nash et al., 2006). The mixing state of organic and inorganic constituents can only be determined with ofﬂine analysis of collected sam- ples rather than in situ bulk aerosol measurements. Transmis- Sea salt aerosols, generated through a bubble bursting pro- cess on the sea surface (Lewis and Schwartz, 2004), have https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing 9391 implications for radiative effects (Murphy et al., 1998) and cloud condensation nuclei (CCN) activity (King et al., 2012). Sea salt aerosols are modiﬁed when they react with sul- fate, nitrate and organic acids, resulting in a Na-rich and Cl-depleted aerosol and emission of gaseous HCl (Gard et al., 1998). There have been studies on the interaction of ur- ban and anthropogenic sources with marine aerosol (Adachi and Buseck, 2015), but single-particle studies of sea salt aerosol (SSA) and variations due to mixing with BB air are scarce. Coastal areas near urban sites show sea salt parti- cles being modiﬁed by anthropogenic sources. Adachi and Buseck found that sea salt particles were modiﬁed by H2SO4 and HNO3 by acid displacement of Cl (2015), and sea salt particles have also been shown to be Cl-depleted by or- ganic acid displacement (Laskin et al., 2012; Kerminen et al., 1998). Pósfai et al. https://doi.org/10.5194/acp-22-9389-2022 2 Method 2 Method 2.1 Filter sampling 2.1 Filter sampling 1 Introduction This is different from previous campaigns such as SAFARI- 2000, which was deployed closer to the burning source, and so TEM results can provide information on processing of aged (2–15 d from emission) BBA. This paper will describe the single-particle analysis in the context of the ancillary data including AMS measurements, back trajectories, cloud pro- cessing, time from source and time in the MBL. Our main questions are as follows: (1) what are the dominant aerosols in the region, and do CLARIFY and ORACLES aerosol dif- fer from each other based on differences in BB plume age? (2) What are the differences observed between MBL and FT aerosol? (3) What are the proposed processes which have acted on the aerosol? We proceed with a description of ﬁl- ter sampling and analysis methods and describe the region’s aerosol types during the two campaigns while comparing and contrasting between the two. Then, we compare aerosol com- position and state in the MBL and FT and discuss possible processing during transport. Size segregation was not performed during particle sam- pling. Most observed particles are in the submicron range. It is possible that morphologies or compositions were altered during collection, as in other aerosol TEM studies. For ex- ample, compositions of hydrate sulfates have been suggested to change in the TEM chamber or during processing (Buseck and Pósfai, 1999), with acidic particles containing more wa- ter spreading more on a TEM grid than neutral species. An- dreae et al. (1986) suggest that CaSO4 observed on ﬁlters without sea salt ions in the marine atmosphere could be from the breakup up sea salt particles containing a gypsum crystal- lite. A sodium chloride core and magnesium chloride coating have been suggested to be due to efﬂorescence of a particle after collection (Ault et al., 2013). Posfai et al. (1994) suggest that an interesting sulfate crystalline rod morphology may be due to water loss within the TEM chamber. Generally, the particles we observed were separated from other particles on the ﬁlter, and so agglomeration and aggregation did not inﬂu- ence organic mixing with adjacent particles. Samples were collected, on average, for approximately 10 min and in dry conditions, which may help to limit any chemical reactions the particles are subject to as the aircraft passes into new air masses. 2.2 Transmission electron microscopy with energy dispersive x-ray analysis (TEM-EDX) 2.2 Transmission electron microscopy with energy dispersive x-ray analysis (TEM-EDX) Meteosat-8. The FRP is produced with a 15 min repeat cycle for pixels which contain active burning (Roberts et al., 2005); hourly data were used to match the time step of the trajec- tory. The age of the BB aerosol is then estimated as the time in days when the trajectory ﬁrst intercepts the FRP points, similar to the method used by Vakkari et al. (2018). The ﬁrst FRP interception point with the back trajectory was chosen, representing the minimum aerosol age, as older aerosol may be more diluted in the plume. The BBA 7 d back trajectory overlaid with MODIS land cover classiﬁcations is included in the Supplement Fig. S1. A JEOL™JEM-2010F FEG-TEM with a ThermoNo- ran™energy dispersive X-ray detector (EDX) was used at Tel Aviv University’s Exact Sciences’ electron microscopy lab- oratory to analyze 14 ﬁlters from CLARIFY (2017) and 16 ﬁlters from the ORACLES (2017, 2018) campaigns. TEM analysis was performed at 200 KeV accelerating voltage and a take-off angle of 15.9◦for X-ray emission from the sam- ple, with an electron beam dwell time of no more than 30 s and spot size 3. The ﬁlter was scanned visually, and repre- sentative particles near the center of the TEM grid were ana- lyzed. TEM analysis has been known to underrepresent par- ticles under 300 nm (Posfai et al., 2003). EDX spectra were collected for each particle, and elemental weight percentage and atomic percentage were found per particle and normal- ized to 100 % using NSS software with the Cliff–Lorimer absorbance correction method. C and O are considered semi- quantitative due to the contribution from the Formvar ﬁlm of C and O from the TEM grid. Ratios of elements such as Na/S and Na/Cl were found by obtaining either the weight or atomic values for individual particles, ﬁnding the ratio of interest and averaging the ratio per ﬁlter. 2.4 Aerosol mass spectrometer, single particle soot photometer and cloud droplet probe The non-refractive chemical composition for submicron particles was measured using two Aerodyne time-of-ﬂight aerosol mass spectrometers (ToF-AMS; Aerodyne Research Inc.), a compact version (C-ToF-AMS) used in CLARIFY (Wu et al., 2020) and a high-resolution version (HR-ToF- AMS) used during ORACLES (Dobracki et al., 2022; Re- demann et al., 2021). The mass concentrations of organ- ics, sulfate, nitrate and ammonium were provided. Organic aerosol fractions including f43 and f44 were also derived from the mass spectra obtained during both campaigns. f43 is the fraction of the measured organic mass at m/z 43 relative to the total organic aerosol (OA) mass concentration and is indicative of non-acid oxygenates (Ng et al., 2011), common of fragments of aldehydes, ketones and acid functionalities. Likewise, f44 is the fraction of the measured organic mass present at m/z 44 relative to the total OA mass concentra- tion. The m/z 44 mass is due to acids or esters (Ng et al., 2011). Since these compound classes are commonly associ- ated with low-volatility organic fractions, a high f44 has been associated with low-volatility aerosol (Aiken et al., 2008). 2.3 Back-trajectory analysis Back trajectories of each sample were generated using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Stein et al., 2015), with the time step set to 1 h. Filter sampling lasted up to approximately 10 min per ﬁlter, and back trajectories were calculated as an ensemble of each minute of ﬁlter sampling time. To improve the ac- curacy of the trajectory, we used the hourly high-resolution ERA5 reanalysis data (ﬁfth-generation atmospheric reanaly- sis data) to drive the calculation. This method captures large- scale movements of air masses and has some inherent uncer- tainty; for example, it cannot capture entrainment. The ERA5 data have a 0.25◦×0.25◦horizontal resolution and include 37 pressure levels. We colocated the cloud liquid water content of ERA5 to the coordinates of trajectories, with a threshold of 0.001 g kg−1 to detect clouds on the trajectory. Two col- locations were performed: one with the 4-D coordinates of the trajectory (time, longitude, latitude and altitude) and an- other one with 3-D coordinates (time, longitude and latitude). Thus, the cloud liquid content points and proﬁles at the tra- jectory are provided, and the mean time of trajectory inside the cloud (cloud liquid water content >0.001 g kg−1), and un- der clear sky (no cloud liquid water above the trajectory) are calculated accordingly. For each sample, we calculated the back trajectories for 1, 2, 3, 5, 7 and 10 d and the in-cloud and clear-sky time correspondingly. The mass concentration of the refractory black carbon (rBC) of particles ranging from 80–650 nm was obtained in both campaigns by single-particle soot photometers (SP2; Droplet Measurement Technologies, Boulder, CO) using laser-induced incandescence. Detailed information on SP2 measurements can be found in Taylor et al. (2020). Cloud droplet probes (CDPs; Droplet Measurement Technologies, Boulder, CO) were used in both campaigns to measure cloud droplet number concentration. C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 2.1 Filter sampling Aerosol sampling was performed with the NASA Ames Re- search Center (ARC) aerosol ﬁlter system (AFS), installed on the P3, and the ﬁlter system operated on the UK Bae-146 air- craft operated by the Facility of Airborne Atmospheric Mea- Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 9392 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing Table 1. Filter IDs, ancillary online aerosol data, location, altitude, back-trajectory-based time from ﬁre as detailed in the text, and in-cloud time over previous 24 h. C. Dang et al.: Biomass burning and marine aerosol processing Filter IDs, ancillary online aerosol data, location, altitude, back-trajectory-based time from ﬁre as detailed in the text, and in-cloud time over previous 24 h. Table 1. Filter IDs, ancillary online aerosol data, location, altitude, back-trajectory-based time from ﬁre as detailed in the text, and in-cloud time over previous 24 h. 3.2 Aerosol classiﬁcations provide additional context for the sampled aerosols. The gaps in the AMS values are due to quality assurance checks which determined that the data for speciﬁc ﬁlters are unreliable. In the “time from ﬁre” column, if back trajectory analysis did not show interception with ﬁre but rather a marine source, “marine” is noted in the column. There are more samples taken above cloud, and generally, BC mass values are higher in above-cloud samples. A total of 6 out of 14 CLARIFY ﬁlters and 3 out of 16 ORACLES ﬁlters were sampled in the MBL, with the remainder sampled in the FT. The ORACLES 2017 and CLARIFY 2017 ﬁlters were sampled from mid- August to early September, while the ORACLES 2018 ﬁlters were collected late September through October. The ORA- CLES samples, in general, represent aged BBA, and CLAR- IFY samples represent extremely aged BBA. The TEM ﬁlters showed a heterogeneous aerosol population with variations in mixing for organics, NaCl salts, potassium salts and black carbon. Approximately 30–70 particles on each ﬁlter were analyzed to determine composition and par- ticle type. The main particle types including potassium salts, sea salt, black carbon and organic aerosol will be described along with the main ﬁndings in the following sections. C. Dang et al.: Biomass burning and marine aerosol processing Campaign Filter Date Particles Latitude Longitude Altitude Org SO4 NO3 NH4 BC BC CO Cloud Above Time and year analyzed (◦) (◦) (m) (µg (µg (µg (µg (µg (particles (ppbv) time in or from cm−3) cm−3) cm−3) cm−3) cm−3) cm−3) 24 h below ﬁre (h) cloud (d) ORACLES RF11Filter5 8/30/2017 47 −9.47 5 3505 20.5 1.5 3.1 1.3 3.8 1162 395 0 Above 2 2017 ORACLES RF02_1 9/30/2018 23 −7.64 5 894 1.1 0.5 0.1 0.1 0.2 56 110 9.35 Below 5 2018 RF02_2 9/30/2018 35 −7.82 5.03 2606 6.6 0.9 0.3 0.2 0.9 273 210 6.55 Above 1 RF03 10/2/2018 59 −7.67 5.5 982 1.2 346 156 18.08 Above 6 RF04 10/3/2018 65 −6.75 7 1195 0.5 0.4 0.0 0.1 0.3 117 120 6.58 Above 6 RF05_1 10/5/2018 55 −9.5 6.17 943 0.7 0.5 0.1 0.2 1 297 154 11.29 Above 6 RF05_2 10/5/2018 64 −9.5 6.21 378 0.2 0.2 0.0 0.1 0.5 119 106 8.42 Below marine RF05_3 10/5/2018 37 −9.5 6.11 3247 6.4 1.2 0.5 0.4 0.9 294 210 0 Above 1 RF06_1 10/7/2018 49 −8.91 5 2444 6 1.1 0.4 0.4 1.3 421 248 0 Above 2 RF06_2 10/7/2018 39 −6.86 5 2570 2.3 0.6 0.1 0.2 0.5 193 173 0 Above 2 RF07_1 10/10/2018 43 −12.77 5.01 1091 0.6 0.3 0.0 0.1 0.5 159 121 2.59 Above 6 RF07_2 10/10/2018 29 −7.39 5 159 0.3 0.2 0.0 0.0 0.4 108 123 0.18 Below marine RF09 10/15/2018 56 −11.35 5 1307 1.2 0.5 0.1 0.1 1 265 158 0.5 Above 7 RF10 10/17/2018 66 −7.18 10.5 1986 18.5 3 2.6 1.4 2.3 807 417 2.25 Above 1 RF11 10/19/2018 62 −7.95 9 3027 3.7 0.6 0.3 0.2 0.9 292 190 0.08 Above 2 RF13 10/23/2018 33 −5.01 −0.68 1127 0.1 0.1 0.0 0.0 0.1 42 118 6.94 Above 4 CLARIFY Gold_1 8/17/2017 49 −8.8 −11.52 323 4.1 1.9 0.2 0.7 0.5 195 108 0 Below marine 2017 Gold_8 8/22/2017 27 −8.46 −13.43 3902 6.9 1.3 1.4 1.0 1.2 380 204 20.77 Above 7 Gold_9 8/23/2017 39 −5.67 −12.42 2813 18.8 2.9 3.1 2.0 3 934 329 0 Above 4 Gold_10 8/24/2017 42 −8.37 −15.24 2918 3.9 0.6 0.3 0.3 0.8 232 158 0 Above 5 Gold_11 8/24/2017 54 −7.7 −13.85 319 0.3 0.3 0.0 0.1 0.1 17 70 0 Below 15 Gold_14 8/28/2017 47 −8.26 −13.74 2845 683 262 0 Above 6 Gold_15 8/28/2017 22 −8.28 −13.66 329 1 287 158 0.3 Below marine Gold_18 8/29/2017 32 −8.69 −12.47 332 0.5 174 119 0 Below marine Gold_19 8/30/2017 57 −8 −17.08 1969 5.7 1.6 0.9 0.8 1.8 535 212 0.38 Above 7 Gold_20 8/30/2017 30 −8.03 −17.3 329 3.6 1.2 0.2 0.5 0.7 225 130 0 Below marine Gold_21 9/7/2017 43 −8.32 −18.48 2357 1.5 436 177 0 Above 6 Gold_22 9/2/2017 43 −5.66 −13 2139 2.1 0.6 0.2 0.3 0.7 208 128 0 Below 9 Gold_23 9/2/2017 44 −6.14 −13.52 3500 12.2 1.3 2.4 1.3 2.5 750 273 0 Above 4 Gold_24 9/4/2017 24 −7.91 −12.72 1950 13.4 2.3 1.9 1.4 3.6 968 331 7.84 Above 4 Table 1. 3.1 Overview of observations Table 1 shows the conditions in which the ﬁlters were col- lected, along with ancillary indicators including latitude and longitude, collection above or below cloud, and AMS data including organics, SO4 NO3 and NH4 mass and fraction of PM1, as well as BC mass and number concentration. Time in cloud in the 24 h prior to ﬁlter collection and time from ﬁre To determine ﬁre locations, ﬁre radiative power (FRP) data were measured by the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) from the geostationary satellite Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 9393 C. Dang et al.: Biomass burning and marine aerosol processing 3.2.1 Organic aerosol The AMS data corresponding to ﬁlter collection times show that 35 % to 70 % of CLARIFY and 18 % to 68 % of ORA- CLES PM1 is organic, by mass; therefore in situ data indicate that a substantial amount of PM1 aerosol is organic in both campaigns. While TEM results show organic aerosol for both CLARIFY and ORACLES ﬁlters, there is signiﬁcantly more organic aerosol present on the ORACLES ﬁlters. We hypoth- esize that this is due to differences in the volatility and vis- cosity of the organic material. Figure 2a shows a comparison of the fraction of particles, by number, with organic particles on each ﬁlter and the AMS organic fraction, by mass, for the corresponding ﬁlter. The majority of CLARIFY ﬁlters do not have any particles with organic material, while the ma- jority of ORACLES ﬁlters have some particles which con- tain organic material. This extends to any organic coatings Figure 1 indicates the location of ﬁlter sampling as well as back trajectories including altitude per ﬁlter. As shown by the back trajectories, the ﬁlter samples covered different BB sources such as savanna, forest and grasses, with ﬁres fo- cused around central and southern Africa. Detailed informa- tion on ORACLES ﬂight and sampling conditions per ﬂight can be found in Redemann et al. (2021), which provides an- cillary data such as CO which will show whether a plume was sampled, with models (Redemann et al., 2021) showing that plumes are often above cloud. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9394 C. Dang et al.: Biomass burning and marine aerosol processin . The location of ﬁlter sampling and back trajectories related to each ﬁlter, including altitude for ORACLES 2017–2018 (a, b) an Y 2017 (c, d). Map colors relate to MODIS land cover types. Figure 1. The location of ﬁlter sampling and back trajectories related to each ﬁlter, including altitude for ORACLES 2017–2018 (a, b) an CLARIFY 2017 (c, d). Map colors relate to MODIS land cover types. Figure 1. The location of ﬁlter sampling and back trajectories related to each ﬁlter, including altitude for ORACLES 2017–2018 (a, b) and CLARIFY 2017 (c, d). Map colors relate to MODIS land cover types. For context, Fig. C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9395 Figure 2. (a) AMS organic fraction vs ﬁlter organic fraction and (b) f44 vs f43 space for ORACLES and CLARIFY campaigns with ﬁlters marked as triangles for ORACLES and squares for CLARIFY. Colors of the marks denoting ﬁlter sampling represent CO concentration, as shown in the color bar. Samples collected below-cloud are outlined with a black border. Figure 2. (a) AMS organic fraction vs ﬁlter organic fraction and (b) f44 vs f43 space for ORACLES and CLARIFY campaigns with ﬁlters marked as triangles for ORACLES and squares for CLARIFY. Colors of the marks denoting ﬁlter sampling represent CO concentration, as shown in the color bar. Samples collected below-cloud are outlined with a black border. UV exposure can work to break down oligomers and low- volatility components in organic aerosol (Wong et al., 2015; Lignell et al., 2014) and may account for lower amounts and/or higher volatility of organic aerosol present on CLAR- IFY ﬁlters. Photooxidation can also lead to fragmentation of organic chains, and oxidation has been observed to change biomass burning organic aerosol (BBOA) volatility in labora- tory studies (Jahn et al., 2021) as well as physical properties (Jahl et al., 2021). Our results of less organic aerosol present for aged samples are consistent with the ﬁndings of Dobracki et al. (2022) and Sedlacek III et al. (2022) of loss of organic aerosol and organic coating with age, although TEM results are caveated by preferential loss of volatile organic material. air data. Low-volatility oxygenated organic aerosol will typ- ically have a lower f43 and higher f44 than semivolatile oxy- genated organic aerosol (SV-OOA) (Ng et al., 2010, 2011). Most of the variation in ﬁlters sampled is in the ORACLES points with higher f44 than the CLARIFY data. As f44 is an indicator of low OOA fraction but not high-volatility frac- tion, the higher ORACLES points with regard to f44 are con- sistent with TEM ﬁndings of lower-volatility organic aerosol on ORACLES ﬁlters. The f43 spread is similar to differences in instrument baselines and therefore should not be overin- terpreted. TEM has been used to differentiate high- and low-contact- angle particles, where the viscosity and volatility of each particle can be qualitatively determined from the particle image. C. Dang et al.: Biomass burning and marine aerosol processing While factors such as surface tension and adhesion forces inﬂuence particle shape, viscosity and volatility can still be qualitatively measured on a comparative basis using electron microscopy images (Reid et al., 2018). Figure 3a shows a progression from left to right of increasingly volatile organic particles as imaged by the TEM. The presence of more rounded, viscous organic particles (Fig. 3, top panel, left image) in ORACLES samples compared to CLARIFY’s low-contact-angle organic particles (Fig. 3, top panel, right image) on the ﬁlters is also indicative of relatively higher volatility of organic particles in CLARIFY ﬁlters. More than 80 % of ORACLES organic particles have a rounded mor- phology, as shown in the left and center panels of Fig. 3a. y p g Tar balls are a type of round organic aerosol unique to biomass aerosol, and as of now, the only way to identify tar balls has been through microscopy. Tar balls are estimated to contribute up to ∼30% of BB aerosol mass (Sedlacek III et al., 2018). They are highly spherical, highly viscous and largely resistant to electron beam damage. SAFARI found a considerable number of tar balls (Pósfai et al., 2003) as well as the Biomass Burning Observation Project (BBOP) (Sedlacek III et al., 2018). Adachi et al. (2019) observed tar ball formation, likely from primary organic particles, within 3 h of emission, with the processing of tar balls possibly re- lated to oligomerization of OA. We did not ﬁnd many tar balls in the CLARIFY and ORACLES campaigns, with the exception of ﬁlters corresponding to RF10 and RF11, which were aged for 1 and 2 d, respectively. RF10 had very viscous aerosol but was mixed with considerable amounts of nitrogen and sulfur. This suggests a removal process, and while there are many unknowns regarding loss processes for tar balls, precipitation near the coast or heterogeneous, photolytically driven processes which may affect the solubility or volatil- ity of tar balls as they are advected west over the ocean may contribute to their removal. Posfai et al. (2003) also reported a dearth of tar balls when sampling in the haze layers repre- Figure 3b and c show the reduction in Org/BC and Org44/BC mass ratios, based on AMS measurements, for both CLARIFY and ORACLES ﬁlters as age from biomass burning source is increased. 3.2.1 Organic aerosol 2b shows the f43 vs f44 space for the entire ORACLES and CLARIFY campaigns, with ﬁlter data overlaid and marked by ﬁlter collection below cloud as well as the CO values marked in the color bar to denote whether the sample is from a BB plume. ORACLES ﬁlters (triangles) are 2–7 d aged, and CLARIFY (squares) are 4–15 d aged. A CO cutoff value of over 120 ppbv is used to denote BB- inﬂuenced air, based on overall campaign data and Fig. 17 in Haywood et al. (2021), which shows the Ascension Is- land CO frequency distribution and which shows that 120 is at the upper end of the Gaussian distribution of the clean as well; ORACLES organic coatings are largely more thick than organic coatings present on CLARIFY ﬁlters. As AMS data show a signiﬁcant amount of organic aerosol present in both campaigns (Wu et al., 2020; Redemann et al., 2021), the differences in visible organic material on ﬁlters can be attributed to loss of volatile organics in the TEM chamber. It is known that volatile species will be lost from particles in a TEM chamber (Pósfai et al., 2003; Hudson et al., 2004), and preferential loss of organics would indicate a comparatively volatile material. as well; ORACLES organic coatings are largely more thick than organic coatings present on CLARIFY ﬁlters. As AMS data show a signiﬁcant amount of organic aerosol present in both campaigns (Wu et al., 2020; Redemann et al., 2021), the differences in visible organic material on ﬁlters can be attributed to loss of volatile organics in the TEM chamber. It is known that volatile species will be lost from particles in a TEM chamber (Pósfai et al., 2003; Hudson et al., 2004), and preferential loss of organics would indicate a comparatively volatile material. Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing 3.2.2 Potassium salts and black carbon p y p The three common black carbon mixing states, BC with salt, BC with organic material and externally mixed BC, are shown as a fractional amount that exists in each campaign and in the boundary layer (BL) or FT in Table 2. Internal mixing refers to a particle which has two or more separate components, whereas externally mixed particles contain one component per particle. The predominant mixing state is BC internally mixed with salt; however, BC mixing with organ- ics is likely underestimated due to volatilization of organics in the chamber. Table 2 shows a difference between BL and FT in all columns, with the sign of the differences being dif- ferent in the two campaigns. It should be noted that of the three ORACLES ﬁlters collected in the BL, two have marine back trajectories, so the BB organic fraction may be under- represented here. For CLARIFY, cloud processing may re- move the more hygroscopic BC containing particles as these are activated and removed by precipitation, and hence the organic/BC ratio is high relative to the FT, but this does not work for ORACLES. The main ﬁnding here is that BC with inorganic salts, as analyzed by TEM, is the most prevalent BC mixing state. More than 60 % of particles, by number, from the two cam- paigns were potassium salts, either externally or internally mixed. Only K salts which appeared solid were counted in this number. If a particle was OA with K present but without a visible K-salt inclusion, this would not be counted as a K salt. If a particle was BC with a K crystal attached, this would be counted as a BC–K-salt internally mixed particle. This is consistent with ﬁndings from (Li et al., 2003), where organic particles and potassium salts were the predominant particle types in the smoke. The salts were often mixed with black carbon, organic aerosols or sulfates. Inorganic salts in BBA can result from volatiles from the burning source depositing inorganics onto particles in the BB plume (Jahn et al., 2020; Li et al., 2003; Gaudichet et al., 1995). senting aged BB plumes, without a clear explanation for their absence. senting aged BB plumes, without a clear explanation for their absence. The source of the coating is not described in that paper; the TEM results show that a common coating type is a hygro- scopic salt, with implications for both absorption enhance- ment and enhanced CCN capability of the particles. C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9396 9396 C. Dang et al.: Biomass burning and marine aerosol processing Figure 3. Example of organic aerosol of different viscosity and volatility, (a) showing more round and viscous particles for ORACLES (on the left) and more volatile particles for CLARIFY (middle and right top panels). (b)Org/BC and (c) Org44/BC ratios are shown with time from ﬁre source. Figure 3. Example of organic aerosol of different viscosity and volatility, (a) showing more round and viscous particles for ORACLES (on the left) and more volatile particles for CLARIFY (middle and right top panels). (b)Org/BC and (c) Org44/BC ratios are shown with time from ﬁre source. C. Dang et al.: Biomass burning and marine aerosol processing Filters where back trajectories did not indicate a BB source are included in the “marine” category. It appears that increased age reduces the organic to BC fraction, similar to the ﬁndings of Dobracki et al. (2022), which found organic aerosol to black carbon mass ratios de- creasing from 14 to 10 as the aerosol aged over the Atlantic. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 3.2.2 Potassium salts and black carbon Variations of Na : Cl, then, can help to determine relative aging for SSA, as has been used for example in Kirpes et al. (2018), Hand et al. (2010) and Young et al. (2016). For context, Na and SO4 weight percent in sea salt, based on the composition of sea water, are 60.31 and 7.68, respectively (Seinfeld and Pandis, 2012). Using the atomic weights of sulfur and oxygen, this leaves an expected Na : S ratio of approximately 16 : 1 in sea salt. A ratio lower than 16 : 1 indicates Cl displacement by S and is an indicator of aerosol aging; therefore based on the ratios in Table 3, our samples are aged sea salt. Prior work has shown variation of up to 13 % in the atomic percent of S in fresh SSA (Ault et al., 2013). Table 2. Black carbon mixing state by campaign in the FT or BL. BC–salt and BC–organic refer to internally mixed particles. BC – salt BC – organic BC – external ORACLES BL 0.78 0.00 0.22 ORACLES FT 0.53 0.31 0.16 CLARIFY BL 0.50 0.29 0.21 CLARIFY FT 0.67 0.07 0.26 BC – salt BC – organic BC – external ORACLES BL 0.78 0.00 0.22 ORACLES FT 0.53 0.31 0.16 CLARIFY BL 0.50 0.29 0.21 CLARIFY FT 0.67 0.07 0.26 Table 4 lists the ORACLES particle percents for Na and Cl as well as ratios. Both Table 3 and Table 4 list altitude, CO and time from ﬁre source to provide context as to whether the air mass is BB-inﬂuenced. As a measure of aging and sea salt conversion, Tables 3 and 4 list Na : Cl and Na : S weight per- cent ratios for particles which have those elements present, averaged per ﬁlter. A comparison of Tables 3 and 4 shows that for ORACLES ﬁlters, average Na and Cl weight percent is less, per particle, and there is a larger variation in the per- centage of particles per ﬁlter containing Na and/or Cl than in CLARIFY. In ORACLES, Na/Cl per ﬁlter is higher than CLARIFY due to the low Cl weight percent particle average, and the Na/S ratio is generally lower due to the lower Na weight percent. For CLARIFY, all SSA on ﬁlters collected in the MBL have NaCl with varying levels of Cl depletion. 3.2.2 Potassium salts and black carbon Different salts will indicate different processes; K salts will form due to evap- oration of potassium in the ﬁre and subsequent near-ﬁeld condensation onto the BC; while this will occur with some S and N as well, co-emitted SO2 and NO2 can oxidize and condense and lead to additional coating in the far ﬁeld. One common particle type was potassium salt internally mixed with BC, where the K salt encapsulates the black carbon in a core–shell conﬁguration. EDX analysis can ablate the salt and leave the refractory black carbon core intact. Another common particle type was organic aerosol with interstitial salts. These two common K-salt mixtures are shown in Fig. 4. The coating of BC gives rise to absorption enhancements as discussed by Taylor et al. (2020), where they found univer- sally thickly coated BC and almost no externally mixed BC. 3.2.3 Marine aerosol CLARIFY and ORACLES aerosol were both inﬂuenced by the marine atmosphere. Most CLARIFY ﬁlters have sea salt aerosols (SSAs) with Na and/or Cl present in varying ratios in the particles, as presented in Table 3. There are also minor amounts of Ca, Mg and K, as would be present in seawater. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 9397 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing Figure 4. Potassium salt in a core–shell morphology around a refractory BC core (a) and organic aerosol with interstitial K salt (b). Note the difference in scale between the two images. Figure 4. Potassium salt in a core–shell morphology around a refractory BC core (a) and organic aerosol with interstitial K salt (b). Note the difference in scale between the two images. Table 2. Black carbon mixing state by campaign in the FT or BL. BC–salt and BC–organic refer to internally mixed particles. since these acids displace the Cl, and these rates will vary by location. The aerosols are processed in the atmosphere, with nitrates and sulfates replacing Cl. S is removed from the atmosphere through oxidation of SO2 in water associated with sea salt particles (Sievering et al., 1991; Miller et al., 1987) as well as cloud processing (Beilke and Gravenhorst, 1978), and N species like HNO3 and NO2 are also available for reactions with sea salt. 3.2.2 Potassium salts and black carbon The presence of Na colocated with Cl on all below-cloud ﬁlters and in only two out of seven of the above-cloud ﬁlters suggests the parti- cles are less aged in the BL samples compared to the FT sam- ples. Gold 23, a ﬁlter sampled in the FT, has a high Na : Cl ra- tio of 20.2 for particles with both Na and Cl, and this suggests that these salts are aged due to the Cl depletion. The other ﬁl- ter sampled above cloud with Cl, Gold 8, has mostly Cl-only particles and also crystals of Na : Cl which appear freshly emitted with a cubic NaCl structure. Gold 8 has particles sim- ilar in morphology to Cl-rich particles present on ﬁlters Gold 14, 15 and 18, which will be described in a later section, the difference being that Gold 14, 15 and 18 ﬁlters did not have any Na-containing particles present. Cl-only particles in the FT suggests mixing of the MBL and FT, as it shows that Cl A schematic of the SSA life cycle, with representative par- ticles from several CLARIFY ﬁlters and example mecha- nisms for Cl depletion, is provided in Fig. 5. Brieﬂy, freshly emitted sea salt, generated from ocean bubbles bursting, has Na and Cl present in a 0.86 : 1 atomic ratio. However, Na and Cl in the sea salt aerosol rarely are in a 0.86 : 1 ratio as would be expected from freshly emitted SSA, indicating that the particles have been processed. Natural variability can be present, with Krueger et al. (2003) ﬁnding that Cl/Na atomic ratio in sea salts increases with particle diameter. The aging timescale of sea salt also varies depending on the production of NO2 and SO2 and its conversion rate to H2SO4 and HNO3 https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9398 Table 3. Na-dominant aerosols on CLARIFY ﬁlters with sampling location, CO levels, time from source and percentage of particles on the ﬁlter with either Na or Cl. The ratios represent the weight percent ratio per particle, averaged across all particles with the elements of interest on each ﬁlter Table 3. Na-dominant aerosols on CLARIFY ﬁlters with sampling location, CO levels, time from source and percentage of particles on the ﬁlter with either Na or Cl. The ratios represent the weight percent ratio per particle, averaged across all particles with the elements of interest on each ﬁlter. hys 22 9389 9412 2022 https://doi org/10 5194/a Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing Figure 5. Schematic showing different stages of sea salt conversion with example mechanisms. The particles are from CLARIFY ﬁlters and range from sea salts which have been freshly emitted to Cl-depleted particles containing nitrates and sulfates. Cl aerosol formation is also shown. Na : Cl ratios showing depletion of Cl with sea salt conversion are shown in the bar above each particle image. Figure 5. Schematic showing different stages of sea salt conversion with example mechanisms. The particles are from CLARIFY ﬁlters and range from sea salts which have been freshly emitted to Cl-depleted particles containing nitrates and sulfates. Cl aerosol formation is also shown. Na : Cl ratios showing depletion of Cl with sea salt conversion are shown in the bar above each particle image. also collected above cloud in BB-inﬂuenced air (329 ppbv of CO; 3.1 µg cm−3 NO3), also has sodium nitrate but to a lesser extent than Gold 24. As these are the two above-cloud CLARIFY ﬁlters with the highest CO levels and are also the two ﬁlters which show some presence of sodium nitrate, this suggests that BB air may inﬂuence the sea salt conversion from NaCl to NaNO3. This is likely due to the emissions of NOX in BB plumes (Jin et al., 2021) to form HNO3 and drive Cl out of the sea salt aerosol. There is also an inﬂuence of marine air on BC. For example, the bottom center image in Fig. 6e shows a black carbon particle mixed with sodium nitrate from the Gold 24 ﬁlter, collected in the FT. The pres- ence of the Na in the FT suggests BB entrainment into the MBL and subsequent mixing of marine air into the FT or, as an as alternate explanation, sea spray mixing into the FT with BC. has reached the FT through turbulent mixing at the top of the MBL. Both Gold 8 and Gold 23 ﬁlters have back trajec- tories which show air masses from the continent which are entirely within the FT. As deep convection does not occur in this region, marine salts which are observed more than a few hundred meters above the BL height in the Ascension Island region of approximately 2250 m (Haywood et al., 2021) may be brought in from outside the region. C. Dang et al.: Biomass burning and marine aerosol processing 9399 Table 4. Na-dominant aerosols on ORACLES ﬁlters with sampling location, CO levels, time from source and percentage of particles on the ﬁlter with either Na or Cl. The ratios represent the weight percent ratio per particle, averaged across all particles containing the elements. g g g 10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389 https://doi.org/10.5194/acp-22-9389-2022 https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 9400 C. Dang et al.: Biomass burning and marine aerosol processing https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing Figure 6f is from Gold 1 which was collected in the MBL, and the particle is BC encapsu- lated in sodium sulfate; this may be due to aqueous process- ing where sodium sulfate forms around a BC core. This sug- gests that while in the FT there is BC and K salts mixed with sodium, BC, if entrained in the MBL, can also be signiﬁ- cantly affected by sea salts and mixed with sodium nitrates and sulfates. elements Si, K, Ca and Mg is hypothesized to be from SSA rather than biomass burning based on the unique morphology and composition of the particles on these ﬁlters. The atmo- spheric implications for the Cl-rich particles are not clear, although their ability to uptake water may be conducive to their ability to act as CCN. y Gold 14 was collected above cloud, with 47 out of 47 of the particles having strong Cl and N peaks and some of the particles showing very minor amounts of Si. The analysis by Wu et al. (2020) of the CLARIFY campaign noted an in- crease in nitrate mass concentration with increasing altitude and found that the nitrate aerosol mostly existed as ammo- nium nitrate in the FT. They suggest that the increased levels of nitrate in the FT may be due to colder temperatures at high altitudes, which would help partition the HNO3–NH3– NH4NO3 system into the aerosol phase (Wu et al., 2020). Particle counts from the condensation particle counter (CPC) show a particle count average of 1437 particles cm−3. As am- monium nitrate is a hygroscopic inorganic salt which can dis- solve into the aqueous aerosol phase, we hypothesize that the Cl aerosols are from HCl in the gas phase which parti- tions into the aerosol water and ammonium nitrate aerosol. Relative humidity at ﬁlter collection times according to back trajectories is in the 60 %–70 % range, which would facili- tate aerosol deliquescence and the subsequent uptake of HCl. Previous work (Semeniuk et al., 2007; Jahn et al., 2021) has also shown BBA to uptake water in this RH range. C. Dang et al.: Biomass burning and marine aerosol processing In the above-cloud CLARIFY samples, all particles were subject to BB-inﬂuenced air based on CO values, where we choose 120 ppbv as a CO level, indicating BB-inﬂuenced air above background levels. In the FT, Cl was mostly not present or depleted. BC often mixed with sodium sulfates, Cl and nitrate, and K salts were often mixed with NaSO4. Filter Gold 24, interestingly, shows sodium nitrate mixed with black carbon. N can be difﬁcult to detect in EDX spec- tra as it is between the C and O peaks and can be difﬁcult to deconvolute; therefore the presence of N in the EDX spec- tra indicates that there is a substantial amount in the parti- cles. Gold 24 was collected above cloud and in highly BB- inﬂuenced air (331 ppbv of CO; 1.9 µg cm−3NO3). Gold 9, There were different morphologies and compositions of the marine salts due to different salt conversion processes: some large rounded NaCl over 1.5 µm in diameter and Ca, Mg and Na sulfates and chlorides. Figure 6 shows a few ex- amples of the many different morphologies of marine salts https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing 9401 C. Dang et al.: Biomass burning and marine aerosol processing Figure 6. Assorted sea salt particles. (a) ∼1.5 µm rounded NaCl, Gold 1; (b) CaSO4, Gold 1; (c) NaCl and Mg, Gold 1; (d) Na2SO4 with K, Gold 1; (e) NaNO3 mixed with BC, Gold 24; and (f) BC encapsulated in Na2SO4 with K, Gold1. Figure 6. Assorted sea salt particles. (a) ∼1.5 µm rounded NaCl, Gold 1; (b) CaSO4, Gold 1; (c) NaCl and Mg, Gold 1; (d) Na2SO4 with K, Gold 1; (e) NaNO3 mixed with BC, Gold 24; and (f) BC encapsulated in Na2SO4 with K, Gold1. found on CLARIFY ﬁlters. Figure 6f is from Gold 1 which was collected in the MBL, and the particle is BC encapsu- lated in sodium sulfate; this may be due to aqueous process- ing where sodium sulfate forms around a BC core. This sug- gests that while in the FT there is BC and K salts mixed with sodium, BC, if entrained in the MBL, can also be signiﬁ- cantly affected by sea salts and mixed with sodium nitrates and sulfates. found on CLARIFY ﬁlters. Table 5. Cl-dominant aerosol on CLARIFY ﬁlters. Table 5. Cl-dominant aerosol on CLARIFY ﬁlters. growth, causing haze to form (Gunthe et al., 2021). The pres- ence of ammonium in the FT and Cl in the samples, as well as strong Cl and N peaks in the EDX spectra, suggests that these particles may be NH4Cl. MBL. If the newly formed particles are aqueous, they can support HCl uptake and may explain the Cl-dominant aerosol observed. The particles could also originate from sea spray without additional processing and as primary particles may be solgel structures of bioorganics as observed in Prather et al. (2013). Gold 15 was collected below cloud at 329 m, and 21 out of 21 of the analyzed particles were composed predominantly of carbon and chlorine with small amounts of silicon and potassium; the particles appeared similar across the ﬁlter so EDX on additional particles was not performed. These parti- cles may show HCl gas uptake onto sea spray aerosol. Gold 18 was collected below cloud at 332 m. A total of 32 out of 32 particles sampled on the TEM ﬁlter were either CaCl2 or MgCl2 with silicon and may have formed through chlorine reacting with Ca and Mg in aerosolized sea water. Prather et al. (2013) observed that long-chain bioorganic species as well as Ca and Mg form stable collapsed structures as solgel materials, and potentially particles from Gold 15 and Gold 18 may be Ca and Mg dispersed within a solgel structure. The lack of an observable counterion in the EDX spectra of Gold 15 particles may be due to a weaker N signal than in Gold 14 or potentially Cl dispersed within a solgel network. The ﬁlters were collected in both above- and below-cloud conditions, which suggests a mixing of the MBL and FT, as Cl, either in the gas or aerosol phase, reached the FT after having being emitted from the ocean; air exchange at the top of the BL due to turbulent mixing may be the cause of this, particularly as there is a transition from stratocumulus to cu- mulus near Ascension Island and a corresponding increase in convection (Gordon et al., 2018). We hypothesize that Cl in the FT may be from HCl in the BL, which gets taken up into cloud droplets and on evaporation at the cloud top is re-released onto more neutral aerosol such as ammonium ni- trate. C. Dang et al.: Biomass burning and marine aerosol processing Hy- drochloric acid partitioning into aerosol water has also been inferred in urban atmospheres, where the highly water ab- sorbing and soluble chloride in the aqueous phase enhances aerosol water uptake through co-condensation and particle Na and Cl are colocated on individual particles in 5 out of 13 above-cloud ﬁlters and 2 out of 3 below-cloud ORACLES ﬁlters. All below-cloud ﬁlters had BC present. RF2_1 BC is not mixed with Na or marine salts. RF5_2 and RF7_2 ﬁlters have BC mixed with salts containing Na, S and K. We did not observe clear crystalline sea salt morphologies in either above-cloud or below-cloud samples as observed in CLAR- IFY ﬁlters and also did not observe Cl-dominant particles. Three ﬁlters from CLARIFY, Gold 14, 15 and 18, were dominated by Cl particles which do not have Na present. Ta- ble 5 shows the three Cl-dominant ﬁlters along with altitude at ﬁlter exposure, CO levels and time from ﬁre, and Fig. S2 provides altitude and particle counts during the ﬁlter expo- sure time. The Cl-rich ﬁlters are interesting because of the high spatial density as well as the uniformity of the parti- cles on the ﬁlter. The uniform particle composition and mor- phology of the particles suggests that, per ﬁlter, the particles have been subject to similar atmospheric processing. Gold 14 was collected in the FT, and back trajectories show intercep- tion with biomass burning 6 d prior to ﬁlter collection. Back trajectories for ﬁlters Gold 15 and 18, both collected in the MBL, show that they do not intercept with the ﬁres and are not as inﬂuenced by BB air as Gold 14. The presence of the https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing 9402 3.3 Marine boundary layer and aqueous-phase processing 3.3 Marine boundary layer and aqueous-phase processing The high density of the chloride particles on the ﬁlters, compared to the other TEM ﬁlters, along with the uniform composition and morphology of these particles, suggests ei- ther condensation onto new particles formed in the FT or condensation onto marine particles formed from spray in the MBL. While the prevailing view is that new particle for- mation rarely occurs over open oceans, work by Zheng et al. (2021) shows that new particle formation in the remote MBL occurs frequently after the passage of a cold front, with factors such as removal of existing particles by pre- cipitation, vertical transport of reactive gases from the ocean and cold temperatures facilitating new particle formation. Io- dine species can also form new particles in pristine regions (He et al., 2021). These conditions do not characterize our sample; however, these studies are part of a growing area of work which supports new particles formation in the remote The MBL had a large effect on BB aerosols as well as on the processing of sea salt aerosols. A comparison of cloud pro- cessing and time in the FT and MBL for the aerosol collected during the ﬁlter sampling time is provided in Fig. 7. The ﬁl- ters are segregated by collection in the MBL and FT, and the fractional time spent in each environment is shown in pan- els (b) (collected in BL) and (d) (collected in FT). Panel (d) shows that aerosols collected in the FT, irrespective of cam- paign, have spent nearly all their time in the FT. Panel (b) shows that of ﬁlters sampled in the BL, CLARIFY aerosols have spent more than 80 % of the fractional time in the BL in the day before ﬁlter sampling compared to ORACLES’ 45 %. The cloud processing intensity, which is the mean cloud liquid water content multiplied by the mean in-cloud time, in the days prior to the air mass reaching the downwind ﬁlter collection location is provided in Fig. 8a. Cloud processing Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing 9403 C. Dang et al.: Biomass burning and marine aerosol processing 9403 Figure 7. (a) BL and FT ORACLES and CLARIFY cloud processing intensity. (b) Time spent in BL and FT for samples collected in the BL in the days prior to ﬁlter collection. 3.3 Marine boundary layer and aqueous-phase processing (c) Fraction of K salt, per ﬁlter, mixed with Na. (d) Fractional time spent in BL and FT for samples collected in the FT in the days prior to ﬁlter collection. Figure 7. (a) BL and FT ORACLES and CLARIFY cloud processing intensity. (b) Time spent in BL and FT for samples collected in the BL in the days prior to ﬁlter collection. (c) Fraction of K salt, per ﬁlter, mixed with Na. (d) Fractional time spent in BL and FT for samples collected in the FT in the days prior to ﬁlter collection. intensity is a metric described in detail in Che et al. (2022). Aerosols collected in the MBL were subject to more cloud processing than those collected in the FT, as the aerosols would need to be entrained into BL. ORACLES aerosol, on average, spent slightly more time in cloud and with clouds with a higher liquid water content than CLARIFY aerosol, particularly apparent for those ﬁlters collected in the MBL. with an 11 : 1 ratio has 13 wt % Na but only 1.2 wt % K; this is a higher Na weight percent than would typically be ex- pected from BBA. However, it should be noted that Na may also be from biomass burning, as sodium has been noted in BB fuel, with the type inﬂuencing the amount of Na in par- ticles (Hudson et al., 2004). The image shows that Na mixes with K salts, organics and sulfur in varying degrees through- out a single aqueous particle, and therefore this is a potential mechanism for Na incorporation into existing K salts. In sea salt, K weight percent is 1.1, assuming sea salt has the composition of seawater and ignoring atmospheric pro- cessing (Seinfeld and Pandis, 2012). The K salt from BB has a signiﬁcantly higher weight percent than the minor amounts of K found in seawater and sea salt. The fraction of K salt mixed with sodium increases with age, as shown in Fig. 7c. This implies that the K salt from biomass burning is pro- cessed in a way which allows Na incorporation into the par- ticle. As Na is not volatile at atmospheric pressures, we hy- pothesize that the mechanism is processing by particle mix- ing, either by cloud drop coalescence (Grabowski and Wang, 2013) or drizzle washout of aerosol and evaporation. 3.4 Elemental mixing in individual particles After more time in cloud, the sulfur / organic content on these particles appears to have a lower viscosity, to be more ﬂat and also to be more affected by the electron beam. Although these are only a few particles, the images imply that cloud processing can affect certain constituents of mixed aerosols more than others; the structure of the K salt and the viscosity of the sul- fur / organic mixture appear to be more affected by aqueous processing compared to BC. BBA is more diluted on CLARIFY than ORACLES ﬁlters, and S/K particle fractions in the BL/FT suggest increased mixing due to detrainment and entrainment for CLARIFY samples. A higher fraction of ORACLES aerosol contains S, 80 %, compared to CLARIFY’s 40 %. S is mixed with K in 34 % of CLARIFY and 73 % of ORACLES aerosols. This suggests that ORACLES S is predominantly from biomass burning as it is colocated with K, and potassium is frequently used as a marker for BBA, consistent with the high amounts of K salts observed in TEM samples. A proportion of 76 % of ORACLES FT aerosol contain colocated S + K, com- pared with 59 % of ORACLES BL aerosol. In comparison, 35 % and 33 % of CLARIFY FT and BL aerosols have colo- cated S+ K, respectively, implying a dilution of BBA in the CLARIFY aerosol population collected on ﬁlters in compar- ison to ORACLES. The similar fraction of S/K particles in the FT and BL during CLARIFY as compared to ORACLES may be due to the transition over the ocean and the increased entrainment and detrainment across the BL top. Higher frac- tions of S-containing particles in ORACLES may also be re- lated to cloud processing (Ervens et al., 2018), with aqueous formation pathways for sulfate in cloud water predicted to be faster than gas-phase formation pathways. More CLARIFY than ORACLES particles show evidence of marine inﬂuence as evidenced by the presence of Na and Cl. A total of 70 % of CLARIFY aerosols and 33 % of OR- ACLES aerosol contain some combination of Cl and/or Na. Of ORACLES aerosols containing Na, only 2 % are mixed with Cl, compared with 20 % of CLARIFY particles. 3.4 Elemental mixing in individual particles 3.4 Elemental mixing in individual particles Figure 8. Aqueous K-salt particle with varying levels of Na throughout the particle, with the Na : K weight ratio designated in red. EDX of individual aerosol particles can provide informa- tion on elemental mixing across the entire ORACLES and CLARIFY sample sets. Figure 10 shows the elemental mix- ing of CLARIFY and ORACLES samples collected in the free troposphere and boundary layer, designated in separate columns, with each row of pie charts indicating whether a set of two elements are colocated on individual particles for all particles of the speciﬁed particle set. All percentages are based on particle numbers rather than mass. The elements S, Na, Cl and K were chosen for elemental mixing analy- sis as they are the elements which most commonly appear in the EDX spectra. C and O are found in the spectra of al- most all particles, and so these elements are not included in this analysis. The Supplement, Figs. S3 and S4, also includes BL and FT Na–S–K and Na–S–Cl ternary diagrams for ad- ditional context, showing that there is more Na and Cl in the CLARIFY samples, particularly the BL. Figure 8. Aqueous K-salt particle with varying levels of Na throughout the particle, with the Na : K weight ratio designated in red. unaffected by exposure to the electron beam. The other im- ages in Fig. 10 show electron beam damage, indicating that these K-salt structures are more susceptible to volatilization and degradation due to the electron beam; this may be due to degradation of the K salt during cloud processing and the amorphization from hydration, dissolution and recrystalliza- tion of the salt during processing. As an alternative explana- tion, nitrates and sulfates have been shown to be affected by electron beam exposure after exposure to acid gases or water vapor (Jahn et al., 2021; Hoffman et al., 2004). The K-salt particles on the right, which were subject to the most time in cloud, have a less distinct morphology, which may indicate periods of water uptake and loss. In Fig. 9, bottom panel, the sulfur / organic constituent of the aerosol which was subject to the least amount of time in cloud appears to have high vis- cosity and was unaffected by the electron beam. C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing 9404 3.3 Marine boundary layer and aqueous-phase processing Single-particle structure and morphology are important as they affect aerosol optical properties and ability to act as cloud condensation nuclei. Figure 9 shows BC + K salt (top panel) and BC + (organic / sulfur) mixtures as a function of time in cloud encountered 1 d before sampling. The spherical nodules are black carbon, and the more reﬂective white areas in the bottom panel are the sulfur / organic mixture. Each par- ticle ID shows the TEM image and associated time in cloud (hours) in the day before sampling, the mean liquid water content of the clouds (g kg−1) in the day before sampling and the weight percent of all elements other than carbon in the particle. From left to right, the time in cloud increases. If the electron beam visibly altered the particle, an “after” image is shown along with an “before” image to indicate the particle’s response to the electron beam. In each particle, the black car- bon is insoluble and appears unaffected by increased cloud processing; however, both the K salts and the sulfur / organic content appear affected by increased time in cloud. In the leftmost panel for Fig. 10, top panel, the K salt is completely The interaction of the MBL with BBA and the effect of BB air on marine aerosol has been shown in previous sec- tions, and we hypothesize the mixing of Na salts with BBA to be due to aqueous processing and particle mixing through, for example, cloud drop coalescence. An aqueous K-salt par- ticle from CLARIFY’s Gold 9 ﬁlter is presented in Fig. 8. It has varying Na : K ratios by weight percent, designated in the red circles, and there is no Cl in the particle. We assume that the Na present is from marine sources because the area https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 igure 9. Potassium salt mixed with BC (top panel) and sulfur / organic material and BC (bottom panel) as a function of time in cloud in th 4 h prior to ﬁlter collection. Figure 9. Potassium salt mixed with BC (top panel) and sulfur / organic material and BC (bottom panel) as a function of time in cloud in the 24 h prior to ﬁlter collection. The reverse trend is observed for Na, with Na particle mix- ing differences between BL and FT more apparent in ORA- CLES than CLARIFY, as evidenced by K+Na and S+Na mixing. A proportion of 44 % of CLARIFY aerosol con- tains Na compared to 30 % of ORACLES aerosol. K+Na is present in 22 % and 50 % of ORACLES FT and BL particles and 35 % and 43 % of CLARIFY FT and BL particles, re- spectively. S+Na is present in 24 % and 46 % of ORACLES FT and BL particles and 28 % and 32 % of CLARIFY FT and BL particles, respectively. Thus while Na is slightly more likely to be mixed with K or S in the CLARIFY BL com- pared to FT, Na is mixed with S or K about 2 times more in the ORACLES BL than FT. This may be due to increased BL and FT entrainment and detrainment for CLARIFY samples and more in-cloud aqueous processing and particle mixing for ORACLES samples, which can deposit Na onto BBA in the BL. Cl in the BL (75 % of particles) versus the FT (31 % of parti- cles), as well as CLARIFY ﬁlters Gold 14, 15 and 18, which were dominated entirely by Cl particles. y y p Cl particle mixing differences between BL and FT dom- inate in CLARIFY aerosol but not in ORACLES, as evi- denced by K+Cl and S+Cl mixing. Cl is present in 6 % and 47 % of ORACLES and CLARIFY particles, respectively. K+Cl is present in 3 % and 8 % of ORACLES FT and BL particles and 10 % and 42 % of CLARIFY FT and BL parti- cles, respectively. S+Cl is present in 3 % and 4 % ORACLES FT and BL particles and 8 % and 27 % of CLARIFY FT and BL particles, respectively. 3.4 Elemental mixing in individual particles Cl can be found in ﬂuids of some vegetation and so can be present in BBA (Liu et al., 2000); thus while ORACLES Cl may be due to either biomass burning or a marine inﬂuence, it is likely that the high fraction of CLARIFY particles with Cl present, relative to the ORACLES population, is indicative of a ma- rine inﬂuence for particles in the BL and lower altitudes in the FT. This is also supported by the difference in CLARIFY Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 9405 CLARIFY Cl is mixed with K or S 3 to 4 times more in the BL than in the FT, while the BL does not have as much of an inﬂuence on Cl mixing with K or S for ORACLES particles. The increased Cl mixing with S/K in the CLARIFY BL may be due to secondary processes which deposit Cl onto BBA or a high fraction of primary SSA, which also includes S and/or K. Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing 9406 Figure 10. Elemental mixing states for select elemental pairs for ORACLES (two left columns) and CLARIFY (two right columns), sepa- rated by ﬁlter collection in the free troposphere and boundary layer. tates for select elemental pairs for ORACLES (two left columns) and CLARIFY (two right columns), sepa- ree troposphere and boundary layer. Figure 10. Elemental mixing states for select elemental pairs for ORACLES (two left columns) and CLARIFY (two right columns), sepa- rated by ﬁlter collection in the free troposphere and boundary layer. 4 Conclusions Dang et al.: Biomass burning and marine aerosol processing our TEM analysis did not ﬁnd tar balls other than on ﬁlters RF10 and RF11, which were aged for approximately 1 and 2 d, respectively. This ﬁnding implies a reduction in tar balls in aged African BB plumes. Tar ball incorporation in BB models (Jacobson, 2014) has been hindered due to lack of data, as tar balls can only be deﬁnitively detected with time- intensive single-particle electron microscopy. While other work shows that tar balls are a signiﬁcant fraction of BB aerosol, with some showing that tar balls outnumber BC by a factor of 10 (Hand et al., 2005; China et al., 2013), our anal- ysis shows a lack of tar balls in the aged BB plumes, con- sistent with Posfai et al. (2003), who also reported a dearth of tar balls in aged plumes as a puzzling phenomenon. Since tar balls are a light-absorbing particle, the absorption from aged plumes is dominated by non-tar-ball components like BC, brown carbon and dust; the absence of tar balls in these aged plumes can help constrain models on radiative forcing in the region. cence or drizzle washout of aerosol and subsequent evapo- ration. Na and Cl were present in a higher fraction of parti- cles in CLARIFY samples as compared to ORACLES. TEM particles show evidence of aqueous processing, with vary- ing ratios of soluble components throughout a single particle. Particle morphology also show evidence of cloud processing, with soluble components more signiﬁcantly affected by time in cloud than insoluble components such as black carbon. The uniformity and ubiquity of Cl-containing aerosol par- ticles on three CLARIFY ﬁlters suggest that particles on these ﬁlters have been uniformly processed and conﬁrm the existence of areas dominated by Cl aerosol. Further, the high density of Cl spatially on these ﬁlters implies HCl condensa- tion onto existing particles which have been recently formed and processed uniformly. Another explanation may be that these are primary particles such as solgel organic particles with Ca, Mg and Cl dispersed. The presence of Cl-rich parti- cles are interesting as they may act as good CCN due to their ability to uptake water. There were differences observed between FT and BL aerosol. 4 Conclusions More BBA was found in the FT, and more SSA was found in the BL, as would be expected; as BBA aerosol aged, there was increased entrainment into the BL for the CLARIFY samples relative to the ORACLES samples. SSA in the MBL was less aged than those sampled in the FT, as measured by Cl depletion. The presence of Cl and, on some ﬁlters, freshly emitted NaCl in the free troposphere sug- gests exchange of the MBL and FT in some regions through turbulent mixing at the top of the BL. For ORACLES, S- containing and K-containing particles were much more likely to be mixed with Na in the BL as compared to the FT. These BL and FT differences with Na–S and Na–K particles were not as evident in CLARIFY. In CLARIFY, the largest compo- sition changes between the FT and BL were that S-containing particles were more likely to be mixed with Cl in the BL as compared to the FT, and similarly, K-containing particles were more likely to be mixed with Cl in the in the BL com- pared to the FT. These Cl trends were not as clear in OR- ACLES, and only a minor fraction of ORACLES particles contain Cl. This may suggest that for ORACLES, aqueous processing is a key driver in depositing Na onto BB parti- cles, for example, by droplet coalescence in clouds and driz- zle evaporating before it reaches the surface. For CLARIFY, secondary processes may be important in depositing Cl onto BB particles. TEM analysis indicates either a loss of organic mate- rial, including as an organic coating, with plume age, or an increase in OA volatility with BB plume age. The re- duced amount of organics observed in CLARIFY compared to ORACLES may be due to less organic aerosol being present, higher-volatility organics which evaporated in the TEM chamber or both. Che et al. (2022) noted secondary organic aerosol (SOA) formation in the ﬁrst ∼70 h of BB aging for ORACLES; our results suggest that the secondary organic aerosol which forms may be more volatile than the initial organic aerosol emitted from biomass burning. There have been different explanations for the vertical structure of single scattering albedo near Ascension Island, with Taylor et al. (2020) and Wu et al. (2020) suggesting a partitioning of inorganic ammonium nitrate onto existing particles at colder temperatures and Dobracki et al. 4 Conclusions increase the particle’s hygroscopicity. This is important as it suggests that the salts formed in the ﬁre via evaporation and recondensation drive the mixing of the carbon aerosol as the secondary inorganic condenses and that the organic fraction is separate. This is consistent with ﬁndings regarding emis- sions of BC and K salts and other salts in the ﬂaming phase of a ﬁre, while organic emissions occur during the pyrolysis or smoldering phases (Haslett et al., 2018). Field conditions in- cluding different burning material and wet and dry conditions can also lead to spatial and temporal variability, which may affect the near-source mixing of fresh BBA. These ﬁndings are caveated due to the loss of organics in the TEM cham- ber, which would artiﬁcially increase the internally mixed and salt-BC fraction, particularly for CLARIFY samples. As CLARIFY sampled older smoke than ORACLES, this study, by comparing the two campaigns in the FT and BL, shows ways in which African BBA smoke is affected by SSA and marine air and reciprocally how SSAs may be affected by mixing with BB plumes. Single-particle analysis revealed considerable heterogeneity in the mixing and processing of CLARIFY and ORACLES aerosols. The main aerosols are BBA and SSA. We found similarities to the previous ma- jor campaign which analyzed African BBA, SAFARI-2000, in that we observed an abundance of potassium salts, black carbon and organic carbon with interstitial potassium salts (Pósfai et al., 2003; Li et al., 2003). Posfai et al. (2003) as well as our analyses show considerable internal mixing be- tween black carbon and salts; therefore, this suggests that the thick coating on CLARIFY and ORACLES BC as measured by the SP2 (Wu et al., 2020; Redemann et al., 2021) is often due to inorganics rather than organics. This has implications for radiative effects due to lensing, as well as CCN capa- bility as the inorganic salts internally mixed with BC would While SAFARI analysis found a large number of KCl salts near the source, both the CLARIFY and ORACLES cam- paigns sampled much more aged aerosol and did not ﬁnd many KCl particles. This is presumably because replacement of Cl by nitrates and sulfates has occurred prior to ORACLES and CLARIFY BBA sampling. While the SAFARI campaign and other recent biomass burning campaigns found tar balls, https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 9407 C. C. Dang et al.: Biomass burning and marine aerosol processing Sedlacek III et al. (2022) of a loss of organic material with age. Increasing oxidation of BB with age, as noted by Wu et al. (2020), and more volatile organics or loss of organics as observed with TEM results imply fragmentation of carbon chains. guided by an independent editor, and the authors also have no other competing interests to declare. Disclaimer. Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Due to the changes in organic aerosol and the noted ef- fects of the MBL on BBA, it appears that aqueous process- ing, oxidation, photolysis, evaporation, condensation and in- teraction with the MBL are key drivers in physical and chem- ical properties such as mixing state and elemental composi- tion of aged BBA. Studies on combustion from different fuel and burn phases show differences in primary particle types in terms of mixing and amount of organic content and el- emental composition, for example, for particles formed in the ﬂaming and smoldering phase (Liu et al., 2017). Our re- sults imply that due to the cloud and aqueous processing that African BBA is subject to, the salt phases present in the BBA will affect the ability of the particles to uptake water, act as CCN and undergo aqueous and cloud processing. Therefore the inorganic salt content of fresh BBA, fuel type and burn- ing conditions, as well as gas-phase oxidation of NOx lead- ing to formation of NO3 as a signiﬁcant pathway for further addition of inorganic salts, are key components for the atmo- spheric aging of BBA in these regions. Special issue statement. This article is part of the special issue “New observations and related modelling studies of the aerosol– cloud–climate system in the Southeast Atlantic and southern Africa regions (ACP/AMT inter-journal SI)”. It is not associated with a conference. Acknowledgements. We thank George Levi, instrument scientist at Tel Aviv University, for his expertise in TEM operation and anal- ysis. Further, we thank the manuscript reviewers for their time and comments. Financial support. The ﬁrst author was supported by the NASA Postdoctoral Fellowship Grant. ORACLES is a NASA EARTH Venture Suborbital-2 investigation, funded by the US National Aeronautics and Space Administrations (NASA)’s Earth Sci- ence Division and managed through the Earth System Science Pathﬁnder Program Ofﬁce (grant no. NNH13ZDA001N-EVS2). CLARIFY-2017 was funded by a Natural Environment Research Council (NERC) Large Grant NE/L013584/1. C. Dang et al.: Biomass burning and marine aerosol processing Haochi Che and Michal Segal-Rozenhaimer are supported by a Department of Energy (DOE) Atmospheric System Research (ASR) grant, DE- SC0020084. Lu Zhang is funded by a Tel Aviv University post- doc fellowship. Paola Formenti is supported by the AErosols, Ra- diatiOn and CLOuds in southern Africa (AEROCLO-sA) project funded by the French National Research Agency under grant agree- ment no. ANR-15-CE01-0014-01, the French national programs LEFE/INSU and PNTS, the French National Agency for Space Studies (CNES), the European Union’s Seventh Framework Pro- gramme (FP7/2014-2018) under EUFAR2 contract no. 312609 and the South African National Research Foundation (NRF) under grant UID 105958. Paquita Zuidema acknowledges additional support from a Department of Energy grant, DE-SC0021250. Data availability. Airborne measurements are available from the Centre for Environmental Data Analysis (https://catalogue. ceda.ac.uk/uuid/38ab7089781a4560b067dd6c20af3769, Facil- ity for Airborne Atmospheric Measurements et al., 2017) for CLARIFY 2017 data. ORACLES science data are available at https://doi.org/10.5067/Suborbital/ORACLES/P3/2017_V1 (ORACLES Science Team, 2019a) and https://doi.org/10.5067/Suborbital/ORACLES/P3/2018_V1 (ORACLES Science Team, 2019b) for the 2017 and 2018 data, respectively. Data availability. Airborne measurements are available from the Centre for Environmental Data Analysis (https://catalogue. ceda.ac.uk/uuid/38ab7089781a4560b067dd6c20af3769, Facil- ity for Airborne Atmospheric Measurements et al., 2017) for CLARIFY 2017 data. ORACLES science data are available at https://doi.org/10.5067/Suborbital/ORACLES/P3/2017_V1 (ORACLES Science Team, 2019a) and https://doi.org/10.5067/Suborbital/ORACLES/P3/2018_V1 (ORACLES Science Team, 2019b) for the 2017 and 2018 data, respectively. Supplement. The supplement related to this article is available online at: https://doi.org/10.5194/acp-22-9389-2022-supplement. Author contributions. MSR, HuC and CD designed the research. JH, HuC, JR, PZ and AN are PIs of the campaigns. JT, MSR, PSW, SP, PZ, AN and SH performed ﬁeldwork or provided support for ﬁlter collection. SH and MSR designed the sampling apparatus. CD performed laboratory analysis. CD and HaC provided ﬁgures. CD, HaC, MSR and LZ analyzed data sets. CD led the paper writing, and all co-authors contributed to ideas and writing. Review statement. This paper was edited by Joshua Schwarz and reviewed by two anonymous referees. 4 Conclusions (2022) hypothesizing that the single scattering albedo differences are due to scatter- ing organic material that is lost from BBA. Sedlacek III et al. (2022) also found a loss of organic material coating BC with plume age in ORACLES; further, the authors deﬁne dif- ferent regimes for BBA where organic coating on black car- bon increases in the ﬁrst few hours after emission, the coat- ing mass then plateaus after a several hours to several days when there are competing chemical physical processes such as photochemistry, SOA production, fragmentation and oxi- dation, and after several days of aging, there is material loss due to cloud processing, volatility and bleaching of brown carbon. While there are different explanations for the vertical structure of single scattering albedo near Ascension Island as well as mechanisms for organic loss with age, our results indicate that higher-volatility organic aerosol is associated with aging; therefore we hypothesize that fragmentation of carbon chains, either through photolysis or oxidation of the organic aerosols, is the predominant mechanism for the ap- parent higher volatility of aged organic aerosols. Our results are compatible with the ﬁndings of Dobracki et al. (2022) and We found evidence of BBA interaction with the MBL as well as marine salts affected by BB air, which has not been reported to date in these campaigns. Mixing of marine air with BB air affects sea salts because of Cl replacement by nitrate, as BB plumes have elevated levels of NOx which replaces the Cl in SSA. Reciprocally, BBA was inﬂuenced through, for example, BC mixing with sodium sulfates and sodium nitrates and the presence of Na and Cl on a large number of BB particles. There is considerable mixing of Na with BBA through what we believe is aqueous processing of particles, through, for example, either cloud drop coales- https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 9408 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing A., and Grassian, V. H.: Size-Dependent Changes in Sea Spray Aerosol Composition and Properties with Different Seawater Conditions, Environ. Sci. Technol., 47, 5603–5612, https://doi.org/10.1021/es400416g, 2013. Gaudichet, A., Echalar, F., Chatenet, B., Quiseﬁt, J. P., Ma- lingre, G., Cachier, H., Buat-Menard, P., Artaxo, P., and Maenhaut, W.: Trace elements in tropical African savanna biomass burning aerosols, J. Atmos. Chem., 22, 19–39, https://doi.org/10.1007/BF00708179, 1995. Gordon, H., Field, P. R., Abel, S. J., Dalvi, M., Grosvenor, D. P., Hill, A. A., Johnson, B. T., Miltenberger, A. K., Yoshioka, M., and Carslaw, K. S.: Large simulated radiative effects of smoke in the south-east Atlantic, Atmos. Chem. Phys., 18, 15261–15289, https://doi.org/10.5194/acp-18-15261-2018, 2018. Beilke, S. and Gravenhorst, G.: Heterogeneous SO2-oxidation in the droplet phase, in: Sulfur in the Atmosphere, edited by: Husar, R. B., Lodge, J. P., and Moore, D. J., Pergamon, 12, 231–239, https://doi.org/10.1016/B978-0-08-022932-4.50025-2, 1978. Grabowski, W. W. and Wang, L.-P.: Growth of Cloud Droplets in a Turbulent Environment, Annu. Rev. Fluid Mech., 45, 293–324, https://doi.org/10.1146/annurev-ﬂuid-011212-140750, 2013. Buseck, P. R. and Pósfai, M.: Airborne minerals and re- lated aerosol particles: Effects on climate and the en- vironment, P. Natl. Acad. Sci. USA, 96, 3372–3379, https://doi.org/10.1073/pnas.96.7.3372, 1999. Gunthe, S. S., Liu, P., Panda, U., Raj, S. S., Sharma, A., Darbyshire, E., Reyes-Villegas, E., Allan, J., Chen, Y., Wang, X., Song, S., Pöhlker, M. L., Shi, L., Wang, Y., Kommula, S. M., Liu, T., Ravikrishna, R., McFiggans, G., Mickley, L. J., Martin, S. T., Pöschl, U., Andreae, M. O., and Coe, H.: Enhanced aerosol par- ticle growth sustained by high continental chlorine emission in India, Nat. Geosci., 14, 77–84, https://doi.org/10.1038/s41561- 020-00677-x, 2021. Che, H., Segal-Rozenhaimer, M. S., Zhang, L., Dang, C., Zuidema, P., Dobracki, A., Sedlacek III, A., Coe, H., Wu, H., Taylor, J., Redemann, J., and Haywood, J.: Cloud Processing and Weeklong ageing signiﬁcantly affect the biomass burning aerosol over the south-eastern Atlantic, Nat. Commun. Earth Environ., accepted, 2022. China, S., Mazzoleni, C., Gorkowski, K., Aiken, A. C., and Dubey, M. K.: Morphology and mixing state of individual freshly emit- ted wildﬁre carbonaceous particles, Nat. Commun., 4, 2122, https://doi.org/10.1038/ncomms3122, 2013. Hand, J. L., Malm, W. C., Laskin, A., Day, D., Lee, T., Wang, C., Carrico, C., Carrillo, J., Cowin, J. P., Collett Jr., J., and Iedema, M. J.: Optical, physical, and chemical proper- ties of tar balls observed during the Yosemite Aerosol Char- acterization Study, J. Geophys. Res.-Atmos., 110, D21210, https://doi.org/10.1029/2004JD005728, 2005. C. Dang et al.: Biomass burning and marine aerosol processing 9409 Chem. Phys. Discuss. [preprint], https://doi.org/10.5194/acp- 2021-1081, 2022. P. R.: Spherical tarball particles form through rapid chem- ical and physical changes of organic matter in biomass- burning smoke, P. Natl. Acad. Sci. USA, 116, 19336–19341, https://doi.org/10.1073/pnas.1900129116, 2019. Ervens, B., Sorooshian, A., Aldhaif, A. M., Shingler, T., Cros- bie, E., Ziemba, L., Campuzano-Jost, P., Jimenez, J. L., and Wisthaler, A.: Is there an aerosol signature of chemi- cal cloud processing?, Atmos. Chem. Phys., 18, 16099–16119, https://doi.org/10.5194/acp-18-16099-2018, 2018. p g p Aiken, A. C., DeCarlo, P. F., Kroll, J. H., Worsnop, D. R., Huff- man, J. A., Docherty, K. S., Ulbrich, I. M., Mohr, C., Kim- mel, J. R., Sueper, D., Sun, Y., Zhang, Q., Trimborn, A., Northway, M., Ziemann, P. J., Canagaratna, M. R., Onasch, T. B., Alfarra, M. R., Prevot, A. S. H., Dommen, J., Duplissy, J., Metzger, A., Baltensperger, U., and Jimenez, J. L.: O/C and OM/OC Ratios of Primary, Secondary, and Ambient Or- ganic Aerosols with High-Resolution Time-of-Flight Aerosol Mass Spectrometry, Environ. Sci. Technol., 42, 4478–4485, https://doi.org/10.1021/es703009q, 2008. Facility for Airborne Atmospheric Measurements, Natural Envi- ronment Research Council, and Met Ofﬁce: CLARIFY: in-situ airborne observations by the FAAM BAE-146 aircraft, Cen- tre for Environmental Data Analysis, http://catalogue.ceda.ac.uk/ uuid/38ab7089781a4560b067dd6c20af3769 (last access: 25 Au- guast 2021), 2017. Gard, E. E., Kleeman, M. J., Gross, D. S., Hughes, L. S., Allen, J. O., Morrical, B. D., Fergenson, D. P., Dienes, T., Gälli, M. E., Johnson, R. J., Cass, G. R., and Prather, K. A.: Direct Obser- vation of Heterogeneous Chemistry in the Atmosphere, Science, 279, 1184–1187, https://doi.org/10.1126/science.279.5354.1184, 1998. Andreae, M. O., Charlson, R. J., Bruynseels, F., Storms, H., Van Grieken, R., and Maenhaut, W.: Internal mixture of sea salt, sili- cates, and excess sulfate in marine aerosols, Science, 232, 1620– 1623, https://doi.org/10.1126/science.232.4758.1620, 1986. Andreae, M. O., Elbert, W., Gabriel, R., Johnson, D. W., Os- borne, S., and Wood, R.: Soluble ion chemistry of the at- mospheric aerosol and SO2 concentrations over the east- ern North Atlantic during ACE-2, Tellus B, 52, 1066–1087, https://doi.org/10.3402/tellusb.v52i4.17087, 2000. Garstang, M., Tyson, P. D., Swap, R., Edwards, M., Kållberg, P., and Lindesay, J. A.: Horizontal and vertical transport of air over southern Africa, J. Geophys. Res.-Atmos., 101, 23721–23736, https://doi.org/10.1029/95JD00844, 1996. Ault, A. P., Moffet, R. C., Baltrusaitis, J., Collins, D. B., Ruppel, M. J., Cuadra-Rodriguez, L. A., Zhao, D., Guasco, T. L., Ebben, C. J., Geiger, F. M., Bertram, T. H., Prather, K. References Adachi, K. and Buseck, P. R.: Changes in Shape and Composition of Sea-Salt Particles upon Aging in an Urban Atmosphere, Atmos. Enviro., 100, 1–9, https://doi.org/10.1016/j.atmosenv.2014.10.036, 2015. Adachi, K., Sedlacek, A. J., Kleinman, L., Springston, S. R., Wang, J., Chand, D., Hubbe, J. M., Shilling, J. E., Onasch, T. B., Kinase, T., Sakata, K., Takahashi, Y., and Buseck, Competing interests. At least one of the (co-)authors is a guest member of the editorial board of Atmospheric Chemistry and Physics for the special issue “New observations and related mod- elling studies of the aerosol–cloud–climate system in the Southeast Atlantic and southern Africa regions”. The peer-review process was Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022 C. Dang et al.: Biomass burning and marine aerosol processing C. Dang et al.: Biomass burning and marine aerosol processing R., and Sipilä, M.: Role of iodine oxoacids in atmospheric aerosol nucleation, Science, 371, 589– 595, https://doi.org/10.1126/science.abe0298, 2021. King, S. M., Butcher, A. C., Rosenoern, T., Coz, E., Lieke, K. I., de Leeuw, G., Nilsson, E. D., and Bilde, M.: Investigating Primary Marine Aerosol Properties: CCN Activity of Sea Salt and Mixed Inorganic–Organic Particles, Environ. Sci. Technol., 46, 10405– 10412, https://doi.org/10.1021/es300574u, 2012. B., Bianchi, F., Brilke, S., Caudillo, L., Chen, D., Chiu, R., Chu, B., Dias, A., Ding, A., Dommen, J., Duplissy, J., Haddad, I. E., Carracedo, L. G., Granzin, M., Hansel, A., Heinritzi, M., Hofbauer, V., Junninen, H., Kangasluoma, J., Kemppainen, D., Kirpes, R. M., Bondy, A. L., Bonanno, D., Moffet, R. C., Wang, B., Laskin, A., Ault, A. P., and Pratt, K. A.: Secondary sul- fate is internally mixed with sea spray aerosol and organic aerosol in the winter Arctic, Atmos. Chem. Phys., 18, 3937– 3949, https://doi.org/10.5194/acp-18-3937-2018, 2018. Kim, C., Kong, W., Krechmer, J. E., Kvashin, A., Laitinen, T., Lamkaddam, H., Lee, C. P., Lehtipalo, K., Leiminger, M., Li, Z., Makhmutov, V., Manninen, H. E., Marie, G., Marten, R., Mathot, S., Mauldin, R. L., Mentler, B., Möhler, O., Müller, T., Nie, W., Onnela, A., Petäjä, T., Pfeifer, J., Philippov, M., Ranjithku- mar, A., Saiz-Lopez, A., Salma, I., Scholz, W., Schuchmann, S., Schulze, B., Steiner, G., Stozhkov, Y., Tauber, C., Tomé, A., Thakur, R. C., Väisänen, O., Vazquez-Puﬂeau, M., Wagner, A. C., Wang, Y., Weber, S. K., Winkler, P. M., Wu, Y., Xiao, M., Yan, C., Ye, Q., Ylisirniö, A., Zauner-Wieczorek, M., Zha, Q., Zhou, P., Flagan, R. C., Curtius, J., Baltensperger, U., Kulmala, M., Kerminen, V.-M., Kurtén, T., Donahue, N. M., Volkamer, R., Kirkby, J., Worsnop, D. R., and Sipilä, M.: Role of iodine oxoacids in atmospheric aerosol nucleation, Science, 371, 589– 595, https://doi.org/10.1126/science.abe0298, 2021. Krueger, B. J., Grassian, V. H., Iedema, M. J., Cowin, J. P., and Laskin, A.: Probing Heterogeneous Chemistry of Individual At- mospheric Particles Using Scanning Electron Microscopy and Energy-Dispersive X-ray Analysis, Anal. Chem., 75, 5170–5179, https://doi.org/10.1021/ac034455t, 2003. W., Onnela, A., Petäjä, T., Pfeifer, J., Philippov, M., Ranjithku- mar, A., Saiz-Lopez, A., Salma, I., Scholz, W., Schuchmann, S., Schulze, B., Steiner, G., Stozhkov, Y., Tauber, C., Tomé, A., Yan, C., Ye, Q., Ylisirniö, A., Zauner-Wieczorek, M., Zha, Q., Laskin, A., Moffet, R. C., Gilles, M. K., Fast, J. D., Zaveri, R. C. Dang et al.: Biomass burning and marine aerosol processing Chou, C., Formenti, P., Maille, M., Ausset, P., Helas, G., Harri- son, M., and Osborne, S.: Size distribution, shape, and com- position of mineral dust aerosols collected during the African Monsoon Multidisciplinary Analysis Special Observation Pe- riod 0: Dust and Biomass-Burning Experiment ﬁeld campaign in Niger, January 2006, J. Geophys. Res.-Atmos., 113, D00C10, https://doi.org/10.1029/2008JD009897, 2008. Hand, V. L., Capes, G., Vaughan, D. J., Formenti, P., Haywood, J. M., and Coe, H.: Evidence of internal mixing of African dust and biomass burning particles by individual particle analysis us- ing electron beam techniques, J. Geophys. Res.-Atmos., 115, D13301, https://doi.org/10.1029/2009JD012938, 2010. p g Haslett, S. L., Thomas, J. C., Morgan, W. T., Hadden, R., Liu, D., Allan, J. D., Williams, P. I., Keita, S., Liousse, C., and Coe, H.: Highly controlled, reproducible measurements of aerosol emis- sions from combustion of a common African biofuel source, At- Dobracki, A., Zuidema, P., Howell, S., Saide, P., Freitag, S., Aiken, A. C., Burton, S. P., Sedlacek III, A. J., Redemann, J., and Wood, R.: Non-reversible aging can increase solar absorption in African biomass burning aerosol plumes of intermediate age, Atmos. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing 9410 mos. Chem. Phys., 18, 385–403, https://doi.org/10.5194/acp-18- 385-2018, 2018. duces ice-active minerals in biomass-burning aerosol and bottom ash, P. Natl. Acad. Sci. USA, 117, 21928–21937, https://doi.org/10.1073/pnas.1922128117, 2020. Haywood, J. M., Abel, S. J., Barrett, P. A., Bellouin, N., Blyth, A., Bower, K. N., Brooks, M., Carslaw, K., Che, H., Coe, H., Cotterell, M. I., Crawford, I., Cui, Z., Davies, N., Dingley, B., Field, P., Formenti, P., Gordon, H., de Graaf, M., Herbert, R., Johnson, B., Jones, A. C., Langridge, J. M., Malavelle, F., Par- tridge, D. G., Peers, F., Redemann, J., Stier, P., Szpek, K., Tay- lor, J. W., Watson-Parris, D., Wood, R., Wu, H., and Zuidema, P.: The CLoud–Aerosol–Radiation Interaction and Forcing: Year 2017 (CLARIFY-2017) measurement campaign, Atmos. Chem. Phys., 21, 1049–1084, https://doi.org/10.5194/acp-21- 1049-2021, 2021. Jahn, L. G., Jahl, L. G., Bowers, B. B., and Sullivan, R. C.: Morphology of Organic Carbon Coatings on Biomass-Burning Particles and Their Role in Reactive Gas Uptake, ACS Earth Space Chem., 5, 2184–2195, https://doi.org/10.1021/acsearthspacechem.1c00237, 2021. Jin, X., Zhu, Q., and Cohen, R. C.: Direct estimates of biomass burning NOx emissions and lifetimes using daily observa- tions from TROPOMI, Atmos. Chem. Phys., 21, 15569–15587, https://doi.org/10.5194/acp-21-15569-2021, 2021. C. Dang et al.: Biomass burning and marine aerosol processing Kerminen, V.-M., Teinilä, K., Hillamo, R., and Pakkanen, T.: Substitution of chloride in sea-salt particles by inor- ganic and organic anions, J. Aerosol Sci., 29, 929–942, https://doi.org/10.1016/S0021-8502(98)00002-0, 1998. He, X.-C., Tham, Y. J., Dada, L., Wang, M., Finkenzeller, H., Stolzenburg, D., Iyer, S., Simon, M., Kürten, A., Shen, J., Rörup, B., Rissanen, M., Schobesberger, S., Baalbaki, R., Wang, D. S., Koenig, T. K., Jokinen, T., Sarnela, N., Beck, L. J., Almeida, J., Amanatidis, S., Amorim, A., Ataei, F., Baccarini, A., Bertozzi, B., Bianchi, F., Brilke, S., Caudillo, L., Chen, D., Chiu, R., Chu, B., Dias, A., Ding, A., Dommen, J., Duplissy, J., Haddad, I. E., Carracedo, L. G., Granzin, M., Hansel, A., Heinritzi, M., Hofbauer, V., Junninen, H., Kangasluoma, J., Kemppainen, D., Kim, C., Kong, W., Krechmer, J. E., Kvashin, A., Laitinen, T., Lamkaddam, H., Lee, C. P., Lehtipalo, K., Leiminger, M., Li, Z., M kh t V M i H E M i G M t R M th t He, X.-C., Tham, Y. J., Dada, L., Wang, M., Finkenzeller, H., Stolzenburg, D., Iyer, S., Simon, M., Kürten, A., Shen, J., Rörup, B., Rissanen, M., Schobesberger, S., Baalbaki, R., Wang, D. S., Koenig, T. K., Jokinen, T., Sarnela, N., Beck, L. J., Almeida, J., Amanatidis, S., Amorim, A., Ataei, F., Baccarini, A., Bertozzi, B., Bianchi, F., Brilke, S., Caudillo, L., Chen, D., Chiu, R., Chu, B., Dias, A., Ding, A., Dommen, J., Duplissy, J., Haddad, I. E., Carracedo, L. G., Granzin, M., Hansel, A., Heinritzi, M., Hofbauer, V., Junninen, H., Kangasluoma, J., Kemppainen, D., Kim, C., Kong, W., Krechmer, J. E., Kvashin, A., Laitinen, T., Lamkaddam, H., Lee, C. P., Lehtipalo, K., Leiminger, M., Li, Z., Makhmutov, V., Manninen, H. E., Marie, G., Marten, R., Mathot, S., Mauldin, R. L., Mentler, B., Möhler, O., Müller, T., Nie, W., Onnela, A., Petäjä, T., Pfeifer, J., Philippov, M., Ranjithku- mar, A., Saiz-Lopez, A., Salma, I., Scholz, W., Schuchmann, S., Schulze, B., Steiner, G., Stozhkov, Y., Tauber, C., Tomé, A., Thakur, R. C., Väisänen, O., Vazquez-Puﬂeau, M., Wagner, A. C., Wang, Y., Weber, S. K., Winkler, P. M., Wu, Y., Xiao, M., Yan, C., Ye, Q., Ylisirniö, A., Zauner-Wieczorek, M., Zha, Q., Zhou, P., Flagan, R. C., Curtius, J., Baltensperger, U., Kulmala, M., Kerminen, V.-M., Kurtén, T., Donahue, N. M., Volkamer, R., Kirkby, J., Worsnop, D. C. Dang et al.: Biomass burning and marine aerosol processing 9411 Savanna Fire Samples, J. Atmos. Chem., 36, 135–155, https://doi.org/10.1023/A:1006387031927, 2000. Savanna Fire Samples, J. Atmos. Chem., 36, 135–155, https://doi.org/10.1023/A:1006387031927, 2000. Prather, K. A., Bertram, T. H., Grassian, V. H., Deane, G. B., Stokes, M. D., DeMott, P. J., Aluwihare, L. I., Palenik, B. P., Azam, F., Seinfeld, J. H., Moffet, R. C., Molina, M. J., Cappa, C. D., Geiger, F. M., Roberts, G. C., Russell, L. M., Ault, A. P., Baltrusaitis, J., Collins, D. B., Corrigan, C. E., Cuadra-Rodriguez, L. A., Ebben, C. J., Forestieri, S. D., Guasco, T. L., Hersey, S. P., Kim, M. J., Lambert, W. F., Modini, R. L., Mui, W., Pedler, B. E., Ruppel, M. J., Ryder, O. S., Schoepp, N. G., Sullivan, R. C., and Zhao, D.: Bringing the ocean into the laboratory to probe the chemical complexity of sea spray aerosol, P. Natl. Acad. Sci. USA, 110, 7550–7555, https://doi.org/10.1073/pnas.1300262110, 2013. Miller, D. F., Lamb, D., and Gertler, A. W.: SO2 oxidation in cloud drops containing NaCl or sea salt as condensation nuclei, Atmos. Environ., 21, 991–993, https://doi.org/10.1016/0004- 6981(87)90096-5, 1987. Murphy, D. M., Anderson, J. R., Quinn, P. K., McInnes, L. M., Brechtel, F. J., Kreidenweis, S. M., Middlebrook, A. M., Pósfai, M., Thomson, D. S., and Buseck, P. R.: Inﬂuence of sea-salt on aerosol radiative properties in the Southern Ocean marine bound- ary layer, Nature, 392, 62–65, https://doi.org/10.1038/32138, 1998. T. L., Hersey, S. P., Kim, M. J., Lambert, W. F., Modini, R. L., Mui, W., Pedler, B. E., Ruppel, M. J., Ryder, O. S., Nash, D. G., Baer, T., and Johnston, M. V.: Aerosol mass spectrom- etry: An introductory review, Int. J. Mass Spectrom., 258, 2–12, https://doi.org/10.1016/j.ijms.2006.09.017, 2006. Redemann, J., Wood, R., Zuidema, P., Doherty, S. J., Luna, B., LeBlanc, S. E., Diamond, M. S., Shinozuka, Y., Chang, I. Y., Ueyama, R., Pﬁster, L., Ryoo, J.-M., Dobracki, A. N., da Silva, A. M., Longo, K. M., Kacenelenbogen, M. S., Flynn, C. J., Pis- tone, K., Knox, N. M., Piketh, S. J., Haywood, J. M., Formenti, P., Mallet, M., Stier, P., Ackerman, A. S., Bauer, S. E., Fridlind, A. M., Carmichael, G. R., Saide, P. E., Ferrada, G. A., How- ell, S. G., Freitag, S., Cairns, B., Holben, B. N., Knobelspiesse, K. D., Tanelli, S., L’Ecuyer, T. S., Dzambo, A. M., Sy, O. O., McFarquhar, G. M., Poellot, M. R., Gupta, S., O’Brien, J. R., Nenes, A., Kacarab, M., Wong, J. P. C. Dang et al.: Biomass burning and marine aerosol processing S., Small-Griswold, J. D., Thornhill, K. L., Noone, D., Podolske, J. R., Schmidt, K. S., Pilewskie, P., Chen, H., Cochrane, S. P., Sedlacek, A. J., Lang, T. J., Stith, E., Segal-Rozenhaimer, M., Ferrare, R. A., Burton, S. P., Hostetler, C. A., Diner, D. J., Seidel, F. C., Platnick, S. E., Myers, J. S., Meyer, K. G., Spangenberg, D. A., Maring, H., and Gao, L.: An overview of the ORACLES (ObseRvations of Aerosols above CLouds and their intEractionS) project: aerosol– cloud–radiation interactions in the southeast Atlantic basin, At- mos. Chem. Phys., 21, 1507–1563, https://doi.org/10.5194/acp- 21-1507-2021, 2021. p g j j Ng, N. L., Canagaratna, M. R., Zhang, Q., Jimenez, J. L., Tian, J., Ulbrich, I. M., Kroll, J. H., Docherty, K. S., Chhabra, P. S., Bahreini, R., Murphy, S. M., Seinfeld, J. H., Hildebrandt, L., Donahue, N. M., DeCarlo, P. F., Lanz, V. A., Prévôt, A. S. H., Dinar, E., Rudich, Y., and Worsnop, D. R.: Organic aerosol components observed in Northern Hemispheric datasets from Aerosol Mass Spectrometry, Atmos. Chem. Phys., 10, 4625– 4641, https://doi.org/10.5194/acp-10-4625-2010, 2010. S., Bahreini, R., Murphy, S. M., Seinfeld, J. H., Hildebrandt, L., Donahue, N. M., DeCarlo, P. F., Lanz, V. A., Prévôt, A. S. ell, S. G., Freitag, S., Cairns, B., Holben, B. N., Knobelspiesse, H., Dinar, E., Rudich, Y., and Worsnop, D. R.: Organic aerosol components observed in Northern Hemispheric datasets from Aerosol Mass Spectrometry, Atmos. Chem. Phys., 10, 4625– 4641, https://doi.org/10.5194/acp-10-4625-2010, 2010. Nenes, A., Kacarab, M., Wong, J. P. S., Small-Griswold, J. D., Thornhill, K. L., Noone, D., Podolske, J. R., Schmidt, K. S., Ng, N. L., Canagaratna, M. R., Jimenez, J. L., Chhabra, P. S., Se- infeld, J. H., and Worsnop, D. R.: Changes in organic aerosol composition with aging inferred from aerosol mass spectra, At- mos. Chem. Phys., 11, 6465–6474, https://doi.org/10.5194/acp- 11-6465-2011, 2011. Pilewskie, P., Chen, H., Cochrane, S. P., Sedlacek, A. J., Lang, T. J., Stith, E., Segal-Rozenhaimer, M., Ferrare, R. A., Burton, S. P., Hostetler, C. A., Diner, D. J., Seidel, F. C., Platnick, S. E., Myers, J. S., Meyer, K. G., Spangenberg, D. A., Maring, H., and Gao, L.: An overview of the ORACLES (ObseRvations of Aerosols above CLouds and their intEractionS) project: aerosol– cloud–radiation interactions in the southeast Atlantic basin, At- mos. Chem. Phys., 21, 1507–1563, https://doi.org/10.5194/acp- 21-1507-2021, 2021. ORACLES Science Team: Moffett Field, CA, NASA Ames Earth Science Project Ofﬁce (ESPO) [data set], https://doi.org/10.5067/Suborbital/ORACLES/P3/2017_V1, 2019a. C. Dang et al.: Biomass burning and marine aerosol processing ORACLES Science Team: Moffett Field, CA, NASA Ames Earth Science Project Ofﬁce (ESPO) [data set], https://doi.org/10.5067/Suborbital/ORACLES/P3/2018_V1, 2019b. Reid, J. P., Bertram, A. K., Topping, D. O., Laskin, A., Martin, S. T., Petters, M. D., Pope, F. D., and Rovelli, G.: The viscosity of atmospherically relevant organic particles, Nat. Commun., 9, 956, https://doi.org/10.1038/s41467-018-03027-z, 2018. Pistone, K., Redemann, J., Doherty, S., Zuidema, P., Burton, S., Cairns, B., Cochrane, S., Ferrare, R., Flynn, C., Freitag, S., Howell, S. G., Kacenelenbogen, M., LeBlanc, S., Liu, X., Schmidt, K. S., Sedlacek III, A. J., Segal-Rozenhaimer, M., Shinozuka, Y., Stamnes, S., van Diedenhoven, B., Van Harten, G., and Xu, F.: Intercomparison of biomass burning aerosol optical properties from in situ and remote-sensing instru- ments in ORACLES-2016, Atmos. Chem. Phys., 19, 9181–9208, https://doi.org/10.5194/acp-19-9181-2019, 2019. Roberts, G., Wooster, M. J., Perry, G. L. W., Drake, N., Re- belo, L.-M., and Dipotso, F.: Retrieval of biomass com- bustion rates and totals from ﬁre radiative power obser- vations: Application to southern Africa using geostationary SEVIRI imagery, J. Geophys. Res.-Atmos., 110, D21111, https://doi.org/10.1029/2005JD006018, 2005. Roberts, G., Wooster, M. J., and Lagoudakis, E.: Annual and diurnal african biomass burning temporal dynamics, Biogeosciences, 6, 849–866, https://doi.org/10.5194/bg-6-849-2009, 2009. Pósfai, M., Anderson, J. R., Buseck, P. R., Shattuck, T. W., and Tindale, N. W.: Constituents of a Remote Paciﬁc Ma- rine Aerosol: A Tem Study, Atmos. Environ., 28, 1747–1756, https://doi.org/10.1016/1352-2310(94)90137-6, 1994. Sanchez-Marroquin, A., Hedges, D. H. P., Hiscock, M., Parker, S. T., Rosenberg, P. D., Trembath, J., Walshaw, R., Burke, I. T., McQuaid, J. B., and Murray, B. J.: Characterisation of the ﬁl- ter inlet system on the FAAM BAe-146 research aircraft and its use for size-resolved aerosol composition measurements, At- mos. Meas. Tech., 12, 5741–5763, https://doi.org/10.5194/amt- 12-5741-2019, 2019. Pósfai, M., Anderson, J. R., Buseck, P. R., and Sievering, H.: Com- positional variations of sea-salt-mode aerosol particles from the North Atlantic, J. Geophys. Res.-Atmos., 100, 23063–23074, https://doi.org/10.1029/95JD01636, 1995. Pósfai, M., Simonics, R., Li, J., Hobbs, P. V., and Buseck, P. R.: Individual aerosol particles from biomass burning in southern Africa: 1. Compositions and size distributions of carbonaceous particles, J. Geophys. Res.-Atmos., 108, 8483, https://doi.org/10.1029/2002JD002291, 2003. Sedlacek III, A., Lewis, E. R., Onasch, T. B., Zuideman, P., Rede- mann, J., Jaffe, D., and Kleinman, L. I.: Black Carbon Particle Mixing State Analysis Allows Classiﬁcation of Biomass Burn Aerosol Lifecycle into Three Aging Regimes, P. Natl. Acad. Sci. USA, in preparation, 2022. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. C. Dang et al.: Biomass burning and marine aerosol processing A., Wang, B., Nigge, P., and Shutthanandan, J.: Tropospheric chem- istry of internally mixed sea salt and organic particles: Surprising reactivity of NaCl with weak organic acids, J. Geophys. Res.- Atmos., 117, D15302, https://doi.org/10.1029/2012JD017743, 2012. Lewis, E. R. and Schwartz, S. E.: Sea Salt Aerosol Production: Mechanisms, Methods, Measurements, and Models- A Critical Review, American Geophysical Union (AGU), ISBN 978-1-118- 66605-0, 2004. Hoffman, R. C., Laskin, A., and Finlayson-Pitts, B. J.: Sodium nitrate particles: physical and chemical proper- ties during hydration and dehydration, and implications for aged sea salt aerosols, J. Aerosol Sci., 35, 869–887, https://doi.org/10.1016/j.jaerosci.2004.02.003, 2004. Li, J., Pósfai, M., Hobbs, P. V., and Buseck, P. R.: Individual aerosol particles from biomass burning in southern Africa: 2, Composi- tions and aging of inorganic particles, J. Geophys. Res.-Atmos., 108, 8484, https://doi.org/10.1029/2002JD002310, 2003. Hudson, P. K., Murphy, D. M., Cziczo, D. J., Thomson, D. S., Gouw, J. A. de, Warneke, C., Holloway, J., Jost, H.-J., and Hübler, G.: Biomass-burning particle measurements: Character- istic composition and chemical processing, J. Geophys. Res.- Atmos., 109, D23S27, https://doi.org/10.1029/2003JD004398, 2004. Lignell, H., Hinks, M. L., and Nizkorodov, S. A.: Ex- ploring matrix effects on photochemistry of organic aerosols, P. Natl. Acad. Sci. USA, 111, 13780–13785, https://doi.org/10.1073/pnas.1322106111, 2014. Jacobson, M. Z.: Effects of biomass burning on climate, account- ing for heat and moisture ﬂuxes, black and brown carbon, and cloud absorption effects, J. Geophys. Res.-Atmos., 119, 8980– 9002, https://doi.org/10.1002/2014JD021861, 2014. Liu, L., Kong, S., Zhang, Y., Wang, Y., Xu, L., Yan, Q., Lin- gaswamy, A. P., Shi, Z., Lv, S., Niu, H., Shao, L., Hu, M., Zhang, D., Chen, J., Zhang, X., and Li, W.: Morphology, composition, and mixing state of primary particles from combustion sources – crop residue, wood, and solid waste, Sci. Rep.-UK, 7, 5047, https://doi.org/10.1038/s41598-017-05357-2, 2017. Jahl, L. G., Brubaker, T. A., Polen, M. J., Jahn, L. G., Cain, K. P., Bowers, B. B., Fahy, W. D., Graves, S., and Sulli- van, R. C.: Atmospheric aging enhances the ice nucleation ability of biomass-burning aerosol, Sci. Adv., 7, eabd3440, https://doi.org/10.1126/sciadv.abd3440, 2021. p g Liu, X., Van Espen, P., Adams, F., Cafmeyer, J., and Maen- haut, W.: Biomass Burning in Southern Africa: Individ- ual Particle Characterization of Atmospheric Aerosols and Jahn, L. G., Polen, M. J., Jahl, L. G., Brubaker, T. A., Somers, J., and Sullivan, R. C.: Biomass combustion pro- https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing Phys., 22, 9389–9412, 2022 C. Dang et al.: Biomass burning and marine aerosol processing 9412 Sedlacek III, A. J., Buseck, P. R., Adachi, K., Onasch, T. B., Springston, S. R., and Kleinman, L.: Formation and evolution of tar balls from northwestern US wildﬁres, Atmos. Chem. Phys., 18, 11289–11301, https://doi.org/10.5194/acp-18-11289- 2018, 2018. Wong, J. P. S., Zhou, S., and Abbatt, J. P. D.: Changes in Secondary Organic Aerosol Composition and Mass due to Photolysis: Rel- ative Humidity Dependence, J. Phys. Chem. A, 119, 4309–4316, https://doi.org/10.1021/jp506898c, 2015. Wu, H., Taylor, J. W., Szpek, K., Langridge, J. M., Williams, P. I., Flynn, M., Allan, J. D., Abel, S. J., Pitt, J., Cotterell, M. I., Fox, C., Davies, N. W., Haywood, J., and Coe, H.: Vertical variability of the properties of highly aged biomass burning aerosol transported over the southeast Atlantic dur- ing CLARIFY-2017, Atmos. Chem. Phys., 20, 12697–12719, https://doi.org/10.5194/acp-20-12697-2020, 2020. Seinfeld, J. H. and Pandis, S. N.: Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, John Wiley & Sons, 1119 pp., ISBN 9781118591505, 2012. Semeniuk, T. A., Wise, M. E., Martin, S. T., Russell, L. M., and Buseck, P. R.: Hygroscopic behavior of aerosol particles from biomass ﬁres using environmental transmis- sion electron microscopy, J. Atmos. Chem., 56, 259–273, https://doi.org/10.1007/s10874-006-9055-5, 2007. Young, G., Jones, H. M., Darbyshire, E., Baustian, K. J., Mc- Quaid, J. B., Bower, K. N., Connolly, P. J., Gallagher, M. W., and Choularton, T. W.: Size-segregated compositional analy- sis of aerosol particles collected in the European Arctic during the ACCACIA campaign, Atmos. Chem. Phys., 16, 4063–4079, https://doi.org/10.5194/acp-16-4063-2016, 2016. Sievering, H., Boatman, J., Galloway, J., Keene, W., Kim, Y., Luria, M., and Ray, J.: Heterogeneous sulfur conversion in sea-salt aerosol particles: the role of aerosol water content and size distribution, Atmos. Environ. A-Gen., 25, 1479–1487, https://doi.org/10.1016/0960-1686(91)90007-T, 1991. Zheng, G., Wang, Y., Wood, R., Jensen, M. P., Kuang, C., Mc- Coy, I. L., Matthews, A., Mei, F., Tomlinson, J. M., Shilling, J. E., Zawadowicz, M. A., Crosbie, E., Moore, R., Ziemba, L., Andreae, M. O., and Wang, J.: New particle formation in the remote marine boundary layer, Nat. Commun., 12, 527, https://doi.org/10.1038/s41467-020-20773-1, 2021. Signorell, R. and Reid, J. (Eds.): Fundamentals and Applications in Aerosol Spectroscopy, CRC Press, https://doi.org/10.1201/b10417, 2011. Stein, A. F., Draxler, R. R., Rolph, G. D., Stunder, B. J. B., Co- hen, M. D., and Ngan, F.: NOAA’s HYSPLIT Atmospheric Transport and Dispersion Modeling System, B. Am. C. Dang et al.: Biomass burning and marine aerosol processing Meteo- rol. Soc., 96, 2059–2077, https://doi.org/10.1175/BAMS-D-14- 00110.1, 2015. Zuidema, P., Sedlacek III, A. J., Flynn, C., Springston, S., Del- gadillo, R., Zhang, J., Aiken, A. C., Koontz, A., and Muradyan, P.: The Ascension Island boundary layer in the remote south- east Atlantic is often smoky, Geophys. Res. Lett., 45, 4456–4465, https://doi.org/10.1002/2017GL076926, 2018. Taylor, J. W., Wu, H., Szpek, K., Bower, K., Crawford, I., Flynn, M. J., Williams, P. I., Dorsey, J., Langridge, J. M., Cotterell, M. I., Fox, C., Davies, N. W., Haywood, J. M., and Coe, H.: Absorption closure in highly aged biomass burning smoke, Atmos. Chem. Phys., 20, 11201–11221, https://doi.org/10.5194/acp-20-11201- 2020, 2020. Vakkari, V., Beukes, J. P., Dal Maso, M., Aurela, M., Josipovic, M., and van Zyl, P. G.: Major secondary aerosol formation in southern African open biomass burning plumes, Nat. Geosci., 11, 580–583, https://doi.org/10.1038/s41561-018-0170-0, 2018. https://doi.org/10.5194/acp-22-9389-2022 Atmos. Chem. Phys., 22, 9389–9412, 2022 Supplement of Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022-supplement © Author(s) 2022. CC BY 4.0 License. Supplement of Biomass burning and marine aerosol processing over the southeast Atlantic Ocean: a TEM single-particle analysis Caroline Dang et al. Correspondence to: Caroline Dang (carolinevandang@gmail.com) and Michal Segal-Rozenhaimer (msegalro@tauex.tau.ac.il) The copyright of individual parts of the supplement might differ from the article licence. Supplement of Atmos. Chem. Phys., 22, 9389–9412, 2022 https://doi.org/10.5194/acp-22-9389-2022-supplement © Author(s) 2022. CC BY 4.0 License. Supplement of Caroline Dang et al. Correspondence to: Caroline Dang (carolinevandang@gmail.com) and Michal Segal-Rozenhaimer (msegalro@tauex.tau.ac.il) The copyright of individual parts of the supplement might differ from the article licence. Supplementary Material Table S1 Filter samples, flight and particles analyzed Table S1 Filter samples, flight and particles analyzed Sample name Flight Date Start time End time Particles Analyzed Total Flow (L) Gold 1 C030 8/17/2017 15:19 15:24 49 198 Gold 8 C033 8/22/2017 12:09 12:21 27 66 Gold 9 C035 8/23/2017 16:52 17:01 39 152 Gold 10 C037 8/24/2017 13:58 14:08 42 157 Gold 11 C036 8/24/2017 9:43 9:48 54 229 Gold 14 C042 8/28/2017 11:26 11:36 47 180 Gold 15 C042 8/28/2017 10:58 11:09 22 279 Gold 18 C044 8/29/2017 11:01 11:17 32 315 Gold 19 C046 8/30/2017 9:59 10:11 57 271 Gold 20 C046 8/30/2017 10:43 10:57 30 536 Gold 21 C055 9/7/2017 16:06 16:27 43 325 Gold 22 C049 9/2/2017 10:08 10:19 43 520 Gold 23 C049 9/2/2017 11:56 12:12 44 289 Gold 24 C050 9/4/2017 15:08 15:33 24 720 RF02_1 RF02 9/30/2018 10:05:00 10:15:34 23 317 RF02_2 RF02 9/30/2018 11:22:17,11:45:22 11:27:24,11:49:16 35 297 RF03 RF03 10/2/2018 13:16:59, 13:21:10 13:18:29,13:30:20 59 315 RF04 RF04 10/3/2018 10:01:52 10:13:00 65 334 RF05_1 RF05 10/5/2018 7:34:19 7:44:30 55 305 RF05_2 RF05 10/5/2018 8:09:30 8:19:50 64 310 RF05_3 RF05 10/5/2018 9:06:21 9:18:10 37 355 RF06_1 RF06 10/7/2018 11:43:09 11:48:45 49 168 RF06_2 RF06 10/7/2018 13:22:50 13:33:40 39 325 RF07_1 RF07 10/10/2018 10:25:05 10:34:52 43 293 RF07_2 RF07 10/10/2018 12:05:41 12:15:49 29 304 RF09 RF09 10/15/2018 11:11:52 11:20:26 56 257 RF10 RF10 10/17/2018 9:59:04 10:06:58 66 237 RF11 RF11 10/19/2018 10:23:38 10:35:27 62 355 RF13 RF13 10/23/2018 10:44:57, 10:50:10, 10:54:46, 11:02:56 10:47:09, 10:51:55, 10:58:28, 11:12:03 33 501 RF11Filter5 RF11 8/30/2017 10:29 10:41 47 360 Figure S1 7-day back trajectories and sources of collected filters. The colors represent different broadleaf forests, urban, unvegetated areas, savannah, crop, grasses, and shrubland as denoted by the MODIS land use vegetation classifications. Figure S1 7-day back trajectories and sources of collected filters. The colors represent different broadleaf forests, urban, unvegetated areas, savannah, crop, grasses, and shrubland as denoted by the MODIS land use vegetation classifications. Supplement of Figure S2 Particle count and altitude for Cl-dominant filters Gold 14, Gold 15, and Gold 18 during filter exposure times Figure S2 Particle count and altitude for Cl-dominant filters Gold 14, Gold 15, and Gold 18 during filter exposure times Figure S3 Na-S-K ternary diagrams for A) ORACLES below cloud B) ORACLES above cloud C) CLARIFY below cloud and D) CLARIFY above cloud. Figure S3 Na-S-K ternary diagrams for A) ORACLES below cloud B) ORACLES above cloud C) CLARIFY below cloud and D) CLARIFY above cloud. Figure S4 Na-S-Cl ternary diagrams for A) ORACLES below cloud B) ORACLES above cloud C) CLARIFY below cloud and D) CLARIFY above cloud. Note that the dearth of particles in A) and B) are due to the majority of particles having no Cl as well as most Cl-containing particles not containing Na or S. Figure S4 Na-S-Cl ternary diagrams for A) ORACLES below cloud B) ORACLES above cloud C) CLARIFY below cloud and D) CLARIFY above cloud. Note that the dearth of particles in A) and B) are due to the majority of particles having no Cl as well as most Cl-containing particles not containing Na or S.
https://openalex.org/W2022996965	https://zenodo.org/records/1617238/files/article.pdf	English	null	PROBLEMS THAT ARISE IN THE TEACHING OF ELEMENTARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE*	School science and mathematics	1,907	public-domain	4,036	ELEMENTARY ALGEBRA ELEMENTARY ALGEBRA 363 Read before the Association of Mathematics Teachers of the Middle States and Maryland, Dec. 1,1906. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE. BY ALICE M. MCKELDEN, PH.D. Philadelphia High School for Girls. In the words of the German proverb I would say "Aller Anfang 1st schwer", and so to the students meeting for the first time, the new concepts found in algebra, the subject will in- deed seem difficult unless presented under the guise of old truths previously learned and digested. Accordingly, upon the teacher who first introduces the subject of algebra to the mind, eager and questioning, but so easily confused, lies a great re- sponsibility. If he will present algebra at first as merely an ex- tension of the arithmetic which has already been assimilated, showing in all detail that there is no vital difference between the consideration of a as a number and 5 as a number, between the process of finding the sum of a and b and of 2 and j, etc.; showing also that powers are only for convenience as 23 expresses more conveniently 2X2X2, and so on; the student will soon realize that to work with a, b, or any other letter involves oper- ations and processes no different from those employed in work with P, ^, or any number. It takes time, but it seems absolutely necessary that at the beginning, each member of the class be made to see clearly the truth of what I have just outlined. I would emphasize especially the fact that the sum of a and b is a-\-b always. Only recently I have noticed that there is a tendency throughout the average class, to write the product of factors as simple as (.r+m) {x-{-n), as x^-^-mnx-^mn; while no mistakes are made when m and n have numerical values. For such reasons as the above, I urge the importance of time spent on the beginning of the subject. This is just wherein lies the first problem of the high school teacher of elementary algebra; he does not have the pupil when the latter first meets the subject. In the first year high school class are found pupils from many different grammar schools, each taught the fundamental and important truths of algebra in a different manner, each feeling that any variation in the method of teaching that the high school teacher introduces is SCHOOL SCIENCE AND MATHEMATICS 364 unnecessary and of no importance. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* This attitude of the class is very disheartening to the high school teacher who, at a loss to know whether the vital points at the beginning of algebra are clearly and definitely understood, finds it necessary to begin again, going thoroughly over ground already covered once in some manner by the grammar school and so of passing interest to the class that feels that the high school, is teaching nothing- new. Is this waste of time and loss of interest necessary? Cannot there be some uniformity in the manner of presentation of the new concepts of algebra so that when the pupil reaches the high school a quick review of essentials is sufficient to bring back the entire subject to his mind and make him ready to comprehend what follows? Another topic I wish to discuss is in regard to work done bv the pupil before entering the high schoolthat more drill work in fundamental definitions and rules found at the beginning of algebra, be given him. It is absolutely necessary for him to know the definitions of algebrathey give him his subject-mat- ter and in addition train his memory. If he gets this drill in the grammar school where there is more time for frequent repe- titions and constant reviews and where his memory is better, in that he is younger, by the time he has reached the high school he can express clearly and concisely what he has learned as the ground work of algebra. As Clarke has said, "The number of new concepts introduced to the mind of the pupil is large; he should have thorough and repeated drill on each before pass- ing to the next. This should not be crowded into the first few weeks of the first high school year." But when the definitions are given they should be clear, brief in so far as is consistent with clearness, and absolutely correct as far as they go. The pupil should learn them once for all exactly as he is to use them throughout his mathematical course, with no change except that required by generalization. These new concepts are difficult for him to master at best; when onc^ learned, there should be no unlearning of them. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* This problem has arisen frequently in my teaching, as for example, in teach- ing that a coefficient is one of the factors of a product, it is at first necessary to efface the definition which has been taught in the grammar school that a coefficient is the number prefixed to a term to denote how many times the term is used. And ELEMENTARY ALGEBRA 365 as accuracy of definition is .essential for clearness of thought, the necessity of thorough drill is perfectly manifest. So far I have considered only what should be taught the pupil before he enters the high school. With such a foundation as outlined, the review, while essential to impress the new concepts upon his mind emphatically by reiteration and to cast on obscure portions new light reflected from another’s manner of presenting the subject, could nevertheless be brief and. interesting. At. this time Pascal’s suggestion that the definition of each term be substituted for the term itself, could be well introduced, for the mind having by previous. drill become familiar with the subject-matter, can make use of this knowledge and will noi become hopelessly confused in the replacing of one expression by another. This method leads naturally to generalization which should ever be kept in evidence. It is necessary as early as possible to reach the stage when the pupil depends less and less upon con- crete illustrations and can reason more and more clearly from the standpoint of mathematics as a purely abstract science. But though this is the desirable ultimate goal, in the beginning moderation must be used; the goal can only be won by pains- taking effort at the first. The pupils that come to the higli school are very immature; they cannot grasp general truths at once. So they should be brought to the realization of the nature of generalization as gradually as possible. I would suggest some such method as the following when a definition, rule or principle is to be learned and applied. Each pupil should be made to illustrate by original examples, first more concrete in form, then more and more general until the definition, rule or principle itself is reached as the limit. For example the definition of a root is to be learned. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* The illustra- tions could be given thus: 8 is the product of 3 equal factors 2; so 2 is the cube root of 8; then, generalizing the root but not the power, a3, where a stands for any number, is the product of 3 equal factors, a, so a is the cube root of a8; then generalizing still more, a^ where m is any number, is the product of m equal factors a, so a is the mth root of a^\ whence comes the general definition that when any number as a is .the product of any number, m, equal factors a, one of these equal factors a is the root of the number ^m. 366 SCHOOL SCIENCE AND MATHEMATICS Unfortunately the limitations of time prevent such lengthy examples in general. With rules or principles which are to be applied, sufficient time must be given to cover all possible mis- understandings. Take for example the application of the law of exponents in multiplication and division to examples contain- ing literal exponents. In general; addition or subtraction of the exponents is a ’simple operation to be performed often indeed mechanically, but addition or subtraction of exponents such as are found in ^m-^n+y+ by x^ presents great difficulties to the average child’s mind, and the result ^m+n+^^m-^ ^ ^3m+3n+2 ^2m-4 ^ ^ likely to occur as the correct product. Why is it that while he can easily multiply correctly x^y^by x^ and add m-{-n-\-i to 3, when the latter occurs among addition examples, why is it that the combination as given above brings confusion and error to the mind of the pupil? It is because he has not yet learned to apply the rule of addition to examples not oc- curring rigidly among addition examples, except in the simplest cases; he has not yet learned that what is true of a rule em- ployed in one division of the subject is just as true of that same rule when it is used elsewhere. He has not grasped the gen- eral truth. Again, if you will pardon a personal experience, in leading up to the general rule for the divisibility of a" =b y1 by a X b I explained the rule at length as applied to definite numbers odd or even, in place of n,, and the class solved the examples with great enthusiasm. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* But when once I ventured into realms of generalization, the enthusiasm of the class died away and one of my brightest pupils voiced the opinion of the whole class when she confessed she could not grasp it; she could not un- derstand it. But it was not that they did not understand; they did not know they understood. I found one of their difficulties to be that they thought they were expected to know at a glance whether n were odd or even when only the type a^ +^11 had been given. Such are the perplexities of the youthful mind in grasp- ing these general truths, with which we have all grown so fa- milar. Hence, though it is necessary indeed that the minds of the class grow from the realms of the concrete to the abstract, it is more necessary that they are not forced to grow too quickly. A -good method by means of which each pupil will be enabled to think and reason in terms more and more general, as he be- comes more and more proficient is to have in every day’s lesson A -good method by means of which each pupil will be enabled to think and reason in terms more and more general, as he be- comes more and more proficient is to have in every day’s lesson ELEMENTARY ALGEBRA 367 simple, little questions such as Wentworth uses m his intro- ductory chapter to problems in equations and such as are also found in Hall and Knight’s Higher Algebra. Also by leading up -to the problems gradually by means of such simple ques- tions as are mentioned above, great facility in the translation of the problems into algebraic language as. well as proficiency in their solution is obtained. The necessity for drill of this kind is easily seen. Go into any average class that has been drilled in the fundamental operations as far as factoring and ask the pupils questions like the following: , (i) What must be added to x to make y? (2) By what must 3 be multiplied to make a? (3) What dividend gives & as a quotient when -^ is the divisor ? (4) If a be one factor of &/ what is the other ? (5) What is the price in cents of loo oranges when x oranges cost 6 cents? PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* (6) By how much does x exceed ^ 30? (7) What is the cost in dollars of 40 books at x dimes each? (7) What is the cost in dollars of 40 books at x dimes each? (8) If x men take 5 days to reap a field, how long will one man take? (8) If x men take 5 days to reap a field, how long will one man take? (9) How many hours will it take to walk x miles at 4 miles an hour? (10) Out of a purse containing a dollars and b dimes a man spends c cents; express in cents the sum left. With hardly an exception the pupils will be puzzled by the general form of the questions, which, were numbers sub- stituted for the letters, they would consider trivial and unim- portant. Out of a class of 36 girls taking the classical- course, almost all very good workers, to whom such a series as part of a test was given, one made 100; 2 made between 90 and 100; 3 between 80 and 90; 6 between 65 and 80; ^ between 50 and 65, and the rest, 17, below 50. That this is not an unusually poor class is evidenced by the fact that only two of the class failed for the quarter, while there were eleven who made between 90 and 100. This experiment only tends to prove the fact that the average high school pupil does not know how to think in general terms, unless he is given careful and conscientious guid- ance at first, over the difficult places where his logic seems ordained to fail him. And as this is one of the vital reasons for the existence of mathematicsto train the mind in logical, SCHOOL SCIENCE AND MATHEMATICS 368 clear thought and in independence and self-relianceany means that can be used to help attain the ultimate object should noi be despised. Having given in brief suggestions as to the kind of algebraic work and methods to be employed in the grammar school and hints as to the necessity of and the moderation to be used in generalization, I shall now take up in the usual order, a series of problems in teaching that have occurred in the class room and seem to merit discussion. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* (1) The process of reasoning leading to the rule for trans- position of terms is generally misunderstood by a class which has hitherto known only that to change sides means to change signs and so has mechanically and blindly followed this rule in solving all equations. To the majority the above rule seems as much a self-evident truth as any one of the axioms and they can see no advantage in substituting a longer for a shorter method of solution. .Infinite pains must be taken here to show clearly the logical dependence of the rule upon the axiom, not only to prove the rule, but especially to awaken at the very beginning a live interest in logical reasoning together with ana- lytical power. A simple but effective illustration that will appeal to and so aid in the cultivation of the imagination is the comparison of the equation with a balance scale and the nec- essary means of preserving the balance. (2) There arises naturally the subject of equations and the means which it offers of solving many practical problems which the student has already met in arithmetic and will meet in physics. But problems of a practical nature are not all that should be offered. There has been a great deal of discussion to the effect of confining all problems to such as will be of future value either in physics or mechanics or in the business of actual life. Why should the more abstract problems be entirely omitted? There is for everyone, whether he be pupil who is going to devote himself hereafter to the study of mathematics or not, pleasure and profit as well that comes from pursuing through its intricate ramifications and finally securing the solution of such a problem as has been designated a mere mathematical - puzzle. No comparable exhilaration and triumph comes from the solution of the more simple problems found in arithmetic and physics. I do not mean to say anything against the many advantages gained by the introduction of physical problems in algebra. In addition to their practicality, they familiarize the ELEMENTARY ALGEBRA 369 pupil with the formulae and terms of science and give them a good drill in solving equations for unknowns represented by letters other than x, y, or z. These physical formulae should by all means be given among the literal equations. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* Still, though I acknowledge that the problems of daily life and science have their place, nevertheless I advise against making too radical a change and swinging the pendulum to the other extreme. (3) As an introduction to factoring, the special rules of multiplication and division should be presented to the class ex- actly as they appear in the type forme. g.the product of the sum and difference of two numbers is equal to the square of the first minus the square of the secondnot the difference of their squares. When the rule is learned in that form, the pupil has no difficulty in projecting in his mind the image of the type form and when he has to reverse the process and find the factors he will recognize the case immediately from the type form of the product. In the special rules for division, the sum and difference of two cubes should be treated under the gen- eral head of the sum and difference of two odd powers and be governed by the same rules, otherwise the pupil looks upon the cubes as entirely different from other odd powers and the teacher loses the opportunity of giving him first a broad gen- eral view of the case before subdividing it into special divisions. When the subject of factoring is reached, all energy and time should of course be spent upon the ordinary cases which are used most frequently later in fractions, quadratic ^ equations, etc. The class should be thoroughly drilled in these so that finally all factoring can be performed with accuracy and skill. But the examples that do not come under any of the more general heads but require special treatment should be left until the pupil finds a need for them later in his mathematical studies. The factor theorem as a theorem is a pretty and useful piece of mathematics but entirely too abstract to be understood by immature first year high school pupils. Later in an algebraic review it may and should be introduced but never to the first year pupils. Because of this omission the method of finding the H. C. F. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* by division should be taught to enable the pupil to reduce to lowest terms fractions that cannot be easily factored; but not much time should be spent on this interesting bit of algebra, for after the principle is understood, the work amounts to a SCHOOL SCIENCE AND MATHEMATICS 370 drill in division merely. I do not think it should be given the important place in high school, algebra it has hitherto occupied. for it should be confined to cases where factoring is seemingly impossible; but it has its uses even in the first year and its principles are easily taught and understood. (4) The subject of algebraic fractions cannot be studied and learned successfully until the underlying principles of arithmetical fractions are first learned and thoroughly assimi- lated. Even though in the graded schools the subject has been taught faithfully and conscientiously, it is absolutely necessary to give a comprehensive review of the ground work of fractions in arithmetic before beginning the work in algebraic fractions. Pupils learn mechanically that i=i, etc. They know almost intuitively that they have not multiplied the fraction i- by 3. But when the fraction has been placed equal to ~~r they will undoubtedly claim that has been multiplied by b. Also they will almost always find the product of by b to be ,’ Mistakes of this kind and mistakes resulting from cancellation of terms are only too frequent. I have found that they can be corrected only by showing how incorrect such methods would be if applied to ordinary arithmetical fractions. So drill in these principles should be given until the pupil understands them thoroughly and can apply them unerringly as well. The three signs of a fraction have been a source of perplexity to the class as a whole. I shall merely suggest here a plan I have found to be effectual. Regarding -, as indicating a pro- ,.. 0 cess in division in which the quotient is to be found by dividing a by b and prefixing the sigh according to the law of signs of divisio, ,e h.,« .(–5). y +(3) = –-, -(^)»y -(3) ’ +? H"« ^(S) = -G) - -CT?) = -(3) and the principles follow. and the principles follow. PROBLEMS THAT ARISE IN THE TEACHING OF ELEMEN- TARY ALGEBRA IN THE FIRST YEAR HIGH SCHOOL COURSE.* (5) In regard to the last part of the algebra taught in the first year, I shall merely suggest, (i) that the omission of cube root entirely is wise on the ground of its infrequent ap’ plication; (2) that the meaning of zero and negative exponents ELEMENTARY ALGEBRA 371 be omitted until the class is prepared to consider the theory of exponents (in some of our text-books the proofs leading to a°=l and cT^^^ A. are given with rules of division); (3) that in ^m quadratic equations one method only of completing the square be presented, the equations to be solved preferably by factoring or by formula. In conclusion the number of new concepts and the amount of theory that the class as a whole can comprehend and apply in a given time, and the amount’of drill that is sufficient to enable its members so to assimilate the theory as to make it a part of their mental equipment and so grow in the mental strength and independence that comes from the exercise of reasoning, both depend largely upon the kind of school, the size of the classes and their personnel, and the number of hours per week allotted to the teaching of the subject. It is this intelligent understanding of each pupil’s individual needs that the teacher should strive to possess, for with that knowledge he can more surely present the subject in such a manner as to best aid the pupil not merely to grow in knowledge but to cultivate and develop his mental power. READING NOTICE. The sixteenth Summer Session of Gornell University will begin July 4, and close August 14, 1907. As last year it will be under the general direction of Professor George P. Bristol, and the plan of the work will be as in former years. Twenty departments of instruction are included, and some of the most noted scholars and most effective teachers of the Cornell faculty will take part in the work. They will be aided by specially competent teachers from other institutions. .A new feature next summer will be a course in General Biology, following the New York State Education Department syllabus. The manual training work, both theory and practice, started last year, will be still further de- veloped. That the advantages which Cornell offers with her extensive plant, her fine laboratories and libraries are appreciated, is proven by the attendance last summer of over 600 persons of whom 300 were teachers in colleges and schools of various kinds.
https://openalex.org/W2116827945	https://fluidsbarrierscns.biomedcentral.com/counter/pdf/10.1186/2045-8118-10-34	English	null	Fingerprint changes in CSF composition associated with different aetiologies in human neonatal hydrocephalus: glial proteins associated with cell damage and loss	Fluids and barriers of the CNS	2,013	cc-by	10,964	Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 FLUIDS AND BARRIERS OF THE CNS FLUIDS AND BARRIERS OF THE CNS RESEARCH Open Access Fingerprint changes in CSF composition associated with different aetiologies in human neonatal hydrocephalus: glial proteins associated with cell damage and loss Irum Naureen1,6, Khawaja A Irfan Waheed2, Ahsen W Rathore2, Suresh Victor3, Conor Mallucci4, John R Goodden5, Shahid N Chohan6 and Jaleel A Miyan1* * Correspondence: j.miyan@manchester.ac.uk 1Faculty of Life Sciences, The University of Manchester, AV Hill Building, Oxford Road, Manchester M13 9PT, UK Full list of author information is available at the end of the article * Correspondence: j.miyan@manchester.ac.uk 1Faculty of Life Sciences, The University of Manchester, AV Hill Building, Oxford Road, Manchester M13 9PT, UK Full list of author information is available at the end of the article © 2013 Naureen et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Full list of author information is available at the end of the article © 2013 Naureen et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2013 Naureen et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background Hydrocephalus is a pathological condition characterised by an abnormal production and/or absorption of cere- brospinal fluid (CSF). The aetiology of hydrocephalus is unclear, even with some evidence of genetic predispos- ition [1]. With a global incidence of 1:500 (NIH, USA), hydrocephalus affects between 1:100–1:5000 live human births with a lower incidence in developed countries achieved through elective terminations. Hydrocephalus can result in deficient/abnormal cerebral cortex develop- ment and lifelong neurological deficits [2-4]. The patho- physiology of hydrocephalus in neonates remains poorly understood since they do not experience raised intracra- nial pressure despite accumulation of CSF in the brain accompanied by ventricular and cranial expansion [5,6]. The extent of ventricular dilation depends on the loca- tion of CSF blockage and length of time of the block [7,8]. Fetal-onset ventriculomegally (enlarged ventricles without raised pressure) is thought to result in more se- vere brain damage [3,4] but recent studies challenge this notion [9] even though ventricular expansion, and con- sequential compression and stretching of the brain must have a damaging effect in severe cases. Currently the only control measure is birth termination and the only treatment option is surgical fluid diversion. Although surgery does address the life-threatening raised intracra- nial pressure, compression and stretching of the brain parenchyma, it does not recover lost or abnormal devel- opment, cellular damage or the life-long neurological is- sues resulting from these [4-6]. The severity of hydrocephalus is usually assessed from clinical signs and symptoms. In very young patients an increase in head circumference usually occurs with only minor symptoms due to the compliance of the immature skull. CT-scanning and intracranial pressure-monitoring have proved to be valuable tools in the assessment of the diagnosis, but many patients still do not show satisfac- tory outcomes after CSF shunt surgery [20] suggesting an underlying pathophysiological process may not have been addressed by the procedure. The outcome, in terms of possible neurological deficits, may reflect these under- lying processes that are not addressed by shunting and thus do not reflect the brain damage that is directly inflicted by hydrocephalus [4,20,21]. These deficits may not be rescued or recovered [22] and may be a conse- quence of poor, or abnormal development of the cerebral cortex [18,19]. (Continued from previous page) (Continued from previous page) Conclusions: Apart from CNPase, which is an enzyme, the markers investigated are intracellular intermediate filaments and would be present in CSF only if the cells are compromised and the proteins released. Raised GFAP observed in post-haemorrhagic hydrocephalus must reflect damage to astrocytes and ependyma. Raised MBP in post-haemorrhagic and spina bifida with hydrocephalus indicates damage to oligodendrocytes and myelin. Vimentin protein detected in some of the late-onset hydrocephalic samples indicates damage to glial and other progenitors and suggests this condition affects periventricular regions. The presence of CNPase in all CSF samples was unexpected and indicates a possible novel role for this enzyme in brain development/myelination. Less CNPase in some cases of late-onset hydrocephalus could therefore indicate changes in myelination in these infants. This study demonstrates differential glial damage and loss in the developing human neonatal hydrocephalic brain associated with different aetiologies. Keywords: Hydrocephalus, Human neonates, CSF, GFAP, Vimentin, CNPase, MBP of raised pressure rather than possible earlier events, due in large part to the view of CSF as a mechanical support fluid rather than a fluid with physiological importance. The latter role of CSF is currently under considerable re- view and renewed research [10-13] with some new in- sights emerging into the unique part played by CSF in development and brain function [9,14,15]. Recent research has focused on the interplay between CSF signals and their role in normal neurodevelopment [11-13], and there is growing evidence that secreted proteins or other factors within the CSF have an age-dependent effect on neural cell precursor proliferation and cortical development [14]. Thus, major alterations in CSF protein composition in neonatal neurological disease states could have direct and significant effects on global neurodevelopment [16]. Alter- natively an imbalance in the levels of specific CSF proteins may directly affect on-going neurodevelopmental pro- cesses, which could lead to the developmental defects as- sociated with post-haemorrhagic hydrocephalus [17] as well as those already associated with CSF composition changes in fetal-onset hydrocephalus [9,18,19]. Abstract Background: In hydrocephalus an imbalance between production and absorption of cerebrospinal fluid (CSF) results in fluid accumulation, compression and stretching of the brain parenchyma. In addition, changes in CSF composition have a profound influence on the development and function of the brain and together, these can result in severe life-long neurological deficits. Brain damage or degenerative conditions can result in release of proteins expressed predominantly in neurons, astroglia, or oligodendroglia into the brain interstitial fluid, CSF and blood. Determination of such products in the CSF might be of value in diagnosing cause, aetiology and/or assessing the severity of the neurological damage in patients with hydrocephalus. We therefore analysed CSF from human neonates with hydrocephalus for these proteins to provide an insight into the pathophysiology associated with different aetiologies. Methods: CSF was collected during routine lumbar puncture or ventricular tap. Samples were categorized according to age of onset of hydrocephalus and presumed cause (fetal-onset, late-onset, post-haemorrhagic or spina bifida with hydrocephalus). Glial fibrillary acidic protein (GFAP), myelin basic protein (MBP), vimentin and 2′ , 3′-cyclic nucleotide 3′-phosphodiesterase (CNPase) were analysed through Western blotting of hydrocephalic CSF samples (n = 17) and compared with data from CSF of normal infants without neurological deficits (n = 8). Results: GFAP was significantly raised only in CSF from post-haemorrhagic hydrocephalus while MBP was significantly raised in post-haemorrhagic and in spina bifida with hydrocephalus infants. Vimentin protein was only detected in some CSF samples from infants with late-onset hydrocephalus but not from other conditions. Surprisingly, CNPase was found in all neonatal CSF samples, including normal and hydrocephalic groups, although it was reduced in infants with late onset hydrocephalus compared with normal and other hydrocephalic groups. (Continued on next page) © 2013 Naureen et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Page 2 of 12 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Materials and methods Patient samples Prior to the study, ethical approval was obtained from the UK NHS REC (ref: 08/H1015/43), Central Manchester NHS Foundation Trust, Alder Hey Children’s Hospital, Leeds Teaching Hospitals NHS Trust, The University of Manchester, and The Children’s Hospital Lahore. Written informed consent for participation and demographic data including sex, age and ethnicity were obtained from each patient’s parents using their home language with verbal explanation where necessary. This study is based on 25 neonates, 16 males and 9 females varying in age from 1 to 44 weeks (Table 1). CSF samples were collected by trained clinical staff either by lumbar puncture from normal in- fants (n = 8), presenting with mild fever and suspected meningitis but sera negative, or through insertion of catheters into the brain lateral ventricle to drain fluid from infants with hydrocephalus. All CSF samples were sera negative for known infective organisms. Any sera positive samples were excluded from this study as were all samples with obvious blood contamination. The procedures and tests were part of the routine clinical We received limited information concerning each sample so that sub-categorisation based on severity of hydrocephalus, neurological deficits etc. was not possible in this initial study. Thus, we categorised hydrocephalic CSF samples according to age of onset and the cause of hydrocephalus into the following groups: 1. Fetal-onset or congenital hydrocephalus (FOH) in infants born with classically enlarged heads and associated neurological symptoms (n = 4); 2. Late-onset hydrocephalus (LOH) in infants developing hydrocephalus some days or weeks after birth due to undetermined factors but probably infections (n = 4); 3. Post-haemorrhagic hydrocephalus (PHH) in infants due to an intraventricular or subarach- noid bleed before, during or soon after birth (n = 5); 4. Spina bifida with hydrocephalus (SB/HC) in infants with a primary neural tube defect (n = 4). 5. Control/normal infants with no neurological condition and sera negative for meningitis (n = 8). Sample details are summarised in Table 1. Background Thus, although appropriate diagnostics have been developed for hydrocephalus, the potential of Improvements in prenatal diagnosis and treatment will raise the quality of life for children with hydrocephalus through decreased mortality and morbidity, but this may only be through a better understanding of the underlying aetiology and biological basis of the condition and the development of additional treatments to shunt surgery. The literature in both human and experimental animal studies is not clear with controversy on cause and with much work targeted at understanding the consequences Page 3 of 12 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 management of the patients and we obtained fluid excess to clinical test requirements. Samples were frozen at -80°C within 30 min of collection. UK samples were transported frozen on dry ice while samples collected from Lahore Children’s Hospital, Pakistan, were lyophilised (freeze dried) prior to shipment for analysis in the UK. Lyophi- lised samples were reconstituted to the original volume using deionised water (Milli-Q) before analysis. For obvi- ous clinical reasons it was not possible to obtain CSF sam- ples from the same site in normal and hydrocephalus infants. Total protein would be higher in lumber CSF but other parameters would be expected to be within normal ranges unless damage, inflammation or infection affected specific regions. Any significant differences between hydro- cephalus groups were directly comparable to each other as they were all ventricular samples, but any differences to normal lumbar CSF samples needed to be interpreted with due regard to reported differences between lumbar and ventricular CSF. Obstruction of CSF outflow may also affect composition but one aim of the current study is to determine if CSF composition is indeed altered, by what- ever mechanism, in hydrocephalus. biological markers to characterise the underlying patho- logical processes has so far been overlooked. The composition of the CSF should reflect physio- logical/biochemical changes happening in the brain par- enchyma, and more particularly in the periventricular white matter and sub-ventricular zones lining the ventri- cles [23,24] where atrophy of the white matter, and to some extent the grey matter and damage to axons were found in various hydrocephalic conditions [3,4,25-28]. Background In the neonatal period a major phase of gliogenesis occurs in the ventricular and sub-ventricular zones and it is likely that there is a significant effect from hydroceph- alus on developing glial cells, but less effect on the post- mitotic neurones produced prenatally [29-32]. The aim of this study was to test whether different aetiologies of neonatal hydrocephalus were correlated with different profiles of proteins associated with glial damage in hu- man neonatal hydrocephalus. Materials and methods Patient samples Table 1 Clinical variables of the normal and hydrocephalic neonates used in the study Table 1 Clinical variables of the normal and hydrocephalic neonates used in the study Patient groups Age (days) Sex ratio (male: female) Site of CSF collection Source (UK: Pakistan) Groups n Age range (days after birth) Mean SEM Normal 8 8-92 24.50 9.72 5:3 Lumber 0:8 FOH 4 11-30 18.25 4.40 4:0 Lateral ventricle 1:3 LOH 4 60-300 153.75 55.58 3:1 Lateral ventricle 0:4 PHH 5 28-132 74.80 22.05 2:3 Lateral ventricle 5:0 SB/HC 4 5-105 41.00 22.79 2:2 Lateral ventricle 4:0 Number of patients, age (days post-partum), gender, and national distribution of the patients studied grouped by known aetiology. Late onset hydrocephalus (LOH) and spina bifida with hydrocephalus (SB/HC) infants are somewhat older than the other groups due to the nature of the condition and the time to receiving treatment. All CSF samples were collected from the lateral ventricle except normal samples that were collected by lumbar puncture. Although there is an age range in each group the low number of samples did not allow analysis of age-related changes. FOH: fetal-onset hydrocephalus, PHH: post-haemorrhagic hydrocephalus. Number of patients, age (days post-partum), gender, and national distribution of the patients studied grouped by known aetiology. Late onset hydrocephalus (LOH) and spina bifida with hydrocephalus (SB/HC) infants are somewhat older than the other groups due to the nature of the condition and the time to receiving treatment. All CSF samples were collected from the lateral ventricle except normal samples that were collected by lumbar puncture. Although there is an age range in each group the low number of samples did not allow analysis of age-related changes. FOH: fetal-onset hydrocephalus, PHH: post-haemorrhagic hydrocephalus. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 4 of 12 Dulbecco’s modified medium (DMEM, Invitrogen, Glasgow, UK) and further dissected under a stereo microscope (Leica M3Z). The meninges were removed and the cho- roid plexus removed from the ventricle. All cleaned hemispheres were then digested with trypsin and gently dissociated through decreasing bore sizes of sterile pi- pettes. The suspension was centrifuged for 10 min at 1000 rpm, the supernatant discarded and the pellet re- suspended in DMEM. After a further centrifugation and resuspension to wash the cells, the pellet was resuspended in 10 ml of DMEM per hemisphere. Astrocyte cell lysates In order to produce astrocyte cell lysates for positive control, 1-day old Sprague–Dawley rat pups (Charles River, Ormiston, Scotland) were killed with an overdose of sodium pentobarbitone given by intraperitoneal injec- tion and then decapitated. 2–4 cerebral hemispheres were dissected under sterile conditions in a laminar flow hood. The cortical hemispheres were placed in ice-cold Materials and methods Patient samples 10 ml of cell suspen- sion was loaded into poly-D-lysine coated 75 cm2 flasks and placed in a humidified incubator for 7–9 days at 37°C with 5% CO2. After 7–9 day incubation, flasks were sealed tight and shaken on an orbital shaker at 230 rpm for 4 hours at 37°C to separate microglial cells from astro- cytes. The media was replaced with 10 ml of fresh DMEM and shaken for a further 1 h. The culture was then washed twice with 5 ml of warm PBS with gentle mixing and then the adherent astrocytes lysed using CelLytic M Cell Lysis Reagent (Sigma) for 15 min at room temperature. Cells were scraped off the flask and the whole media centri- fuged to remove cell debris. Protein concentration was measured using a Bradford Assay and 5-10 μg total protein loaded on gels as a control for GFAP and vimentin. Glial fibrillary acidic protein (GFAP) A 53 KDa, GFAP-positive protein band present in the CSF from PHH and SB/HC infants was not detected in normal, FOH and LOH samples (Figure 1a). This GFAP-positive protein was significantly increased in PHH (P < 0.05) and showed a non-significant increase in SB/HC CSF when compared with CSF from normal infants (Figure 1b). Western blotting All CSF samples were compared through analysis of equal volumes. This was the most meaningful approach since we expected to see changes in total protein as well as in individual components. The specifications for each anti- body on the manufacturer’s web sites also listed previous publications characterising the antibody as well as its spe- cificity and utility for identifying the target protein. All samples were diluted 1:1 by volume in Laemmli sample buffer (BioRad 161–0737, Hemel Hempstead, UK) con- taining mercaptoethanol and heated for 7 min at 95°C. Protein standards were included in each gel as well as a molecular weight standard (Biorad161-0375). Rat (post- natal day 1) astrocyte cell lysates (5–10 μg, see below) were used as a positive control for GFAP and vimentin antibodies. The CNPase antibody was well characterised by previous studies of oligodendrocytes and cortical white matter [33,34]. Furthermore, we had previous data from mass spectrometry that CNPase was indeed present in CSF samples [35]. 5 μl of each CSF sample were separated together with control proteins and molecular weight markers on a 4-12% SDS-PAGE gel and transferred to a PVDF membrane using an iBlot system with blotting kits (Invitrogen, Glasgow, UK). The membranes were blocked using 5% skimmed-milk (Marvel) or 5% fish skin gelatine (Sigma Poole, UK) in 0.1% phosphate- buffer solution (pH 7.4) with 0.1% Tween (Sigma) for 1 hour at room temperature (RT). Primary antibodies were optimised to the following dilutions for 1:4000 (0.25 μg/ml) for rabbit anti-human GFAP (Abcam 7779, Cambridge, UK ), 1:1000 (0.1 μg/ml) for rabbit anti-human Vimentin (Cell Signaling-3932, New Eng- land Biolabs, Hitchin, UK), 1:10,000 (0.1 μg/ml) for rabbit anti-human MBP (Thermo Fisher PA1-18324 Lutterworth, UK), 1:1000 (0.1 μg/ml) for rabbit anti- human CNPase (Sigma C9743) were diluted in blocking solution and membranes incubated overnight at 4°C with- out agitation. Membranes were then washed and incubated in a 1:3000 dilution of the secondary antibody, horseradish peroxidise-conjugated anti-rabbit IgG (Cell signalling 7074) in blocking solution for 1h at RT with agitation, washed and the signals detected using an enhanced chemiluminescence substrate (Amersham Hyper film ECL, GE Healthcare Little Chalfont, UK) and exposed to film. Densitometry and statistical analysis Films were scanned and the relative density of bands quantified using Image J (Version 1.44p, NIH). Where no bands were obvious a measure of the staining (pre- sumably background) where the band would have been was used for those CSF samples. Data are presented as mean ± SEM. Results of the experiments were entered into Graphpad Prism V software and the data analysed using one way analysis of variance (ANOVA) and post hoc (Tukey) testing to identify significant differences between groups. The value p < 0.05 was considered sig- nificant. No distribution tests were carried out and the data was assumed to fall within a normal distribution. Vimentin Ten-la g y g y p ( ) p g lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 53 KDa band was labelled using the antibody to GFAP and only found in PHH and SB/HC samples as shown in this representative gel. All samples were run on at least 3 gels with a total number in each group as noted for the semi-quantitative analysis. FOH: fetal-onset hydrocephalus, LOH: late-onset hydrocephalus, PHH: post-haemorrhagic hydrocephalus, SB/HC: spina bifida with hydrocephalus. b. Semi-quantitative analysis of the 53 KDa band labelled with anti-GFAP antibody. Values are Mean ± SEM of Normal n = 8, FOH n = 4, LOH n = 4, PHH n = 5, SB/HC n = 4. Only PHH showed a significant difference from normal which had undetectable levels so that the difference was effectively from zero protein (P < 0.05 indicated by asterisk ). SB/HC also shows an increase but this was not significantly different from normal. Rat (postnatal P1) astrocytes cell lysates were used as a positive control for the GFAP antibody and these are shown in the last lane. different LOH infants gave a non-significant (P > 0.05) statistical test (Figure 2b) against normal. Two of four LOH samples have obviously raised levels of vimentin compared to any other individual tested. HC infants compared to normal. The elevation of MBP in PHH and SB/HC infants showed little variation be- tween different samples analysed. Vimentin ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software A single 53 KDa band was labelled using the antibody to GFAP and only found in PHH and SB/HC samples as shown in this representative (a) SB/HC (b) (b) SB/HC Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 53 KDa band was labelled using the antibody to GFAP and only found in PHH and SB/HC samples as shown in this representative gel. All samples were run on at least 3 gels with a total number in each group as noted for the semi-quantitative analysis. FOH: fetal-onset hydrocephalus, LOH: late-onset hydrocephalus, PHH: post-haemorrhagic hydrocephalus, SB/HC: spina bifida with hydrocephalus. b. Semi-quantitative analysis of the 53 KDa band labelled with anti-GFAP antibody. Values are Mean ± SEM of Normal n = 8, FOH n = 4, LOH n = 4, PHH n = 5, SB/HC n = 4. Only PHH showed a significant difference from normal which had undetectable levels so that the difference was effectively from zero protein (P < 0.05 indicated by asterisk ). SB/HC also shows an increase but this was not significantly different from normal. Rat (postnatal P1) astrocytes cell lysates were used as a positive control for the GFAP antibody and these are shown in the last lane. Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Ten-lane precast gels were used. Each l h d i di id l 5 l l f CSF f h i f i h Af i i i SDS PAGE h bl d Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Vimentin A 55 KDa vimentin-positive protein band was observed in the CSF from two infants with LOH but was barely de- tected in normal, FOH, PHH or SB/HC infants (Figure 2a). The presence or absence of the vimentin positive band in Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 5 of 12 (a) SB/HC (b) Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 53 KDa band was labelled using the antibody to GFAP and only found in PHH and SB/HC samples as shown in this representative gel. All samples were run on at least 3 gels with a total number in each group as noted for the semi-quantitative analysis. FOH: fetal-onset hydrocephalus, LOH: late-onset hydrocephalus, PHH: post-haemorrhagic hydrocephalus, SB/HC: spina bifida with hydrocephalus. b. Semi-quantitative analysis of the 53 KDa band labelled with anti-GFAP antibody. Values are Mean ± SEM of Normal n = 8, FOH n = 4, LOH n = 4, PHH n = 5, SB/HC n = 4. Only PHH showed a significant difference from normal which had undetectable levels so that the difference was effectively from zero protein (P < 0.05 indicated by asterisk ). SB/HC also shows an increase but this was not significantly different from normal. Rat (postnatal P1) astrocytes cell lysates were used as a positive control for the GFAP antibody and these are shown in the last lane. (a) SB/HC (b) Figure 1 Western blot analyses for glial fibrillary acidic protein (GFAP). a. Western blots for GFAP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. Myelin basic protein (MBP) Semi-quantitative analysis of the 55 KDa band labelled with anti-vimentin antibody relative to the background stain of normal samples where this protein was not detected. Values are Mean ± SEM of Normal n = 8, FOH n = 4, LOH n = 4, PHH n = 4, SB/ HC n = 4. Rat (postnatal P1) astrocyte cell lysates were used as a positive control for the vimentin antibody. No significant differences were found. Myelin basic protein (MBP) Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 55 KDa band was labelled using the antibody to vimentin and only detected in 2 of the 4 LOH samples or glial cell lysates (last lane except for LOH where it was not run on this particular gel) as shown in these representative gels. All samples were run a minimum of 3 times with a total number of samples as given for semi-quantitative analysis. No significant differences to normal were detected in any group even LOH where two samples had this protein and 2 not. A greater sample number is needed to test if vimentin is generally present in LOH or not. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 55 KDa band labelled with anti-vimentin antibody relative to the background stain of normal l h hi i d d V l M SEM f N l 8 FOH 4 LOH 4 PHH 4 SB/ HC 4 R ( l (a) (b) SB/HC (b) SB/HC (b) Figure 2 Western blot analyses for vimentin. a. Western blots for vimentin. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 55 KDa band was labelled using the antibody to vimentin and only detected in 2 of the 4 LOH samples or glial cell lysates (last lane except for LOH where it was not run on this particular gel) as shown in these representative gels. All samples were run a minimum of 3 times with a total number of samples as given for semi-quantitative analysis. No significant differences to normal were detected in any group even LOH where two samples had this protein and 2 not. A greater sample number is needed to test if vimentin is generally present in LOH or not. Patient groups as for Figure 1. b. Myelin basic protein (MBP) A 45 KDa CNPase-positive protein band was detected in all samples of CSF analysed (Figure 4a). CNPase was significantly lower in LOH (P < 0.05) although one in- fant had normal levels. CNPase was reduced but with- out significance in FOH compared with CSF from normal, PHH and SB/HC infants (Figure 4b). A 40 KDa MBP-positive protein band was observed in the CSF from infants with PHH and SB/HC conditions but was below the detection limit in CSF from normal, FOH and LOH infants (Figure 3a). MBP showed a highly significant (P > 0.0001) increase in CSF of PHH and SB/ Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 6 of 12 (b) (a) SB/HC Figure 2 Western blot analyses for vimentin. a. Western blots for vimentin. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 55 KDa band was labelled using the antibody to vimentin and only detected in 2 of the 4 LOH samples or glial cell lysates (last lane except for LOH where it was not run on this particular gel) as shown in these representative gels. All samples were run a minimum of 3 times with a total number of samples as given for semi-quantitative analysis. No significant differences to normal were detected in any group even LOH where two samples had this protein and 2 not. A greater sample number is needed to test if vimentin is generally present in LOH or not. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 55 KDa band labelled with anti-vimentin antibody relative to the background stain of normal samples where this protein was not detected. Values are Mean ± SEM of Normal n = 8, FOH n = 4, LOH n = 4, PHH n = 4, SB/ HC n = 4. Rat (postnatal P1) astrocyte cell lysates were used as a positive control for the vimentin antibody. No significant differences were found. (b) (a) SB/HC Figure 2 Western blot analyses for vimentin. a. Western blots for vimentin. Ten-lane precast gels were used. Discussion a. Western blots for MBP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 40 KDa band was labelled using the antibody to myelin basic protein and only detected in PHH and SB/HC samples as shown in this representative gel. All samples were run a minimum of three times with a total number of samples as given for semi-quantitative analysis. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 40 KDa band labelled with anti-MBP antibody. MBP was increased significantly (P < 0.001) in PHH and SB/HC compared to undetectable levels in normal, FOH and LOH infants. Values are Mean ± SEM of Normal n = 4, FOH n = 4, LOH n = 4, PHH n = 4, SB/HC n = 4. Values in PHH are SB/HC are significantly different to the undetectable level in normal, so effectively from the zero protein level, P < 0.001 indicated by asterisks *. Figure 3 Western blot analyses for myelin basic protein (MBP). a. Western blots for MBP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 40 KDa band was labelled using the antibody to myelin basic protein and only detected in PHH and SB/HC samples as shown in this representative gel. All samples were run a minimum of three times with a total number of samples as given for semi-quantitative analysis. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 40 KDa band labelled with anti-MBP antibody. MBP was increased significantly (P < 0.001) in PHH and SB/HC compared to undetectable levels in normal, FOH and LOH infants. Values are Mean ± SEM of Normal n = 4, FOH n = 4, LOH n = 4, PHH n = 4, SB/HC n = 4. Discussion study is urgently needed. There are clearly sub-categories within each group that show significant changes and so further clinical data may unravel these issues. The proteins we analysed are intermediate filaments found in astrocytes, ependymal cells (GFAP) early glial progenitors and dif- ferentiating astrocytes (vimentin) and oligodendrocytes (MBP) and an enzyme associated with the formation of myelin in oligodendrocytes (CNPase). Their presence in CSF is thus thought to indicate cell compromise, The present study determined CSF levels of GFAP, vimen- tin, MBP and CNPase in human neonatal hydrocephalus with different aetiologies compared to CSF from neonates with no neurological abnormalities. Although significant differences as well as non-significant indications of differ- ences have been found, the low number of samples avail- able for analysis limited clear conclusions. The study does show that differences are likely to exist and that a larger Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 7 of 12 (a) SB/HC (b) * * Figure 3 Western blot analyses for myelin basic protein (MBP). a. Western blots for MBP. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 40 KDa band was labelled using the antibody to myelin basic protein and only detected in PHH and SB/HC samples as shown in this representative gel. All samples were run a minimum of three times with a total number of samples as given for semi-quantitative analysis. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 40 KDa band labelled with anti-MBP antibody. MBP was increased significantly (P < 0.001) in PHH and SB/HC compared to undetectable levels in normal, FOH and LOH infants. Values are Mean ± SEM of Normal n = 4, FOH n = 4, LOH n = 4, PHH n = 4, SB/HC n = 4. Values in PHH are SB/HC are significantly different to the undetectable level in normal, so effectively from the zero protein level, P < 0.001 indicated by asterisks . (a) SB/HC (b) * (a) (b) Figure 3 Western blot analyses for myelin basic protein (MBP). Discussion Semi-quantitative analysis of the 45 KDa band labelled with anti-CNPase antibody shows similar levels of CNPase to that in normal CSF in FOH, PHH and SB/HC but in the LOH group where a significant reduction (P < 0.05 indicated by asterisk ) was observed compared to normal and other hydrocephalic groups. In the LOH gel 3 of the 4 samples show reduced CNPase compared to normal and one sample not. Increased numbers will provide better resolution of this particular group. Values are Mean ± SEM of Normal n = 4, FOH n = 4, LOH n = 4, PHH n = 4, SB/HC n = 4. (a) (a) (b) (b) Figure 4 Western blot analyses for 2′ , 3′-cyclic nucleotide 3′-phosphodiesterase (CNPase). a. Western blots for CNPase. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 45 KDa band was labelled using the antibody to CNPase and was detected in all samples as shown in these representative gels. All samples were run a minimum of three times with a total number of samples as shown for semi-quantitative analysis. Patient groups as for Figure 1. b. Semi-quantitative analysis of the 45 KDa band labelled with anti-CNPase antibody shows similar levels of CNPase to that in normal CSF in FOH, PHH and SB/HC but in the LOH group where a significant reduction (P < 0.05 indicated by asterisk ) was observed compared to normal and other hydrocephalic groups. In the LOH gel 3 of the 4 samples show reduced CNPase compared to normal and one sample not. Increased numbers will provide better resolution of this particular group. Values are Mean ± SEM of Normal n = 4, FOH n = 4, LOH n = 4, PHH n = 4, SB/HC n = 4. interesting to test CNPase on the differentiation of oligo- dendrocytes from cortical stem/progenitor cells in in vitro cell culture as well as to investigate the factors stimulating secretion of this enzyme from oligodendrocytes. hydrocephalus, the latter associated with neonatal and postnatal infections, do not present with raised GFAP but that PHH and SB/HC neonates do. Discussion Values in PHH are SB/HC are significantly different to the undetectable level in normal, so effectively from the zero protein level, P < 0.001 indicated by asterisks *. breakdown and release of intracellular components. In this study we found that FOH, the congenital form of hydro- cephalus, is not associated with any appearance of cell breakdown products in agreement with previous studies demonstrating absence of cell death [36] and decreased cell proliferation with a cell cycle arrest in the H-Tx rat model of FOH [18,19]. LOH had only vimentin detectable in CSF while PHH and SB/HC had GFAP and MBP but not vimentin in CSF. Interestingly, CNPase has never been de- scribed in normal/developmental CSF but we found it in all samples of CSF including those from our non- hydrocephalic control infants. We found it reduced in 3 of 4 samples of LOH compared to other samples and it is in- teresting to speculate whether the 4th sample might have had a non-infectious cause compared to the normal cause in this category of hydrocephalus. Its presence in all neo- natal CSF samples analysed suggests a possible role in the immediate post-natal acceleration of gliogenesis. The source of CSF CNPase must surely be oligodendrocytes and their precursors since no other cell has been shown to express this enzyme. This potential role as a secreted molecule remains to be investigated. It would be very Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 8 of 12 (a) (b) ′ ′ ′ (a) * (b) Figure 4 Western blot analyses for 2′ , 3′-cyclic nucleotide 3′-phosphodiesterase (CNPase). a. Western blots for CNPase. Ten-lane precast gels were used. Each lane had an individual 5 μl sample of CSF from the categories of patients shown. After protein separation using SDS PAGE they were blotted onto a PVDF membrane for probing with antibodies. ECL detection was used to visualise the labelled bands and for semi-quantitative analysis using Image-J software. A single 45 KDa band was labelled using the antibody to CNPase and was detected in all samples as shown in these representative gels. All samples were run a minimum of three times with a total number of samples as shown for semi-quantitative analysis. Patient groups as for Figure 1. b. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 If confirmed the finding of CNPase in non-myelin producing cells presents a further mystery regarding its functions along with its presence in CSF found in the current study. Absence of CNPase results in serious conditions. CNPase-knockout mice show severe symptoms, including convulsions and ataxia, and most homozygous mice die between 6 and 12 months of age [64]. Decrease in CNPase expression is also observed in chronic schizophrenic patients [65] indicating a possible involvement in the underlying developmental pathology. In kaolin-induced hydrocephalic rats Del Bigio et al. [43] reported decreased CNPase in the corpus callo- sum/supraventricular white matter, fimbria, medulla, and spinal cord indicating a sensitivity of myelin in these regions to raised intracranial pressure. Furthermore, a reduction in the number of CNPase immunoreactive oli- godendrocytes both in the subependymal layer and the cerebral cortex of hydrocephalic rat brains correlated with the severity of the hydrocephalus [66]. CNPase has not been described in neonatal CSF before this study and was an unexpected finding particularly its presence in all neonatal CSF samples tested. LOH had reduced CNPase in CSF compared to normal and other conditions. As CNPase is thought to be involved in oligodendrocyte surface membrane expansion and migration during early stages of axonal ensheathment [67], its presence in CSF may indicate a potential role in global brain myelination MBP in CSF is an indicator of damage specifically asso- ciated with myelination as this protein is only known to be present in oligodendrocytes [45,46]. MBP is an import- ant marker to study in hydrocephalus since the degree of pathology and functional neurological deficits are largely associated with lack of, or damage to myelination, e.g. of the corpus callosum and periventricular white matter. We found significantly elevated levels of MBP in the CSF of infants with PHH and SB/HC indicating that demyelin- ation, and/or a failure of myelination is likely to be occur- ring in these conditions, but not in FOH or LOH. The intraventricular or subarachnoid bleeding in PHH is likely to contribute directly to brain damage [47] and to the de- myelination of the periventricular white matter reported in PHH. Sutton et al., [48] reported the presence of MBP protein in the CSF of hydrocephalic patients with different aetiologies and suggested that active hydrocephalus, asso- ciated with progressive ventricular dilatation, would pro- duce periventricular demyelination through mechanical stretching of the brain parenchyma. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 9 of 12 astrogliosis and microgliosis were observed and corre- lated with the severity of hydrocephalus and increasing age [42]. Moreover, Del Bigio et al. reported signifi- cantly increased GFAP levels in hydrocephalic rats which decreased after shunting suggesting an effect of fluid accumulation and/or raised intracranial pressure on these cells [43]. Similar increased GFAP RNA and protein levels were observed in 10-day old kaolin- induced hydrocephalic kittens which also decreased after shunting [44]. Ependymal damage is well docu- mented in hydrocephalus as is astrogliosis so it is per- haps not surprising to find GFAP in CSF as a consequence of specific insults. The finding that GFAP is not present in CSF from FOH and LOH must there- fore indicate protection of these cells from damage in these forms of hydrocephalus. Thus, the current study indicates early-stage pathology differs depending on etiology. also express vimentin, specifically after brain damage and/ or activation through local inflammatory mediators [57]. In hydrocephalus vimentin-positive cells increase around dis- rupted areas of ependyma suggesting that reactive microglia and proliferating immature glial cells are associated with areas of ependymal cell loss [58]. Our data would further suggest that these cells may be compromised by fluid accu- mulation and raised intracranial pressure. The increased levels of vimentin in the LOH infants must be a direct con- sequence of infection affecting the brain but the lack of GFAP in LOH indicates that the source is unlikely to be astroglial and may be microglial, endothelial or leptomenin- geal since many patients present with meningitis. This requires further analysis as it indicates a very different pathophysiology in these patients compared to those with other aetiologies of hydrocephalus. CNPase is a protein accounting for approximately 4% of myelin protein content and is considered an index of myelin formation [59,60] where the amount of immuno- reactive CNPase correlates with the thickness of the myelin sheath in the central nervous system [61,62]. Wu and colleagues [63] reported that they also found CNPase expressed in prenatal and early postnatal microglial cells in rat brain with a gradual decrease with age, essentially undetectable after birth. They suggested that downregula- tion of CNPase related to the transformation of microglia from the mobile and amoeboid type to the ramified type during development. Discussion GFAP was reported in the CSF of elderly normal pressure hydrocephalus (NPH) patients compared with neurologically healthy age- matched controls [40,41] suggesting a common patho- physiology involving astroglial damage in these types of hydrocephalus, but astroglial protection in FOH and LOH in the neonate. This protection may not be present in older untreated individuals. The latter point is supported by findings in hydrocephalic H-Tx rats in which reactive GFAP is an intermediate filament expressed in mature astrocytes [37,38] and immature ependymal cells [39]. Thus the GFAP seen in hydrocephalic CSF may originate from either source as the ependyma is likely to be matur- ing in the neonatal period. Our findings support those re- ported for experimental neonatal acquired hydrocephalus [3,4] and also show that congenital and late-onset Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Increased intracranial pressure has a greater impact on early cognitive develop- ment than increased CSF volume, and the negative effect is partially reversible through early ventricular shunting [49] giving improved myelination. The fact that MBP is not detected in FOH and LOH again indicates a lack of direct pathology in these specific aetiologies of hydroceph- alus and the possibility to prevent loss of myelination through early intervention. Vimentin is a cytoskeletal intermediate filament in- volved in maintaining cell integrity in many different cells types including fibroblasts and endothelial cells [50] as well as progenitor cells and early neuronal and glial cells [51-53]. Vimentin is found in neuronal stem and progenitor cells and astrocytes during the early postnatal period and re-expressed in reactive astrocytes in cases of central nervous system injury [54-56]. Microglial cells Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 10 of 12 not previously considered. This requires further study but may prove to be an important tool in promoting myelin- ation in affected brains. differentiated by CSF analysis. This presents an import- ant step in understanding this devastating neurological condition. These results suggest that early-stage changes are associated, not with raised pressure, but with currently unknown physiological changes that affect dif- ferent cell types in different locations within the hydro- cephalic brain. Shunting will only address the common end point of these different aetiologies and is unlikely to address the underlying differences in physiology. It is therefore important to understand these physiological processes in order to prevent irreversible cell damage and lasting neurological deficits. This study has a number of limitations. Firstly, to ob- tain CSF free of contamination excluded many collected samples so that a much larger study is required to investi- gate our current findings in more detail. Secondly, it is practically impossible to obtain CSF samples through ventricular catheterisation from healthy neonates due to technical limitations and ethical issues. Similarly, it is unusual to obtain lumbar CSF from hydrocephalic pa- tients. Although some reports indicate that ventricular CSF may be more stable and accurate for measuring pathological changes [68], others contradict this [69]. For our study, all normal samples were taken by lumbar puncture while all hydrocephalic CSF came from lateral ventricles. Authors’ contributions JAM and IN conceived and designed the study. IN carried out all of the lab work. KAIW, AWR, SV, CM and JRG were involved in providing clinical samples, discussing the study design and data collected. SNC coordinated the work in Pakistan. All authors contributed to various drafts of the manuscript and have read and approved the final version. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 We assumed that only obstructive hydro- cephalus would produce a change in ventricular CSF that was not reflected in lumbar drainage but also ac- knowledge reported differences in total protein between lumbar and ventricular CSF which we believe would not affect the validity of our findings since we are looking for changes in specific proteins associated with path- ology. Of note here is the finding of CNPase in both normal and affected neonatal CSF at equal levels (except in LOH) that supports an argument that the measure- ments are comparable in this study. Competing interests The authors declare they have no competing interests. Acknowledgements p y Taken together, the findings of this study indicate that different aetiologies leading to hydrocephalus are associ- ated with different pathophysiological mechanisms af- fecting different cell types, at least in the initial stages where intracranial pressure may not be pathologically raised. The congenital form of hydrocephalus has no de- tectable pathological markers in CSF but is known to be affected by a physiological block of cell cycle as well as a folate block [18,19]. Differential diagnosis is therefore possible and the potential for more effective treatment of these different conditions may emerge from further re- search. Further studies are clearly needed and, with more detailed clinical data, e.g., measurements of ventriculomeg- ally and additional radiographical data on brain measures, developmental and neurological signs and symptoms, as well as pre- and post-surgical data, a more robust differential diagnostic is likely to emerge which would feed into research for differential treatment depending on the characterisation of the underlying pathology. Furthermore, we have focused on glial cell effects but additional data on neuronal status is also needed in dif- ferentiating between the different aetiologies leading to hydrocephalus. Financial support was provided by The Charles Wolfson Charitable Trust (JAM). IN was supported through a British Council INSPIRE grant (SP055) to The University of Manchester, UK and COMSATS Institute for Information Technology, Islamabad, Pakistan that also supported the work carried out in Islamabad and Lahore. We thank Dr Nadeem Malik (Neurosurgeon) and the Neonatology registrars at the Children’s Hospital Lahore for CSF collections in Pakistan; Sasha Burn, Dawn Williams, Clare Jennings, Anna Hendrickson, Jennifer Hill for CSF collections in the UK; Professor M. Akhtar, FRS and Dr Naeem of the School of Biological Sciences, University of Punjab, Lahore, for access to laboratory facilities for sample preparation and lyophilisation. Abbreviations CNP ′ ′ Abbreviations CNPase: 2′, 3′-cyclic nucleotide 3′-phosphodiesteras; CSF: Cerebrospinal fluid; FOH: Fetal-onset hydrocephalus; GFAP: Glial fibrillary Acidic protein; H-Tx: Hydrocephalic Texas rat; LOH: Late onset hydrocephalus; MBP: Myelin basic protein; PHH: Post-haemorrhagic hydrocephalus; SB/HC: Spina bifida with hydrocephalus. 1. Zhang J, Williams MA, Rigamonti D: Genetics of human hydrocephalus. J Neurol 2006, 253:1255–1266. 2. Pattisapu JV: Etiology and clinical course of hydrocephalus. Neurosurg Clin N Am 2001, 12:651–659. vii. 3. McAllister JP, Chovan P: Neonatal hydrocephalus. Mechanisms and consequences. Neurosurg Clin N Am 1998, 9:73–93. 4. McAllister JP 2nd: Pathophysiology of congenital and neonatal hydrocephalus. Semin Fetal Neonatal Med 2012, 17:285–294. Author details 1 1Faculty of Life Sciences, The University of Manchester, AV Hill Building, Oxford Road, Manchester M13 9PT, UK. 2Department of Neonatology, The Children’s Hospital and Institute of Child Health, Ferozepur Road, Lahore, Pakistan. 3Institute of Human Development, Manchester Academic Health Sciences Centre, The University of Manchester, Oxford Road, Manchester M13 9WL, UK. 4Neurosurgical Unit, Alder Hey Children’s Hospital, Eaton Road, Liverpool L12 2AP, UK. 5Department of Neurosurgery, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK. 6Department of Biosciences, COMSATS Institute of Information and Technology, Islamabad, Pakistan. Received: 2 September 2013 Accepted: 14 December 2013 Published: 18 December 2013 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Owen-Lynch PJ, Draper CE, Mashayekhi F, Bannister CM, Miyan JA: Defective cell cycle control underlies abnormal cortical development in the hydrocephalic Texas rat. BRAIN 2003, 126:623–631. 19. Mashayekhi F, Draper CE, Bannister CM, Pourghasem M, Owen-Lynch PJ, Miyan JA: Deficient cortical development in the hydrocephalic Texas (H-Tx) rat: a role for CSF. BRAIN 2002, 125:1859–1874. 43. Del Bigio MR, Kanfer JN, Zhang YW: Myelination delay in the cerebral white matter of immature rats with kaolin-induced hydrocephalus is reversible. J Neuropathol Exp Neurol 1997, 56:1053–1066. 44. Eskandari R, Harris CA, McAllister JP 2nd: Reactive astrocytosis in feline neonatal hydrocephalus: acute, chronic, and shunt-induced changes. Childs Nerv Syst 2011, 27:2067–2076. 20. Hale PM, Mcallister JP, Katz SD, Wright LC, Lovely TJ, Miller DW, Wolfson BJ, Salotto AG, Shroff DV: Improvement of cortical morphology in infantile hydrocephalic animals after ventriculoperitoneal shunt placement. Neurosurgery 1992, 31:1085–1096. 45. Levin SD, Hoyle NR, Brown JK, Thomas DG: Cerebrospinal fluid myelin basic protein immunoreactivity as an indicator of brain damage in children. Dev Med Child Neurol 1985, 27:807–813. 21. Mori K, Shimada J, Kurisaka M, Sato K, Watanabe K: Classification of hydrocephalus and outcome of treatment. Brain Dev 1995, 17:338–348. 46. Miller A, Glass-Marmor L, Abraham M, Grossman I, Shapiro S, Galboiz Y: Bio-markers of disease activity and response to therapy in multiple sclerosis. Clin Neurol Neurosurg 2004, 106:249–254. 22. Hawkins D, Bowers TM, Bannister CM, Miyan JA: The functional outcome of shunting H-Tx rat pups at different ages. Eur J Pediatr Surg 1997, 7(Suppl 1):31–34. 47. Yung YC, Mutoh T, Lin ME, Noguchi K, Rivera RR, Choi JW, Kingsbury MA, Chun J: Lysophosphatidic acid signaling may initiate fetal hydrocephalus. Sci Transl Med 2011, 3:99ra87. 23. McKeating EG, Andrews PJ, Tocher JI, Menon DK: The intensive care of severe head injury: a survey of non-neurosurgical centres in the United Kingdom. Br J Neurosurg 1998, 12:7–14. 47. Yung YC, Mutoh T, Lin ME, Noguchi K, Rivera RR, Choi JW, Kingsbury MA, Chun J: Lysophosphatidic acid signaling may initiate fetal hydrocephalus Sci Transl Med 2011, 3:99ra87. 24. Wikkelso C, Blomstrand C: Cerebrospinal fluid proteins and cells in normal-pressure hydrocephalus. J Neurol 1982, 228:171–180. 48. Sutton LN, Wood JH, Brooks BR, Barrer SJ, Kline M, Cohen SR: Cerebrospinal fluid myelin basic protein in hydrocephalus. J Neurosurg 1983, 59:467–470. 25. Del Bigio MR: Neuropathological changes caused by hydrocephalus. Acta Neuropathol 1993, 85:573–585. 49. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 The University of Manchester, Faculty of Life Sciences; 2006. 12. Lehtinen MK, Zappaterra MW, Chen X, Yang YJ, Hill AD, Lun M, Maynard T, Gonzalez D, Kim S, Ye P, et al: The cerebrospinal fluid provides a proliferative niche for neural progenitor cells. Neuron 2011, 69:893–905. 36. Miyan JA, Khan MI, Kawarada Y, Sugiyama T, Bannister CM: Cell death in the brain of the HTx rat. Eur J Pediatr Surg 1998, 8(Suppl 1):43–48. g 13. Zappaterra MW, Lehtinen MK: The cerebrospinal fluid: regulator of neurogenesis, behavior, and beyond. Cell Mol Life Sci 2012, 69:2863–2878. 37. Bartosik-Psujek H, Stelmasiak Z: Biochemical markers of damage of the central nervous system in multiple sclerosis. Ann Univ Mariae Curie Sklodowska Med 2001, 56:389–392. 14. Miyan JA, Zendah M, Mashayekhi F, Owen-Lynch PJ: Cerebrospinal fluid supports viability and proliferation of cortical cells in vitro, mirroring in vivo development. Cerebrospinal Fluid Res 2006, 3:2. 38. Malmestrom C, Haghighi S, Rosengren L, Andersen O, Lycke J: Neurofilament light protein and glial fibrillary acidic protein as biological markers in MS. Neurology 2003, 61:1720–1725. 15. Nabiuni M, Rasouli J, Parivar K, Kochesfehani HM, Irian S, Miyan JA: In vitro effects of fetal rat cerebrospinal fluid on viability and neuronal differentiation of PC12 cells. Fluids Barriers CNS 2012, 9:8. 39. Roessmann U, Velasco ME, Sindely SD, Gambetti P: Glial fibrillary acidic protein (GFAP) in ependymal cells during development. An immunocytochemical study. Brain Res 1980, 200:13–21. 16. Schutzer SE, Liu T, Natelson BH, Angel TE, Schepmoes AA, Purvine SO, Hixson KK, Lipton MS, Camp DG, Coyle PK, et al: Establishing the proteome of normal human cerebrospinal fluid. PLoS One 2010, 5:e10980. 40. Albrechtsen M, Sorensen PS, Gjerris F, Bock E: High cerebrospinal fluid concentration of glial fibrillary acidic protein (GFAP) in patients with normal pressure hydrocephalus. J Neurol Sci 1985, 70:269–274. 17. Morales DM, Townsend RR, Malone JP, Ewersmann CA, Macy EM, Inder TE, Limbrick DD Jr: Alterations in protein regulators of neurodevelopment in the cerebrospinal fluid of infants with posthemorrhagic hydrocephalus of prematurity. Mol Cell Proteomics 2012, 11:M111 011973. 41. Tullberg M, Rosengren L, Blomsterwall E, Karlsson JE, Wikkelso C: CSF neurofilament and glial fibrillary acidic protein in normal pressure hydrocephalus. Neurology 1998, 50:1122–1127. 42. Miller JM, McAllister JP 2nd: Reduction of astrogliosis and microgliosis by cerebrospinal fluid shunting in experimental hydrocephalus. Cerebrospinal Fluid Res 2007, 4:5. 18. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Page 11 of 12 5. Vachha B, Adams R: Language sample analysis in children with myelomeningocele and shunted hydrocephalus. Eur J Pediatr Surg 2003, 13(Suppl 1):S36–S37. 30. Murray K, Noble M: In vitro studies on the comparative sensitivities of cells of the central nervous system to diphtheria toxin. J Neurol Sci 1985, 70:283–293. 31. Khan OH, Enno TL, Del Bigio MR: Brain damage in neonatal rats following kaolin induction of hydrocephalus. Exp Neurol 2006, 200:311–320. 6. Hetherington R, Dennis M, Barnes M, Drake J, Gentili F: Functional outcome in young adults with spina bifida and hydrocephalus. Childs Nervous System 2006, 22:117–124. 32. Del Bigio MR, Zhang YW: Cell death, axonal damage, and cell birth in the immature rat brain following induction of hydrocephalus. Exp Neurol 1998, 154:157–169. 7. Milhorat TH: Hydrocephalus and the Cerebrospinal Fluid. Baltimore: Williams and Wilkins Co; 1972. 8. Milhorat TH, Davis DA, Hammock MK: Localization of ouabin-sensitive Na-K ATPase in frog, rabbit and rat choroid plexus. Brain Res 1975, 99:170–174. 33. Ding YX, Wei LC, Liu YH, Duan L, Jiao XY, Xia Y, Chen LW: Midbrain neural stem cells show unique cell survival, neuronal commitment and neurotrophic properties with therapeutic potential for parkinson’s disease. J Alzheimers Dis Res 2012, S10:001. doi:10.4172/2161-0460. 9. Cains S, Shepherd A, Nabiuni M, Owen-Lynch PJ, Miyan J: Addressing a folate imbalance in fetal cerebrospinal fluid can decrease the incidence of congenital hydrocephalus. J Neuropathol Exp Neurol 2009, 68:404–416. 34. Stephenson DT, O’Neill SM, Narayan S, Tiwari A, Arnold E, Samaroo HD, Du F, Ring RH, Campbell B, Pletcher M, et al: Histopathologic characterization of the BTBR mouse model of autistic-like behavior reveals selective changes in neurodevelopmental proteins and adult hippocampal neurogenesis. Mol Autism 2011, 2:7. 10. Gato A, Desmond ME: Why the embryo still matters: CSF and the neuroepithelium as interdependent regulators of embryonic brain growth, morphogenesis and histiogenesis. Dev Biol 2009, 327:263–272. 10. Gato A, Desmond ME: Why the embryo still matters: CSF and the neuroepithelium as interdependent regulators of embryonic brain growth, morphogenesis and histiogenesis. Dev Biol 2009, 327:263–272. 11. Lehtinen MK, Walsh CA: Neurogenesis at the brain-cerebrospinal fluid interface. Annu Rev Cell Dev Biol 2011, 27:653–679. 11. Lehtinen MK, Walsh CA: Neurogenesis at the brain-cerebrospinal fluid interface. Annu Rev Cell Dev Biol 2011, 27:653–679. 35. Nabiuni M: Analysis of protein content of cerebrospinal fluid in developing hydrocephalic texas rat, PhD Thesis. Conclusions This study has demonstrates that different aetiologies leading to neonatal hydrocephalus can likely be 4. McAllister JP 2nd: Pathophysiology of congenital and neonatal hydrocephalus. Semin Fetal Neonatal Med 2012, 17:285–294. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Hanlo PW, Gooskens RJ, van Schooneveld M, Tulleken CA, van der Knaap MS, Faber JA, Willemse J: The effect of intracranial pressure on myelination and the relationship with neurodevelopment in infantile hydrocephalus. Dev Med Child Neurol 1997, 39:286–291. 26. Del Bigio MR, Bruni JE, Vriend JP: Monoamine neurotransmitters and their metabolites in the mature rabbit brain following induction of hydrocephalus. Neurochem Res 1998, 23:1379–1386. 50. Franke WW, Schmid E, Osborn M, Weber K: Different intermediate-sized filaments distinguished by immunofluorescence microscopy. Proc Natl Acad Sci U S A 1978, 75:5034–5038. 27. Del Bigio MR, da Silva MC, Drake JM, Tuor UI: Acute and chronic cerebral white matter damage in neonatal hydrocephalus. Can J Neurol Sci 1994, 21:299–305. 51. Kamei Y, Inagaki N, Nishizawa M, Tsutsumi O, Taketani Y, Inagaki M: Visualization of mitotic radial glial lineage cells in the developing rat brain by Cdc2 kinase-phosphorylated vimentin. Glia 1998, 23:191–199. 28. Harris NG, Jones HC, Patel S: Ventricle shunting in young H-Tx rats with inherited congenital hydrocephalus: a quantitative histological study of cortical grey matter. Childs Nerv Syst 1994, 10:293–301. discussion 301. 52. Kalman M, Szekely AD, Csillag A: Distribution of glial fibrillary acidic protein and vimentin-immunopositive elements in the developing chicken brain from hatch to adulthood. Anat Embryol (Berl) 1998, 198:213–235. 29. Kondziella D, Qu H, Ludemann W, Brinker T, Sletvold O, Sonnewald U: Astrocyte metabolism is disturbed in the early development of experimental hydrocephalus. J Neurochem 2003, 85:274–281. Page 12 of 12 Page 12 of 12 Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 53. Sancho-Tello M, Valles S, Montoliu C, Renau-Piqueras J, Guerri C: Developmental pattern of GFAP and vimentin gene expression in rat brain and in radial glial cultures. Glia 1995, 15:157–166. 54. Mucke L, Eddleston M: Astrocytes in infectious and immune-mediated diseases of the central nervous system. FASEB J 1993, 7:1226–1232. 55. Eddleston M, de la Torre JC, Oldstone MB, Loskutoff DJ, Edgington TS, Mackman N: Astrocytes are the primary source of tissue factor in the murine central nervous system. A role for astrocytes in cerebral hemostasis. J Clin Invest 1993, 92:349–358. 56. Eddleston M, Mucke L: Molecular profile of reactive astrocytes–implications for their role in neurologic disease. Neuroscience 1993, 54:15–36. 57. Graeber MB, Streit WJ, Kreutzberg GW: The microglial cytoskeleton: vimentin is localized within activated cells in situ. J Neurocytol 1988, 17:573–580. 58. Takano T, Rutka JT, Becker LE: Overexpression of nestin and vimentin in ependymal cells in hydrocephalus. Acta Neuropathol 1996, 92:90–97. 59. Reinikainen KJ, Pitkanen A, Riekkinen PJ: 2′,3′-cyclic nucleotide-3′- phosphodiesterase activity as an index of myelin in the post-mortem brains of patients with Alzheimer’s disease. Neurosci Lett 1989, 106:229–232. 60. Vogel US, Thompson RJ: Molecular structure, localization, and possible functions of the myelin-associated enzyme 2′,3′-cyclic nucleotide 3′-phosphodiesterase. J Neurochem 1988, 50:1667–1677. p p 61. Braun PE, Sandillon F, Edwards A, Matthieu JM, Privat A: Immunocytochemical localization by electron microscopy of 2′3′-cyclic nucleotide 3′-phosphodiesterase in developing oligodendrocytes of normal and mutant brain. J Neurosci 1988, 8:3057–3066. 62. Trapp BD, Bernier L, Andrews SB, Colman DR: Cellular and subcellular distribution of 2′,3′-cyclic nucleotide 3′-phosphodiesterase and its mRNA in the rat central nervous system. J Neurochem 1988, 51:859–868. 63. Wu CY, Lu J, Cao Q, Guo CH, Gao Q, Ling EA: Expression of 2′,3′-cyclic nucleotide 3′-phosphodiesterase in the amoeboid microglial cells in the developing rat brain. Neuroscience 2006, 142:333–341. 64. Lappe-Siefke C, Goebbels S, Gravel M, Nicksch E, Lee J, Braun PE, Griffiths IR, Nave KA: Disruption of Cnp1 uncouples oligodendroglial functions in axonal support and myelination. Nat Genet 2003, 33:366–374. 65. Hakak Y, Walker JR, Li C, Wong WH, Davis KL, Buxbaum JD, Haroutunian V, Fienberg AA: Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia. Proc Natl Acad Sci U S A 2001, 98:4746–4751. 66. Olopade FE, Shokunbi MT, Siren AL: The relationship between ventricular dilatation, neuropathological and neurobehavioural changes in hydrocephalic rats. Fluids Barriers CNS 2012, 9:19. 67. Naureen et al. Fluids and Barriers of the CNS 2013, 10:34 http://www.fluidsbarrierscns.com/content/10/1/34 Yin X, Peterson J, Gravel M, Braun PE, Trapp BD: CNP overexpression induces aberrant oligodendrocyte membranes and inhibits MBP accumulation and myelin compaction. J Neurosci Res 1997, 50:238–247. 68. Talab R, Valis M, Rehak S, Krejsek J: Abnormalities of tau-protein and beta-amyloid in brain ventricle cerebrospinal fluid. Neuro Endocrinol Lett 2009, 30:647–651. 69. Moghekar A, Goh J, Li M, Albert M, O’Brien RJ: Cerebrospinal fluid Abeta and tau level fluctuation in an older clinical cohort. Arch Neurol 2012, 69:246–250. doi:10.1186/2045-8118-10-34 Cite this article as: Naureen et al.: Fingerprint changes in CSF composition associated with different aetiologies in human neonatal hydrocephalus: glial proteins associated with cell damage and loss. Fluids and Barriers of the CNS 2013 10:34. doi:10.1186/2045-8118-10-34 Cite this article as: Naureen et al.: Fingerprint changes in CSF composition associated with different aetiologies in human neonatal hydrocephalus: glial proteins associated with cell damage and loss. Fluids and Barriers of the CNS 2013 10:34. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission
https://openalex.org/W4293704590	https://arrow.tudublin.ie/context/engscheleart2/article/1371/viewcontent/Impact_of_Distributed_Energy_Resources_in_Smart.pdf	English	null	Impact of Distributed Energy Resources in Smart Homes and Community-Based Electricity Market	IEEE transactions on industry applications	2,023	cc-by-sa	10,209	Articles Follow this and additional works at: https://arrow.tudublin.ie/engscheleart2 Part of the Electrical and Computer Engineering Commons Part of the Electrical and Computer Engineering Commons Technological University Dublin Technological University Dublin ARROW@TU Dublin ARROW@TU Dublin School of Electrical and Electronic Engineering Articles Recommended Citation Recommended Citation Saif, Aziz; Khadem, Shafi K.; Conlon, Michael; and Norton, Brian, "Impact of Distributed Energy Resources in Smart Homes and Community-Based Electricity Market" (2023). Articles. 343. https://arrow.tudublin.ie/engscheleart2/343 This Article is brought to you for free and open access by the School of Electrical and Electronic Engineering at ARROW@TU Dublin. It has been accepted for inclusion in Articles by an authorized administrator of ARROW@TU Dublin. For more information, please contact arrow.admin@tudublin.ie, aisling.coyne@tudublin.ie, vera.kilshaw@tudublin.ie. This work is licensed under a Creative Commons Attribution-Share Alike 4.0 International License. This work is licensed under a Creative Commons Attribution-Share Alike 4.0 International License. Funder: BEYOND project; ERA-Net Smart Energy Systems European Union's Horizon 2020 research and innovation programme (Grant Number: 775970); 10.13039/501100001603-Sustainable Energy Authority of Ireland (Grant Number: 91/RDD/578) This work is licensed under a Creative Commons Attribution-Share Alike 4.0 International License Funder: BEYOND project; ERA-Net Smart Energy Systems European Union's Horizon 2020 research programme (Grant Number: 775970); 10.13039/501100001603-Sustainable Energy Authority of Ir Number: 91/RDD/578) Authors Authors Aziz Saif, Shafi K. Khadem, Michael Conlon, and Brian Norton Authors Authors Aziz Saif, Shafi K. Khadem, Michael Conlon, and Brian Norton Impact of Distributed Energy Resources in Smart Homes and Community-Based Electricity Market Khadem , Senior Member, IEEE, Michael F. Conlon, Member, IEEE, and Brian Norton Aziz Saif , ShaﬁK. Khadem , Senior Member, IEEE, Michael F. Conlon, Member, IEEE energy-positive smart homes into energy-active smart homes, thus building a smart community. This transformation will expe- dite if the residential customers are provided with a better choice of supply, access to reliable energy prices, possibility to produce and sell their own electricity with increased transparency and better regulation for more involvement in the energy system and respond to the price signals [1]. Smart community-based electricitymarket(SCEM),centredaroundalocalsmartcommu- nity, is an emerging and consumer-centric market approach that empowers consumers with smart homes to become more active through participation in the trading of green electricity among smart homes within the community or beyond. As Europe is rolling out smart electricity meters at a promising pace [2] along with widespread deployment of DERs and energy management systems [3], the abovementioned trend is becoming more emi- nent in future. Abstract—The transformation of passive to energy-active con- sumers in smart homes has been enabled by the proliferation of dis- tributed energy resources (DERs) and demand-side management technologies. Building a smart community-based electricity market (SCEM) centred around a local energy community has the potential to expedite this transformation by tapping the ﬂexibility associated with peer-to-peer energy transactions inside the community. The article presents a systematic approach to quantifying the beneﬁts of smart homes, starting from the energy-passive to energy-active homes under SCEM with intermediate stages identifying smart homes with DERs. The investigation also includes the impact of seasonal variations with contrasting characteristics. Smart homes with solar PV and energy storage under SCEM achieve maximum savings of 50% and 36.6% for the summer and winter months, respectively, and SCEM boosts consumption of localized green energy by a further 31% in the summer month. ES leverages the smart homes gain signiﬁcantly through self-consumption and energy arbitrage. However, the operation of ES under SCEM in the winter month reduces the network’s voltage stability. The article is conducted based on real-life measurements from an energy com- munity in Ireland. Recommendations are made further to boost the transition of smart homes toward the decarburization of smart grid networks. Currently, residential households only engage in the retail electricity market (REM), where consumers have long-term con- tractswithelectricityretailers.ThebusinessmodelofREMisde- signed for traditional energy-passive residential households [4]. The energy transition is motivating to maximize self-sufﬁciency and minimize energy expenditure. 0093-9994 © 2022 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See https://www.ieee.org/publications/rights/index.html for more information. Authors Authors This article is available at ARROW@TU Dublin: https://arrow.tudublin.ie/engscheleart2/343 This article is available at ARROW@TU Dublin: https://arrow.tudublin.ie/engscheleart2/343 59 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 Impact of Distributed Energy Resources in Smart Homes and Community-Based Electricity Market The business model, which facilitates homes with DERs, involves energy retailers buying surplus electricity through support schemes, e.g., feed-in-tariff or net metering [5]. These support schemes have been suc- cessful in the rapid integration of DERs. Nevertheless, support schemes do not have any connection to the market price. As a result, it risks being market inefﬁcient and burdened with a cost that is socialized across end users’ electricity bills. Over the last decade, the remuneration under such support schemes has been drastically reduced or terminated in most countries worldwide [6]. As subsidy-based support schemes are seeing a limited future, the subsequent progression of the energy-passive households towards energy-active smart homes with the inclu- sion of technologies, such as home energy management system (HEMS), energy storage (ES) system etc., enabling demand- side management (DSM). This reduces smart homes’ electricity bills by maximizing the self-consumption of locally generated electricity. Smart homes with DSM capabilities (such as peak shaving, shifting etc.) still operate under retail pricing structures. Though the feed-in of surplus energy is reduced with DSM in place, it still introduces cost recovery problems and cross- subsidization among smart passive homes [7], [8]. This leads to the beneﬁt of smart homes with DSM and DERs, depending on various factors, including self-consumption policy, retail tariff design and cost-recovery design of distribution networks [9]. Index Terms—Distributed energy resources (DERs), distribution grid, energy community, local market, peer-to-peer (P2P), smart homes, transactive energy. Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 gy y [ ] The different types of DER assets, inﬂuencing the beneﬁts of smart homes from a technoeconomic perspective are inves- tigated in [11], [12], [13], and [14]. The retail pricing scheme leverages the smart homes’ beneﬁts and [15] provides a com- parative study of ﬁve tariff structures for four combinations of DER assets. Rahimpour et al. [16], Keerthisinghe et al. [17], and Luna et al. [18] worked with different solution techniques for DER scheduling, e.g., mixed-integer linear programming, dynamic programming, and particle swarm optimization. Re- cently, several research works have been conducted on the community-basedelectricitymarketwithanemphasisonarange of aspects of the market design: market clearing mechanism, bidding strategy, and interaction with the wholesale market. Different market-clearing mechanisms having centralized and decentralized approaches and present the impact of such clearing mechanisms on the community along with other metrics, such as scalability andconvergence of the local market are worked in [19], [20], [21], [22], and [23]. Another important aspect of the community-based market is the strategic and non-strategic bidding of the market players which have been studied in [24], [25], [26], and [27]. The stochastic nature of DERs, being one of the key features of DERs, has been incorporated into the study of community-based electricity markets using two different broad approaches, e.g., robust optimization [28] and stochastic programming [29], [30]. The data-driven approach is gaining attention in the scenario generation of stochastic programming [30]. A segment of literature on the community-based electricity market is often oblivious to the electricity network hosting the community. However, the network constraints must be respected with a certain degree of freedom. Electricity network constraints have been incorporated in this market formulation implement- ing a range of techniques, ac optimal power ﬂow [31], [32], linearized dc optimal power ﬂow [33], [34], network loss [35], [36], constraint-based sensitivity factors [37] and decoupled approach [38], [39], [40], [41] and studied the impact of network constraints on market outcome. 2) A systematic approach is then presented to analyze the beneﬁts of smart homes, starting from the initial stage of energy-passive homes towards energy-active homes under SCEM with intermediate steps identifying smart homes with DER assets (mainly PV and ES). 3) A comparative study on different stages has been con- ducted on energy-passive and active smart homes for a short-term operational timeframe extending from hours to a month. It provides insights into different constituents working under the SCEM from an operational horizon. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 60 DER assets. Several studies have investigated case studies of SCEM for different categories of DER assets. However, those studiespresenttheﬁndingfromacommunityperspective.Onthe other hand, the studies under DER and DSM integration on resi- dential households have paid attention from a technoeconomical viewpointwithalimitedfocusontheoperationalperspectiveand their collective impact on the distribution network. Therefore, an integrated study is required to build synergy among different elements contributing to the beneﬁts of energy-active smart homes under SCEM and the network performance. Detailed examination of these synergies is of paramount importance for fostering a smart community from conception to realization, promoting the proliferation of energy-active smart homes. The novel contribution of the study are enumerated as follows. SCEM is an advanced approach to extend the periphery of self-consumptiontothecommunityscalewheresmarthomesen- gageinenergytradinginsideacommunity-basedelectricitymar- ket, minimizing the supply from REM. This results in economic beneﬁtsforsmarthomesaslocal,peer-to-peer(P2P)transactions inside SCEM offer better pricing for buyers and sellers than energy retailers. This bottom-up, community-centred approach of SCEM provides market power to residential customers and facilitates rapid uptake of DERs in residential households. Apart from empowering residential customers, SCEM offers a coordi- nated, granular, market-based mechanism for smart community promoting local balancing of generation and consumption close to real-time. This introduces a decline in renewable curtailment, less usage of transmission networks, and other positive notions toward a decarbonized energy system [10]. SCEM is an advanced approach to extend the periphery of self-consumptiontothecommunityscalewheresmarthomesen- gageinenergytradinginsideacommunity-basedelectricitymar- ket, minimizing the supply from REM. This results in economic beneﬁtsforsmarthomesaslocal,peer-to-peer(P2P)transactions inside SCEM offer better pricing for buyers and sellers than energy retailers. This bottom-up, community-centred approach of SCEM provides market power to residential customers and facilitates rapid uptake of DERs in residential households. Apart from empowering residential customers, SCEM offers a coordi- nated, granular, market-based mechanism for smart community promoting local balancing of generation and consumption close to real-time. This introduces a decline in renewable curtailment, less usage of transmission networks, and other positive notions toward a decarbonized energy system [10]. 1) The article presents the SCEM as a simpliﬁed and de- terministic linear programming optimization model, in- cluding realistic ES constraints and P2P transactions. The performance quality of this holistic approach is preserved by implementing the market and network model in a cascaded and decoupled fashion. This accommodates a high volume of data for seasonal study with signiﬁcantly reduced computational time. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 4) Power-ﬂow-based quantitative assessment has also been carried out on a realistic low-voltage distribution network (LVDN) hosting the smart community. 5) The impact of the time-of-use (ToU) tariff on the operation of ES is investigated, and thus, its impact on the HEMS and SCEM operations and the performance of LVDN is examined. 6) The seasonal variations (summer and winter) with maxi- mum load demand and clean energy generation conditions are analyzed further to understand the possible extreme impacts of this energy transition at the community level. The rest of the article is organized as follows: the business model of SCEM in Section II. Section III presents the modeling approach of HEMS and SCEM along with LVDN details. Case studies and descriptions of the scenarios have been presented in Section IV. Section V discusses and analyzes the results. Finally, Section V presents the conclusion of this article and provides future research directions. The research work related to the energy-passive homes’ transformation to energy-active ones and the community-based electricity market found in the previously published articles; the authors observed a lack of comprehensive study which evaluates the beneﬁts of the residential households under SCEM with a comparison of different transitional stages of an energy-passive home. The stages are identiﬁed as the gradual incorporation of I. INTRODUCTION T T OWARD the achievement of decarbonised European smart grid network by 2050, the European Commission (EC) has given high importance to its “Clean Energy for all Euro- peans Package (CEP)” by empowering individuals and groups of consumers to participate in this energy transition. Such en- ergy transition demands residential households to transform the Manuscript received 9 February 2022; revised 20 June 2022; accepted 12 August 2022. Date of publication 30 August 2022; date of current version 19 January 2023. This work was supported by the BEYOND project, funded by joint programming initiative ERA-Net Smart Energy Systems, co-funded by the European Union’s Horizon 2020 research and innovation programme under Grant 775970, and in part by Sustainable Energy Authority of Ireland under Grant 91/RDD/578. Paper 2022-SBSC-0071.R1, approved for publication in the IEEE Transactions on Industry Applications by the Smart Buildings for Smart Cities of the IEEE Industry Applications Society. (Corresponding author: Aziz Saif.) Aziz Saif and Brian Norton are with the International Energy Research Centre, TyndallNationalInstitute,T12R5CPCork,Ireland,andalsowiththeTechnolog- ical University Dublin, D07 EWV4 Dublin, Ireland (e-mail: aziz.saif@ierc.ie; brian.norton@ierc.ie). ShaﬁK. Khadem is with the International Energy Research Centre, Tyndall National Institute, T12 R5CP Cork, Ireland (e-mail: shaﬁ.khadem@ierc.ie). Michael F. Conlon is with the Technological University Dublin, D07 EWV4 Dublin, Ireland (e-mail: michael.conlon@tudublin.ie). ShaﬁK. Khadem is with the International Energy Research Centre, Tyndall National Institute, T12 R5CP Cork, Ireland (e-mail: shaﬁ.khadem@ierc.ie). Michael F. Conlon is with the Technological University Dublin, D07 EWV4 Dublin, Ireland (e-mail: michael.conlon@tudublin.ie). Color versions of one or more ﬁgures in this article are available at https://doi.org/10.1109/TIA.2022.3202756. Digital Object Identiﬁer 10.1109/TIA.2022.3202756 ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. TABLE I BUSINESS MODEL FRAMEWORK require to ensure that value propositions for smart homes, utility suppliers, network operators, and other relevant stakeholders are well maintained. The deﬁnition of a business model is still changing to accommodate the rapid innovation undergoing in businesses. A short and concise deﬁnition of a business model in [42] is “A business model describes how you create, distribute and capture value.” The challenges of introducing an innovative business model in the energy system lie in effectively capturing value propositions for different actors and deﬁning those complex values in the current environment. Brown et al. [43] has investigated the busi- ness model for prosumers in the U.K. and classiﬁed the business model into seven archetypes. The authors have taken the insights of those models in the article to develop a business model for smart homes under SCEM with different DER portfolios. p The focal point of this article is on smart homes with residen- tial electricity end-users. The SCEM introduces smart homes with the possibility to conduct P2P transactions within the local energy community sphere. It is motivated to boost the com- munity’s collective self-consumption, reducing dependency on REM electricity purchases. The presence of REM is necessary for the business model to ensure the security of supply. It is obvious that the smart homes under SCEM will not have collective energy self-sufﬁciency for each trading period on the operational horizon. Therefore, it requires a provision to transact deﬁcit/surplus electricity with the central electricity market as it has not been utilized in the P2P transaction in SCEM. The role of the balance responsible party also needs to be addressed. The above-mentioned reasons persuade the presence of electricity retailers in the business case, with REM being the point of connection for SCEM to the central electricity market. The other key actors are the SCEM operator and distribution system operator (DSO). The role of the SCEM operator involves managing the P2P transactions among the market participants to reachthegoaloftheSCEM.MarketparticipantsintheSCEMare the electricity customers: producers, prosumers and consumers. DSO ensures the P2P transactions in the SCEM operation adhere to the network’s technical constraints. As deﬁned by [44], four basic components constitute the business model: value proposition, customer interface, supply chainandﬁnancialmodel.TableIgivesonthecomponentsunder the framework of the business model proposed in the article, especially from the perspective of smart homes and the SCEM operator. SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET 61 Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE II. BUSINESS MODEL The development of innovative business models is of paramount importance to rolling out smart homes with DERs. Future smart grid networks will possess complex architecture with the presence of stochastic behavior of DERs and the en- trance of new actors in the energy transition. Business models ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 g P2P transaction constraint P2P transaction constraint p q̸=p P P2P buy q→p,t = μloss p q̸=p P P2P sell p→q,t . (4) (4) Equation (4) ensures the total electricity purchased through P2P transactions should be equal to electricity sold in P2P transactions at each trading period t ∈T. SCEM mode (8) CEM mode min P Im p,t , P Ex p,t t p λIm t P Im p,t − p λEx t P Ex p,t ΔT where, ηch p and ηdis p are the charging and discharging efﬁciency of the ES, respectively. (2) (2) III. MODELING APPROACH The modeling approach presented in the article is a two-stage, cascaded approach where the model of the study is comprised of two models: the HEMS/SCEM model and the LVDN model. The former model schedules the smart homes’ ﬂexible DER assets to meet the scenarios’ deﬁned objective (see Section IV- C). HEMS/SCEM model has two scheduling modes: HEMS and SCEM, to capture the business model of the study. The HEMS mode only schedules ﬂexible DER assets of individual smart homes separately without providing P2P transactions. In contrast, the SCEM mode schedules the ﬂexible DER assets based on the objective of the market having provision of P2P transactions among smart homes. This article considers that only one of the modes under HEMS/SCEM is in operation, and residential ES is the only ﬂexible DER asset modeled in the HEMS/SCEM. The second model features the network topology and characteristics of the network assets describing the distribution test feeder hosting the smart homes under SCEM. It synthesizesitsinputdatasetfromtheoutputoftheHEMS/SCEM model and conducts network performance analysis based on the dispatch of the DER assets under HEMS/SCEM. The two- stage, cascaded modeling approach enables the extraction of outcomes from the two models separately, namely dispatch outcome from network-unrestrained HEMS/SCEM model and HEMS mode p, p, Equation (7) presents the constraint on state-of-charge, Ep,t of Equation (7) presents the constraint on state of charge, Ep,t of ES units with upper, Ep and lower-level, Ep threshold. Lastly, the state-of-charge dynamics of ES is expressed by the following constraint (1) s.t. energy balance constraint DER operational constraint Ep,t = Ep,t−1 + ηch p P ch p,tΔT −P dis p,t 1 ηdis p ΔT (8) SCEM mode A. HEMS/SCEM Model The HEMS/SCEM model is a linear multi-period optimiza- tion model that has been formulated for a set of smart homes, P = {1, 2, . . . .., Np} across a market horizon with a trading period denoted by t ∈T having duration, ΔT. Both modes of the HEMS/SCEM model are formulated to minimize the procurement cost of electricity and to maximize the revenue from exporting energy to the electricity retailer. The objective function of the HEMS model (1) is centred around each smart home separately and individually, whereas the SCEM model (2) operates for the entire smart community collectively with the provision of P2P transactions DER operational constraints This article considers only residential ES as a ﬂexible DER asset and thus, requires to be represented in the optimisation modelling. The operational constraints of the residential ES can be expressed as P ch p,t ≤P ch,max p (5) P dis p,t ≤P dis,max p (6) Ep ≤Ep,t ≤Ep. (7) P ch p,t ≤P ch,max p (5) P dis p,t ≤P dis,max p (6) Ep ≤Ep,t ≤Ep. (7) (5) (6) Ep ≤Ep,t ≤Ep. (7) (7) Equations (5) and (6) enforce upper limit constraints to charg- ing power, P ch p,t and discharging power, P dis p,t of residential ES. Equation (7) presents the constraint on state-of-charge, Ep,t of ES units with upper, Ep and lower-level, Ep threshold. Lastly, the state-of-charge dynamics of ES is expressed by the following constraint Equations (5) and (6) enforce upper limit constraints to charg- ing power, P ch p,t and discharging power, P dis p,t of residential ES. HEMS mode min P Im p,t , P Ex p,t t λIm t P Im p,t −λEx t P Ex p,t ΔT s.t. energy balance constraint DER operational constraint HEMS mode min P Im p,t , P Ex p,t t λIm t P Im p,t −λEx t P Ex p,t ΔT s.t. energy balance constraint DER operational constraint ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 62 network performance outcome from LVDN model, which is useful for the study. Two different software platforms have been used for the separate models. HEMS/SCEM model is developed in the MATLAB environment using open-source op- timization modeling language, YALMIP and MOSEK being the optimization solver. LVDN model is developed in open source distribution system simulator (OpenDSS), which can conduct a time-series simulation of the complex, unbalanced, multi-phase distribution network. Further details on the modeling approach, HEMS/SCEM model and LVDN model can be found in the authors’ previous work [45]. This article HEMS/SCEM model follows a similar methodological approach as presented in [41] and [19]. However, the contribution of the article is not in terms of methodology, rather the quantitative, comprehensive analysis of beneﬁts brought by SCEM to residential smart homes compared with other transitional stages. network performance outcome from LVDN model, which is useful for the study. Two different software platforms have been used for the separate models. HEMS/SCEM model is developed in the MATLAB environment using open-source op- timization modeling language, YALMIP and MOSEK being the optimization solver. LVDN model is developed in open source distribution system simulator (OpenDSS), which can conduct a time-series simulation of the complex, unbalanced, multi-phase distribution network. Further details on the modeling approach, HEMS/SCEM model and LVDN model can be found in the authors’ previous work [45]. This article HEMS/SCEM model follows a similar methodological approach as presented in [41] and [19]. However, the contribution of the article is not in terms of methodology, rather the quantitative, comprehensive analysis of beneﬁts brought by SCEM to residential smart homes compared with other transitional stages. Equation (3) refers to the energy balance constraint for each smart home operating under HEMS or SCEM mode. Here, superscript ch and dis are used to represent charging and dis- charging of the residential ES, whereas, P dem p,t and P gen p,t indicate the load and self-generated electricity of smart homes, p ∈P during the trading period, t ∈T. The terms associated with P P2P buy q→p,t and P P2P sell p→q,t in Equation (3) only applies to the SCEM mode but not to the HEMS mode. P P2P buy q→p,t represents the electricity procured by smart home p from peer q in the SCEM and P P2P sell p→q,t represents vice-versa. μloss is a co-efﬁcient denoting network loss factor afﬁliated with P2P transactions. licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Res Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. B. Test Network The IEEE European low voltage test feeder has been used as an LVDN test network for the study. It has the radial topol- ogy, a typical European low voltage distribution network. The study uses a modiﬁed version of the test feeder with a 200 kVA, 11 KV/0.416 kV transformer to align the parameters with the Irish network. The test feeder consists of 906 buses and 55 customer connection points for single-phase residential customers. All of the 55 smart homes are located at different connection points. The smart homes are modeled as constant PQ loads. The power ﬂow simulation uses the connection point at MV/LV substation as a slack bus. In alignment with the temporal resolution of the HEMS/SCEM model, the power ﬂow has also beenconductedonhourlyresolution. Fig.1showstheLVDNtest feeder’s schematic diagram with the smart homes’ placement. C. Scenarios’ Descriptions The scenarios have been developed considering the notion of the article to show the beneﬁts of DER assets and SCEM on smart homes. Therefore, the proposed scenarios consider the gradual integration of DER assets and P2P trading provision to a passive home. Five scenarios have been evaluated for the study described in given in Table II. Under each scenario, every smart home’s DER asset portfolio is considered identical. Consequently, the smart homes’ con- sumption and generation proﬁles have been assumed consistent across all the cases. hourly time series data have been implemented for the selected months. A. Exchange With the Energy Retailer A. Exchange With the Energy Retailer B. LVDN Model LVDN model is capable to conduct detailed network studies, e.g., power ﬂow solution, fault calculation, harmonic analysis. It takes power injection proﬁles of each smart home in the LVDN, which is calculated by (9) and runs the power ﬂow. Different power ﬂow solution algorithms exist to solve power ﬂow for distribution networks [46], however, this article uses the default built-in power ﬂow solution algorithm in OpenDSS [47] where λIm t is the ToU retail electricity tariff, λEx t is the feed-in tariff, P Im p,t represents the amount of electricity procured from the retailer and P Ex p,t represents electricity sold to the retailer. The ﬁrst term of the objective function represents the cost function related to procuring electricity from REM under a ToU tariff scheme. The second term refers to the revenue function denoting electricity exported to the grid at a feed-in-tariff rate. The constraints of the HEMS/SCEM model are elaborated as where λIm t is the ToU retail electricity tariff, λEx t is the feed-in tariff, P Im p,t represents the amount of electricity procured from the retailer and P Ex p,t represents electricity sold to the retailer. The ﬁrst term of the objective function represents the cost function related to procuring electricity from REM under a ToU tariff scheme. The second term refers to the revenue function denoting electricity exported to the grid at a feed-in-tariff rate. The constraints of the HEMS/SCEM model are elaborated as P inj p,t = P Im p,t + q̸=p P P2P buy q→p,t −P Ex p,t − q̸=p P P2P sell p→q,t (9) (9) Energy balance constraint Energy balance constraint where P P2P buy q→p,t represents the electricity procured by smart home p from peer q in the SCEM and P P2P sell p→q,t represents vice- versa. These two terms in (9) have only been considered in the SCEM mode but not in the HEMS mode. P Im p,t + q̸=p P P2P buy q→p,t + P dis p,t + P gen p,t = P Ex p,t + μloss q̸=p P P2P sell p→q,t + P ch p,t + P dem p,t . (3) (3) Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET 63 Fig. 1. Schematic diagram of test network identifying location of smart homes. IV. SYSTEM SETUP Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE A. Case Study The case study is presented for a real neighbourhood located in the Dingle area in Ireland [48]. The smart home’s time series measured smart meter data has been used. DER assets considered in the study are roof-top solar PV and residential ES. The capacities of the roof-top PV of the smart homes are ranged between 2.0–2.2 kWp. The lithium-ion battery is considered a residential ES with a capacity of 10kWh/3.3kW peak. Data of 55 smart homes are used for two different months: January (winter) and June (summer) 2020, to understand the impact of seasonal variation in the best and worst conditions. The community self- sufﬁciency of the neighbourhood, deﬁned as a percentage ratio of aggregated solar PV generation and aggregated consumption for the neighbourhood, for above mentioned months is 12.6% and 62.5%, respectively. As described in Section III-A, the HEMS/SCEM model takes day-night electricity retail prices as an exogenous price signal from the existent static ToU tariff schemes from REM in Ireland. The day and night rates are 20.07 and 9.91 c€/kWhr, respectively, for 2020. The feed-in tariff has a ﬁxed rate of 9.0 c€/kWhr. The DER scheduling and P2P trading in HEMS/SCEM model is considered to operate in hourly resolution. Fig. 1. Schematic diagram of test network identifying location of smart homes. V. SIMULATION RESULTS Extensive simulation studies have been performed for all the scenarios, along with network performance analysis. In addition, To understand the extreme impact of the different scenarios on smart homes DER scheduling and subsequently their interaction echnological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IE IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 64 Fig. 2. Net supply proﬁles for all scenarios for the day with maximum demand in winter day. Fig. 3. Net supply proﬁles for all scenarios for the day with maximum PV generation in summer day. Fig. 3. Net supply proﬁles for all scenarios for the day with maximum PV generation in summer day. Fig. 2. Net supply proﬁles for all scenarios for the day with maximum demand in winter day. Fig. 3. Net supply proﬁles for all scenarios for the day with maximum PV generation in summer day. with REM, Fig. 2 illustrates the average net supply of electricity for all the scenarios for January 16, 2020, the day with a maxi- mum aggregated demand in the representative winter month. Net supply is calculated by subtracting the smart home’s sold energy fromtheprocuredenergyforeachmarkettimeinterval.SincePV generation is low compared to the demand in the winter month, scenarios with no ES (Base, Base+PV and SCEM-PV) do not show any signiﬁcant change in net energy supply. However, the scenarios with ES (HEMS-PV+ES and SCEM-PV+ES) have shifted in the net supply of energy across the day where most of the energy is now consumed in the low tariff hours (from midnight to hour 09:00). The base scenario (average demand proﬁle of smart homes) has two peaks, with the morning peak occurring at 10:00 and the evening peak at 19:00. With the introduction of ES (HEMS-PV+ES and SCEM-PV+ES), smart homes have the ﬂexibility of shifting their consumption; hence, the peak of the day occurs at 09:00, the last hour of the low tariff time band. It can be observed that compared to HEMS-PV+ES, SCEM-PV+ES has further increased net supplied energy in the low tariff hours timeband and reduced consumption in the high tariff hours. This is due to the energy exchange/P2P transaction provision opening up the energy arbitrage with neighbouring smart homes. V. SIMULATION RESULTS Since all the smart homes have PV facilities with nearly the same capacity and due to the high PV generation during the mid-day, the homes achieve self-sufﬁciency. The possible P2P trading/energy exchange options are also nearly zero. Hence, the SCEM-PV scenario does not show any reduction in the feed-in energy. With the introduction of ES (HEMS-PV+ES and SCEM- PV+ES), the stored PV energy has also covered the net REM supplied energy required in Base+PV and SCEM-PV scenarios. Moreover, the SCEM-PV+ES scenario has demonstrated the highest performance with no energy exchange with the retailer after the low tariff time band as smart homes with energy deﬁcit at certain hours meets their demand from other peers with excess energy through the P2P trading. Though Figs. 2 and 3 show the average value of the net supply, it is also important to observe the extreme net supply conditions of smart homes. Hence, the net supply proﬁle of the individual smart homes with the highest and lowest values in comparison to averaged proﬁle for both winter and summer days are presented in Fig. 4. This is presented for the most prospective scenario for the smart community with greater ﬂexibility, SCEM-PV+ES. For the winter day, the smart homes with the highest and lowest- demand are considered, whereas, for the summer day, smart homes with maximum and minimum PV generation are shown. In continuation, Fig. 3 depicts the net supply of June 21, 2020, the day with a maximum aggregated PV generation for the summer month. High PV generation in the summer month diminishes the need for net supply energy from REM during the mid-day for all the scenarios with solar PV. It can be seen that the Base+PV and SCEM-PV scenarios have high feed-in as smart homes do not have the ﬂexibility to store excess energy. ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET 65 Fig. 6. Averaged (a) charging and (b) discharging proﬁle of the smart homes for the summer day. Fig. 4. Net supply proﬁles of two individual smart homes along with averaged proﬁle for the community (SCEM-PV+ES scenario). Fig. 4. Net supply proﬁles of two individual smart homes along with averaged proﬁle for the community (SCEM-PV+ES scenario). Fig. 5. Averaged (a) charging and (b) discharging proﬁle of the smart homes for the winter day. Fig. 4. Net supply proﬁles of two individual smart homes along with averaged proﬁle for the community (SCEM-PV+ES scenario). Fig. 4. Net supply proﬁles of two individual smart homes along with averaged proﬁle for the community (SCEM-PV+ES scenario). Fig. 6. Averaged (a) charging and (b) discharging proﬁle of the smart homes for the summer day. during the high tariff hours. This observation is also coherent with Fig. 2, where a similar pattern is observed over low tariff hours, but the net supply is reduced in high tariff hours. Though energy arbitrage boosted by P2P trading brings beneﬁts to the smart homes under SCEM, it results in higher charging peaks in different time horizons, which is detrimental to the distribution network. It needs to be noted that the ﬁgures shown in Fig. 5 is average proﬁle and therefore, charging proﬁles of a number of smart homes will be higher than that which deteriorates the voltage of the network nodes connecting these smart homes. The averaged discharging proﬁle shown in Fig. 5 indicates the discharging of ES is taking place to cover the demand of smart homes at high tariff hours or excess PV generation (hours 12:00- 14:00) to reduce the bills. In the summer month, the charging action is primarily from excess PV energy to cover the demand avoiding procurement from the REM. It can be seen in Fig. 6 that signiﬁcant charging action is taking place at mid-day which is later discharged to meet the demand after hours 19:00. In contrast to the HEMS-PV+ES scenario, SCEM-PV+ES ob- serves lower ES charging at low tariff hours and higher ES charging during mid-day. As the HEMS-PV+ES scenario does not serve P2P trading, smart homes, with their ES being charged to full capacity from excess PV energy, exports the surplus PV generated electricity to the retailer (see Fig. 3). B. Operation of ES As observed in the previous Section V-A, the ﬂexibility introduced by the ES plays a crucial role in the operation of smart homes to achieve the objective set by the scenarios. Fig. 5 uthorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplo E. Network Impact As described in Section V-B, the inclusion of DER assets, especially ES inﬂuenced by the SCEM has signiﬁcantly altered the smart homes’ daily proﬁle. The LVDN hosting the smart homes is usually designed to be in a “ﬁt-and-forget” approach and is not generally equipped with monitoring and control devices. Therefore, it is imperative to understand how the change of proﬁles due to the integration of DERs impacts the LVDN and the homes connected to it consecutively. This article only presents the voltage proﬁle at nodes connecting smart homes as it directly affects the network stability. Fig. 9 depicts the voltage proﬁle of 10 consecutive days at LVDN nodes for scenarios- SCEM-PV+ES in winter and SCEM-PV in summer months. SCEM-PV+ES scenario on winter month has certain nodes experiencingunder-voltagesituationsduetothehighchargingof ES. On the other hand, SCEM-PV shows overvoltage conditions at nodes resulting from high surplus PV injection. However, incorporatingEShaseliminatedtheovervoltageproblem,shown in the SCEM-PV+ES scenario. C. P2P Transactions P2P transactions can be considered as an indication to assess the factors impacting SCEM operation. Different DER assets under two scenarios, SCEM-PV and SCEM-PV+ES, give in- sights into the role of ES on SCEM operation. In Fig. 7, for the winter month, the contrast between the two scenarios implies that the P2P transactions are mostly contributed from the energy arbitrage. The SCEM-PV scenario has P2P transactions only for a few hours at mid-day when smart homes with surplus PV trades it with their peers. The smart homes, equipped with ES, store their surplus PV for later use rather than selling it to other peers under SCEM or exporting it to the REM (Fig. 3, HEMS-PV+ES scenario). Besides, ES opens up the prospect of energy arbitrage and thus, contributes to signiﬁcant P2P transactions and is initiated just after the low tariff time band (after hour 09:00). This demonstrates the consolidated impact of ES and ToU tariff on P2P transactions, especially at times when community self-sufﬁciency is low. On the other hand, the P2P transaction is driven by surplus PV generation with insigniﬁcant energy arbitrage for days with high community self-sufﬁciency, as seen in Fig. 8 for a summer day. Therefore, both scenarios are closely alike. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 66 Fig. 7. Averaged P2P transaction proﬁle of the smart homes for the winter day. TABLE III MONTHLY AVERAGED RESULTS OF A SMART HOME UNDER DIFFERENT SCENARIOS Fig. 7. Averaged P2P transaction proﬁle of the smart homes for the winter day. TABLE III MONTHLY AVERAGED RESULTS OF A SMART HOME UNDER DIFFERENT SCENARIOS Fig. 7. Averaged P2P transaction proﬁle of the smart homes for the winter day. Fig. 8. Averaged P2P transaction proﬁle of the smart homes for the summer day. of the selected parameters for the smart homes under the ﬁve scenarios given in Table II. It can be seen that, compared to the Base+PV (b) scenario, the SCEM-PV+ES (e) scenario has the maximum reduction in net supply cost with 36.6% and 50% (marked in blue) for winter and summer month respectively. With the introduction of ES in the smart homes’ portfolio {HEMS-PV+ES (c) and SCEM-PV+ES (e)}, the localised con- sumption of locally generated electricity, usually green energy in nature, is maximised as indicated by the reduction of REM exported energy for smart homes in the summer month (marked in green) compared with scenario (b) (91 kWhr) with homes having only PV, but no ES. The impact of P2P energy exchange provision on smart homes’ consumption of green energy is visible in the summer month (month with higher community self-sufﬁciency), which exhibits a reduction of average REM exported energy by 27-1 = 26 kWhr in the SCEM-PV+ES (e) scenario compared with the HEMS-PV+ES (c) scenario. The P2P transaction is also boosted by the presence of ES (SCEM- PV+ES scenario) driven by energy arbitrage in the winter month and surplus PV generation in the summer month, as evident in the SCEM-PV+ES scenarios (e) (marked with orange) while comparing with SCEM scenario without ES, SCEM-PV (d). Fig. 8. Averaged P2P transaction proﬁle of the smart homes for the summer day. SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET On the contrary, the P2P trading arrangement in the SCEM-PV+ES scenario allows smart homes to trade PV generated electricity with their peers in need of energy resulting in higher community self-sufﬁciency. Hence, electricity feed-in to the energy retailer diminishes at high tariff hours for SCEM-PV+ES scenario in Fig. 3 and similarly, discharging is not taken place at mid-day in Fig. 6. Fig. 5. Averaged (a) charging and (b) discharging proﬁle of the smart homes for the winter day. shows the charging and discharging proﬁle of the ES averaged over all the smart homes on a winter day. It can be seen that the charging is occurred primarily at a low tariff timeband, up to hour 10:00. SCEM-PV+ES scenario observes more charging at low tariff hours compared with the HEMS-PV+ES scenario. This is due to the fact the winter day has low PV generation and high demand. As a result, smart homes involve in charging ES facilities procuring electricity from REM at low tariff hours to meet the demand for the rest of the day. Since SCEM provides energy exchange/trading possibility, this creates an opportu- nity for smart homes to engage in energy arbitrage, procuring electricity from REM at low tariff hours and selling it under SCEM at a lower price to other smart homes with energy deﬁcits ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. VI. CONCLUSION The transition of energy-passive homes to energy-active ones through incorporating DERs and demand response capabilities can further be augmented with the introduction of SCEM. The research contributes to the discussion of the transitional stages of residential households by investigating from a short-term op- erational perspective different types of smart homes categorised based on DER assets, ﬂexibilities and participation in SCEM. The result shows that the smart homes with PV and ES under SCEM achieve the highest beneﬁts in extreme conditions for both typical summer and winter. The presence of ES facilities in homes’ premises plays a crucial role as the ability to store allows the smart homes to maximise the consumption of locally generated, green energy and energy arbitrage. Fig. 9. Voltage proﬁles of the nodes connecting smart homes for scenarios. (a) SCEM-PV+ES on winter. (b) SCEM-PV and (c) SCEM-PV+ES on summer month. SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET 67 Fig. 9. Voltage proﬁles of the nodes connecting smart homes for scenario (a) SCEM-PV+ES on winter. (b) SCEM-PV and (c) SCEM-PV+ES on summe month. F. Recommendations R lt f thi ti l i f l d The study has found that the SCEM operation is intertwined with the grid tariff design, as an exogenous price signal to the SCEM, especially in the presence of ES. Grid tariff is determined by the regulatory authority and is required to adhere few design principles, e.g., cost-reﬂectivity, nondistortionary, cost recovery, nondiscriminatory, etc [49]. Therefore, regulatory authority requires to carefully design the future grid tariff that fosters the SCEM and hence, energy activism of customers in the community. The P2P transaction in SCEM only boosts when the pricing of P2P energy exchange is capped by the ToU and feed-in tariffs, as assumed in the article. Hence, in designing future grid tariffs, the locational dimension of grid tariff may appear relevant consid- ering the localised, P2P energy exchange nature of the SCEM. Energy arbitrage among customers in the energy community is dominant in the P2P transaction during winter, in the pres- ence of residential ES and static ToU tariff. This endangers the retailer’s revenue under the existing business model as a number of households are buying stored energy (stored from retailer-supplied energy at low tariff hours) from their peers through energy arbitrage rather than buying directly from the retailer. This opens up the necessity of investigating adaptation of a localized, community-based market (e.g., SCEM) to the REM and the need for changes in the retailer’s business model. The network performance study shows SCEM has resulted in poor voltage performance. Though the analysis has been performed for the extreme case where the entire community under the same substation participates in the SCEM with the same DER portfolio, the detailed network hosting capacity of the SCEM can provide insights into the penetration level of SCEM under a single substation for secured network operation. Till now, the R&D projects on SCEM are taking place in the regulatory sandbox due to the absence of clear direction on SCEM in regulation. The abovementioned concerns derived from the results are relevant to be addressed for the development of existing and/or emerging regulations, grid codes, standards, legal framework, business model and central electricity mar- ket arrangement, which facilitates real-life roll-out of energy community-centred local market. D. Smart Home’s Beneﬁt The results have been presented in previous sections as av- eraged proﬁles of smart homes for representative winter and summer days. Table III summarises the monthly average values o: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 68 smarthomeswithESonwinterdays.Conversely,thedrivingfac- tor on summer days involves primarily maximal consumption of locally generated electricity. The provision of P2P transactions under the SCEM opens the opportunity of energy arbitrage for smart homes with ES and further boosts the local consumption of locally generated electricity (compared to the HEMS scenario). However, this leads to another issue, heightening of high demand peak further in wintertime resulting in an under-voltage situation in the network. The ﬁndings in the study identify, quantify and synergise the underlying factors constituting the gradual shift of residential households under various scenarios across different seasons while exploring from a short-term operational horizon. [15] M. Avau, N. Govaerts, and E. Delarue, “Impact of distribution tariffs on prosumer demand response,” Energy Policy, vol. 151, pp. 1–13, 2021. [16] Z. Rahimpour et al., “Energy management of buildings with phase change materials based on dynamic programming,” in Proc. IEEE Milan Power Tech, 2019, pp. 1–6. [17] C. Keerthisinghe et al., “PV and demand models for a markov decision process formulation of the home energy management problem,” IEEE Trans. Ind. Electron., vol. 66, no. 2, pp. 1424–1433, Feb. 2019. [18] A. C. Luna, N. L. Diaz, M. Graells, J. C. Vasquez, and J. M. Guer- rero, “Mixed-integer-linear-programming-based energy management sys- tem for hybrid PV-wind-battery microgrids: Modeling, design, and ex- perimental veriﬁcation,” IEEE Trans. Power Electron., vol. 32, no. 4, pp. 2769–2783, Apr. 2017. [19] R. Faia, J. Soares, T. Pinto, F. Lezama, Z. Vale, and J. M. Corchado, “Optimal model for local energy community scheduling considering peer to peer electricity transactions,” IEEE Access, vol. 9, pp. 12420–12430, 2021. Envisaged future work includes the following extensions. Impact of variety of dynamic tariff schemes on smart homes under SCEM. Inclusion of uncertainty associated with gener- ation and consumption. Cyclic degradation of ES is crucial to be acknowledged to quantify the beneﬁts of ES under SCEM properly. Future research will also extend on the study of SCEM operation under different penetration of households and DER capacity. Detailed analysis of network performance, substation congestion, and network unbalance study, after inclusion of certain network constraints in the HEMS/SCEM model, will be carried out to understand the hosting capacity of SCEM in the residential network. [20] J. L. C. Vazquez, T. AlSkaif, Á. M. González-Rueda, and M. REFERENCES [1] A. Nouicer et al., “The EU clean energy package,” Euro. Univ. Inst., Florance, 2020. [26] A. Paudel, K. Chaudhari, C. Long, and H. B. Gooi, “Peer-to-Peer energy trading in a prosumer-based community microgrid: A game-theoretic model,” IEEE Trans. Ind. Electron., vol. 66, no. 8, pp. 6087–6097, Aug. 2019. [2] European Commission, “Benchmarking smart metering deployment in the EU-28: Final report,” Publications Ofﬁce Euro. Union, Luxembourg, Germany, 2020. [3] “ Innovation landscape brief: Market integration of distributed energy resources,” Int. Renew. Energy Agency, 2019. g [27] A. Yu, C. Zhang, and Y. -J. A. Zhang, “Optimal bidding strategy of prosumers in distribution-level energy markets,” IEEE Trans. Power Syst., vol. 35, no. 3, pp. 1695–1706, May 2020. [4] C. Defeuilley, “Retail competition in electricity markets,” Energy Policy, vol. 37, no. 2, pp. 377–386, 2009. [28] J. Yang and C. Su, “Robust optimization of microgrid based on renew- able distributed power generation and load demand uncertainty,” Energy, vol. 223, pp. 1–13, 2021. [5] “Medium-term renewable energy market report 2016,” Int. Energy Agency, Paris, France, 2016. [6] International Energy Agency, “Market report series: Renewables 2019,” IEA, Paris, France, 2019. [29] Y. Li, M. Han, Z. Yang, and G. Li, “Coordinating ﬂexible demand re- sponse and renewable uncertainties for scheduling of community inte- grated energy systems with an electric vehicle charging station: A bi-level approach,” IEEE Trans. Sustain. Energy, vol. 12, no. 4, pp. 2321–2331, Oct. 2021. [7] A. Picciariello et al., “Electricity distribution tariffs and distributed genera- tion: Quantifying cross-subsidies from consumers to prosumers,” Utilities Policy, vol. 37, pp. 23–33, 2015. [8] C. Eid, J. R. Guillén, P. F. Marín, and R. Hakvoort, “The economic effect of electricity net-metering with solar PV: Consequences for network cost recovery, cross subsidies and policy objectives,” Energy Policy, vol. 75, pp. 244–254, 2014. [30] Y. Li et al., “Hierarchical stochastic scheduling of multi-community inte- grated energy systems in uncertain environments via stackelberg game,” Appl. Energy, vol. 308, pp. 1–17, 2022. [31] T. AlSkaif and G. van Leeuwen, “Decentralized optimal power ﬂow in distribution networks using blockchain,” in Proc. Int. Conf. Smart Energy Syst. Technol., Porto, Portugal, 2019, pp. 1–6. [9] GfK Belgium consortium, “Study on “Residential Prosumers in the Euro- pean Energy Union,” Belgium, 2017. [10] E. Mengelkamp et al., “Increasing the efﬁciency of local energy markets through residential demand response,” Energy Inform., vol. 1, no. 1, pp. 1–18, 2018. [32] E. Münsing, J. Mather, and S. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 59, NO. 1, JANUARY/FEBRUARY 2023 Gibescu, “A community-based energy market design using decentralized decision- making under uncertainty,” IEEE Trans. Smart Grid, vol. 12, no. 2, pp. 1782–1793, Feb. 2021. pp [21] T. Morstyn and M. D. McCulloch, “Multiclass energy management for peer-to-peer energy trading driven by prosumer preferences,” IEEE Trans. Power Syst., vol. 34, no. 5, pp. 4005–4014, Sep. 2019. [22] F. Moret and P. Pinson, “Energy collectives: A community and fairness based approach to future electricity markets,” IEEE Trans. Power Syst., vol. 34, no. 5, pp. 3994–4004, Sep. 2019. [23] W. Tushar et al., “Energy storage sharing in smart grid: A modiﬁed auction- based approach,” IEEE Trans. Smart Grid, vol. 7, no. 3, pp. 1462–1475, May 2016. y [24] S. Ø. Ottesen, A. Tomasgard, and S. -E. Fleten, “Multi market bidding strategies for demand side ﬂexibility aggregators in electricity markets,” Energy, vol. 149, pp. 120–134, 2018. [25] T. Pinto, R. Faia, M. A. F. Ghazvini, J. Soares, J. M. Corchado, and Z. Vale, “Decision support for small players negotiations under a transactive energy framework,” IEEE Trans. Power Syst., vol. 34, no. 5, pp. 4015–4023, Sep. 2019. F. Recommendations Results from this article premises for several recommenda- tions that are crucial for harnessing the beneﬁts brought forward by SCEM in the transition of residential customers. Results also demonstrate that the differential tariff scheme (static ToU tariff) contributes signiﬁcantly to the operation of Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE REFERENCES He is a Research Active in smart grid networks with a special focus on grid stability, power quality, microgrids, virtual power plants, local energy systems, and markets. He is the National Lead of the Global Observatory IEA Task on P2P, Community Self-Consumption and Transactive Energy Models. Mr. Khadem is active in IEEE Power and Energy Society, smart grids and smart cities societies and is an Associate Editor for IEEE ACCESS and Journal of Modern Power System and Clean Energy. [38] R. Faia, T. Pinto, Z. Vale, and J. M. Corchado, “A local electricity market model for DSO ﬂexibility trading,” in Proc. 16th Int. Conf. Eur. Energy Market, 2019, pp. 1–5. [39] T. Orlandini, T. Soares, T. Sousa, and P. Pinson, “Coordinating consumer- centric market and grid operation on distribution grid,” in Proc. 16th Int. Conf. Eur. Energy Market, Ljubljana, Slovenia, 2019, pp. 1–6. [40] B. P. Hayes, S. Thakur, and J. G. Breslin, “Co-simulation of electricity distribution networks and peer to peer energy trading platforms,” Int. J. Elect. Power Energy Syst., vol. 115, pp. 1–10, 2020. [41] M. F. Dynge, P. C. del Granado, N. Hashemipour, and M. Korpås, “Impact of local electricity markets and peer-to-peer trading on low-voltage grid operations,” Appl. Energy, vol. 301, pp. 1–14, 2021. [42] A. Osterwalder, “The business model ontology—A proposition in a design science approach,” Ph.D. Thesis, Univ. Lausanne, Vaud, Switzerland, 2004. Michael F. Conlon (Member, IEEE) received the Dip. E.E., B.Sc. degree from Dublin Institute of Tech- nology, Dublin, Ireland, in 1982, and the M.Eng.Sc. and Ph.D. degrees from the University College, Gal- way, Ireland, in 1984 and 1987, respectively, all in electrical engineering. [43] D. Brown, S. Hall, and M. E. Davis, “Prosumers in the post subsidy era: An exploration of new prosumer business models in the UK,” Energy Policy, vol. 135, pp. 1–17, 2019. pp [44] F. Boons and F. Lüdeke-Freund, “Business models for sustainable inno- vation: State-of-the-art and steps towards a research agenda,” J. Cleaner Prod., vol. 45, pp. 9–19, 2013. He was with Monash University, Melbourne, VIC, Australia, and VENCorp, Melbourne. He is cur- rently the Head of the School of Electrical and Elec- tronic Engineering with the Tehnological University Dublin, where he is also the Director of the Electrical Power Research Centre. e limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions appl Authorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE REFERENCES Moura, “Blockchains for decentralized optimization of energy resources in microgrid networks,” in Proc. IEEE Conf. Control Technol. Appl., 2017, pp. 2164–2171. pp [11] A. Amer, K. Shaban, A. Gaouda, and A. Massoud, “Home energy man- agement system embedded with a multi-objective demand response op- timization model to beneﬁt customers and operators,” Energies, vol. 14, no. 2, 2021, Art. no. 257. [33] J. Qin, R. Rajagopal, and P. Varaiya, “Flexible market for smart grid: Coordinated trading of contingent contracts,” IEEE Trans. Control Netw. Syst., vol. 5, no. 4, pp. 1657–1667, Apr. 2018. [34] A. Masood, J. Hu, A. Xin, A. R. Sayed, and G. Yang, “Transactive energy for aggregated electric vehicles to reduce system peak load con- sidering network constraints,” IEEE Access, vol. 8, pp. 31519–31529, 2020. [12] S. Quoilin, K. Kavvadias, A. Mercier, I. Pappone, and A. Zucker, “Quan- tifying self-consumption linked to solar home battery systems: Statistical analysis and economic assessment,” Appl. Energy, vol. 182, pp. 58–67, 2016. [35] S. Lilla, C. Orozco, A. Borghetti, F. Napolitano, and F. Tossani, “Day- ahead scheduling of a local energy community: An alternating direction method of multipliers approach,” IEEE Trans. Power Syst., vol. 35, no. 2, pp. 1132–1142, Mar. 2020. [13] H. T. Dinh et al., “A home energy management system with renewable energy and energy storage utilizing main grid and electricity selling,” IEEE Access, vol. 8, pp. 49436–49450, 2020. [14] M. Youseﬁ, A. Hajizadeh, M. N. Soltani, and B. Hredzak, “Predictive home energy management system with photovoltaic array, heat pump, and plug-in electric vehicle,” IEEE Trans. Ind. Inform., vol. 17, no. 1, pp. 430–440, Jan. 2021. pp [36] M. L. Di Silvestre, P. Gallo, M. G. Ippolito, E. R. Sanseverino, and G. Zizzo, “A technical approach to the energy blockchain in microgrids,” IEEE Trans. Ind. Inform., vol. 14, no. 11, pp. 4792–4803, Nov. 2018. uthorized licensed use limited to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplo ted to: Technological University Dublin. Downloaded on June 06,2023 at 10:45:20 UTC from IEEE Xplore. Restrictions apply. SAIF et al.: IMPACT OF DISTRIBUTED ENERGY RESOURCES IN SMART HOMES AND COMMUNITY-BASED ELECTRICITY MARKET 69 ShaﬁK. Khadem (Senior Member, IEEE) has been a Senior Researcher leading the Intelligent Grid (iG) Research Group at the International Energy Research Centre (IERC), since 2015. REFERENCES He is a Research Active in smart grid networks with a special focus on grid stability, power quality, microgrids, virtual power plants, local energy systems, and markets. He is the National Lead of the Global Observatory IEA Task on P2P, Community Self-Consumption and Transactive Energy Models. Mr. Khadem is active in IEEE Power and Energy Society, smart grids and smart cities societies and is an Associate Editor for IEEE ACCESS and Journal of Modern Power System and Clean Energy. [37] A. Paudel, L. P. M. I. Sampath, J. Yang, and H. B. Gooi, “Peer-to-peer energy trading in smart grid considering power losses and network fees,” IEEE Trans. Smart Grid, vol. 11, no. 6, pp. 4727–4737, Nov. 2020. ShaﬁK. Khadem (Senior Member, IEEE) has been a Senior Researcher leading the Intelligent Grid (iG) Research Group at the International Energy Research Centre (IERC), since 2015. He is a Research Active in smart grid networks with a special focus on grid stability, power quality, microgrids, virtual power plants, local energy systems, and markets. He is the National Lead of the Global Observatory IEA Task on P2P, Community Self-Consumption and Transactive Energy Models. Mr. Khadem is active in IEEE Power and Energy Society, smart grids and smart cities societies and is an Associate Editor for IEEE ACCESS and Journal of Modern Power System and Clean Energy. ShaﬁK. Khadem (Senior Member, IEEE) has been a Senior Researcher leading the Intelligent Grid (iG) Research Group at the International Energy Research Centre (IERC), since 2015. He is a Research Active in smart grid networks with a special focus on grid stability, power quality, microgrids, virtual power plants, local energy systems, and markets. He is the National Lead of the Global Observatory IEA Task on P2P, Community Self-Consumption and Transactive Energy Models. ShaﬁK. Khadem (Senior Member, IEEE) has been a Senior Researcher leading the Intelligent Grid (iG) Research Group at the International Energy Research Centre (IERC), since 2015. He is a Research Active in smart grid networks with a special focus on grid stability, power quality, microgrids, virtual power plants, local energy systems, and markets. He is the National Lead of the Global Observatory IEA Task on P2P, Community Self-Consumption and Transactive Energy Models. ShaﬁK. Khadem (Senior Member, IEEE) has been a Senior Researcher leading the Intelligent Grid (iG) Research Group at the International Energy Research Centre (IERC), since 2015. REFERENCES His current research interests include power systems analysis and control applications, power systems economics, integration of wind energy in power networks, analysis of distribution networks, and quality of supply and reliability assessment. [45] A. Saif, S. K. Khadem, M. Conlon, and B. Norton, “Hosting a community- based local electricity market in a residential network,” IET Energy Syst. Integr., vol. 4, pp. 1–12, 2022. [46] K. P. Schneider et al., “Analytic considerations and design basis for the IEEE distribution test feeders,” IEEE Trans. Power Syst., vol. 33, no. 3, pp. 3181–3188, May 2018. [47] R. C. Dugan and T. E. McDermott, “An open source platform for collab- orating on smart grid research,” in Proc. IEEE Power Energy Soc. Gen. Meeting, 2011, pp. 1–7. [48] ESB Networks, “The dingle project.” Dublin, Ireland, Accessed: Feb. 3, 2022. [Online]. Available: https://www.esbnetworks.ie/who-we-are/ innovation/esb-networks’-dingle-project g p j [49] “Electricity distribution tariffs supporting the energy transition,” Council Eur. Energy Regulators, Brussels, Belgium, 2020. Brian Norton is the Head of Energy Research with Tyndall National Institute, Cork, Ireland, a Research Professor with the University College Cork, Cork, Ireland, and a Professor of solar energy applications with Technological University Dublin, Dublin, Ire- land. Past President, Dublin Institute of Technology. He successfully drove institutional change, led major award-winning campus development and developed innovative learning and teaching; research institutes and business incubators; leading to designation as Technological University Dublin in 2019. He is a Leading Researcher in energy and the environmental with an emphasis on solar energy, daylight and energy in buildings. Mr. Norton is an Associate Editor for journal Solar Energy. He serves on eight other editorial boards. Aziz Saif received the joint M.Sc. degree in power engineering from the Royal Institute of Technology, Stockholm, Sweden, and the Eindhoven University of Technology (TU/e), Eindhoven, The Netherlands, in 2014. Aziz Saif received the joint M.Sc. degree in power engineering from the Royal Institute of Technology, Stockholm, Sweden, and the Eindhoven University of Technology (TU/e), Eindhoven, The Netherlands, in 2014. He is currently a Research Associate with the In- ternational Energy Research Centre, Tyndall National Institute. His research interests include the operation of power grid with high renewable generation, local electricity markets in distribution grids, power system dynamics and stability, and demand response.
https://openalex.org/W4389897401	https://www.e3s-conferences.org/articles/e3sconf/pdf/2023/102/e3sconf_icimece2023_01016.pdf	English	null	Design of a Semi-Automatic Hydraulic Extruder Machine for the Manufacture of Pastry Bricks Using VDI 2221	E3S web of conferences	2,023	cc-by	3,821	* Corresponding author: 71063730@continental.edu.pe Design of a Semi-Automatic Hydraulic Extruder Machine for the Manufacture of Pastry Bricks Using VDI 2221 Cristhian Luis Nieto-Yali 1*, Rodolfo Alberto Ceras-Eguavil 1, Amarildo Kevin Albornoz-Zevallos 1, Manuel Michael Beraún-Espíritu 2, Carlos Alberto Coaquira-Rojo 1, and Samuel Turpo-Ccoa 1 1 Universidad Continental, School of Mechanical Engineering, Huancayo, Perú 2 Universidad Continental, School of Mechatronics Engineering, Huancayo, Perú 1 Universidad Continental, School of Mechanical Engineering, Huancayo, Perú 2 Universidad Continental, School of Mechatronics Engineering, Huancayo, Perú 1 Universidad Continental, School of Mechanical Engineering, Huancayo, Perú 2 Universidad Continental, School of Mechatronics Engineering, Huancayo, Perú Abstract. The main objective of this research work is to design a semi-automatic hydraulic extruder machine to produce pastry bricks for Bricks Imperio. The aim is to achieve a production rate of 700 bricks per hour at a maximum cost of 12,000 dollars. For this purpose, an interpretative theoretical framework of similar products is carried out, and simulation software is used to test the theoretical concepts in real-world conditions. Following the methodology recommended by the German Engineering Association VDI 2221, a detailed list of requirements is drawn up, existing technology is analyzed, and mechanical, electrical, and hydraulic functions are designed. The design is verified with CAD software, and a technoeconomic analysis is carried out. As a result, the successful design of the semi-automatic hydraulic extruder machine for the manufacture of pastry bricks is achieved, with a cost of 11,992.00 dollars and a production capacity of 700 bricks per hour. © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http ://creativecommons.org/licenses/by/4.0/). s ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 2 Methods and Materials In the city of Cusco, the district of San Jerónimo, and the community of Sucso Aucaylle, industrial and artisan brick factories are located in a total of more than 120 establishments. It is these factories that supply bricks, blockers, and tiles to meet the requirements of the brick industry [1]. The design will be based on the methodology recommended by the German Association of Engineers (VDI), VDI 2221, entitled "Mechanical Engineering Design Methodology", from which we will extract the structure, guidelines, recommendations, and others for the realisation and obtaining of an optimal solution that has to fulfil the client's requests [4]. The VDI 2221 methodology is used to help determine the best solution design so that the result meets expectations [5]. Within this requirement is the need to manufacture pastry bricks from the company Ladrillos Imperio, which to date are manufactured manually. Given the increase in orders, the brick businessman has chosen to hire a greater amount of labour to comply with the existing demand. However, by mid2020, the demand continued to grow and production was limited to the amount that one person could produce. Once again, in his desire to provide a solution, the brick businessman, in his desire to provide a solution, began to pay overtime to staff and increase benefits. This solution was viable for a while, but the staff got tired and began to ask for more benefits, which became unsustainable over time [2–3]. The design process using the VDI 2221 method consists of several stages; among the most important is the development of a morphological matrix where functions must be fulfilled for each mechanism. In addition, mathematical calculations are used to understand and know the forces and reactions. Involved in the production of the brick. Likewise, a mechanical and electrical-hydraulic design is carried out to obtain an optimal solution for the design, as shown in Fig. 1. Then arises the need for technological equipment that replaces labour and has an acceptable efficiency compared to the stated needs: a machine capable of forming a pastry brick of commercial measures, 20 cm long, 20 cm wide, and 3 cm high, which exceeds the production of 700 bricks per hour, is easy to transport, is comfortable in its maintenance, is semiautomatic in such a way that it can be operated by a single person, has security elements that prevent accidents, and its cost does not exceed 12,000 dollars. 2.3 Mathematical Calculations To understand and learn about the forces and reactions involved in the production of pastry bricks, it was necessary to observe in detail the work of the personnel involved in this activity and analyse step-by-step occurrences of the process of forming pastry bricks. For this reason, the design of the mould will take into account the measurements of the brick, the pressure to which the metal material will be subjected, with which the brick mould will be shaped, and then the characteristics that the material must have to guarantee its correct functioning over time. Fig. 2. Morphological Matrix Fig. 2. Morphological Matrix 2 Methods and Materials ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 Fig. 1. VDI 2221 Methodology black box is that it doesn't care about events, processes, or how. We are only interested in the results obtained. Fig. 3. Black Box black box is that it doesn't care about events, processes, or how. We are only interested in the results obtained. Fig. 3. Black Box black box is that it doesn't care about events, processes, or how. We are only interested in the results obtained. Fig. 3. Black Box Likewise, there is a description of the black box. First, there is the description of hydraulic energy and kinetic energy, as shown in Table I. Fig. 1. VDI 2221 Methodology 2.1 Morphological Matrix Table 1. Description Of The Electrical Signal Entrance Exit Hydraulic energy Kinetic energy It is the force that will be injected into the extruder machine so that it can do the work of pressing and shaping a brick. The energy released causes rectilinear displacement, friction, heat, and pressure. Thirdly, there is the description of the material and he bricks according to the black box. The morphological matrix is a popular conceptual design tool. Although concept design methods based on morphological matrices are effective in creating the concept design, it is difficult to determine the optimal concept design by combining these solution function principles [6–7]. The matrix is quantized in such a way that each decision principal uses decision variables and formulates an optimisation problem [8], as shown in Fig. 2 2. 2. Fig. 2. Morphological Matrix Thirdly, there is the description of the material and the bricks according to the black box. Thirdly, there is the description of the material and the bricks according to the black box. Table 2. Description Des Briques Entrance Exit Subject Bricks The raw material is a mixture of clay and sand with percentage proportions of 70 and 30, respectively. The material that enters the machine is processed and transformed inside. Final product of good texture and with the appropriate measurements. Table 2. Description Des Briques 2.2 Black Box We present a new approach for black-box simulation models called the probability distribution-based posterior box model [9–10], as shown in Fig. 2. An element that allows us to study the system from the point of view of the inputs that affect it and the inputs that respond to those inputs, the interesting thing about the Table 3. Description Des Briques Data Measurements of a pastry brick 20cm x 20cm x 3cm Table 3. Description Des Briques Data Measurements of a pastry brick 20cm x 20cm x 3cm 2 2 ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 System pressure 2.5Mpa stipulated in the requirements, it must have a design that allows efficient assembly and maintenance, as shown in Fig. 5. stipulated in the requirements, it must have a design that allows efficient assembly and maintenance, as shown in Fig. 5. With this data, it is assumed that the inner dimension of the mould should be 20 cm wide and 3 cm high, and the steel to be used should withstand at least 2.5 MPa. Then a drawing of the brick mould is made, followed by a free body diagram, as shown in Fig. 4. g Fig. 5. Brick Dimensions Fig. 5. Brick Dimensions y g , g Fig. 4. Brick Dimensions Developing we have. Fig. 5. Brick Dimensions 2.4 Mechanical Design Fig. 4. Brick Dimensions The following is a graphical representation of the optimal mechanical design solution. This graph does not include measurements or dimensions, but some parts can be described, which are listed as follows: (1) the metal base of the hydraulic extruder machine, (2) the hopper, (3) the extrusion chamber, (4) the double-acting hydraulic cylinder, (5) the brick mould, and the electrical and hydraulic systems, as shown in Fig. 6. Developing we have. Table 4. Mathematical Data Pressure = 2.5 MPa Distance B = 20 cm (0.2 m) Width (C) = 10 cm (0.1 m) Fig. 6. Mechanical design 2.5 Considerations of the brick baker's brick making machine Formula 1: Area=BxC=0.02 m2 Area=BxC=0.02 m2 (1) (1) We will rely on the formula. We will rely on the formula. P= F A (2) (2) Fig. 6. Mechanical design Replace the values in the formula (1), where 2.5 x 106 N/m2𝑥 = 𝐹 /0.02 m2 and we obtain the results of the force Fb = 50000 𝑁. With this result we can find the value of the thickness of the metal plate. Calculating the thickness of the plate with the following formula. 2.5 Considerations of the brick baker's brick making machine 2.5 Considerations of the brick baker's brick making machine For the execution of our design, it is important to recognise the elements that have to be manufactured and acquired. Then, TABLE VI and TABLE VII are made, where we name those elements that have to be manufactured inside the workshop and those elements that have to be bought in the market either locally or nationally. Quantitative quantities are considered that will serve as a reference for a list of acquisitions in the future. σ=MxC/I (3) (3) Where: Where: 𝜎 = Effort 𝑀 = Bending moment C = Neutral axis distance I = Moment of Inertia 𝑀 = Bending moment C = Neutral axis distance I = Moment of Inertia By finding the bending moment with the following formula. Table 5. Item To Be Manufactured Elements to be Manufactured Quantity Mould for pastry bricks. 1 Extrusion chamber 1 Feeding hopper 1 Base, extruder machine support 1 Table 5. Item To Be Manufactured Elements to be Manufactured Quantity Mould for pastry bricks. 1 Extrusion chamber 1 Feeding hopper 1 Base, extruder machine support 1 Table 5. Item To Be Manufactured Mb=Fxd (4) Airtight 30 x 30 x 10 cm metal cabinet 1 Airtight 30 x 30 x 10 cm metal cabinet 1 Table 6. Selected Element Elements to Select Quantity Hydraulic equipment (incl. electric motor, hydraulic pump, 4/3 valve) 1 Elements to Select Quantity Double-acting hydraulic cylinder 1 High pressure hose 2 Electric timer 2 Electrical power and control components 1 Bolts and Turks 25 Washers 25 3 Results Within this chapter, we will use the technology to check the calculations that have been previously performed using the finite element analysis method in SolidWorks 2020 SP5 [11]. We will check whether or not the material that has been chosen is correct for the applications required by the semiautomatic hydraulic extruder machine for the manufacture of pastry bricks. SolidWorks Corp. describes its software as a 3D CAD design tool, which in Spanish translates as "computer-aided design" [\|12], in which it is possible to model and assemble parts in 3 dimensions and drawings in 2 dimensions. This platform offers the possibility to design, simulate, manufacture, publish, and manage the data from the design process. Mb=Fxd Mb=Fxd (4) Replacing the values in the formula to obtain the bending moment, where the result is Mb = 50000 N x 0.1m and the bending moment is Mb = 5000 N x m. For this mould to be correctly mounted on the machine, it must have a clamping mechanism. This mechanism must have the exact geometry of the extrusion chamber to which it will be attached. Also, as 3 ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 2.6 Electrical Elements 4 4 ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 Fig. 8. Brick-formed mould We also find in Fig. 9, the sample of the pastry brick mould, subjected to an internal pressure of 2.5 MPa, the results obtained by the software show that the Von Misess stress has an elastic limit of 5.3 x 108 N/m2, the designed system works with 8.8 x 107 N/m2, therefore, the choice of the material is good. Fig. 10. Flow Analysis 1 Fig. 8. Brick-formed mould Fig. 8. Brick-formed mould Fig. 10. Flow Analysis 1 Fig. 10. Flow Analysis 1 Fig. 8. Brick-formed mould Fig. 8. Brick-formed mould Finally, Fig. 11 shows the value of the pressure with which the brick is going to come out: 102457.94 Pa as the maximum value and 97545.50 Pa as the minimum value. With these values, we have improved the compression with which the pastry brick is going to be formed; in this way, the qualities of the final product are improved. We also find in Fig. 9, the sample of the pastry brick mould, subjected to an internal pressure of 2.5 MPa, the results obtained by the software show that the Von Misess stress has an elastic limit of 5.3 x 108 N/m2, the designed system works with 8.8 x 107 N/m2, therefore, the choice of the material is good. We also find in Fig. 9, the sample of the pastry brick mould, subjected to an internal pressure of 2.5 MPa, the results obtained by the software show that the Von Misess stress has an elastic limit of 5.3 x 108 N/m2, the designed system works with 8.8 x 107 N/m2, therefore, the choice of the material is good. g Fig. 9. Brick mould p Fig. 11. Flow Analysis 2 Fig. 9. Brick mould Fig. 11. Flow Analysis 2 In the following Fig. 10, the machine is subjected to a flow analysis where we first study the speed with which the material will come out; this speed must be according to the needs and congruent with the amount of bricks to be produced. According to our calculations, we need to advance 38 cm in a time of 5 seconds to ensure 700 bricks per hour. In the calculation made by the software, we can see the average advance of 16 cm per second, which added in 5 seconds would be a total of 80 cm. 2.6 Electrical Elements We will also use Festo's FluidSim 4.2 software to simulate the semi-automatic sequences to be carried out by the hydraulic and electrical systems. Due to the customer's requirements, the machine must operate semi-automatically, which implies selecting control elements that allow the operator to give start and stop signals, elements that can detect position and interact with the system, and elements that can close or open contact and control large loads. The company Festo, known worldwide for providing pneumatic electric, hydraulic electric, or pneumatic or hydraulic automation solutions, is the creator of the FluidSim software, a simulator of pneumatic or hydraulic processes or a combination of the two options with the electrical and/or electronic system, thus guaranteeing the correct selection of the components to be acquired. The simulations performed are shown below [13–14]. Likewise, the selection of electrical elements was carried out, as shown in Table 7. Table 8. Electrical Elements Description of electrical elements Quantity Thermomagnetic key Schneider Acti 9 IC60N bipolar 1 Schneider thermal magnetic motor guard type GV2ME 1 Three-pole contactor D series + auxiliary contact CAD50M7 1 Timer to NO + NC contact connection 1 Schneider run push button series 9001K green colour NO contact 2 Schneider operating push button series 9001K red colour NC contact 1 Emergency mushroom head push button Schneider 9001K series NC contact 1 GPT cable N° 16 AWG indeco flexible red colour 1 NLT 3x14 AWG indeco flexible cable 1 ½ inch cable gland 2 ½ inch hose clamp for flexible pipe 2 Signalling lamps 3 Table 8. Electrical Elements In Fig. 7, the simulation in CAD SolidWorks 2020 SP5 shows the extrusion chamber subjected to a pressure of 2.5 MPa in the internal part of the part. In the results obtained by the software, it is observed that the Von Mises stress has an elastic limit of 5.3 x 108 N/m2. The designed system works at 3.188 x 107 N/m2. Therefore, the choice of material is correct. Fig. 7. Extrusion chamber Fig. 7. Extrusion chamber Fig. 7. Extrusion chamber Fig. 7. Extrusion chamber Also, in Fig. 8, the cake brick mould is subjected to an internal pressure of 2.5 MPa, the yield strength of the material is 5.3 x 108 N/m2, and the maximum stress to which the brick is subjected is 1.03 x 108 N/m2. Therefore, the choice of material that has been made is correct. References 1. V. C. Montufar and E. L. Rojas, “Relationship Between Environmental Management and Social Responsibility in the Tile and Brick Sector,” Prod. y Limpia, vol. 17, no. 1, pp. 20–34, (2022), doi: 10.22507/PML.V17N1A2. Fig. 13. Electrical and Hydraulic System Fi 13 El t i l d H d li S t 2. Y. Huachaca, “Determinación de la conductividad térmica de ladrillos utilizando vidrio reciclado para zonas alto andinas,” Dir. Gen. Investig. - Univ. Peru. Unión, pp. 1–93, 2020, [Online]. Available: https://tesis.pucp.edu.pe/repositorio/bitstream/hand le/20.500.12404/14615/Moran_Paucar_Estrés_aca démico_apoyo social1.pdf?sequence=1&isAllowed=y 3. Z. Zhang, S. Lin, Q. Lian, and C. Jin, “RepStore : A Self-Managing and Self-Tuning Storage Backend with Smart Bricks 2 . Background : P2P DHT The common terminology referring to a peer in,” Management, (2003) Fig. 13. Electrical and Hydraulic System 2.6 Electrical Elements Therefore, the design is capable of achieving the goal of 700 bricks per hour. The respective drawings of the hydraulic and electrical circuits are made using the tools of the FluidSim 4.2 hydraulic software. Once the circuit is completed, the corresponding simulation is made. Fig. 12, shows the start of the machine in a vacuum state. For this, it is necessary to unlock the emergency button represented by the letter E. Then, it is necessary to press the button P1, which activates the contactor K1. In this way, the motor of the hydraulic group is energized. At this stage, the piston does not produce any movement since the tandem valve is not yet activated and there is no passage of fluid into the cylinder. Fig. 12. Electrical and Hydraulic System Fig. 12. Electrical and Hydraulic System 5 ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 In the next step, in Fig. 13, the contactor K2 is activated; for this, it was necessary to press the start button M1, which gives the signal and allows self- locking with a normally open contact, thus giving way to a timer F, which energizes one of the coils of the tandem valve in this way, opening a passage to the fluid so that it enters the cylinder and the extrusion process occurs. The F timer has the function of controlling the travel time that the piston will have; once the time has elapsed, it cuts off the fluid supply and gives way to the return process. and costs, facilitating the modification of parts, increasing and decreasing dimensions, positioning and simulating functions, and using CAD software. 4 Discussion 4. J. Jänsch and H. Birkhofer, “The development of the guideline VDI 2221 - The change of direction,” 9th Int. Des. Conf. Des. 2006, pp. 45– 52, (2006) The design of a semi-automatic hydraulic extruder represents a significant advance in improving the production of pastry bricks. Despite its benefits, it is important to recognize that there are opportunities for future research, such as advanced automation, materials innovation, and continuous process optimization. In addition, constraints such as cost, access to components, specialized maintenance, and adaptability to changes in the industry must be addressed. A balanced approach in these aspects will allow to maximize the potential of this technology in the future. 5. K. N. Santoso et al., “Modification design of Melanger machine with reverse engineering method and VDI 2221,” Proc. Int. Conf. Ind. Eng. Oper. Manag., pp. 2711–2721, (2021) 6. W. Zeiler, “Morphology in conceptual building design,” Technol. Forecast. Soc. Change, vol. 126, no. xxxx, pp. 102–115, 2018, doi: 10.1016/j.techfore.2017.06.012. 7. T. D. L. Gontarski and R. K. Scalice, “Educational tool based on a morphological matrix for design alternative generation for use in railway wagon design,” Procedia CIRP, vol. 100, pp. 822–827, 2021, doi: 10.1016/j.procir.2021.05.037. 5 Conclusion The design of the semi-automatic hydraulic extruder machine for the manufacture of baking bricks in the company bricks Imperio meets the request to produce 700 baking bricks in one hour; it is semi-automatic, reduces labour, and processes a better product than the handmade one. Its approximate cost is US$11,992.00. 8. J. Ölvander, B. Lundén, and H. Gavel, “A computerized optimization framework for the morphological matrix applied to aircraft conceptual design,” CAD Comput. Aided Des., vol. 41, no. 3, pp. 187–196, 2009, doi: 10.1016/j.cad.2008.06.005. Customer requirements, safety standards, ergonomics, and the environment present a list of demands that guarantee an optimal and friendly design in an environment where people and industry can coexist. The morphological matrix was also used to identify the many possible solutions and alternatives, which in turn were subjected to a rigorous technical and economic analysis, resulting in an ideal optimum solution. 9. S. Nidhra, “Black Box and White Box Testing Techniques - A Literature Review,” Int. J. Embed. Syst. Appl., vol. 2, no. 2, pp. 29–50, 2012, doi: 10.5121/ijesa.2012.2204. 10. X. Zhang, L. Wu, Y. Fang, and H. Jiang, “A Study of FR Video Quality Assessment of Real Time Video Stream,” Int. J. Adv. Comput. Sci. Appl., vol. 3, no. 6, pp. 1–7, 2012, doi: 10.14569/ijacsa.2012.030601. Calculations were carried out to find the material characteristics for the brick mould, the extrusion chamber, the hopper, and the selection of electrical and hydraulic components. Finally, with the knowledge and mastery of CAD software, enormous advantages were found in the desire to design machines and electrical and/or hydraulic circuits, reducing construction times 11. O. G. Woge, C. O. G. Morán, and A. L. Chau, “Introducción al método del elemento finito: Solidworks y Matlab,” Ideas en Ciencias la Ing., vol. 1, no. 1, pp. 27–47, 2020, [Online]. Available: 6 ICIMECE 2023 E3S Web of Conferences 465, 01016 (2023) https://doi.org/10.1051/e3sconf/202346501016 https://ideasencienciasingenieria.uaemex.mx/article/view/1458 9 12. O. Rojas and L. Rojas, “Diseño de productos asistidos,” 2006, [Online]. Available: https://www.redalyc.org/articulo.oa?id=76820304 12. O. Rojas and L. Rojas, “Diseño de productos asistidos,” 2006, [Online]. Available: https://www.redalyc.org/articulo.oa?id=76820304 13. M. Orošnjak, M. Jocanović, and V. Karanović, “Simulation and Modeling of A Hydraulic System in FluidSim,” XVII Int. Sci. Conf. Ind. Syst., no. November, pp. 50–53, (2017) 14. G. Silva, “Open FluidSim: uma ferramenta multiplataforma para sistemas hidráulicos e pneumáticos,” p. 65, (2018) 7 7
https://openalex.org/W1972726807	https://europepmc.org/articles/pmc3745352?pdf=render	English	null	Translocation of Activator of G-protein Signaling 3 to the Golgi Apparatus in Response to Receptor Activation and Its Effect on the trans-Golgi Network	Journal of biological chemistry/The Journal of biological chemistry	2,013	cc-by	12,011	Translocation of Activator of G-protein Signaling 3 to the Golgi Apparatus in Response to Receptor Activation and Its Effect on the trans-Golgi Network* ecember 13, 2012, and in revised form, June 13, 2013 Published, JBC Papers in Press,June 14, 2013, DOI 10.1074/jbc.M112.44450 Sukru S. Oner, Ali Vural1, and Stephen M. Lanier2 From the Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medi Charleston, South Carolina 29425 Sukru S. Oner, Ali Vural1, and Stephen M. Lanier2 From the Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medical University of South Carolina Charleston, South Carolina 29425 Sukru S. Oner, Ali Vural1, and Stephen M. Lanier2 From the Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medical University of South Carolina, Charleston, South Carolina 29425 Background: The AGS3Gi complex is regulated by a GPCR, but downstream signaling events are unknown. Results: Upon receptor activation, AGS3 translocates to the Golgi apparatus, where it regulates events at the TGN. Conclusion: The AGS3Gi complex serves as a signal transducer for GPCRs. Background: The AGS3Gi complex is regulated by a GPCR, but downstream signaling events are unknown. Results: Upon receptor activation, AGS3 translocates to the Golgi apparatus, where it regulates events at the TGN. Conclusion: The AGS3Gi complex serves as a signal transducer for GPCRs. Significance: The regulated translocation of AGS3 offers unexpected mechanisms for modulating protein secretion and/or endosome recycling events at the TGN. i p g Significance: The regulated translocation of AGS3 offers unexpected mechanisms for modulating protein secretion and/or endosome recycling events at the TGN. Group II activators of G-protein signaling play diverse func- tional roles through their interaction with Gi, Gt, and Go via a G-protein regulatory (GPR) motif that serves as a docking site for G-GDP. We recently reported the regulation of the AGS3- Gi signaling module by a cell surface, seven-transmembrane receptor. Upon receptor activation, AGS3 reversibly dissociates from the cell cortex, suggesting that it may function as a signal transducer with downstream signaling implications, and this question is addressed in the current report. In HEK-293 and COS-7 cells expressing the 2A/D-AR and Gi3, receptor activa- tion resulted in the translocation of endogenous AGS3 and AGS3-GFP from the cell cortex to a juxtanuclear region, where it co-localized with markers of the Golgi apparatus (GA). The agonist-induced translocation of AGS3 was reversed by the 2-AR antagonist rauwolscine. Translocation of Activator of G-protein Signaling 3 to the Golgi Apparatus in Response to Receptor Activation and Its Effect on the trans-Golgi Network* The TPR domain of AGS3 was required for agonist-induced translocation of AGS3 from the cell cortex to the GA, and the translocation was blocked by per- tussis toxin pretreatment or by the phospholipase C inhibitor U73122. Agonist-induced translocation of AGS3 to the GA altered the functional organization and protein sorting at the trans-Golgi network. The regulated movement of AGS3 between the cell cortex and the GA offers unexpected mecha- nisms for modulating protein secretion and/or endosome recy- cling events at the trans-Golgi network. signal transfer from receptor to G-protein, guanine nucleotide binding and hydrolysis and/or G-protein subunit interactions and/or serve as alternative binding partners for G and G independent of the classical heterotrimeric G. AGS and RGS proteins have revealed unexpected functional diversity for the “G-switch” signaling mechanism, expanding the functional roles of G-protein subunits and perhaps identifying new oppor- tunities for therapeutic manipulation of G-protein signaling (1–6). RGS proteins fall into several subgroups, and there are more than 30 distinct proteins that contain an RGS or an RGS- like domain in mammalian cells (2). AGS proteins generally fall into three subgroups (1, 5). Group I AGS proteins and related entities, such as Ric-8A, GIV (Girdin), and the Saccharomyces cerevisiae protein Arr4, act as non-receptor guanine nucleotide exchange factors for G and/or G (7–9). Group III AGS proteins interact with G or perhaps heterotrimer (1, 10). Sato et al. (11) also recently identified three transcription factors as AGS11, -12, and -13 with apparent selectivity for G16; how- ever, the biochemical properties of the G-protein regulation have not yet been fully characterized, and thus it is difficult to place these three AGS proteins into Group I, II, or III. Group II AGS proteins are defined by the presence of one or more guanine nucleotide regulatory (GPR) motifs (12), also known as GoLoco or LGN motifs (13, 14), that bind G-GDP free of G. Group II AGS proteins consist of seven proteins (AGS3 (GPSM1), LGN (GPSM2, AGS5), AGS4 (GPSM3), RGS12 (AGS6), Rap1Gap (transcript variant 1), RGS14, and PCP2/L7), each of which contains 1–4 GPR motifs for docking of G serving as alternative binding partners for specific sub- types of G. There are three types of Group II AGS proteins that are distinguished by the number of GPR motifs and/or the presence of defined regulatory protein domains (1, 5). * Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth Grant DA025896 (to S. M. L.). 1 Presentaddress:BCellMolecularImmunologySection,LaboratoryofImmu- noregulation, NIAID, National Institutes of Health, Bethesda, MD 20892. 2 To whom correspondence should be addressed: Dept. of Pharmacology, Medical University of South Carolina, 179 Ashley Ave., Charleston, SC 29425. Tel.: 843-792-7134; E-mail: lanier@musc.edu. 3 The abbreviations used are: AGS, activator(s) of G-protein signaling; RGS, regulator(s) of G-protein signaling; GPR, G-protein regulatory; GA, Golgi apparatus; ST, sialyltransferase; BRET, bioluminescence resonance energy transfer; EGFP, enhanced GFP; TGN, trans-Golgi network. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 33, pp. 24091–24103, August 16, 2013 © 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 33, pp. 24091–24103, August 16, 2013 © 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 288, NO. 33, pp. 24091–24103, August 16, 2013 © 2013 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A. EXPERIMENTAL PROCEDURES Materials—UK-14304, U0126, and pertussis toxin were obtained from Sigma. Rauwolscine hydrochloride was obtained from Carl Roth GmbH (Karlsruhe, Germany). Polyethyleneimine (PEI) was obtained from Polysciences, Inc. (Warrington, PA). All cell culture materials were obtained from Invitrogen. HEK-293 and COS-7 cells (CRL-1573 and CRL-1651, respectively) were obtained from American Type Culture Collection (Manassas, VA). Benzyl-coelenterazine was obtained from NanoLight Technology. 96-well gray OptiPlates were obtained from PerkinElmer Life Sciences. GM130 antibody (610823) was obtained from BD Biosciences. TGN46 antibody (110- 40769) was obtained from Novus Biologicals (Littleton, CO). Mannosidase II antibody (ab12277) and LAMP2 antibody (ab25631) were obtained from Abcam (Cambridge, MA). AGS3 antisera were generated in the laboratory of Dr. Dzwokai Zach Ma (University of California, Santa Barbara, CA) by immuniza- tion of rabbits with a glutathione S-transferase (GST)-AGS3 fusion protein encoded by the GPR domain (Ala461–Ser650) of AGS3 (26). Sialyltransferase-EGFP (ST-GFP) was generated in the laboratory of Dr. Jennifer Lippincott-Schwartz (37). All other reagents and materials were obtained as described elsewhere (34). FIGURE 1. Influence of cell surface receptor activation on the interaction of AGS3 with Gi1 over time and the subcellular distribution of AGS3. A, real-time BRET between AGS3-Rluc and Gi1-YFP. HEK-293 cells expressing phRluc::AGS3 (10 ng), pcDNA3::Gi1-YFP (750 ng), and pcDNA3::2A/D-AR (750 ng) were processed for BRET measurements as described under “Exper- imental Procedures.” The 2A/D-AR agonist UK-14304 and antagonist rau- wolscine (Rauw) were added as indicated. These data are expressed as means S.E. (error bars) (n 3). B, cells were transfected with AGS3-GFP (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) and incubated with vehicle or UK-14304 (10 M) for 10 min and then processed for immuno- fluorescence image analysis as described under “Experimental Procedures.” The image presented (10 objective) is representative of multiple experi- ments as described under “Experimental Procedures.” Cell Culture—HEK-293 and COS-7 cells were maintained in DMEM high glucose medium supplemented with 5 or 10% fetal bovine serum, respectively, in the presence of 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were grown in a humidified incubator in the presence of 5% CO2 at 37 °C. Cells were transfected using PEI as described previously (34). cells/well). Luminescence was measured in the 480- and 530-nm emission windows at 2.5-s intervals beginning imme- diately after the addition of coelenterazine H to the sample wells. Translocation of Activator of G-protein Signaling 3 to the Golgi Apparatus in Response to Receptor Activation and Its Effect on the trans-Golgi Network* AGS3 and LGN (AGS5) have four GPR motifs downstream of a tet- ratricopeptide repeat domain, whereas RGS12 (AGS6), RGS14, and Rap1GAP have one GPR motif plus other defined domains that act to accelerate G-GTP hydrolysis (1, 5). Activators of G-protein signaling (AGS)3 and regulators of G-protein signaling (RGS) proteins were both discovered in functional screens for G-protein signaling modulators and essentially define a panel of biological regulators that influence The GPRG complex is involved in an increasingly interest- ing set of regulatory functions that we are just beginning to understand, including roles in cell division, neuronal plastic- 24091 JOURNAL OF BIOLOGICAL CHEMISTRY AUGUST 16, 2013•VOLUME 288•NUMBER 33 FIGURE 1. Influence of cell surface receptor activation on the interaction of AGS3 with Gi1 over time and the subcellular distribution of AGS3. A, real-time BRET between AGS3-Rluc and Gi1-YFP. HEK-293 cells expressing phRluc::AGS3 (10 ng), pcDNA3::Gi1-YFP (750 ng), and pcDNA3::2A/D-AR (750 ng) were processed for BRET measurements as described under “Exper- imental Procedures.” The 2A/D-AR agonist UK-14304 and antagonist rau- wolscine (Rauw) were added as indicated. These data are expressed as means S.E. (error bars) (n 3). B, cells were transfected with AGS3-GFP (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) and incubated with vehicle or UK-14304 (10 M) for 10 min and then processed for immuno- fluorescence image analysis as described under “Experimental Procedures.” The image presented (10 objective) is representative of multiple experi- ments as described under “Experimental Procedures.” Regulated Trafficking of AGS3 to the Golgi ity, autophagy, protein trafficking, hearing, morphogenesis, and the renal response to injury (15–30). Although the majority of G within the cell probably exists as part of the G het- erotrimer at the cell cortex, a subpopulation of Gi is appar- ently complexed with GPR proteins. This GPRG signaling module is regulated by non-receptor guanine nucleotide exchange factors, such as Ric-8A and GIV (24, 31–33), with the latter entity processing signals from cell surface growth factor receptors (24). The GPRG complex (AGS3Gi, AGS4Gi, RGS14Gi) at the cell cortex is also regulated by a subgroup of G-protein-coupled receptors (34–36). Upon receptor activa- tion, AGS3 reversibly dissociated from Gi at the cell cortex (34), but the trafficking of AGS3 and its regulation following receptor activation in this system is not known. In this work, we report that AGS3 reversibly translocates from the cell cortex to the Golgi apparatus (GA), and this is associated with altered functional integrity of the trans-Golgi network. The regulated translocation of AGS3 between the cell cortex and the GA offers unexpected mechanisms for modulating protein secre- tion, endosome recycling, and Golgi dynamics. EXPERIMENTAL PROCEDURES At 150 and 270 s, UK-14304 (1 M) and rauwolscine (10 M) were added by injection, respectively. Signal measure- ments were continued for 390 s following the addition of antagonist. Real-time Bioluminescence Resonance Energy Transfer (BRET) Measurements—HEK-293 cells were seeded on 6-well plates and culturedovernightat37 °C.Cellsweretransientlytransfectedwith donor (phRluc::AGS3), acceptor (pcDNA3::Gi1-YFP), and pcDNA3::2A/D-adrenergic receptor and processed for BRET measurements as described previously (34, 35). Cells were dis- tributed into gray 96-well OptiPlates in triplicate (1 105 Cell Imaging—HEK-293 and COS-7 cells were seeded onto sterile 25-mm polylysine-D-coated coverslips 24 h following transfection. Cells were then processed for imaging 24 h later as described elsewhere (38). Primary antibodies were used at a dilution of 1:200 in cell washing solution (137 mM NaCl, 2.6 mM KCl, 1.8 mM KH2PO4, 10 mM Na2HPO4, pH 7.4), and secondary 24092 VOLUME 288•NUMBER 33•AUGUST 16, 2013 24092 JOURNAL OF BIOLOGICAL CHEMISTRY Regulated Trafficking of AGS3 to the Golgi oat anti-rabbit AlexaFluor488 or mouse Alexa- hl d b d M l l P b ) d processed by deconvolution for higher resolution using Simple PCI i 6 6 0 0 (H t C S i kl PA) I nce of cell surface receptor activation on the subcellular distribution of AGS3. HEK-293 (A) or COS-7 (B) cells expressing pEGFP::AGS3 (25 - COS-7), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed icroscopy. The Golgi marker GM130 (red) was detected by immunofluorescence as described under “Experimental Procedures.” The images sentative of 10 (HEK) or three (COS-7) separate experiments. The arrowheads indicate juxtanuclear localization of AGS3 following receptor 0 m. Bottom panels, the relative subcellular distributions of AGS3-GFP and GM130 were quantified by fluorescence intensity scans (green, M130) through the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” FIGURE 2. Influence of cell surface receptor activation on the subcellular distribution of AGS3. HEK-293 (A) or COS-7 (B) cells expressing pEGFP::AGS3 (25 ng - HEK, 100 ng - COS-7), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy. The Golgi marker GM130 (red) was detected by immunofluorescence as described under “Experimental Procedures.” The images shown are representative of 10 (HEK) or three (COS-7) separate experiments. The arrowheads indicate juxtanuclear localization of AGS3 following receptor activation. Bar, 10 m. EXPERIMENTAL PROCEDURES Bottom panels, the relative subcellular distributions of AGS3-GFP and GM130 were quantified by fluorescence intensity scans (green, AGS3-GFP; red, GM130) through the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” antibodies (goat anti-rabbit AlexaFluor488 or mouse Alexa- Fluor594, highly cross-adsorbed, Molecular Probes) were used at a dilution of 1:2000 in cell washing solution. All antibody dilutions were centrifuged at 10,000 g for 10 min prior to use. The nucleus was stained with 1 g/ml DAPI at the last washing step. Slides were then mounted with glass coverslips and ana- lyzed with a Leica CTR5500 deconvolution fluorescence micro- scope using a 40 or 63 oil immersion objective as described elsewhere (38). All images were obtained from approximately the middle plane of the cells. Selected enlarged images were processed by deconvolution for higher resolution using Simple PCI version 6.6.0.0 (Hamamatsu Corp., Sewickley, PA). Images were evaluated for colocalization of GA marker proteins and AGS3 by three different approaches. First, each experiment was visually examined by at least two individuals to identify and count the population of cells exhibiting clear juxtanuclear enrichment of AGS3 and overlap of the AGS3 with GA marker proteins. Cells undergoing division were not counted to avoid any nonclarity in regard to data interpretation and the GA frag- mentation that occurs during mitosis. 100–200 individual AUGUST 16, 2013•VOLUME 288•NUMBER 33 JOURNAL OF BIOLOGICAL CHEMISTRY 24093 AUGUST 16, 2013•VOLUME 288•NUMBER 33 24093 Regulated Trafficking of AGS3 to the Golgi Regulated Trafficking of AGS3 to the Golgi cells expressing AGS3-GFP were examined in each experiment, In cells transfected with pEGFP::ST, the relativ FIGURE 3. Influence of cell surface receptor activation on the subcellular distribution of AGS3 and the GA marker proteins TGN46 and HEK-293 cells transfected with pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or U or 10 min and processed for fluorescent microscopy. The Golgi markers TGN46 (trans-Golgi network) (A) and Mann-II (luminal GA) (B) w mmunofluorescence as described under “Experimental Procedures.” The images shown are representative of 5–10 separate experiments. relative subcellular distributions of AGS3-GFP and TGN46 and Mann-II were quantified by fluorescence intensity scans (green, AGS3-GFP; red, the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” Regulated Trafficking of AGS3 to the Golgi FIGURE 3. EXPERIMENTAL PROCEDURES Influence of cell surface receptor activation on the subcellular distribution of AGS3 and the GA marker proteins TGN46 and mannosidase II. HEK-293 cells transfected with pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy. The Golgi markers TGN46 (trans-Golgi network) (A) and Mann-II (luminal GA) (B) were detected by immunofluorescence as described under “Experimental Procedures.” The images shown are representative of 5–10 separate experiments. Bar, 10 m. The relative subcellular distributions of AGS3-GFP and TGN46 and Mann-II were quantified by fluorescence intensity scans (green, AGS3-GFP; red, GM130) through the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” In cells transfected with pEGFP::ST, the relative distribution of ST-GFP, which resides in the trans-Golgi network, was determined by counting the numbers of ST-GFP puncta in the presence and absence of agonist. Similar experiments were conducted in cells expressing ST-GFP alone or together with pcDNA3::Gi3 and pcDNA3::2A/D-AR. cells expressing AGS3-GFP were examined in each experiment, and each experiment was repeated at least three times. Sec- ond, selected images were examined by scanning of fluores- cence intensity across the cell using ImageJ (39). Third, images were also evaluated for colocalization using the Coloc_2 module within the Fiji image processing package (40), which provides a thresholded Pearson’s correlation coefficient for pixel overlap based on the approach devel- oped by Costes et al. (41). cells expressing AGS3-GFP were examined in each experiment, and each experiment was repeated at least three times. Sec- ond, selected images were examined by scanning of fluores- cence intensity across the cell using ImageJ (39). Third, images were also evaluated for colocalization using the Coloc_2 module within the Fiji image processing package (40), which provides a thresholded Pearson’s correlation coefficient for pixel overlap based on the approach devel- oped by Costes et al. (41). Data Analysis—Data were analyzed by analysis of vari- ance, and significant differences between groups were deter- mined by the Tukey a posteriori test using GraphPad Prism 24094 JOURNAL OF BIOLOGICAL CHEMISTRY B, quantitative colocalization analysis was conducted using the Fiji image processing suite to generate thresholded Pearson’s correlation coefficients for vehicle and for cells incubated with the agonist UK-14304 as described under “Experimental Procedures.” Two general subpopulations of cells were identified following agonist incubation, and each population was evaluated for colocalization of AGS3-GFP and GA marker proteins. The white bars correspond to the subpopulation of cells exhibiting a robust, juxtanuclear localization of AGS3 after agonist treatment. The red bars correspond to the second subpopulation of cells as described under “Experimental Procedures.” C, the agonist-induced translocation of AGS3-GFP to the Golgi apparatus was determined by visual examination of spectral overlap between AGS3 and the GA marker proteins as described under “Experimental Procedures” (n 3). Data are expressed as means S.E. (error bars) as described under “Experimental Procedures.” , p 0.0001 versus vehicle. #, p 0.0001. The numbers above the bars in parentheses indicate the number of AGS3-GFP-expressing cells examined. version 4.03 for Windows (GraphPad Software, San Diego, CA). results were also obtained in both HEK and COS7 cells express- ing Gi1.4 The juxtanuclear localization of AGS3-GFP observed in the presence of agonist also overlapped with the trans-Golgi marker TGN46 and a marker of the luminal GA (mannosidase II) (Fig. 3). To further define the juxtanuclear localization of AGS3 observed upon the addition of agonist, we also scanned fluorescence intensities across cells for AGS3-GFP and GA marker proteins (Figs. 2 and 3). The fluorescence intensity scans indicated that AGS3 was generally enriched at the cell cortex in the absence of agonist. However, upon receptor acti- vation, many cells exhibited a robust accumulation of AGS3 at the juxtanuclear region and spatial overlap with the GA marker proteins (Figs. 2 and 3). 24094 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 33•AUGUST 16, 2013 24094 JOURNAL OF BIOLOGICAL CHEMISTRY Regulated Trafficking of AGS3 to the Golgi FIGURE 4. Quantitative analysis of AGS3 translocation. HEK-293 cells expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy. A, representative image of cells incubated with the agonist UK-14304, illustrating two populations of cell response. The white arrowheads indicate cells exhibiting a robust, juxtanuclear localization of AGS3-GFP after agonist treatment that was visually obvious. The red arrowheads indicate cells exhibiting a more diffuse intracellular distribution of AGS3-GFP. B, quantitative colocalization analysis was conducted using the Fiji image processing suite to generate thresholded Pearson’s correlation coefficients for vehicle and for cells incubated with the agonist UK-14304 as described under “Experimental Procedures.” Two general subpopulations of cells were identified following agonist incubation, and each population was evaluated for colocalization of AGS3-GFP and GA marker proteins. The white bars correspond to the subpopulation of cells exhibiting a robust, juxtanuclear localization of AGS3 after agonist treatment. The red bars correspond to the second subpopulation of cells as described under “Experimental Procedures.” C, the agonist-induced translocation of AGS3-GFP to the Golgi apparatus was determined by visual examination of spectral overlap between AGS3 and the GA marker proteins as described under “Experimental Procedures” (n 3). Data are expressed as means S.E. (error bars) as described under “Experimental Procedures.” , p 0.0001 versus vehicle. #, p 0.0001. The numbers above the bars in parentheses indicate the number of AGS3-GFP-expressing cells examined. FIGURE 4. Quantitative analysis of AGS3 translocation. HEK-293 cells expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy. A, representative image of cells incubated with the agonist UK-14304, illustrating two populations of cell response. The white arrowheads indicate cells exhibiting a robust, juxtanuclear localization of AGS3-GFP after agonist treatment that was visually obvious. The red arrowheads indicate cells exhibiting a more diffuse intracellular distribution of AGS3-GFP. 4 S. S. Oner and S. M. Lanier, unpublished observations. RESULTS AND DISCUSSION Comparative subcellular distribution for AGS3 and markers for early endosomes or lysosomes following receptor activation expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) processed for fluorescent microscopy as described under “Experimental Procedures.” The early endosome marker EEA1 (A) and lysosome ma were detected by immunofluorescence as described under “Experimental Procedures.” The images shown are representative of three separat The relative subcellular distributions of AGS3-GFP and EEA1 and LAMP2 were quantified by fluorescence intensity scans (green, AGS3-GFP; red, through the cell as indicated by the lines in the images and as described under “Experimental Procedures.” Regulated Trafficking of AGS3 to the Golgi Regulated Trafficking of AGS3 to the Golgi Regulated Trafficking of AGS3 to the Golgi Regulated Trafficking of AGS3 to th FIGURE 5. Comparative subcellular distribution for AGS3 and markers for early endosomes or lysosomes following receptor activation. HEK-293 cells expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” The early endosome marker EEA1 (A) and lysosome marker LAMP2 (B) were detected by immunofluorescence as described under “Experimental Procedures.” The images shown are representative of three separate experiments. The relative subcellular distributions of AGS3-GFP and EEA1 and LAMP2 were quantified by fluorescence intensity scans (green, AGS3-GFP; red, EEA1 or LAMP2) through the cell as indicated by the lines in the images and as described under “Experimental Procedures.” enrichment of AGS3 overlapping with GA marker proteins as defined by visual examination and/or thresholded Pearson’s correlation coefficients greater than 0.7, which represented 50% of the cells expressing transfected AGS3 (Fig. 4, B and C). Analysis of enlarged fluorescent images with the three GA marker proteins and AGS3-GFP by deconvolution indicated that the translocated AGS3-GFP generally had strong spatial colocalization with TGN46, exhibiting a tubero-vesicular appearance perhaps related to protein secretion at the trans face of the GA (Figs. 2 and 3). Although there is clear overlap of AGS3-GFP with the GA marker proteins, there are distinct areas where the GA marker proteins GM130 and Mann-II do not overlap with AGS3, indicating that AGS3 may translocate to specific subcompartments within the GA (Figs. 2 and 3). RESULTS AND DISCUSSION We established a platform to monitor the interaction of GPR proteins and G in the intact cell by BRET. This platform involves an energy donor (AGS3-Rluc) and acceptor (Gi-YFP). Because AGS3-Rluc may bind up to four Gi-YFP, the BRET signal is particularly robust (34). At least three different types of GPRGi complexes (AGS3Gi, AGS4Gi, and RGS14Gi) are regulated by activation of a cell surface, G-protein-coupled receptor as well as by the non-receptor guanine nucleotide exchange factor Ric-8A and/or GIV (24, 33–36). Activation of the 2-AR rapidly decreased AGS3-RlucGi1-YFP BRET to a new steady state, which was rapidly reversed by the addition of the 2-AR antagonist rauwolscine (Fig. 1A). Based upon previ- ous studies, it appears that the AGS3-Rluc is “released” from its membrane site, whereas Gi1-YFP remains tethered to the plasma membrane (34). The degree of colocalization of AGS3 with GA marker proteins following agonist treatment was quantitatively examined in AGS3-transfected cells using the Fiji image anal- ysis suite. The robust, juxtanuclear localization of AGS3 after agonist treatment was visually obvious in many cells, whereas some cells exhibited a more diffuse intracellular distribution of AGS3, which was clearly different from the predominantly cor- tical distribution in the absence of agonist (Fig. 4A). Analysis of this mixed population of cells observed in the presence of ago- nist yielded a thresholded Pearson’s correlation coefficient for colocalization analysis with GM130 and AGS3-GFP of 0.58 0.05 (n 34) with agonist incubation versus 0.08 0.04 (n We hypothesized that the tethered AGS3 was released to play a role in signal propagation. As a first approach to test this hypothesis, we determined the subcellular location of AGS3 after its release from the cell cortex. Upon agonist treatment in HEK cells, the marked cortical distribution of AGS3 was lost in 90% of the cells (Fig. 1B), and many cells exhibited a robust juxtanuclear enrichment of AGS3 that appeared to overlap with the cis-Golgi apparatus-resident protein GM130 (Fig. 2A). Sim- ilar results were obtained in COS-7 cells (Fig. 2B). Similar AUGUST 16, 2013•VOLUME 288•NUMBER 33 24095 JOURNAL OF BIOLOGICAL CHEMISTRY 24095 AUGUST 16, 2013•VOLUME 288•NUMBER 33 FIGURE 5. RESULTS AND DISCUSSION Middle, influence of the phospholipase C inhibitor U73122 on agonist-induced translocation of AGS3 to the Golgi apparatus as determined by co-localization with GM130. HEK-293 cells expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were pretreated with U73122 (10 M) or an inactive analog U73343 (10 M) for 20 min prior to the addition of the 2A/D-AR agonist UK-14304. Incubation with agonist was continued for 10 min before the cells were processed for image analysis. Data are expressed as means S.E. (n 3). Right, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localizationwithGM130.HEK-293cellsexpressingpEGFP::AGS3(25ng),pcDNA3::Gi3(750ng),andpcDNA3::2A/D-AR(750ng)wereincubatedwithvehicle or brefeldin A (BFA) (10 g/ml) for 30 min, and agonist (UK-14304, 1 M) was then added with incubation continued for 10 min. GM130 (red) was detected by immunofluorescence. Data are expressed as means S.E. (n 3). , p 0.0001 compared with control value in the absence of agonist. #, p 0.0001 compared with value obtained in the presence of agonist. C, left, agonist-induced translocation of AGS3-GFP to the juxtanuclear area. HEK-293 cells expressing pEGFP::AGS3(25ng),pcDNA3::Gi3(750ng),andpcDNA3::2A/D-AR(750ng)werepretreatedwithvehicleorU0126(20M)for1h.Cellswerethentreatedwith UK-14304 for 10 min and processed for image analysis. Right, subcellular distribution of an AGS3 construct in which 24 serine/threonine residues in the GPR domain were mutated to alanine (AGS3-PM). HEK-293 cells expressing pEGFP::AGS3-WT (25 ng) or pEGFP::AGS3-PM (25 ng) in the presence of pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were treated with the agonist UK-14304 for 10 min and processed for image analysis. Data are expressed as means h h d h b b h b h d h b f ll d FIGURE6.TimecourseandmechanismsinvolvedwiththetranslocationofAGS3totheGolgiapparatus.A,HEK-293cellsexpressingpEGFP::AGS3(25ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with UK-14304 (1 M) for different time periods and processed for fluorescent microscopy.Theimagesshownarerepresentativeofthreeseparateexperiments.ThearrowheadsindicatejuxtanuclearlocalizationofAGS3followingreceptor activation. Red, GM130. Bottom, the relative subcellular distributions of AGS3-GFP and GM130 were quantified by fluorescence intensity scans (green, AGS3- GFP; red, GM130) through the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” B, left, agonist-induced translo- cation of AGS3-GFP to the Golgi apparatus as determined by co-localization with GM130. Reversibility of the UK-14304-induced translocation of AGS3 was assessedbytheadditionoftheantagonistrauwolscine(Rauw)(10M)followinga10-minincubationofthecellswithUK-14304(1M).Rauwolscineincubation was continued for 10 min, and then the cells were processed for fluorescent microscopy as described under “Experimental Procedures.” Another group of cells were treated with vehicle or 100 ng/ml pertussis toxin for 16 h prior to incubation with agonist. Data are expressed as means S.E. RESULTS AND DISCUSSION (error bars) (n 3–7). Middle, influence of the phospholipase C inhibitor U73122 on agonist-induced translocation of AGS3 to the Golgi apparatus as determined by co-localization with GM130. HEK-293 cells expressing pEGFP::AGS3 (25 ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were pretreated with U73122 (10 M) or an inactive analog U73343 (10 M) for 20 min prior to the addition of the 2A/D-AR agonist UK-14304. Incubation with agonist was continued for 10 min before the cells were processed for image analysis. Data are expressed as means S.E. (n 3). Right, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localizationwithGM130.HEK-293cellsexpressingpEGFP::AGS3(25ng),pcDNA3::Gi3(750ng),andpcDNA3::2A/D-AR(750ng)wereincubatedwithvehicle or brefeldin A (BFA) (10 g/ml) for 30 min, and agonist (UK-14304, 1 M) was then added with incubation continued for 10 min. GM130 (red) was detected by immunofluorescence. Data are expressed as means S.E. (n 3). , p 0.0001 compared with control value in the absence of agonist. #, p 0.0001 compared with value obtained in the presence of agonist. C, left, agonist-induced translocation of AGS3-GFP to the juxtanuclear area. HEK-293 cells expressing pEGFP::AGS3(25ng),pcDNA3::Gi3(750ng),andpcDNA3::2A/D-AR(750ng)werepretreatedwithvehicleorU0126(20M)for1h.Cellswerethentreatedwith UK-14304 for 10 min and processed for image analysis. Right, subcellular distribution of an AGS3 construct in which 24 serine/threonine residues in the GPR domain were mutated to alanine (AGS3-PM). HEK-293 cells expressing pEGFP::AGS3-WT (25 ng) or pEGFP::AGS3-PM (25 ng) in the presence of pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were treated with the agonist UK-14304 for 10 min and processed for image analysis. Data are expressed as means S.E. (n 3). PM, phosphomutant. In B and C, the numbers above the bars in parentheses indicate the number of AGS3-GFP-expressing cells examined. (thresholded Pearson’s correlation coefficient 0.88 0.02, n 4) (0.5 min), 39.4 1.2% (thresholded Pearson’s correlation coefficient 0.87 0.03, n 5) (1 min), 49 2% (thresholded Pearson’s correlation coefficient 0.84 0.03, n 4) (2 min), and 46.5 2.8%, (Pearson’s correlation coefficient 0.81 0.33, n 6) (30 min). The spatial overlap of AGS3 with the GA marker protein GM130 was also observed by fluorescence intensity scanning across individual cells (Fig. 6A, bottom). (LAMP2), and the endoplasmic reticulum (calnexin). The sub- cellular distribution of these three organelle markers is much broader than the juxtanuclear localization of AGS3 observed following agonist treatment (Fig. RESULTS AND DISCUSSION Although there are many different organelles and functional compartments that position themselves at juxtanuclear sites, the clear spatial overlap with the three GA marker proteins is consistent with translocation of AGS3 to the GA following receptor activation. We also examined the potential distribu- tion of AGS3 among early endosomes (EEA1), lysosomes 34) with vehicle. Further colocalization analysis of these two general subpopulations of cells with the GA marker GM130 indicated that the cells with robust, juxtanuclear localization of AGS3 following agonist treatment exhibited a thresholded Pearson’s correlation coefficient of 0.81 0.02 (n 16), whereas cells exhibiting a more diffuse distribution of AGS3 without any clear enrichment at the GA exhibited a thresh- olded Pearson’s correlation coefficient of 0.37 0.04 (n 18) (Fig. 4B) This analysis indicates that agonist treatment results in clear, statistically distinct, colocalization of AGS3 with the GA marker proteins. Similar results were obtained when colo- calization analysis was conducted for AGS3 and the GA marker proteins TGN46 and mannosidase II (Fig. 4B). The population of agonist-treated cells exhibiting the lower thresholded Pearson’s correlation coefficient (Fig. 4, A and B) probably reflects a graded response that may represent an inter- mediate step in translocation of AGS3 to the juxtanuclear region or represent other trafficking pathways for AGS3 (19, 23, 24, 38). For the purposes of this current study, we focused on the population of cells exhibiting the robust, juxtanuclear 24096 JOURNAL OF BIOLOGICAL CHEMISTRY 24096 VOLUME 288•NUMBER 33•AUGUST 16, 2013 Regulated Trafficking of AGS3 to the Golgi FIGURE6.TimecourseandmechanismsinvolvedwiththetranslocationofAGS3totheGolgiapparatus.A,HEK-293cellsexpressingpEGFP::AGS3(25ng), pcDNA3::Gi3 (750 ng), and pcDNA3::2A/D-AR (750 ng) were incubated with UK-14304 (1 M) for different time periods and processed for fluorescent microscopy.Theimagesshownarerepresentativeofthreeseparateexperiments.ThearrowheadsindicatejuxtanuclearlocalizationofAGS3followingreceptor activation. Red, GM130. Bottom, the relative subcellular distributions of AGS3-GFP and GM130 were quantified by fluorescence intensity scans (green, AGS3- GFP; red, GM130) through the cell, as indicated by the lines in the images and as described under “Experimental Procedures.” B, left, agonist-induced translo- cation of AGS3-GFP to the Golgi apparatus as determined by co-localization with GM130. Reversibility of the UK-14304-induced translocation of AGS3 was assessedbytheadditionoftheantagonistrauwolscine(Rauw)(10M)followinga10-minincubationofthecellswithUK-14304(1M).Rauwolscineincubation was continued for 10 min, and then the cells were processed for fluorescent microscopy as described under “Experimental Procedures.” Another group of cells were treated with vehicle or 100 ng/ml pertussis toxin for 16 h prior to incubation with agonist. Data are expressed as means S.E. (error bars) (n 3–7). Regulated Trafficking of AGS3 to the Golgi Furthermore, the agonist- induced translocation of AGS3 to the juxtanculear region was not observed following treatment of cells with brefeldin A, which disrupts the structure of the GA (Fig. 6B). The precise mechanism by which phospholipase C may regulate this transport event is not clear. Although U73122 is typically used as a phospholipase C inhibitor, the molecule may have a num- ber other actions that contribute to its pharmacological profile (42). Nevertheless, U73122 should be a useful tool to dissect the mechanisms involved in AGS3 translocation going forward. FIGURE 7. Influence of the TPR domain of AGS3 on the agonist-induced translocation of AGS3 to the Golgi apparatus. A, left, schematic represen- tation of AGS proteins. AGS3-Q/A is an AGS3 mutant with single residue mutations in each of the four GPR motifs (GPR1-Q488A, GPR2-Q541A, GPR3- Q589A, and GPR4-Q623A) that disrupt Gi binding to the individual GPR motif. Right, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localizationwithGM-130.HEK-293cellsexpressingpcDNA3::Gi3(750ng), pcDNA3::2A/D-AR (750 ng) and pEGFP::AGS3 (25 ng), pEGFP::AGS3-Q/A (25 ng), pEGFP::AGS4 (25 ng), or pEGFP::AGS3-SHORT (AGS3-SH-GFP) (25 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” ,p0.0001comparedwiththecontrolvalueintheabsenceofagonist.Band C, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localiza- tion with GM-130. HEK-293 cells expressing pcDNA3::Gi3 (750 ng), pcDNA3::2A/D-AR (750 ng), and pEGFP::AGS3-WT or AGS3 constructs (25 ng) with progressive deletion of individual TPR motifs (B) or with single amino acid mutations of conserved residues within individual TPR motifs (C) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” GM130 (red) was detected by immunofluorescence. In B, 800 AGS3-GFP-ex- pressing cells were examined for subcellular distribution determination for eachAGS3-GFPconstruct.DataareexpressedasmeansS.E.(errorbars)(n 3–5). , p 0.0001 compared with UK-14304-treated control. For C, G50F (TPR1), G90F (TPR2), G130F (TPR3), G270F (TPR6). For A and C, the numbers above the bars in parentheses indicate the number of AGS3-GFP-expressing cells examined. The subcellular distribution of AGS3 and its interaction with G may be influenced by AGS3 phosphorylation (26, 63). How- ever, a similar agonist-induced translocation of AGS3 to the GA was observed for an AGS3 construct in which 24 serine or thre- onine residues in the GPR domain were mutated to alanine to disrupt phosphorylation (Fig. 6C) (26). The agonist-induced translocation of AGS3 to the GA was also not altered by the MEK inhibitor U0126 (Fig. 6C). Regulated Trafficking of AGS3 to the Golgi FIGURE 7. Influence of the TPR domain of AGS3 on the agonist-induced translocation of AGS3 to the Golgi apparatus. A, left, schematic represen- tation of AGS proteins. AGS3-Q/A is an AGS3 mutant with single residue mutations in each of the four GPR motifs (GPR1-Q488A, GPR2-Q541A, GPR3- Q589A, and GPR4-Q623A) that disrupt Gi binding to the individual GPR motif. Right, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localizationwithGM-130.HEK-293cellsexpressingpcDNA3::Gi3(750ng), pcDNA3::2A/D-AR (750 ng) and pEGFP::AGS3 (25 ng), pEGFP::AGS3-Q/A (25 ng), pEGFP::AGS4 (25 ng), or pEGFP::AGS3-SHORT (AGS3-SH-GFP) (25 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” ,p0.0001comparedwiththecontrolvalueintheabsenceofagonist.Band C, quantitative analysis of cells exhibiting juxtanuclear AGS3-GFP co-localiza- tion with GM-130. HEK-293 cells expressing pcDNA3::Gi3 (750 ng), pcDNA3::2A/D-AR (750 ng), and pEGFP::AGS3-WT or AGS3 constructs (25 ng) with progressive deletion of individual TPR motifs (B) or with single amino acid mutations of conserved residues within individual TPR motifs (C) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” GM130 (red) was detected by immunofluorescence. In B, 800 AGS3-GFP-ex- pressing cells were examined for subcellular distribution determination for eachAGS3-GFPconstruct.DataareexpressedasmeansS.E.(errorbars)(n 3–5). , p 0.0001 compared with UK-14304-treated control. For C, G50F (TPR1), G90F (TPR2), G130F (TPR3), G270F (TPR6). For A and C, the numbers above the bars in parentheses indicate the number of AGS3-GFP-expressing cells examined. antagonist rauwolscine (10 M) was added for 10 min with sub- sequent processing of the cells for visualization. As indicated in Fig. 6B, the GA translocation of AGS3 was reversed by the receptor antagonist. We also asked if pertussis toxin treatment, which ADP-ribosylates Gi, effectively uncoupling it from the receptor, altered the release and translocation of AGS3. The agonist-induced translocation of AGS3 to the GA was com- pletely blocked by pertussis toxin treatment (Fig. 6B). These data are consistent with the reversibility of the effect of receptor activation on AGS3-RlucGi1-YFP BRET (Fig. 1) and its sensi- tivity to pertussis toxin (34). y p We then further investigated various signaling pathways to define the mechanisms of AGS3 translocation. The agonist- induced translocation of AGS3 to the GA was completely blocked by inhibition of phospholipase C with U73122 (Fig. 6B). The effect of U73122 was specific, because its inactive ana- log U73343 had no effect (Fig. 6B). RESULTS AND DISCUSSION 5).4 We initiated a series of experiments to understand the mech- anism of AGS3 translocation and to further test our hypothesis. We first determined the time course for the translocation of AGS3 to the GA. AGS3-GFP translocation to the GA was observed within 30 s of the agonist addition and was maintained in GA for at least 30 min (Fig. 6A). The percentages of AGS3- GFP-expressing cells exhibiting agonist-induced translocation of AGS3 to the GA at different time points were 31.6 2.7% We then asked if agonist-induced translocation of AGS3 to the GA was reversible. In this series of experiments, cells were first incubated with the 2-AR UK-14304 (1 M) for 10 min to induce AGS3 release and translocation. At this point, the 2-AR AUGUST 16, 2013•VOLUME 288•NUMBER 33 24097 JOURNAL OF BIOLOGICAL CHEMISTRY 24097 AUGUST 16, 2013•VOLUME 288•NUMBER 33 STRY 24097 Regulated Trafficking of AGS3 to the Golgi We thus addressed the role of individual TPR motifs in GA translocation of AGS3 by deter- 24098 24098 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 33•AUGUST 16, 2013 Regulated Trafficking of AGS3 to the Golgi FIGURE 8. Subcellular distribution of endogenous AGS3 after receptor activation. HEK-293 cells expressing pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” The images shown are representative of over five separate experiments. Endogenous AGS3 was detected with affinity-purified antibody generated against a GST-AGS3 fusion protein encoding the GPR domain (Ala461–Ser650) of AGS3. GM130 was detected by immunofluorescence as described under “Experimental Procedures.” The arrowheads indicate juxtanuclear localization of AGS3 following receptor activation. The relative subcellular distributions of endogenous AGS3 and GM130 were quantified by fluorescence intensity scans (green, AGS3-GFP; red, GM130) through the cell as indicated by the lines in the images and as described under “Experimental Procedures.” FIGURE 8. Subcellular distribution of endogenous AGS3 after receptor activation. HEK-293 cells expressing pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or UK-14304 (10 M) for 10 min and processed for fluorescent microscopy as described under “Experimental Procedures.” The images shown are representative of over five separate experiments. Endogenous AGS3 was detected with affinity-purified antibody generated against a GST-AGS3 fusion protein encoding the GPR domain (Ala461–Ser650) of AGS3. GM130 was detected by immunofluorescence as described under “Experimental Procedures.” The arrowheads indicate juxtanuclear localization of AGS3 following receptor activation. The relative subcellular distributions of endogenous AGS3 and GM130 were quantified by fluorescence intensity scans (green, AGS3-GFP; red, GM130) through the cell as indicated by the lines in the images and as described under “Experimental Procedures.” for engagement with regulatory binding partners for AGS3 in the GA or in the context of the transport process. protein trafficking from the GA to the plasma membrane. siRNA-mediated knockdown of AGS3 also led to dispersal of TGN-localized proteins, including TGN46 and the mannose 6-phosphate receptor. These data suggest that AGS3 dynami- cally regulates events at the TGN. We next asked if endogenous AGS3 also translocated to the GA. For endogenous AGS3, 200 cells were examined in each of three experiments for localization of AGS3 and the GA marker protein GM130 with and without agonist treatment (Fig. 8). Regulated Trafficking of AGS3 to the Golgi Upon receptor activation, endogenous AGS3 was clearly translocated to a juxtanuclear region, where it also local- ized with GM130 (Fig. 8). Each of the fluorescence intensity scans for Fig. 8 actually spans two cells, one of which exhibited AGS3 translocation to the GA and one of which did not. Of the 735 cells examined, 47.4 2.4% exhibited GA localization of AGS3 following agonist treatment. The colocalization of endogenous AGS3 with GM130 was also examined by quanti- tative image analysis. The thresholded Pearson’s correlation coefficient for AGS3 and GM130 in responsive cells was 0.9 0.02 (n 10). We thus examined the functional consequences of agonist- induced translocation of endogenous AGS3 to the GA by mon- itoring the distribution of the specific trans-Golgi-resident pro- tein sialyltransferase tagged with GFP (ST-GFP). In the absence of agonist, ST-GFP was tightly focused in the GA. However, agonist treatment (3-h incubation) disrupted the focused local- ization of ST-GFP with apparent vesiculation and fragmenta- tion of the trans-Golgi network (Fig. 9, A and B). We also exam- ined the distribution of AGS3 and the GA marker proteins GM130 and TGN46 following incubation of cells with agonist for 3 h. In general, the juxtanuclear localization of endogenous AGS3 or transfected AGS3-GFP was also observed following a 3-h incubation with agonist (Fig. 10). However, in some instances, the juxtanuclear localization of AGS3 following lon- ger incubation with agonist was more heterogeneous and not as compact as observed with shorter incubation times, consistent with an effect on overall all GA architecture. In addition, the immunofluorescence intensity for TGN46 in the juxtanuclear region was consistently lower in cells exhibiting apparent GA accumulation of AGS3 following a 3-h agonist treatment (Fig. 10). This is most evident by comparing the intensity of jux- tanuclear TGN46 in cells exhibiting apparent juxtanuclear accumulation of AGS3 with surrounding cells in Fig. 10A. The magnitude of the decrease in immunofluorescence intensity for These data support the previous observations that the AGS3- Gi signaling module at the cell surface senses receptor activa- tion, and then the two binding partners dissociate (34). It is not known if the receptor directly couples to AGS3-Gi or whether the coupling is indirect within the context of a larger signaling complex (see discussion in Ref. 34). Subsequent to release from the cell cortex, AGS3 is found in the GA. Regulated Trafficking of AGS3 to the Golgi g To further dissect the mechanism of agonist-induced trans- location of AGS3 to the GA, we examined the role of the AGS3 TPR and GPR domains. Both the TPR and GPR domains of AGS3 are important determinants of its subcellular distribu- tion (34, 38, 43). Interaction of Gi with the GPR domains sta- bilizes AGS3 and other GPR proteins at the cell cortex, where it senses activation of a cell surface receptor (34–36). We first asked if GPR proteins lacking TPR domains (Fig. 7A) exhibited a similar agonist-induced translocation to the GA. The tran- script variant AGS3-SHORT lacks the TPR domain but con- tains three GPR motifs (43). AGS4 is a group II AGS protein that has three GPR motifs and lacks a TPR domain (44). Although both AGS3-SH and AGS4 interact with Gi and are stabilized at the cell cortex, where they sense receptor activa- tion (34, 35), neither protein was translocated to the GA upon receptor activation (Fig. 7A). Agonist-induced translocation of AGS3 to the GA was not observed for AGS3-Q/A-GFP (Figs. 7A), which cannot bind Gi and thus is not stabilized at the cell surface, where it can sense receptor activation (34, 38). mining the subcellular distribution of AGS3-GFP constructs lacking individual TPR motifs. AGS3 lacking TPR1 exhibited agonist-induced translocation similar to wild type AGS3. How- ever, deletion of both TPR1 and TPR2 markedly reduced the agonist-induced accumulation of AGS3 in the GA (Fig. 7B). Similar results were observed with the progressive deletion of TPR3–7 (Fig. 7B). These data suggested that TPR2 or perhaps broader structural organization within the TPR domain was a key factor in the regulated translocation of AGS3. To further address this question, we generated AGS3-GFP constructs with single amino acid mutations of conserved residues within indi- vidual TPR motifs (38). Mutation of a single conserved G within TPR2 (G90F), but not TPR1 (G50F), TPR3 (G130F), or TPR6 (G270F), markedly reduced the agonist-induced accumulation of AGS3 in the GA (Fig. 7C). Thus, TPR2 appears to be a key site As noted above, neither AGS4 or AGS3-SHORT translo- cated to the GA upon receptor activation, suggesting that the TPR domain was required for accumulation of AGS3 at the GA upon the addition of agonist. Regulated Trafficking of AGS3 to the Golgi Receptor-mediated translocation of AGS3 is of great interest because it may have functional consequences with respect to protein secretion and trafficking through the GA, as suggested earlier by Groves et al. (18). Groves et al. reported that AGS3 overex- pression redistributed TGN-localized proteins and modulated AUGUST 16, 2013•VOLUME 288•NUMBER 33 24099 JOURNAL OF BIOLOGICAL CHEMISTRY 2409 AUGUST 16, 2013•VOLUME 288•NUMBER 33 Regulated Trafficking of AGS3 to the Golgi FIGURE 9. AGS3 translocation to the Golgi apparatus and distribution of the trans-Golgi protein ST-GFP. HEK-293 cells expressing pEGFP::ST (sialyl- transferase) (ST-GFP) (50 ng) alone or together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with agonist UK-14304 (10 M) for 3 h and processed for fluorescent microscopy as described under “Experimental Procedures.” The images shown are representative of 3–5 sep- arate experiments. B, quantitative analysis of percentage of the cells with the indicated vesicle numbers. Data are expressed as means S.E. (error bars) (n 3). , p 0.0001 compared with control value in the absence of agonist The numbers above the bars indicate the number of ST-GFP-expressing cells examined. tion of the TGN or interference with the recycling endosome pathway, the accumulation of TGN46 at the trans-Golgi net- work is not observed because the protein is apparently dis- persed through the cell, resulting in a loss of the robust immu- nofluorescence signal at the TGN. The reduction in the accumulation of TGN46 at the TGN following 3-h incubation with the 2-AR agonist UK-14304 probably reflects an effect on vesicle cycling events at the TGN rather than a reduction of TGN46 protein in the cell. (TGN46 dispersal also occurs with cell division, brefeldin A, and other interventions that disrupt the TGN. TGN46 immunoblots from vehicle- or agonist- treated cells indicated no marked change in the levels of TGN46 following agonist treatment, consistent with its apparent dis- persal rather than actual loss of the protein).4 These data are consistent with the hypothesis that the AGS3Gi complex acts as a signal transducer mediating down- stream functional responses to activation of cell surface recep- tors. A primary downstream target of AGS3 is the trans-Golgi network, where it appears to modulate protein secretion and/or endosome recycling. The first suggestion of a direct effect of AGS3 on GA function was the report by Groves et al. Regulated Trafficking of AGS3 to the Golgi (18), indicating that either a reduction or an increase in AGS3 expression resulted in altered GA structural organization. In addition, Groves et al. (18) reported regulation of the trafficking of a specific subgroup of proteins to the cell surface following siRNA-mediated reduction of AGS3. These data and the pres- ent work suggest that the receptor activation-induced dissoci- ation of an AGS3-Gi complex with the subsequent docking of AGS3 at the GA provides an unexpected platform for regula- tion of events at the GA. However, although altered AGS3 expression may influence GA structure and/or the movement of proteins through the secretory pathway, the observed effect of agonist on the TGN organization reported in the present paper may not be solely due to AGS3 translocation to the GA. The regulation of protein processing and movement in the secretory pathway is complex and involves multiple regulatory mechanisms and checkpoints. Specific subtypes of Gi and G and various proteins that interact and regulate G-proteins are found in the GA, where they may influence basal and regulated processing of proteins through the secretory pathway (45–61). Several different fac- tors may act in concert to influence membrane dynamics and vesicle recycling, sorting, and budding at the TGN. FIGURE 9. AGS3 translocation to the Golgi apparatus and distribution of FIGURE 9. AGS3 translocation to the Golgi apparatus and distribution of the trans-Golgi protein ST-GFP. HEK-293 cells expressing pEGFP::ST (sialyl- transferase) (ST-GFP) (50 ng) alone or together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with agonist UK-14304 (10 M) for 3 h and processed for fluorescent microscopy as described under “Experimental Procedures.” The images shown are representative of 3–5 sep- arate experiments. B, quantitative analysis of percentage of the cells with the indicated vesicle numbers. Data are expressed as means S.E. (error bars) (n 3). , p 0.0001 compared with control value in the absence of agonist. The numbers above the bars indicate the number of ST-GFP-expressing cells examined. FIGURE 9. AGS3 translocation to the Golgi apparatus and distribution of the trans-Golgi protein ST-GFP. HEK-293 cells expressing pEGFP::ST (sialyl- transferase) (ST-GFP) (50 ng) alone or together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with agonist UK-14304 (10 M) for 3 h and processed for fluorescent microscopy as described under “Experimental Procedures.” The images shown are representative of 3–5 sep- arate experiments. Regulated Trafficking of AGS3 to the Golgi B, quantitative analysis of percentage of the cells with the indicated vesicle numbers. Data are expressed as means S.E. (error bars) (n 3). *, p 0.0001 compared with control value in the absence of agonist. The numbers above the bars indicate the number of ST-GFP-expressing cells examined. y g g g Activation of M2 and M3 muscarinic receptors, which cou- ple to heterotrimeric Gi/o and Gq, respectively, leads to translo- cation of specific G subtypes to the GA as G complexes (57). As reported for AGS3 in the present paper, the translocation of G to the GA leads to alterations in TGN organization and the processing of proteins through the secretory pathway (57). Pro- tein kinase D and phospholipase C, both of which are acti- vated by G, play central roles in transport vesicle budding at the trans-Golgi network. G-mediated regulation of protein kinase D leads to GA fragmentation (49, 59–61). The G- mediated regulation of GA fragmentation may involve activa- tion of phospholipase C as it is blocked by U73122 (57, 60, 61). It is not clear if the phospholipase C involved resides at the plasma membrane or within the GA, but the diacylglycerol gen- erated upon its activation is postulated to be important for pro- TGN46 in cells exhibiting GA translocation of AGS3-GFP after a 3-h incubation with agonist varied as indicated in the different panels shown in Fig. 10A. The loss of TGN46, which is a marker of recycling endosomes, would suggest an effect of AGS3 on the cycling process and/or vesicle sorting at the TGN. In contrast to the observations with TGN46, the immunofluorescence inten- sity for GM130 was similar across all cells (Fig. 10, A and B). The distribution of TGN46 within the cell is actually in apparent constant flux as it moves within the recycling endo- some pathway. The kinetics of this flux are such that under standard conditions, the protein is accumulated at the trans- Golgi, providing a robust signal as a TGN marker. Regulated Trafficking of AGS3 to the Golgi (2008) Activator of G protein signaling 3 null mice. I. Unexpected alterations in metabolic and cardiovascular function. Endocrinology 149, 3842–3849 Acknowledgments—We thank Dr. Dzwokai Zach Ma (University of California, Santa Barbara, CA) for AGS3 antisera and the AGS3 construct containing GPR Ser/Thr mutations (26), Dr. Jennifer Lip- pincott-Schwartz (National Institutes of Health, Bethesda, MD) for pEGFP::ST plasmid, Dr. Joe Blumer for pEGFP::AGS4 and for helpful discussions, and Dr. Michel Bouvier (University of Montreal) for sup- port and collaboration on the development of the BRET assay for monitoring GPR-Gi interactions. We also thank Dr. Gregory G. Tall (University of Rochester School of Medicine and Dentistry) for pcDNA3.1::Gi1-YFP (pos. 122) plasmid, which was generated by Dr. Scott Gibson (62). We also appreciate the continued technical as- sistance provided by Heather Bainbridge. Acknowledgments—We thank Dr. Dzwokai Zach Ma (University of California, Santa Barbara, CA) for AGS3 antisera and the AGS3 construct containing GPR Ser/Thr mutations (26), Dr. Jennifer Lip- pincott-Schwartz (National Institutes of Health, Bethesda, MD) for pEGFP::ST plasmid, Dr. Joe Blumer for pEGFP::AGS4 and for helpful discussions, and Dr. Michel Bouvier (University of Montreal) for sup- port and collaboration on the development of the BRET assay for monitoring GPR-Gi interactions. We also thank Dr. Gregory G. Tall (University of Rochester School of Medicine and Dentistry) for pcDNA3.1::Gi1-YFP (pos. 122) plasmid, which was generated by Dr. Scott Gibson (62). We also appreciate the continued technical as- sistance provided by Heather Bainbridge. 16. Bowers, M. S., Hopf, F. W., Chou, J. K., Guillory, A. M., Chang, S. J., Janak, P. H., Bonci, A., and Diamond, I. (2008) Nucleus accumbens AGS3 expres- sion drives ethanol seeking through G. Proc. Natl. Acad. Sci. U.S.A. 105, 12533–12538 17. Bowers, M. S., McFarland, K., Lake, R. W., Peterson, Y. K., Lapish, C. C., Gregory, M. L., Lanier, S. M., and Kalivas, P. W. (2004) Activator of G protein signaling 3. A gatekeeper of cocaine sensitization and drug seek- ing. Neuron 42, 269–281 18. Groves, B., Gong, Q., Xu, Z., Huntsman, C., Nguyen, C., Li, D., and Ma, D. (2007) A specific role of AGS3 in the surface expression of plasma mem- brane proteins. Proc. Natl. Acad. Sci. U.S.A. 104, 18103–18108 19. Nadella, R., Blumer, J. B., Jia, G., Kwon, M., Akbulut, T., Qian, F., Sedlic, F., Wakatsuki, T., Sweeney, W. E., Jr., Wilson, P. D., Lanier, S. M., and Park, F. Regulated Trafficking of AGS3 to the Golgi by cell-surface receptors and an intracellular exchange factor. Curr. Biol. 18, 211–215 tein kinase D activation in the GA and the subsequent shuttling of proteins out of the trans-Golgi network. 10. Sato, M., Cismowski, M. J., Toyota, E., Smrcka, A. V., Lucchesi, P. A., Chilian, W. M., and Lanier, S. M. (2006) Identification of a receptor- independent activator of G protein signaling (AGS8) in ischemic heart and its interaction with G. Proc. Natl. Acad. Sci. U.S.A. 103, 797–802 Thus, AGS3, G, and G can all influence the structure and function of the GA. The role of AGS3 in these regulatory events is of particular interest. Via its GPR motifs, translocated AGS3 may regulate the interaction of Golgi-resident G with trans- located G, increasing the availability of free G for effector engagement in the GA. Alternatively, the translocated AGS3 may interact with specific TPR-binding partners in the GA to regulate GA function independent of any GA-resident G or G translocated to the GA. 11. Sato, M., Hiraoka, M., Suzuki, H., Bai, Y., Kurotani, R., Yokoyama, U., Okumura, S., Cismowski, M. J., Lanier, S. M., and Ishikawa, Y. (2011) Identification of transcription factor E3 (TFE3) as a receptor-independent activator of G16. Gene regulation by nuclear G subunit and its activa- tor. J. Biol. Chem. 286, 17766–17776 12. Takesono, A., Cismowski, M. J., Ribas, C., Bernard, M., Chung, P., Hazard, S., 3rd, Duzic, E., and Lanier, S. M. (1999) Receptor-independent activa- tors of heterotrimeric G-protein signaling pathways. J. Biol. Chem. 274, 33202–33205 The regulated positioning of AGS3 in the GA adds an unex- pected additional mechanism for dynamic and reversible regu- lation of protein trafficking through the secretory pathway. As noted above, AGS3 is implicated in various functional responses in several different tissues or cell types, suggesting that it is regulating a basic cellular process. Perhaps the regula- tion of GA dynamics and the secretory pathway is a target for AGS3 shared across multiple tissues. 13. Ponting, C. P. (1999) Raf-like Ras/Rap-binding domains in RGS12- and still-life-like signalling proteins. J. Mol. Med. 77, 695–698 14. Siderovski, D. P., Diverse´-Pierluissi, M., and De Vries, L. (1999) The GoLoco motif. A Gi/o binding motif and potential guanine-nucleotide exchange factor. Trends Biochem. Sci. 24, 340–341 g 15. Blumer, J. B., Lord, K., Saunders, T. L., Pacchioni, A., Black, C., Lazar- tigues, E., Varner, K. J., Gettys, T. W., and Lanier, S. M. REFERENCES 20. Yao, L., McFarland, K., Fan, P., Jiang, Z., Inoue, Y., and Diamond, I. (2005) Activator of G protein signaling 3 regulates opiate activation of protein kinase A signaling and relapse of heroin-seeking behavior. Proc. Natl. Acad. Sci. U.S.A. 102, 8746–8751 1. Blumer, J. B., Smrcka, A. V., and Lanier, S. M. (2007) Mechanistic path- ways and biological roles for receptor-independent activators of G-pro- tein signaling. Pharmacol. Ther. 113, 488–506 2. Hollinger, S., and Hepler, J. R. (2002) Cellular regulation of RGS proteins. Modulators and integrators of G protein signaling. Pharmacol. Rev. 54, 527–559 21. Yao, L., McFarland, K., Fan, P., Jiang, Z., Ueda, T., and Diamond, I. (2006) Adenosine A2a blockade prevents synergy between -opiate and canna- binoid CB1 receptors and eliminates heroin-seeking behavior in addicted rats. Proc. Natl. Acad. Sci. U.S.A. 103, 7877–7882 3. Regner, K. R., Nozu, K., Lanier, S. M., Blumer, J. B., Avner, E. D., Sweeney, W. E., Jr., and Park, F. (2011) Loss of activator of G-protein signaling 3 impairs renal tubular regeneration following acute kidney injury in ro- dents. FASEB J. 25, 1844–1855 22. Willard, F. S., Zheng, Z., Guo, J., Digby, G. J., Kimple, A. J., Conley, J. M., Johnston, C. A., Bosch, D., Willard, M. D., Watts, V. J., Lambert, N. A., Ikeda, S. R., Du, Q., and Siderovski, D. P. (2008) A point mutation to Gi selectively blocks GoLoco motif binding. direct evidence for GGoLoco complexes in mitotic spindle dynamics. J. Biol. Chem. 283, 36698–36710 4. Sato, M., Jiao, Q., Honda, T., Kurotani, R., Toyota, E., Okumura, S., Takeya, T., Minamisawa, S., Lanier, S. M., and Ishikawa, Y. (2009) Activa- tor of G protein signaling 8 (AGS8) is required for hypoxia-induced apo- ptosis of cardiomyocytes. Role of G and connexin 43 (CX43). J. Biol. Chem. 284, 31431–31440 23. Pattingre, S., De Vries, L., Bauvy, C., Chantret, I., Cluzeaud, F., Ogier- Denis, E., Vandewalle, A., and Codogno, P. (2003) The G-protein regulator AGS3 controls an early event during macroautophagy in human intestinal HT-29 cells. J. Biol. Chem. 278, 20995–21002 5. Blumer, J. B., Sadik Oner, S., and Lanier, S. M. (2012) Acta Physiol. 204, 202–218 24. Garcia-Marcos, M., Ear, J., Farquhar, M. G., and Ghosh, P. (2011) A GDI (AGS3) and a GEF (GIV) regulate autophagy by balancing G protein ac- tivity and growth factor signals. Mol. Biol. Cell 22, 673–686 6. Kimple, A. J., Bosch, D. E., Gigue`re, P. Regulated Trafficking of AGS3 to the Golgi Upon disrup- 24100 VOLUME 288•NUMBER 33•AUGUST 16, 2013 24100 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 33•AUGUST 16, 2013 Regulated Trafficking of AGS3 to the Golgi S3 translocation to the Golgi apparatus following 3-h incubation with vehicle or the 2-adrenergic receptor agonist UK xpressing pEGFP::AGS3 (25 ng) together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle M) for 3 h and processed for fluorescence microscopy. The Golgi marker protein TGN46 was detected by immunofluorescence. Thre e presented for cells incubated with agonist. B, HEK-293 cells expressing pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were agonist UK-14304 (10 M) for 3 h and processed for fluorescence microscopy. Endogenous AGS3 (red) and the Golgi marker prot tectedbyimmunofluorescenceasdescribedunder“ExperimentalProcedures.”Twodifferentimagefieldsarepresentedforcellsincu wer image panel was generated in parallel from HEK-293 cells transfected with pEGFP::AGS3 (25 ng) together with pcDNA3::Gi3 (7 -AR (750 ng). The images shown in A and B are representative of three separate experiments. cation to the Golgi apparatus following 3-h incubation with vehicle or the 2-adrenergic recepto pEGFP::AGS3 (25 ng) together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated nd processed for fluorescence microscopy The Golgi marker protein TGN46 was detected by immunofluor FIGURE 10. AGS3 translocation to the Golgi apparatus following 3-h incubation with vehicle or the 2-adrenergic receptor agonist UK-14304. A, HEK-293 cells expressing pEGFP::AGS3 (25 ng) together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or agonist UK-14304 (10 M) for 3 h and processed for fluorescence microscopy. The Golgi marker protein TGN46 was detected by immunofluorescence. Three different image fields are presented for cells incubated with agonist. B, HEK-293 cells expressing pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng) were incubated with vehicle or agonist UK-14304 (10 M) for 3 h and processed for fluorescence microscopy. Endogenous AGS3 (red) and the Golgi marker protein GM130 (green)weredetectedbyimmunofluorescenceasdescribedunder“ExperimentalProcedures.”Twodifferentimagefieldsarepresentedforcellsincubatedwith agonist. The lower image panel was generated in parallel from HEK-293 cells transfected with pEGFP::AGS3 (25 ng) together with pcDNA3::Gi3 (750 ng) and pcDNA3::2A/D-AR (750 ng). The images shown in A and B are representative of three separate experiments. JOURNAL OF BIOLOGICAL CHEMISTRY 24101 AUGUST 16, 2013•VOLUME 288•NUMBER 33 Regulated Trafficking of AGS3 to the Golgi (2010) Activator of G protein signaling 3 promotes epithelial cell prolifer- ation in PKD. J. Am. Soc. Nephrol. 21, 1275–1280 24102 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 33•AUGUST 16, 2013 Regulated Trafficking of AGS3 to the Golgi G., Adhikari, A., and Sprang, S. R. (2008) Ric-8A catalyzes guanine nucleotide exchange on Gi1 bound to the GPR/ GoLoco exchange inhibitor AGS3. J. Biol. Chem. 283, 23150–23160 51. Denker, S. P., McCaffery, J. M., Palade, G. E., Insel, P. A., and Farquhar, M. G. (1996) Differential distribution of subunits and subunits of heterotrimeric G proteins on Golgi membranes of the exocrine pancreas. J. Cell Biol. 133, 1027–1040 33. Oner, S. S., Maher, E. M., Gabay, M., Tall, G. G., Blumer, J. B., and Lanier, S. M. (2013) Regulation of the G-protein regulatory-Gi signaling com- plex by nonreceptor guanine nucleotide exchange factors. J. Biol. Chem. 288, 3003–3015 52. Wilson, B. S., Palade, G. E., and Farquhar, M. G. (1993) Endoplasmic reticulum-through-Golgi transport assay based on O-glycosylation of na- tive glycophorin in permeabilized erythroleukemia cells. Role for Gi3. Proc. Natl. Acad. Sci. U.S.A. 90, 1681–1685 34. Oner, S. S., An, N., Vural, A., Breton, B., Bouvier, M., Blumer, J. B., and Lanier, S. M. (2010) Regulation of the AGS3Gi signaling complex by a seven-transmembrane span receptor. J. Biol. Chem. 285, 33949–33958 35. Oner, S. S., Maher, E. M., Breton, B., Bouvier, M., and Blumer, J. B. (2010) Receptor-regulated interaction of activator of G-protein signaling-4 and Gi. J. Biol. Chem. 285, 20588–20594 53. Chatterjee, T. K., and Fisher, R. A. (2000) Cytoplasmic, nuclear, and Golgi localization of RGS proteins. Evidence for N-terminal and RGS domain se- quences as intracellular targeting motifs. J. Biol. Chem. 275, 24013– 24021 36. Vellano, C. P., Maher, E. M., Hepler, J. R., and Blumer, J. B. (2011) G protein-coupled receptors and resistance to inhibitors of cholinester- ase-8A (Ric-8A) both regulate the regulator of g protein signaling 14 RGS14Gi1 complex in live cells. J. Biol. Chem. 286, 38659–38669 54. Akgoz, M., Kalyanaraman, V., and Gautam, N. (2004) Receptor-mediated reversible translocation of the G protein complex from the plasma membrane to the Golgi complex. J. Biol. Chem. 279, 51541–51544 37. Patterson, G. H., Hirschberg, K., Polishchuk, R. S., Gerlich, D., Phair, R. D., and Lippincott-Schwartz, J. (2008) Transport through the Golgi apparatus by rapid partitioning within a two-phase membrane system. Cell 133, 1055–1067 55. Azpiazu, I., Akgoz, M., Kalyanaraman, V., and Gautam, N. (2006) G pro- tein 11 complex translocation is induced by Gi, Gq, and Gs coupling receptors and is regulated by the subunit type. Cell. Signal. 18, 1190–1200 38. Regulated Trafficking of AGS3 to the Golgi 28. Hofler, C., and Koelle, M. R. (2011) AGS-3 alters Caenorhabditis elegans behavior after food deprivation via RIC-8 activation of the neural G pro- tein Go. J. Neurosci. 31, 11553–11562 branes regulates the secretion of a heparan sulfate epithelial cells. J. Cell Biol. 114, 1113–1124 branes regulates the secretion of a heparan sulfate proteoglycan in LLC-PK1 epithelial cells. J. Cell Biol. 114, 1113–1124 branes regulates the secretion of a heparan sulfate proteoglycan in LLC-PK1 epithelial cells. J. Cell Biol. 114, 1113–1124 46. Leyte, A., Barr, F. A., Kehlenbach, R. H., and Huttner, W. B. (1992) Mul- tiple trimeric G-proteins on the trans-Golgi network exert stimulatory and inhibitory effects on secretory vesicle formation. EMBO J. 11, 4795–4804 29. Gigue`re, P. M., Laroche, G., Oestreich, E. A., Duncan, J. A., and Siderovski, D. P. (2012) Regulation of the subcellular localization of the G-protein subunit regulator GPSM3 through direct association with 14-3-3 protein. J. Biol. Chem. 287, 31270–31279 47. Schwaninger, R., Plutner, H., Bokoch, G. M., and Balch, W. E. (1992) Multiple GTP-binding proteins regulate vesicular transport from the ER to Golgi membranes. J. Cell Biol. 119, 1077–1096 30. Doherty, D., Chudley, A. E., Coghlan, G., Ishak, G. E., Innes, A. M., Lemire, E. G., Rogers, R. C., Mhanni, A. A., Phelps, I. G., Jones, S. J., Zhan, S. H., Fejes, A. P., Shahin, H., Kanaan, M., Akay, H., Tekin, M., FORGE Canada Consortium, Triggs-Raine, B., and Zelinski, T. (2012) GPSM2 mutations cause the brain malformations and hearing loss in Chudley-McCullough syndrome. Am. J. Hum. Genet. 90, 1088–1093 48. Jamora, C., Takizawa, P. A., Zaarour, R. F., Denesvre, C., Faulkner, D. J., and Malhotra, V. (1997) Regulation of Golgi structure through heterotri- meric G proteins. Cell 91, 617–626 49. Jamora, C., Yamanouye, N., Van Lint, J., Laudenslager, J., Vandenheede, J. R., Faulkner, D. J., and Malhotra, V. (1999) G-mediated regulation of Golgi organization is through the direct activation of protein kinase D. Cell 98, 59–68 31. Tall, G. G., and Gilman, A. G. (2005) Resistance to inhibitors of cholines- terase 8A catalyzes release of Gi-GTP and nuclear mitotic apparatus protein (NuMA) from NuMA/LGN/Gi-GDP complexes. Proc. Natl. Acad. Sci. U.S.A. 102, 16584–16589 50. Yamaguchi, T., Nagahama, M., Itoh, H., Hatsuzawa, K., Tani, K., and Tagaya, M. (2000) Regulation of the Golgi structure by the subunits of heterotrimeric G proteins. FEBS Lett. 470, 25–28 32. Thomas, C. J., Tall, G. Regulated Trafficking of AGS3 to the Golgi Vural, A., Oner, S., An, N., Simon, V., Ma, D., Blumer, J. B., and Lanier, S. M. (2010) Distribution of activator of G-protein signaling 3 within the aggresomal pathway. Role of specific residues in the tetratricopeptide re- peat domain and differential regulation by the AGS3 binding partners Gi and mammalian inscuteable. Mol. Cell Biol. 30, 1528–1540 56. Akgoz, M., Kalyanaraman, V., and Gautam, N. (2006) G protein com- plex translocation from plasma membrane to Golgi complex is influenced by receptor subunit interaction. Cell. Signal. 18, 1758–1768 57. Saini, D. K., Karunarathne, W. K., Angaswamy, N., Saini, D., Cho, J. H., Kalyanaraman, V., and Gautam, N. (2010) Regulation of Golgi structure and secretion by receptor-induced G protein complex translocation. Proc. Natl. Acad. Sci. U.S.A. 107, 11417–11422 39. Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012) NIH Image to ImageJ. 25 years of image analysis. Nat. Methods 9, 671–675 40. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pi- etzsch, T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B., Tinevez, J. Y., White, D. J., Hartenstein, V., Eliceiri, K., Tomancak, P., and Cardona, A. (2012) Fiji. An open-source platform for biological-image analysis. Nat. Methods 9, 676–682 58. Hewavitharana, T., and Wedegaertner, P. B. (2012) Non-canonical signal- ing and localizations of heterotrimeric G proteins. Cell. Signal. 24, 25–34 59. Dı´az Anel, A. M., and Malhotra, V. (2005) PKC is required for 12/ 32- and PKD-mediated transport to the cell surface and the organiza- tion of the Golgi apparatus. J. Cell Biol. 169, 83–91 41. Costes, S. V., Daelemans, D., Cho, E. H., Dobbin, Z., Pavlakis, G., and Lockett, S. (2004) Automatic and quantitative measurement of protein- protein colocalization in live cells. Biophys. J. 86, 3993–4003 60. Dı´az An˜el, A. M. (2007) Phospholipase C 3 is a key component in the G/PKC/PKD-mediated regulation of trans-Golgi network to plasma membrane transport. Biochem. J. 406, 157–165 42. Klein, R. R., Bourdon, D. M., Costales, C. L., Wagner, C. D., White, W. L., Williams, J. D., Hicks, S. N., Sondek, J., and Thakker, D. R. (2011) Direct activation of human phospholipase C by its well known inhibitor u73122. J. Biol. Chem. 286, 12407–12416 61. Irannejad, R., and Wedegaertner, P. B. (2010) Regulation of constitutive cargo transport from the trans-Golgi network to plasma membrane by Golgi-localized G protein subunits. J. Biol. Chem. 285, 32393–32404 62. Gibson, S. REFERENCES M., and Siderovski, D. P. (2011) Regu- latorsofG-proteinsignalingandtheirGsubstrates.Promisesandchallenges in their use as drug discovery targets. Pharmacol. Rev. 63, 728–749 25. Chauhan, S., Jelen, F., Sharina, I., and Martin, E. (2012) The G-protein regulator LGN modulates the activity of the NO receptor soluble guany- late cyclase. Biochem. J. 446, 445–453 7. Tall, G. G., Krumins, A. M., and Gilman, A. G. (2003) Mammalian Ric-8A (synembryn) is a heterotrimeric G protein guanine nucleotide exchange factor. J. Biol. Chem. 278, 8356–8362 26. Groves, B., Abrahamsen, H., Clingan, H., Frantz, M., Mavor, L., Bailey, J., and Ma, D. (2010) An inhibitory role of the G-protein regulator AGS3 in mTOR-dependent macroautophagy. PLoS One 5, e8877 8. Garcia-Marcos, M., Kietrsunthorn, P. S., Pavlova, Y., Adia, M. A., Ghosh, P., and Farquhar, M. G. (2012) Functional characterization of the guanine nucleotide exchange factor (GEF) motif of GIV protein reveals a threshold effect in signaling. Proc. Natl. Acad. Sci. U.S.A. 109, 1961–1966 27. Zheng, Z., Zhu, H., Wan, Q., Liu, J., Xiao, Z., Siderovski, D. P., and Du, Q. (2010) LGN regulates mitotic spindle orientation during epithelial mor- phogenesis. J. Cell Biol. 189, 275–288 9. Lee, M. J., and Dohlman, H. G. (2008) Coactivation of G protein signaling 24102 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 288•NUMBER 33•AUGUST 16, 2013 Regulated Trafficking of AGS3 to the Golgi K., and Gilman, A. G. (2006) Gi and G subunits both define selectivity of G protein activation by 2-adrenergic receptors. Proc. Natl. Acad. Sci. U.S.A. 103, 212–217 43. Pizzinat, N., Takesono, A., and Lanier, S. M. (2001) Identification of a truncated form of the G-protein regulator AGS3 in heart that lacks the tetratricopeptide repeat domains. J. Biol. Chem. 276, 16601–16610 63. Blumer, J. B., Bernard, M. L., Peterson, Y. K., Nezu, J., Chung, P., Dunican, D. J., Knoblich, J. A., and Lanier, S. M. (2003) Interaction of activator of G-protein signaling 3 (AGS3) with LKB1, a serine/threonine kinase in- volved in cell polarity and cell cycle progression: phosphorylation of the G-protein regulatory (GPR) motif as a regulatory mechanism for the in- teraction of GPR motifs with Gi. J. Biol. Chem. 278, 23217–23220 44. Cao, X., Cismowski, M. J., Sato, M., Blumer, J. B., and Lanier, S. M. (2004) Identification and characterization of AGS4. A protein containing three G-protein regulatory motifs that regulate the activation state of Gi. J. Biol. Chem. 279, 27567–27574 45. Stow, J. L., de Almeida, J. B., Narula, N., Holtzman, E. J., Ercolani, L., and Ausiello, D. A. (1991) A heterotrimeric G protein, Gi-3, on Golgi mem- JOURNAL OF BIOLOGICAL CHEMISTRY 24103 24103 AUGUST 16, 2013•VOLUME 288•NUMBER 33
https://openalex.org/W2905281498	https://ojs.lib.unideb.hu/IJHS/article/download/1136/1134	English	null	Farm economic evaluation of elderberry production	International journal of horticultural science	2,014	cc-by	3,416	Introduction besides the national average yield of 2 to 3 tons per hectare. In the most up-to-date farms in good years yields of even 13 to 15 tons per hectare may be realized in case of irrigation and using fertigation. Its cultivation is relatively simple, its inputs are low, and it does not have any significant production risk if the production site has been chosen properly. The elderberry production in Hungary started at the second half of the 1990’ies. Its development accelerated at the turn of the millennium, by the years 2006 to 2007 the growing area reached 2500 hectares. Its rate of development ceased this time; however, as after the extremely favourable selling prices at the beginning, the market collapsed, the price did not make even maintaining the profit to a minimum level possible. After the year 2010, because the selling prices turned to favourable again, further tendencies for the establishment occurred, thus today an orchard surface of 4500 hectares may be found in Hungary. It must be noted that the extreme wet weather of the year 2010 contributed to the normalization of the market conditions as well, because of this due to the permanent inland pressure several orchard surface was damaged or significantly suffered from thinning, which reduced the basis of production by 20 to 40%. Apáti, F. University of Debrecen, Faculty of Economics and Business, Institute of Management Scien H-4032 Debrecen, Böszörményi Street 138.; e-mail: fapati@agr.unideb.hu University of Debrecen, Faculty of Economics and Business, Institute of Management Sciences H-4032 Debrecen, Böszörményi Street 138.; e-mail: fapati@agr.unideb.hu Summary: In this present study the efficiency as well as the farm economic advantages and disadvantages of elderberry production are examined. Our objective is to determine the fact that under what conditions the elderberry production may be profitable regarding the present economic and market situations. Our analysis was carried out by a simulation model based on a farm-level data gathering in production enterprises. The total investment cost of an up-to-date, elderberry orchard of traditionally cultivated without irrigation is between 1000 to 1700 thousand HUF per one hectare and turning to productivity is expected within 4-5th years. These orchards are able to produce yields of 8,0 to 9,0 tons per one hectare in the average of the productive years, which makes reaching a revenue of 800 to 1000 thousand HUF possible regarding a per kg average selling price of 80 to 110 HUF. By this a net profit of 200 to 400 thousand HUF may be realized in case of a per hectare production cost of 600 to 700 thousand HUF. At the end of the lifetime of the orchard (12-15th year) an internal rate of return of 10 to 14%, an NPV of 1500 to 2000 thousand HUF per one hectare are typical in an average case, and the payback may be expected in the 6th to 8th year. From the farm economic aspect the elderberry may be considered as an extensive sector, which advantages are low capital and labour need, early recovery, good-acceptable profit on capital and cost to profit ratios, but its disadvantage is low per hectare profit comparing to intensive fruit species and orchards. In this way in general farms of capital-extensive and avoiding risks choose elderberry production. Keywords: elderberry, elderberry production, farm economic analysis, elderberry economics Farm economic evaluation of elderberry production Apáti, F. International Journal of Horticultural Science 2014, 20 (3–4): 57–60. Agroinform Publishing House, Budapest, Printed in Hungary ISSN 1585-0404 International Journal of Horticultural Science 2014, 20 (3–4): 57–60. Agroinform Publishing House, Budapest, Printed in Hungary ISSN 1585-0404 Materials and methods An elderberry orchard cultivated in a traditional way does not need any support system, and even irrigation is not typical in most of the orchards. The costs of the stock and planting, as well as of soil preparation constitute the two major items of the establishment costs. Comparing to other fruit species, the establishment cost is very low, it is only 1000 to 1700 thousand HUF per one hectare thank to the low number of plants, to the relatively cheap propagating material and the dispensable support system (Table 1). The economic issues of elderberry orchards in good conditions, in a traditionally cultivated system cultivated on a high standard are examined in this study. The parameters of the characterized orchard types are the following: •• trunk-stem crown form •• distance between rows is 5.0 meters, distance within a row is 3.0 meters (666 stems per one hectare) •• the ’Haschberg’ is the only one cultivar in the cultivar structure, Table 1: The Investment Cost of the Characterized Elderberry Orchard Denomination Cost (thousand HUF/ha) In an interval In average Site and soil preparation 400-800 600 Establishing the support equipment 0 0 Stock and planting 400-550 480 Establishing irrigation system 0 0 Other 200-350 270 Total establishing costs 1.000-1.700 1.350 Cultivation cost in the period of turning to productivity (3 years) 700-1.000 850 Total investment cost 1.700-2.700 2.200 Revenue in the period of turning to productivity (3 years) 400-700 550 Clear investment cost 1.300-2.000 1.650 Annual depreciation cost in the period of productivity 110-200 130 Source: own calculation Table 1: The Investment Cost of the Characterized Elderberry Orchard •• there is not any irrigation and support system, •• the technology of harvesting is carried out by hand in one or two courses, •• the reachable yield level in good years is 10 to 12 tons per one hectare, in the average of the productive years it is 8 to 9 tons per one hectare, •• the sold products are fruits in plastic bins, immediately sold after harvesting without post-harvest operations (cost of storage, selection and packaging do not arise). Such an orchard may be characterized by yields exceeding the national average, good product quality (size, ripeness), high inputs, production technology of good standard and strict technological discipline. It must be highlighted that these parameters reflect not the Hungarian average but the best orchards belonging to the upper third. Materials and methods The prices of the utilized inputs (materials, labour, mechanical work) and the prime cost reflect a price standard in 2012-2013 years. The price of materials lacks the VAT, while the wages of labour is considered altogether with benefits. The time wage was 700 HUF per hour, and it was calculated to every labour hour regardless the fact that whether paid or unpaid family labour worked. The selling prices are represented by a longer-term (3 to 5 years) average. Data gathering basing the analysis happened in production enterprises. It must be noted that during the past several years there were water-saving micro-irrigation systems (with drip or micro-spray) established in up-to-date farms which already constitutes the essential part of the technology increasing the establishment cost significantly by 700 to 1100 thousand HUF per one hectare. This is, however, expected to become a profitable and recovering investment during 2 to 4 years because of the higher yields by 25 to 30%, better yield safety, greater size of berries and panicles (lower specific picking cost) as well as better orchard condition (plantation deterioration occurs later). As long-term experiences and data-series are available on orchards which are not irrigated, the economic aspects of these kinds of orchards are examined in this study. In case of the examined model (orchard) good production standard and strict technological discipline were set, thus the calculation refers not to the national average but to up-to- date farms producing on good standard, which take up of 30 to 40% of the 4.500 hectares elderberry orchard. The utilized analysing methods are cost-benefit analysis and investment analysis on return. In the latter case the dynamic methods were chosen as they give a more valid and precise results from the professional aspects. They differ from static methods in the fact that they count with the time value of the money (Illés, 2002). There are several indicators for dynamic investment analysis, from which NPV (Net Present Value), DPP (Discounted Payback Period), IRR (Internal Rate of Return, return on capital) are calculated (Flock, 2000; Brealey, 2006). The return on capital of alternative investment is reflected by the calculative interest rate, which value is 6%, being the interest need of the investment capital. The clear investment cost is 1300 to 2000 thousand HUF per one hectare calculating even the cultivation cost and the reachable revenue of the three-year-period of turning to productivity. Results which may contribute to making good decisions for the establishment and production. which may contribute to making good decisions for the establishment and production. Objectives The main objective of this present study is to determine the fact that what cost and profit relations are typical to elderberry production, and under what conditions it may be profitable if it is characterized by a traditional cultivation, a high technological standard without irrigation, which constitutes most of our orchards yet. The specific objectives relating to this are the followings: •• investigation of inputs and production cost during the establishment and in the productive period, According to Csizmadia (2011) additionally 1000 to 1500 hectares orchard may be established in the country, as besides the present production of 8 to 10 thousand tons, further quantity of 10 to 12 thousand tons may be sold. This reason is the fact that the demand for food colorings of natural-based has been increasing all over Europe. •• determination of yield, selling price and production value,ii •• evaluation of profit generation, profitability and efficiency, •• characterization of recovery calculations on the whole lifetime of the orchard, •• determination of the major conditions of profitable elderberry production by carrying out sensitivity analysis of all these factors. The typical cultivation system of elderberry is trunk- stem crown, but the natural rounded broad crown is used more and more frequently. The very heterogenic standard of the national elderberry production may be characterized by yields of 8 to 10 tons per hectare in farms of high standard On these bases, the most important advantages and disadvantages of elderberry production are identified, Apáti, F. 58 Yield, Production Value, Profit and Profitability in the Productive Years Table 3: Revenue, Profit and Profitability of the Elderberry Orchard Cultivated on a Good Standard Relating to the Whole Productive Years, in Case of Different Yield Levels Denomination Unit Yield level (average yield of many years) 6,5 t/ha 9,0 t/ha 11,0 t/ha Yield t/ha 6,5 9,0 11,0 Average selling price HUF/kg 105,0 105,0 105,0 Total revenue thousand HUF/ha 682,5 945,0 1.155,0 Direct production cost thousand HUF/ha 658,0 695,0 702,0 Contribution thousand HUF/ha 24,5 250,0 453,0 Overhead cost thousand HUF/ha 60,0 60,0 60,0 Net profit* thousand HUF/ha -35,5 190,0 393,0 Cash flow thousand HUF/ha 94,5 260,0 523,0 Profit to cost ratio % -4,9 25,1 51,5 Source: own data gathering and calculation * profit before taxes Table 3: Revenue, Profit and Profitability of the Elderberry Orchard Cultivated on a Good Standard Relating to the Whole Productive Years, in Case of Different Yield Levels •• Depreciation and overhead cost are fixed costs; the per hectare values are independent from the yields. The per hectare production cost including depreciation and overhead cost hardly changes within the examined yield interval, regarding the fact that in this relevant yield lane, almost every cost item is fixed costs except for the cost of harvesting. •• On the basis of the facts mentioned above the prime cost per one kilogram ranges from 70 to 110 HUF depending on the yields reachable in the average of many years in case of technologies of good standard. Regarding profit and profitability, in the average of many years the elderberry produces deficit only in an extremely pessimistic case, in the average of good years it reaches or exceeds 50% profitability. Comparing to other, more intensive fruit varieties and orchards (intensive apple, semi-intensive apricot), it realizes similar parameters with respect to cost to profit ratio, at the same time regarding the absolute profit volume realized on one hectare it lags behind significantly. On the basis of its farm economic parameters, it may be classified into the extensive cultures. Yield, Production Value, Profit and Profitability in the Productive Years Table 2 summarizes the production cost of elderberry, which is calculated to several yield levels. This reason is the fact that the yield in the average of many years ranges from 6,5 to 11,0 tons per one hectare even in an orchard of good standard having a developed production technology mainly because of frost, hail and drought damages. Regarding the production cost of elderberry, the following major conclusions may be defined: The calculation of production value and profit is summarized in Table 3. The selling prices ranged within a wide interval (30 to 200 HUF/kg) during the last 10 years. In the average of many years the selling price is 105 HUF/ kg. The annual revenue is 700 to 1200 thousand HUF/ hectare depending on the yield level. The realizable revenue is around zero in case of an average yield of 6 to 7 tons per a hectare regarding average selling prices of many years. It means that it is the yield level which shows the turning point of the profitability, though cash flow still reflects a positive value. In case of an average yield of 10 tons per one hectare or higher than that the cost to profit ratio is favourable, it is about 50%, though it is not excellent in permanent cultures. •• The annual level of the production technological inputs are low thus the orchard may be cultivated with a production cost of 400 thousand HUF per one hectare. With the change of the yields, this per hectare production cost does not change basically, but its per kg value decrease proportionally. •• The cost of harvest shows an opposite tendency, its value ranges from 10 to 30 HUF/kg depending on many factors; it is 20 HUF/kg in average. The picking cost is a variable cost generally, thus the per hectare value changes proportionally with the yield, while the per kg value ranges only within a very narrow interval. Yield, Production Value, Profit and Profitability in the Productive Years Table 2: The Production Cost of the Elderberry Orchard Cultivated on a Good Standard Relating to the Whole Productive Years, in Case of Different Yield Levels Cultivation phases Yield level (average yield of many years) Yield: 6,5 t/ha Yield: 9,0 t/ha Yield: 11,0 t/ha Cost (thousand HUF/ha) Cost (HUF/kg) Cost thousand HUF/ha) Cost (HUF/kg) Cost (thousand HUF/ha) Cost (HUF/kg) Pruning 79 12,2 79 8,8 79,0 7,2 Soil and spacing cultivation 62 9,5 62 6,9 62,0 5,6 Fertilizing 104 16,0 104 11,5 104,0 9,5 Plant protection 98 15,1 98 10,9 98,0 8,9 Other cultivations and costs 42 6,5 42 4,7 42,0 3,8 Total production technological costs 385 59,3 385 42,8 385,0 35,0 Harvesting 143 22,0 180 20,0 187,0 17,0 Depreciation of the orchard 130 20,0 130 14,4 130,0 11,8 Total direct cost 658 101,3 695 77,2 702,0 63,8 Overhead cost 60 9,2 60 6,7 60,0 5,5 Total production cost 718 110,5 755 83,9 762,0 69,3 From which: operation costs 588 90,5 625 69,5 632,0 57,5 Source: own data gathering and calculation Table 2: The Production Cost of the Elderberry Orchard Cultivated on a Good Standard Relating to the Whole Productive Years, in Case of Different Yield Levels Farm economic evaluation of elderberry production 59 indicators, w Source: own data gathering and calculation Inputs and Production Cost in the Productive Period Yield, Production Value, Profit and Profitability in the Productive Years Materials and methods This is resulted in a depreciation cost of 110 to 200 thousand HUF projected to 10 to 12 productive years. All in all, elderberry may be classified into fruit varieties of low capital need at the beginning. Farm economic evaluation of elderberry production Recovery Conditions of the Investment during the Lifetime of the Orchard Due to limitation on length, we strive to summarize only the major results here. In calculating investment recovery indicators, we used a calculative interest 60 Apáti, F. rate of 6% as a basis for comparison. According to our results, regardless subsidies for investment (in case of establishment from completely own sources) and direct production subsidies, the payback period of the investment is 6 to 8 years in a realistic case, while it is 11 to 13 years in a pessimistic case, which means that the investment will recover till the end of the lifetime of the orchard. At the end of the lifetime of the orchard (15th year) the realized NPV is 1,5 to 2,0 million HUF per one hectare in a realistic case, the internal rate of return is 10 to 14% (by this it lags behind the more intensive fruit species). Regarding subsidies for investment and direct production, the payback period shortens to 5 to 7 years, while the value of the NPV improves to 2,6 to 2,8 million HUF per one hectare, and the IRR to 18 to 23%. rate of 6% as a basis for comparison. According to our results, regardless subsidies for investment (in case of establishment from completely own sources) and direct production subsidies, the payback period of the investment is 6 to 8 years in a realistic case, while it is 11 to 13 years in a pessimistic case, which means that the investment will recover till the end of the lifetime of the orchard. At the end of the lifetime of the orchard (15th year) the realized NPV is 1,5 to 2,0 million HUF per one hectare in a realistic case, the internal rate of return is 10 to 14% (by this it lags behind the more intensive fruit species). Regarding subsidies for investment and direct production, the payback period shortens to 5 to 7 years, while the value of the NPV improves to 2,6 to 2,8 million HUF per one hectare, and the IRR to 18 to 23%. The major farm economic disadvantages of elderberry are: •• low per hectare profit, significantly lagging behind the more intensive fruit cultures, by which the elderberry will be suitable for farms operating in an extensive way and taking few risks. Conclusions This research was supported by the European Union and the State of Hungary, co-financed by the European Social Fund in the framework of TÁMOP-4.2.4.A/ 2-11/1-2012- 0001 ‘National Excellence Program’. The identification of the application is: A2-MZPD-13-0358 On the basis of the results, the major farm economic advantages and disadvantages of elderberry may be summarized as follows. •• low capital need for the establishment at the beginning, •• low input and labour need in the years of productivity, •• relatively low production risk (frost, danger of hail damages, drought, discontinuous yielding etc.) comparing to other fruit species, Recovery Conditions of the Investment during the Lifetime of the Orchard •• Good profitably parameters may only be realized if yields reach or exceed 10 tons per one hectare, for which expertise and the inputs detailed above are essential, that is, elderberry cannot be handled as a modest culture. These farm economic features shows that elderberry has a relatively low capital and input need, small production risk and is a more extensive fruit variety when comparing to other fruit species, which is able to realize a not so high profit, but a favourable profitability in general due to the low production costs. References Brealey, R. A., Myers, S. C. & Allen, F. (2006): Corporate Finance. Eight Edition, McGraw-Hill, 84-143. p., 244-272. p. Brealey, R. A., Myers, S. C. & Allen, F. (2006): Corporate Finance. Eight Edition, McGraw-Hill, 84-143. p., 244-272. p. Brealey, R. A., Myers, S. C. & Allen, F. (2006): Corporate Finance. Eight Edition, McGraw-Hill, 84-143. p., 244-272. p. Flock, C. (2000): Betriebszweigabrechnung in der Landwirtschaft. Verlag Pflug und Feder. Sankt Augustin, 11-158. p. Illés M. (2002): A beruházások gazdaságossága. [In: Vezetôi gazdaságtan.] Szerk.: Illés M. Kossuth Kiadó,115-162. p. •• the relatively low profit may be acceptable and it may result in a satisfactory and good profit to cost ratio because of low production costs, Flock, C. (2000): Betriebszweigabrechnung in der Landwirtschaft. Verlag Pflug und Feder. Sankt Augustin, 11-158. p. Flock, C. (2000): Betriebszweigabrechnung in der Landwirtschaft. Verlag Pflug und Feder. Sankt Augustin, 11-158. p. •• in case of total yield damage, the realized loss is low, because the level of production technological inputs is low, Illés M. (2002): A beruházások gazdaságossága. [In: Vezetôi gazdaságtan.] Szerk.: Illés M. Kossuth Kiadó,115-162. p. Illés M. (2002): A beruházások gazdaságossága. [In: Vezetôi gazdaságtan.] Szerk.: Illés M. Kossuth Kiadó,115-162. p. •• presently good market position, favourable selling opportunities. Csizmadia Gy. (2011): Bodza – 800 ezer Ft a bevétel. Haszon. Agrár Magazin, Budapest, 6. p.
https://openalex.org/W1599828951	https://seer.ufu.br/index.php/biosciencejournal/article/download/22212/15855	Portuguese	null	Crescimento da cultura da cenoura após aplicações de resíduos de curtume e carboniferos no solo	Bioscience journal	2,015	cc-by	6,592	INTRODUÇÃO substâncias químicas e da geração de grande quantidade de efluentes líquidos e de resíduos sólidos (KRAY et al., 2008). A serragem cromada é obtida na etapa de rebaixamento de couro e representa 75% dos resíduos sólidos gerados. É um resíduo sólido altamente tóxico, pois está impregnado de sais de cromo, sendo classificada como Resíduo de Classe I, segundo a NBR-100004, (CLASS; MAIA, 1994), e sua aplicação deve levar em conta as diretrizes descritas na Política Nacional de Resíduos Sólidos. A disposição correta dos resíduos de curtume tem causado discordância entre as indústrias e os órgãos ambientais. O uso do solo para o descarte destes resíduos tem se mostrado, em outros estudos, uma alternativa viável devido à sua capacidade de degradar, complexar e inativar os componentes presentes nestes materiais (KRAY et al., 2008; GIANELLO et al., 2011; SEGATO et al., 2012). A presença de nutrientes e/ou a capacidade de neutralização da acidez de alguns resíduos têm mostrado benefícios ao solo e às plantas (FERREIRA et al., 2003; KRAY et al., 2008; SEGATO et al., 2012; LAUSCHNER et al., 2012); entretanto, a presença de metais pesados, entre outros fatores, pode tornar limitante à sua utilização. Apesar dos inúmeros estudos existentes quanto ao comportamento do cromo, não há registros de estudos que apresentam a recuperação de todo o cromo contido no material descartado no solo. Kray et al. (2008), adicionando ao solo lodo de curtume e serragem cromada determinaram recuperações de cromo que variaram de 23 a 43%. Castilhos (1998), também estudando o efeito da aplicação de lodo de curtume ao solo, recuperou valores de 52, 67 e 88% do cromo adicionado ao solo, em três diferentes solos do Estado do Rio Grande do Sul, após 70 dias da aplicação. Ainda, o cromo pode ficar indisponível ligado ao solo, e as plantas podem precipitar, complexar e não absorver o cromo existente no solo, mesmo que altas concentrações sejam adicionadas ao mesmo. Assim, objetivou-se avaliar o efeito residual de adições De acordo com estudo realizado por Kray et al. (2008), o lodo de curtume pode ser utilizado para a correção do pH de solos ácidos e como fonte de nitrogênio para as plantas. Além disso, a frequência de aplicação e as doses utilizadas são limitadas pelo valor de neutralização da acidez, concentração de sais (principalmente de sódio) e quantidades de metais pesados presentes nos resíduos. 127 127 Original Article CRESCIMENTO DA CULTURA DA CENOURA APÓS APLICAÇÕES DE RESÍDUOS DE CURTUME E CARBONIFEROS NO SOLO GROWTH OF CARROT CROP AFTER APPLICATION OF TANNERY WASTE AND COAL MINING IN SOIL Maurizio Silveira QUADRO1; Marino José TEDESCO2; Clesio GIANELLO2; Amauri Antunes BARCELOS1; Robson ANDREAZZA1; Leandro BORTOLON3 1. Centro de Engenharias; Universidade Federal de Pelotas; Pelotas, RS, Brasil. robsonandreazza@yahoo.com.br; 2. Departamento de Solos, Faculdade de Agronomia; Universidade Federal do Rio Grande do Sul; Porto Alegre, RS, Brasil; 3. Embrapa Pesca e Aquicultura; EMBRAPA; TO, Brazil. RESUMO: Com a industrialização, a produção de resíduos tem aumentado durante os anos. Além disso, a disposição destes resíduos é uma posição discutida entre os órgãos ambientais. Em virtude disto, o objetivo foi estudar o efeito residual de adições sucessivas de resíduos de curtume e carboníferos sobre as propriedades químicas do solo e o acumulo de metais pesados nas plantas de cenoura. Os tratamentos aplicados a campo foram T1 = Controle, somente solo; T2 = Adubação com NPK + calcário para atingir pH 6,0; T3 = Lodo de curtume em quantidade adequada para atingir pH 6,0 + PK; T4 = Duas vezes a quantidade de lodo de curtume utilizada no tratamento 3 + PK; T5 = Resíduo carbonífero + NPK + calcário em quantidade adequada para atingir pH 6,0; T6 = Resíduo carbonífero + lodo de curtume em quantidade adequada para atingir pH 6,0 + PK; T7 = Serragem cromada + NPK + calcário em quantidade adequada para atingir pH 6,0; T8 = Cr mineral + lodo de curtume em quantidade adequada para atingir pH 6,0 + PK. O experimento foi conduzido em delineamento inteiramente casualizado, com três repetições por tratamento. A adição de cromo, tanto via mineral, quanto via resíduos não afetaram o crescimento da cenoura. Além disso, os teores encontrados na parte aérea, radicular, e córtex da raiz são considerados baixos, demonstrando um baixo potencial de contaminação destes resíduos. PALAVRAS-CHAVES: Descarte de resíduos. Daucus carota. Cromo. Absorção de metais pesados. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 INTRODUÇÃO O processo de transformar pele crua ou salgada em couro é considerado altamente poluidor devido a utilização de um grande volume de água, Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 Received: 25/03/13 Accepted: 07/07/14 128 QUADRO, M. S. et al. Crescimento da cultura da cenoura... para atingir pH 6,0; T3 = Lodo de curtume em quantidade adequada para atingir pH 6,0 + PK; T4 = Duas vezes a quantidade de lodo de curtume utilizada no tratamento 3 + PK; T5 = Resíduo carbonífero + NPK + calcário em quantidade adequada para atingir pH 6,0; T6 = Resíduo carbonífero + lodo de curtume em quantidade adequada para atingir pH 6,0 + PK; T7 = Serragem cromada + NPK + calcário em quantidade adequada para atingir pH 6,0; T8 = Cr mineral + lodo de curtume em quantidade adequada para atingir pH 6,0 + PK. A área experimental de onde foi coletado o solo estava em pousio desde a safra 1999/2000 até a data de implantação do experimento (11/2005). A quantidade de resíduos e materiais aplicados na área experimental é apresentada na Tabela 1. A adubação de NPK foi adicionada em cada tratamento de acordo com as recomendações estabelecidas pela COMISSÃO DE QUÍMICA E FERTILIDADE DO SOLO (2004) para a cultura da cenoura. sucessivas de resíduos de curtume e carboníferos sobre as propriedades químicas do solo e o acumulo de metais pesados nas plantas de cenoura. MATERIAL E MÉTODOS Foi iniciado em 1996 um experimento de campo de longa duração na Estação Experimental Agronômica da UFRGS (EEA/UFRGS), no município de Eldorado do Sul (RS), região fisiográfica da Depressão Central, nas coordenadas geográficas 30º05’76’’ S de latitude e 51º40’67’’ W de longitude. O solo da área experimental é classificado como ARGISSOLO Vermelho distrófico típico (STRECK et al., 2002), com declividade do terreno menor que 5%. Na área experimental foram aplicados resíduos de curtume (lodo de estações de tratamento e serragem cromada) e de mineração de carvão (carbonífero) nas safras agrícolas de 1996/97 e 1999/2000. Os tratamentos aplicados a campo foram T1 = Controle, somente solo; T2 = Adubação com NPK + calcário Tabela 1. Tratamentos, quantidades de materiais e de cromo aplicados na 1ª aplicação (12/1996) e na 2ª aplicação (01/2000). 1ª Aplicação 2ª Aplicação Cr total adicionado Tratamentos † Resíduos Cr Resíduos Cr t ha-1 kg ha-1 t ha-1 kg ha-1 kg ha-1 T1 - - - - - T2 3,4 ca* - 6,3 ca - - T3 21,25 LC 172 (2) 22,4 LC 498(2) 670 T4 42,5 LC 343 (2) 44,8 LC 994(2) 1336 T5 106,0 RC - 56,0 RC 20,0 ca - - T6 106,0 RC 21,25 LC 172 (2) 56,0 RC 34,4 LC 764(2) 936 T7 29,4 SC 3,4 ca 617 (2) 30,0 SC 8,6 ca 720(2) 1337 T8 21,24 LC 172 (2) 125 (1) 22,4 LC 498 (2) 500 (3) 1295 † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; ca significa calcário necessário para atingir pH 6,0; LC significa lodo de curtume; RC significa resíduo carbonífero; SC significa serragem cromada;(1)Na forma de Cr2(SO4)3, (2)Adicionado pelo resíduo. (3)Adicionado na forma mineral. Tabela 1. Tratamentos, quantidades de materiais e de cromo aplicados na 1ª aplicação (12/1996) e na 2ª aplicação (01/2000). Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 MATERIAL E MÉTODOS As características físico-químicas dos resíduos e do solo foram realizadas de acordo com as metodologias descritas por TEDESCO et al. (1995). As características físico-químicas dos resíduos são apresentadas na Tabela 2, e do solo na Tabela 3. A camada superficial do solo (0-20 cm) de cada tratamento, da área experimental foi coletado, com pá-de-corte, tamisado e acondicionados em vasos de PVC. Os vasos com capacidade de 10,5 L (cilindros de PVC com 20 cm de diâmetro e 33 cm de altura, com a parte inferior fechada por um disco de madeira revestida por resina epóxi) tendo um Tabela 3. Características físico-químicas da camada superficial (zero a 20 cm) do solo do experimento de longa duração após a aplicação dos diferentes tratamentos. Tratamentos1 pH (H2O) N Total P disp. K disp. M.O. Al troc. Ca troc. Mg troc. 1:1 g kg-1 --- mg dm-3 -- g kg-1 -------- cmolc dm-3 ------- T1 4,8 0,9 1,8 118 24,5 0,8 1,9 1,1 T2 6,4 1,0 3,9 125 25,3 0,0 4,8 2,3 T3 6,1 1,1 5,2 120 27,1 0,0 5,1 1,4 T4 6,6 1,2 7,8 130 27,6 0,0 6,4 1,7 T5 4,9 0,9 3,0 78 29,3 0,4 4,4 1,7 T6 4,7 1,4 6,2 77 32,1 0,7 4,0 0,9 T7 5,7 2,0 3,1 97 33,1 0,0 3,6 1,8 T8 5,9 1,2 5,9 154 29,3 0,0 4,8 1,1 1T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK. bela 3. Características físico-químicas da camada superficial (zero a 20 cm) do solo do exp longa duração após a aplicação dos diferentes tratamentos. rísticas físico-químicas da camada superficial (zero a 20 cm) do solo do experimento de duração após a aplicação dos diferentes tratamentos. MATERIAL E MÉTODOS † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; ca significa calcário necessário para atingir pH 6,0; LC significa lodo de curtume; RC significa resíduo carbonífero; SC significa serragem cromada;(1)Na forma de Cr2(SO4)3, (2)Adicionado pelo resíduo. (3)Adicionado na forma mineral. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 Tabela 2. Características físico-químicas dos resíduos utilizados neste trabalho. Parâmetro Unidade Lodo de curtume Serragem cromada Resíduo carbonífero pH em água 6,7 4,5 7,0 Carbono orgânico g kg-1 239,3 331,0 184,9 Nitrogênio total g kg-1 36,5 172,9 2,5 Relação C/N 6,5 1,9 74,0 Tabela 2. Características físico-químicas dos resíduos utilizados neste trabalho. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 129 Crescimento da cultura da cenoura... QUADRO, M. S. et al. NH4 + mg L-1 4,0 ND (1) ND NO3 - + NO2 - mg L-1 1,9 ND ND Fósforo total g kg-1 1,1 0,3 0,1 Potássio total g kg-1 0,1 0,1 3,7 Cálcio total g kg-1 41 8,5 1,1 Magnésio total g kg-1 7,8 0,75 1,2 Enxofre total g kg-1 18 22 25 Cobre total mg kg-1 8,27 30 15 Zinco total mg kg-1 112 0 48 Sódio total g kg-1 9,2 3,2 0,16 Cromo total g kg-1 34 20 0,11 Cádmio total mg kg-1 0,18 0,03 15,3 Níquel total mg kg-1 5,47 8,6 24 Chumbo total mg kg-1 11,2 1,39 8,72 Poder de neutralização % 10% ND ND (1)ND significa não determinado forma. Crescimento da cultura da cenoura... QUADRO, M. S. et al. orifício lateral para drenagem do excesso de água da chuva. As características físico-químicas dos resíduos e do solo foram realizadas de acordo com as metodologias descritas por TEDESCO et al. (1995). As características físico-químicas dos resíduos são apresentadas na Tabela 2, e do solo na Tabela 3. orifício lateral para drenagem do excesso de água da chuva. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 MATERIAL E MÉTODOS ) , ; ) ç p g p , ( ); ) ( ) q adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK. Após a caracterização do solo, foram acondicionados 9 kg de solo (seco ao ar), por vaso, com três repetições por tratamento. Foram mantidos os tratamentos especificados, sem reaplicação dos resíduos na área experimental não coberta. Posteriormente, foi aplicada adubação mineral baseado nas análises do solo e interpretação e as recomendações conforme a COMISSÃO... (2004) para a cultura da cenoura. A cultivar Brasília foi semeada, sendo mantidas três plantas por vaso até o final do experimento. A irrigação foi realizada com água destilada para manter o solo com 75% da capacidade de campo. As plantas foram colhidas aos 90 dias, separando-se a parte aérea das raízes. Estas foram lavadas com água destilada e quantificadas a massa úmida radicular, e após secagem em estufa a 65ºC, a massa seca da parte aérea também foi determinada. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 130 QUADRO, M. S. et al. QUADRO, M. S. et al. Crescimento da cultura da cenoura... As aplicações de resíduos contendo cromo, no solo não reduziram a produção de massa úmida e seca radicular, e a massa seca da parte aérea das plantas (Tabela 4). O tratamento controle (T1) não apresentou produção devido ao baixo teor de fósforo no solo (Tabela 2). maior produção de massa foi obtida no tratamento com adição de serragem cromada. Entretanto, os rendimentos entre os tratamentos, foram semelhantes ao tratamento controle T2. Deste modo, não foram verificados efeitos nocivos de aplicações anteriores de resíduos de curtume, sobre a cultura da cenoura. Outros estudos com a aplicação de resíduos de curtume na cultura do rabanete não apresentaram diferenças significativas entre tratamentos, devido a baixa translocação do metal na planta (Castilhos, 1998; Domaszak, 2000). Isto pode ser devido ao potencial de redução e complexação do cromo no solo, não ficando em formas tóxicas ou biodisponíveis as plantas. MATERIAL E MÉTODOS O córtex radicular (com aproximadamente 1 mm de espessura) foi removido, sendo seco em estufa a 65ºC, juntamente com a parte interna e a parte aérea. Foram determinados os teores de N, P, K, Ca, Mg e Cr extraídos na parte aérea e parte interna das raízes, e os teores de Cr no córtex radicular das plantas, conforme metodologia descrita por Tedesco et al. (1995). O córtex radicular (com aproximadamente 1 mm de espessura) foi removido, sendo seco em estufa a 65ºC, juntamente com a parte interna e a parte aérea. Foram determinados os teores de N, P, K, Ca, Mg e Cr extraídos na parte aérea e parte interna das raízes, e os teores de Cr no córtex radicular das plantas, conforme metodologia descrita por Tedesco et al. (1995). O experimento foi conduzido em delineamento inteiramente casualizado, com três repetições por tratamento. A análise estatística dos dados foi feita com o software de análise estatística WINSTAT (Machado, 2001), utilizando-se a análise da variância (teste F) conforme recomendações de Silva (1997), e as diferenças significativas foram determinadas pelo teste de comparação múltiplas de Tukey. Produção de massa Tabela 4. Matéria úmida (MU) e matéria seca (MS) das raízes e da parte aérea das plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. Tratamentos† MU Raiz MS Raiz MS parte aérea ---------------------------- g vaso-1 ------------------------------ T1** 0,0 0,0 0,0 T2 126,1 a* 17,2 a 12,8 a T3 123,6 a 16,9 a 12,8 a T4 118,1 a 18,7 a 12,2 a T5 122,6 a 16,8 a 13,7 a T6 90,2 a 12,4 a 9,5 a T7 145,5 a 19,7 a 15,6 a T8 116,0 a 14,8 a 13,6 a CV (%) 20 17 18 † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade de erro. Sem crescimento de plantas. ria úmida (MU) e matéria seca (MS) das raízes e da parte aérea das plantas de cenoura após 90 de crescimento nos diferentes tratamentos. † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade de erro. Sem crescimento de plantas. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 Teores de macronutrientes (N, P, K, Ca e Mg) O teor de fósforo na parte aérea da cenoura não diferiu estatisticamente do T7, entretanto o primeiro foi maior no tratamento controle positivo (T2), do que nos tratamentos com aplicações anteriores de lodo de curtume (T3 e T4). Os teores de Mg na parte aérea da cenoura foram maiores nos tratamentos com adição de calcário (T2; T5 e T7), o qual foi utilizado o calcário com teores de ±12% de MgO apresentando este elemento em sua composição. Isto pode ter ocorrido pelo baixo teor desse nutriente nos resíduos utilizados neste experimento. Após 90 dias de crescimento das plantas de cenoura não foram observadas diferenças significativas entre os tratamentos para os teores de N, K e Ca na parte aérea (Tabela 5). Os teores médios de N, P, K, Ca e Mg na parte aérea da cenoura foram: 16,9; 2,9; 20,2; 28,4 e 5,6 g kg-1, respectivamente. Conforme a COMISSÃO...(2004), os teores de N e K são considerados baixos, podendo ter limitado o crescimento das plantas, contudo, foram semelhantes ao tratamento controle T2. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 131 Crescimento da cultura da cenoura... QUADRO, M. S. et al. Tabela 5. Teores de macronutrientes na parte aérea das plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. Teores de macronutrientes (N, P, K, Ca e Mg) Tratamentos† N P K Ca Mg ------------------------------------g kg-1----------------------------- T1 - - - - - T2 15,7 a* 4,3 a 18,6 a 28,5 a 7,9 a T3 16,5 a 3,0 bc 24,5 a 31,4 a 4,8 c T4 15,3 a 2,3 bc 22,3 a 32,1 a 4,2 c T5 16,3 a 3,0 bc 17,5 a 26,1 a 7,1 ab T6 17,1 a 1,9 c 16,2 a 23,8 a 4,5 c T7 19,9 a 3,2 ab 21,3 a 25,6 a 6,5 b T8 17,7 a 2,5 bc 20,9 a 31,0 a 4,3 c CV (%) 16,9 2,9 20,2 28,4 5,6 †T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. Sem crescimento de plantas. Tabela 5. Teores de macronutrientes na parte aérea das plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. macronutrientes na parte aérea das plantas de cenoura após 90 dias de crescimento nos ratamentos †T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. Sem crescimento de plantas. significativamente entre os tratamentos. Os teores de K e Ca na raiz da cenoura variaram de 14,4 a 17,7 e de 2,8 a 3,3 g kg-1, respectivamente. Teores de macronutrientes (N, P, K, Ca e Mg) O teor de nitrogênio na raiz da cenoura foi maior no tratamento T8, mas não foi determinada diferença significativa entre este e os tratamentos T2 e T4 (Tabela 6). Os teores de K e Ca não diferiram Tabela 6. Teores de macronutrientes na raiz das plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. Tratamentos† N P K Ca Mg ------------------------------------- g kg-1 --------------------------------- T1 - - - - - T2 12,6 ab* 4,3 a 14,4 a 2,8 a 2,1 ab T3 9,0 b 3,0 b 16,3 a 3,3 a 1,6 b T4 12,9 ab 2,6 b 15,5 a 3,0 a 1,5 b T5 11,4 b 3,4 ab 17,7 a 2,8 a 2,0 ab T6 11,2 b 2,8 b 14,0 a 2,8 a 1,8 ab T7 10,7 b 3,6 ab 15,2 a 2,8 a 2,3 a T8 16,7 a 3,1 b 15,5 a 2,9 a 1,8 ab CV (%) 12,1 3,2 15,5 2,9 1,9 † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. Sem crescimento de plantas. O t d C t é d t 19 6 31 1 k -1 100 di O de macronutrientes na raiz das plantas de cenoura após 90 dias de crescimento nos diferentes t Tabela 6. Teores de macronutrientes na raiz das plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. Tabela 6. Teores de macronutrientes na raiz das plantas de cenoura após 90 dias de cresciment tratamentos. Teores de macronutrientes (N, P, K, Ca e Mg) † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. *Sem crescimento de plantas. cenoura entre 19,6 a 31,1 g kg-1 aos 100 dias. Os teores de P nas raízes foram maiores na tetemunha, demonstrando que os resíduos diminuíram os teores de P nas raízes. Para o Mg, os resíduos também afetaram a absorção. Embora a absorção tenha sido afetada, não representou diferenças significativas na produção de massa seca de raízes. Isto pode ser devido a imobilização temporária dos microorganismos do solo, pois quando é estimulado o crescimento, o P e N, principalmente, são cenoura entre 19,6 a 31,1 g kg-1 aos 100 dias. Os teores de P nas raízes foram maiores na tetemunha, demonstrando que os resíduos diminuíram os teores de P nas raízes. Para o Mg, os resíduos também afetaram a absorção. Embora a absorção tenha sido afetada, não representou diferenças significativas na produção de massa seca de raízes. Isto pode ser devido a imobilização temporária dos microorganismos do solo, pois quando é estimulado o crescimento, o P e N, principalmente, são y p p O teor de Ca na parte aérea da cenoura variou de 23,8 a 32,1 g kg-1; e de 2,8 a 3,3 g kg-1 na raiz (Tabelas 5 e 6). Conforme Sediyama et al. (1998) o teor de Ca na parte aérea de plantas de cenoura colhidas aos 45 dias variou de 12,2 a 16,6 g kg-1. Conforme o mesmo autor, os teores de Ca variaram entre 2,7 a 3,8 g kg-1 aos 90 dias. O teor padrão de Ca em raízes de cenoura, conforme Watt; Merril (1975) que é de 3,1 g kg-1. RICCI et al. Teores de cromo Teor de cromo nas plantas de cenoura após 90 dias de crescimento nos diferentes t Tratamentos† Parte Aérea Raiz Cór 1 de cromo nas plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. tamentos† Parte Aérea Raiz Córtex † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. Sem crescimento de plantas. molecular, que possuem baixa permeabilidade em membranas. Se considerado o teor de cromo das raízes descascadas de cenoura do tratamento T2 (Tabela 7), uma pessoa adulta deveria consumir 2 kg dia-1 de raízes cruas para obter o suprimento de 0,1 mg de Cr. No caso de consumir as cenouras do tratamento T3 (com aplicação de lodo de curtume nas taxas agronomicamente adequadas, no caso com adição cumulativa de 840 kg de Cr ha-1), a necessidade seria de 1 kg dia-1 de raízes cruas de cenoura Teores de macronutrientes (N, P, K, Ca e Mg) (2006) determinaram teores de Ca na parte aérea de O teor de Ca na parte aérea da cenoura variou de 23,8 a 32,1 g kg-1; e de 2,8 a 3,3 g kg-1 na raiz (Tabelas 5 e 6). Conforme Sediyama et al. (1998) o teor de Ca na parte aérea de plantas de cenoura colhidas aos 45 dias variou de 12,2 a 16,6 g kg-1. Conforme o mesmo autor, os teores de Ca variaram entre 2,7 a 3,8 g kg-1 aos 90 dias. O teor padrão de Ca em raízes de cenoura, conforme Watt; Merril (1975) que é de 3,1 g kg-1. RICCI et al. (2006) determinaram teores de Ca na parte aérea de Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 132 QUADRO, M. S. et al. Crescimento da cultura da cenoura... imobilizados nos constituintes celulares, deixando indisponíveis temporariamente para as plantas e solo. Efeitos assim são descritos e discutidos por outros autores (MOREIRA; SIQUEIRA, 2006). situam-se na faixa de teores considerados normais (não tóxicos) para consumo (PAIS; JONES, 1997; KABATA-PENDIAS; PENDIAS, 1986). A concentração máxima de cromo aceitável em resíduos como o lodo de esgoto para aplicação na agricultura é de 1000 mg kg-1 (CONAMA, 2006). Embora as concentrações de cromo nos resíduos sejam maiores que a permitida pela resolução, os teores de cromo adsorvidos na cenoura foi muito baixo,não demonstrando efeito de contaminação nas plantas de cenoura. situam-se na faixa de teores considerados normais (não tóxicos) para consumo (PAIS; JONES, 1997; KABATA-PENDIAS; PENDIAS, 1986). A concentração máxima de cromo aceitável em resíduos como o lodo de esgoto para aplicação na agricultura é de 1000 mg kg-1 (CONAMA, 2006). Embora as concentrações de cromo nos resíduos sejam maiores que a permitida pela resolução, os teores de cromo adsorvidos na cenoura foi muito baixo,não demonstrando efeito de contaminação nas plantas de cenoura. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 Teores de cromo Os teores de cromo nas plantas de cenoura são apresentados na Tabela 7. O tratamento com aplicações anteriores de serragem cromada (T7) apresentou maior teor de cromo na parte aérea das plantas do que os tratamentos com adição de calcário (T2 e T5). Os tratamentos com aplicações anteriores de lodo de curtume não apresentaram diferenças significativas entre si, com valores intermediários. Desde a década de 1960, sabe-se que o cromo é um mineral essencial para os mamíferos (MERTZ, 1969). O mesmo participa como co-fator na atividade da insulina, no metabolismo dos carboidratos, reduzindo o colesterol e triglicerideos. A deficiência desse elemento é importante na patogênese da arterioescherose e de coronariopatias. A “Internecional Union of Nutritional Sciences” (1993) recomenda a ingestão diária de 0,05 a 0,2 mg de cromo por pessoa adulta (SILVA, 1989). As aplicações anteriores de resíduos de curtume proporcionaram aumentos dos teores de cromo tanto na parte aérea como nas raízes das plantas. Os teores de cromo na raiz sem córtex (parte comestível), entretanto, são menores que na parte aérea e no córtex radicular (1 mm) (Tabela 7). Os teores de cromo determinados, mesmo na raiz, nos tratamentos com aplicação dos resíduos, Tabela 7. Teor de cromo nas plantas de cenoura após 90 dias de crescimento nos diferentes tratamentos. Tratamentos† Parte Aérea Raiz Córtex ----------------------------------- mg kg-1 ------------------------------ T1 - - - T2 0,25 b* 0,32 a 1,0 b T3 0,51 ab 0,63 a 1,5 b T4 1,05 ab 0,75 a 1,6 ab T5 0,23 b 0,35 a 1,0 b T6 1,15 ab 0,75 a 1,2 b T7 1,61 a 0,58 a 1,7 ab T8 1,44 ab 0,83 a 3,5 a CV (%) 55 49 25 † T1) Controle, somente solo; T2) Adubação com NPK + calcário para atingir pH 6,0 (ca); T3) Lodo de curtume (LC) em quantidade adequada para atingir pH 6,0 + PK; T4) = Duas vezes a quantidade de LC utilizada no T3 + PK; T5) Resíduo carbonífero (RC) + NPK + ca; T6) RC + LC em quantidade adequada para atingir pH 6,0 + PK; T7) Serragem cromada (SC) + NPK + ca); T8) Cr mineral + LC em quantidade adequada para atingir pH 6,0 + PK; Médias com letras iguais na mesma coluna não apresentam diferença significativa pelo teste de Tukey a 5% de probabilidade. *Sem crescimento de plantas. Tabela 7. CONCLUSÃO A adição de resíduos de lodo de curtume, resíduo carbonífero e serragem cromada não afetam o crescimento de plantas de cenoura nas concentrações apresentadas neste trabalho. Além disso, não há indícios da redução de absorção de Shivas (1978) observou que 85% do cromo absorvido permanece na parte externa da raiz (com 1 mm de espessura) devido à reação com proteínas e outros colóides, formando compostos com alto peso Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 133 Crescimento da cultura da cenoura... QUADRO, M. S. et al. QUADRO, M. S. et al. QUADRO, M. S. et al. das plantas e na raiz são pequenas, não promovendo grandes problemas no consumo de cenoura. nutrientes com o crescimento da cenoura com altas concentrações de cromo no solo. Embora as concentrações de cromo no solo serem altas, a concentração e acumulação de cromo na parte aérea ABSTRACT: With the industrialization, waste production has been increased over the years. Moreover, the disposition of these wastes is a position discussed among environmental agencies. So, the aim of this study was to evaluate the residual effect of successive additions of tannery waste and coal on the chemical properties of the soil and the accumulation of heavy metals in carrot plants. The treatments were applied to field: T1 = Control, only solo; T2 = fertilization with NPK + lime to reach pH 6.0; T3 = tannery sludge in adequate quantity to achieve pH 6.0 + PK; T4 = Twice the amount tannery sludge used in treatment 3 + PK; T5 = waste coal + NPK + lime in adequate quantity to achieve pH 6.0; T6 = waste coal + tannery sludge in adequate quantity to achieve pH 6.0 + PK; T7 = Sawdust = chrome + NPK + lime in adequate quantity to reach pH 6.0; T8 = Cr + mineral tannery sludge in adequate quantity to achieve pH 6.0 + PK. The experiment was conducted in a completely randomized design, with three replicates per treatment. The results demonstrate that carrot plants grew normally in the treatments with high chromium concentrations, either addition with residues or mineral. Furthermore, the levels found in the shoots, roots and roots cortex were low, indicating a low potential of these waste contamination. KEYWORDS: Residue discharge. Daucus carota. Chromium. Heavy metals uptake. REFERÊNCIAS CASTILHOS, D. D. Alterações químicas e biológicas devidas à adição de resíduos de curtume e de cromo hexavalente ao solo. 1998. 160f. Tese (Doutorado) – Programa de Pós-Graduação em Ciência do Solo, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Porto Alegre, 1998. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 MERTZ, W. Chromium occurrence and function in biological systems. Physiological Reviews, Methesda, v. 49, n. 2, p. 163-239, 1969. MOREIRA, M. S.; SIQUEIRA, J. O. Microbiologia e bioquímica do solo. 2ª Edição. Lavras: UFLA, 2006. 729p. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 CLAAS, I. C.; MAIA, R. A. M. Manual básico de resíduos industriais de curtume. Porto Alegre: SENAI/RS, 1994. 664 p. CLAAS, I. C.; MAIA, R. A. M. Manual básico de resíduos industriais de curtume. Porto Alegre: SENAI/RS, 1994. 664 p. COMISSÃO DE QUÍMICA E FERTILIDADE DO SOLO. Manual de adubação e de calagem para os estados do Rio Grande do Sul e de Santa Catarina.. Porto Alegre: Evangraf. 2004. 394 p. CONAMA. Ministério do Meio Ambiente. Resolução n. 375, de 29 de Agosto de 2006. Define critérios e procedimentos, para o uso agrícola de lodos de esgoto gerados em estações de tratamento de esgoto sanitário e seus produtos derivados, e dá outras providências. 2002. Diário Oficial da União, n. 375, de 29 de Agosto de 2006. DOMASZAK, S. C. Efeito imediato e residual da aplicação de resíduos de curtume nas plantas em três solos. 2000. 107f. Dissertação (Mestrado) – Programa de Pós-Graduação em Ciência do Solo, Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Porto Alegre, 2000. FERREIRA, A. S.; CAMARGO, F. A. O; TEDESCO, M. J.; BISSANI. C. A. Alterações de atributos químicos e biológicos de solo e rendimento de milho e soja pela utilização de resíduos de curtume e carbonífero. Revista Brasileira de Ciência do Solo, Viçosa, v. 27, p. 755-763, 2003. GIANELLO, C.; DOMASZAK, S. C.; BORTOLON, L.; KRAY, C. H.; MARTINS, V. Viabilidade do uso de resíduos da agroindústria coureiro-calçadista no solo. Ciência Rural, Santa Maria, v. 41, p. 242-245, 2011. http://dx.doi.org/10.1590/S0103-84782011005000007 KABATA-PENDIAS, A.; PENDIAS, H. Trace elements in soils and plants. 4.ed. Florida: CRC Press, 1986. 315 p. KRAY, C. H. Efeitos de duas aplicações de resíduo de curtume e carbonífero no solo e nas plantas. 2001. 90f. Dissertação (Mestrado em Ciência do Solo) – Faculdade de Agronomia, Universidade de Federal do Rio Grande do Sul, Porto Alegre, 2001. Biosci. J., Uberlândia, v. 31, n. 1, p. 127-134, Jan./Feb. 2015 134 QUADRO, M. S. et al. Crescimento da cultura da cenoura... KRAY, C. H.; TEDESCO, M. J.; BISSANI, C. A.; GIANELLO, C.; SILVA, K. J. Tannery and coal mining waste disposal on soil. Revista Brasileira de Ciência do Solo, Viçosa, v. 32, p. 2877-2882, 2008. LAUSCHNER, M. H.; TEDESCO, M. J.; GIANELLO, C.; BORTOLON, L.; ANDREAZZA, R.; KRAY, C. H.; BARBOSA, D. B. P. Avaliação da acidez dos resíduos da agroindústria fumageira após aplicação em diferentes solos. Pesquisa Agropecuária Gaúcha, Porto Alegre, v. 18, p. 12-24, 2012. MACHADO, A. A. Sistema de análise estatística para Windows (WINSTAT). MOREIRA, M. S.; SIQUEIRA, J. O. Microbiologia e bioquímica do solo. 2ª Edição. Lavras: UFLA, 2006. 729p. PACHECO, J. W. F. Curtumes. São Paulo: CETESB, 2005. 76 p. PAIS, I.; JONES, J. B. The handbook of trace elements. Boca Raton : St. Lucie Press, 1997. 323 p RICCI, M. S. F.; OLIVEIRA, F. F.; MIRANDA, S. C.; COSTA, J. R. Produção da cenoura e efeito na fertilidade do solo e nutrição decorrente da solarização do solo para controle da tiririca. Bragantia, Campinas, v. 65, n. 4, p. 607 -614, 2006. http://dx.doi.org/10.1590/S0006-87052006000400011 SHIVAS, S. A. J. The environmental effects of chromium in tannery effluents. Journal of American Leather Chemistry Association, Lubbock, v. 73, n. 71, p. 370-377, 1978. SEDIYAMA, M. A. N.; VIDIGAL, S. M.; PEREIRA, P. R. G.; GARCIA, N. C. P.; LIMA, P. C. Produção e composição mineral de cenoura adubada com resíduos orgânicos. Bragantia, Campinas, v. 57, n. 2, p. 379-386, 1998. http://dx.doi.org/10.1590/S0006-87051998000200019 SEGATTO, M. P.; ANDREAZZA, R.; BORTOLON, L.; SANTOS, V. P.; GIANELLO, C.; CAMARGO, F. A. O. Decomposição de resíduos industriais no solo. Ciência e Natura, Santa Maria, v. 34, p. 49–62, 2012. SILVA, S. Agricultural use of leather working residues as means of energy saving and environmental protection. Milão: UNIC, 1989. 60 p. CLAAS, I. C.; MAIA, R. A. M. Manual básico de resíduos industriais de curtume. Porto Alegre: SENAI/RS, 1994. 664 p. Pelotas : Universidade Federal de Pelotas, 2001. MERTZ, W. Chromium occurrence and function in biological systems. Physiological Reviews, Methesda, v. 49, n. 2, p. 163-239, 1969. SILVA, J. G. C. Análise estatística de experimentos. Pelotas: UFPel, 1997. 263 p. SILVA, J. G. C. Análise estatística de experimentos. Pelotas: UFPel, 1997. 263 p STRECK, E. V.; KÄMPF, N.; DALMOLIN, R. S. D.; KLAMT, E.; NASCIMENTO, P. C.; SCHNEIDER, P. Solos do Rio Grande do Sul. Porto Alegre: EMATER/RS: Departamento de Solos da UFRGS, 2002. 222 p. TEDESCO, M. J.; GIANELLO, C.; BISSANI, C. A. et al. Análises de solo, plantas e outros materiais. 2. ed. Porto Alegre: Departamento de Solos da UFRGS, 1995. 174 p. (Boletim Técnico 5). WATT, B. K.; MERRIL, A. L. Handbook of the nutrition contents of foods. New York : United States Department of Agriculture. Dover Publication, 1975. 190 p.
https://openalex.org/W4254326185	https://jyx.jyu.fi/bitstream/123456789/74511/1/fenrg-08-606742.pdf	English	null	Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction	null	2,020	cc-by	14,072	This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Verma, Anand M.; Honkala, Karoliina; Melander, Marko M. Title: Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction Title: Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ © 2021 Verma, Honkala and Melander. Published version Verma, A. M., Honkala, K., & Melander, M. M. (2021). Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction. Frontiers in Energy Research, 8, Article 606742. https://doi.org/10.3389/fenrg.2020.606742 2021 This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Verma, Anand M.; Honkala, Karoliina; Melander, Marko M. This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Verma, Anand M.; Honkala, Karoliina; Melander, Marko M. This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Verma, Anand M.; Honkala, Karoliina; Melander, Marko M. Please cite the original version: Verma, A. M., Honkala, K., & Melander, M. M. (2021). Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction. Frontiers in Energy Research, 8, Article 606742. https://doi.org/10.3389/fenrg.2020.606742 ORIGINAL RESEARCH published: 02 February 2021 doi: 10.3389/fenrg.2020.606742 Keywords: electrocatalysis, graphene, electrosorption, proton-coupled electron transfer, density functional theory Keywords: electrocatalysis, graphene, electrosorption, proton-coupled electron transfer, density functional theory Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction Anand M. Verma, Karoliina Honkala* and Marko M. Melander* Department of Chemistry, Nanoscience Center, University of Jyväskylä, Jyväskylä, Finland The electrocatalytic CO2 reduction reaction (CO2RR) is considered as one of the most promising approaches to synthesizing carbonaceous fuels and chemicals without utilizing fossil resources. However, current technologies are still in the early phase focusing primarily on identifying optimal electrode materials and reaction conditions. Doped graphene-based materials are among the best CO2RR electrocatalysts and in the present work we have performed a computational screening study to identify suitable graphene catalysts for CO2RR to CO under alkaline conditions. Several types of modiﬁed- graphene frameworks doped with metallic and non-metallic elements were considered. After establishing thermodynamically stable electrodes, the electrochemical CO2RR to CO is studied in the alkaline media. Both concerted proton-coupled electron transfer (PCET) and decoupled proton and electron transfer (ETPT) mechanisms were considered by developing and using a generalization of the computational hydrogen electrode approach. It is established that the CO2 electrosorption and associated charge transfer along the ETPT pathway are of utmost importance and signiﬁcantly impact the electrochemical thermodynamics of CO2RR. Our study suggests an exceptional performance of metal- doped nitrogen-coordinated graphene electrodes, especially 3N-coordinated graphene electrodes. 1 INTRODUCTION Specialty section: This article was submitted to Electrochemical Energy Conversion and Storage, a section of the journal Frontiers in Energy Research Received: 15 September 2020 Accepted: 11 December 2020 Published: 02 February 2021 Citation: Verma AM, Honkala K and Melander MM (2021) Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction. Front. Energy Res. 8:606742. doi: 10.3389/fenrg.2020.606742 Specialty section: This article was submitted to Electrochemical Energy Conversion and Storage, a section of the journal Frontiers in Energy Research Received: 15 September 2020 Accepted: 11 December 2020 Published: 02 February 2021 The extensive use of fossil resources has escalated the emission of green house gases, particularly CO2, and disrupted the atmospheric carbon balance. An appealing approach for maintaining this balance is to utilize renewable energy resources but their intermittent nature presents a serious obstacle in the energy conversion and storage applications (Vasileff et al., 2017; Jia C. et al., 2019). In this regard, converting renewable electrical energy into chemical energy through the electrochemical CO2 reduction reaction (CO2RR) has been identiﬁed as an efﬁcient solution (Tian et al., 2018; Jia C. et al., 2019). Recently, an integrated electrocatalytic CO2RR process has drawn appreciable attention due to its extra-ordinary features in enabling atmospheric sequestration of CO2 followed by ambient CO2RR to synthesize chemicals or fuels (MacDowell et al., 2010; Vasileff et al., 2017; Jia C. et al., 2019). Further ﬂexibility is obtained by using the electrode potential and reaction conditions (pH, electrolyte) to control reaction thermodynamics and kinetics as well as activity and selectivity. Edited by: Kai S. Exner, Soﬁa University, Bulgaria Reviewed by: Axel Gross, University of Ulm, Germany Michael Busch, Aalto University, Finland Correspondence: Karoliina Honkala karoliina.honkala@jyu.ﬁ Marko M. Melander marko.m.melander@jyu.ﬁ Edited by: Kai S. Exner, Soﬁa University, Bulgaria Reviewed by: Axel Gross, University of Ulm, Germany Michael Busch, Aalto University, Finland Correspondence: Karoliina Honkala karoliina.honkala@jyu.ﬁ Marko M. Melander marko.m.melander@jyu.ﬁ Reviewed by: Axel Gross, University of Ulm, Germany Michael Busch, Aalto University, Finland Citation: Verma AM, Honkala K and Melander MM (2021) Computational Screening of Doped Graphene Electrodes for Alkaline CO2 Reduction. Front. Energy Res. 8:606742. doi: 10.3389/fenrg.2020.606742 However, achieving efﬁcient CO2RR is a challenging task and requires optimization of the electrode material as well as the reaction conditions. There are multiple reasons for this. Firstly, CO2 is a highly stable molecule as reﬂected by its high negative reduction potential (−1.90 V vs. SHE) February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 1 Modiﬁed Graphene for CO2 Reduction Verma et al. electrochemical PCET steps but the piling evidence (Ringe et al., 2019; Lee et al., 2020; Vijay et al., 2020) indicates that the non- PCET CO2 adsorption step is limiting CO2RR. The reason for calculations focusing on electrochemical PCET steps is that the computational hydrogen electrode (CHE) model (Nørskov et al., 2004), in its original form can only account for PCET steps. The importance of non-PCET steps, such as CO2 adsorption, can be highlighted by comparing two recent CO2RR studies (Li et al., 2019; Guo et al., 2020). Considering only the PCET steps, led to the conclusion that CO2RR on a Fe_4N is thermodynamically facile and a potential as small as ∼0.1 V vs. RHE is sufﬁcient to produce COOH (Guo et al., 2020). However, accounting for non- PCET steps led to a very different conclusion as the CO2 adsorption itself is the potential limiting step that has a high thermodynamic barrier of ∼0.9 eV (Li et al., 2019). required to drive the ﬁrst-electron transfer process. To circumvent this, active electrocatalysts need to be developed, where CO2 can be converted to several different products through sequences of complex, multistep proton-coupled electron transfer (PCET) steps. In addition, as many CO2 derived products have similar thermodynamic stability, the developed electrocatalyst has to be selective. Controlling selectivity while retaining high activity is the primary goal in CO2RR electrocatalysis and requires exquisite control over the complex PCET chemistry, which depends sensitively on the electrode material, electrode potential, pH, and electrolyte. Therefore, optimizing the electrocatalytic performance presents a serious challenge to CO2RR-based applications (Schneider et al., 2012; Jia C. et al., 2019). Citation: In the search for ideal CO2RR catalysts, numerous metallic electrodes have been thoroughly explored experimentally and computationally (Back et al., 2015a; Back et al., 2015b; Mistry et al., 2014; Peterson et al., 2010; Peterson and Nørskov, 2012; Shi et al., 2014; Zhu et al., 2014; Hori, 2008; DeWulf et al., 1989; Kim et al., 1988; Hori et al., 1986; Hori et al., 1989; Lu et al., 2014; Chen et al., 2012; Gattrell et al., 2006; Akhade et al., 2014; Bagger et al., 2019). However, most of them suffer from one or several of the following shortcomings: low faradic efﬁciency and selectivity, high overpotential, CO poisoning, high cost, and/or low abundance. The extended metal surfaces are also subject to the d-band center theory and intrinsic thermodynamic scaling relationships between the reaction energies of CO2RR intermediates. These features together with the Sabatier principle set constraints on the achievable electrocatalytic performance of metallic CO2RR catalysts and hampers the search for efﬁcient electrode materials (Peterson and Nørskov, 2012; Hansen et al., 2013; Back et al., 2015b; Li et al., 2015). In addition to considering the electrode material, it should be recognized that the electrocatalytic CO2RR activity and selectivity are inherent properties of electrochemical interfaces. As the interfacial properties depend sensitively on the electrode material, the electrolyte, pH, and electrode potential (Lu and Jiao, 2016; Nitopi et al., 2019), they can be used for controlling the reaction environment and the electrocatalytic CO2RR performance (Pérez-Gallent et al., 2017; Aljabour et al., 2018; Gao et al., 2018; Guo et al., 2018; Ringe et al., 2019). Recent studies have demonstrated that the CO2 adsorption step is sensitive to these properties and it also determines the CO2RR activity and selectivity (Ringe et al., 2019; Lee et al., 2020; Vijay et al., 2020). Furthermore, the faradaic efﬁciency of CO2RR is higher in alkaline conditions where the competing hydrogen evolution reaction (HER) is suppressed. Utilizing highly alkaline conditions facilitates obtaining high current densities at lower overpotentials (Gabardo et al., 2018) and the issue of carbonate formation can be circumvented using gas diffusion electrodes (Malkhandi and Yeo, 2019). However, one has to consider both bulk and local pH which may differ as H+/OH− are consumed or formed at the interface resulting in a pH gradient between the electrolyte and electrode interface (Bohra et al., 2019; Varela, 2020). Citation: Furthermore, these local pH changes are found to be sensitive to the used electrode potential and current density (Lu et al., 2020). To circumvent these inherent limitations of current CO2RR electrocatalysts, several unorthodox or innovative CO2RR electrocatalysts have been suggested recently (Qu et al., 2010; Wang et al., 2015; Lu and Jiao, 2016; Sun et al., 2017; Kibria et al., 2019; Nitopi et al., 2019). Carbon-based electrocatalysts and in particular doped graphene sheets, are among the most promising materials and have been widely investigated as potential CO2RR electrodes. Graphene electrocatalysts have several attractive physical properties such as high surface area, high electron mobility, high thermal conductivity, high Young’s modulus etc. Furthermore, their geometrical ﬂexibility and electronic structure have been suggested to enable escaping the scaling relationships (Kim et al., 2014; Li et al., 2015). While pure graphene can facilitate outer-sphere electron transfer reactions, edges or other defects are usually required for appreciable electrocatalytic activity (Brownson et al., 2012). Even higher activity and selectivity can be achieved by doping pristine or defected graphene with metal or non-metallic elements (Varela et al., 2018; Jia M. et al., 2019; Wu et al., 2019). Given the advantages of pH control and highly alkaline conditions, it is surprising that previous computational CO2RR studies have solely focused on the acidic CO2RR. A crucial difference between acidic and alkaline CO2RR is the proton donor: in acidic conditions, the proton is donated by the hydronium ion or some other acid, whereas, under alkaline conditions, the solvent (water) is more likely the hydrogen donor. This difference can introduce changes to the reaction mechanism: A fully coupled PCET mechanism is preferred in acidic conditions, whereas, under alkaline conditions, a decoupled transfer of an electron and a proton may become the dominant mechanism (Koper, 2013a). In order to understand alkaline CO2RR, one has to examine the possibility of decoupled electron transfer/proton transfer (ETPT) steps but this requires going beyond the original CHE (Nørskov et al., 2004) model, which is applicable to coupled PCET reactions only. This limitation can be circumvented using more general approaches Given the promising CO2RR performance of such graphene- based materials (Varela et al., 2018; Jia M. et al., 2019; Wu et al., 2019), several experimental and computational investigations have explored the origin of their electrocatalytic behavior and to search for new electrocatalysts. Frontiers in Energy Research \| www.frontiersin.org Citation: Most computational studies have focused on identifying the optimal reaction energies of the February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 2 Modiﬁed Graphene for CO2 Reduction Verma et al. such as general grand canonical DFT methods (Mermin, 1965; Melander et al., 2019; Melander, 2020) and the decoupled CHE (Lindgren et al., 2020). While these methods are applicable to ETPT pathways as well, they are currently computationally too demanding for large scale computational screening studies and more tractable methods need to be developed. structures can be experimentally realized using, e.g., selective bombardment followed by ion deposition (Wang et al., 2012), or more reﬁned chemical synthesis (Varela et al., 2018). These vacancies or defects can host a variety of dopants (Wang et al., 2012; He et al., 2014), which we modeled by introducing foreign atom(s) into the created vacancy. These atoms include platinum group metals (Rh, Pd, Pt, and Ru), coinage metals (Ag, Au, and Cu), base metals (Al, Fe, Ni, Mo, Co, Mn, Zn, and Cr), and non-metals (B and N). Structural information for all the studied systems is provided as a Supplementary Material set. In this study, we address the CO2RR-to-CO on several graphene electrodes in alkaline conditions and account for both PCET and ETPT pathways. To enable this, the commonly applied computational hydrogen electrode method is extended to study the decoupled ETPT reactions. In particular, we consider the effect of partial ET and the potential-dependency of CO2 adsorption. On this basis a computational screening approach is developed and applied to identify promising doped graphene electrocatalysts for alkaline CO2RR. A four-level selection criteria is introduced to rank and select catalysts according to their thermodynamic stability, ability to bind and activate CO2, electrosorption properties, and theoretical limiting and overpotentials. These principles allow us to identify thermodynamically stable doped graphene electrodes with low limiting and overpotentials in alkaline environments. Our study outlines that highly exergonic CO2 adsorption associated with high degree of electron transfer is detrimental for catalytic performance. We ﬁnd that N-coordinated Pt and Ru-doped electrodes are promising candidates as pH universal CO2RR electrodes. y p pp y We consider three different vacancy structures: single-vacancy (SV), di-vacancy (DV), and tri-vacancy (TV), by removing one, two, and three central carbon atoms, respectively. Three SV systems are investigated namely, M_SV, M2_SV, and MPt_SV (see Figures 1A–C). Citation: The M_SV structure results from the deposition of a single metal on a single-vacant structure and presents the simplest doped structure. In the case of M2_SV and MPt_SV structures, homo- and heteroatom dimers are placed on SV as models for nucleated reaction centers (Ferrante et al., 2016). We also investigated an experimentally inspired M2_2SV structure (He et al., 2014), where two neighboring single vacancies are both ﬁlled with identical atomic dopants, see Figure 1D. The M_DV structure, shown in Figure 1F has a single dopant in a di-vacancy. The trapping of dopant pairs over tri-vacant surface is energetically more favourable than a single metal dopant and have been observed experimentally (He et al., 2014), and for this reason the M2_TV structure is considered as well (Figure 1H). Frontiers in Energy Research \| www.frontiersin.org 2 METHODS The N-coordinated, heteroatom-doped carbon frameworks are the most experimentally (Li et al., 2017; Cheng et al., 2019; Wang et al., 2019; Zhang et al., 2019; Koshy et al., 2020) and computationally (Tripkovic et al., 2013; Li et al., 2015; Ju et al., 2017; Pan et al., 2018; Varela et al., 2019) studied doped graphene electrodes for oxygen reduction reaction (ORR) and CO2RR. In these N-coordinated structure, the reaction-center is modiﬁed by replacing the coordinating carbon atoms with nitrogens. Such modiﬁcations of the M_SV and M_DV structures lead to the M_3N and M_4N electrodes having either three or 4 N atoms (Figures 1E,G), respectively. 2.1 Modeling of Graphene Sheets 2.1 Modeling of Graphene Sheets Pristine graphene is modeled using a simpliﬁed non-periodic structure with 42 carbon atoms forming a honeycomb structure with 14 benzene rings. The dangling carbon atoms at the edges of the sheet are terminated with hydrogen atoms (Verma et al., 2018). The ﬂake-based graphene models have been previously shown to be perform well compared to their respective periodic models (Lazar et al., 2013; Verma et al., 2018; Shang et al., 2020). This is attributed to the accounts of non-clustering of carbon atoms and negligible edge effects due to hydrogen-terminations, which make the graphene-ﬂake models trustworthy. In addition, terminating the dangling edge carbon atoms with hydrogen lead to uniform and delocalized distribution of charge and orbitals all over the catalyst surface (Tachikawa and Kawabata, 2011), which prevents any possible artiﬁcial localization effects. The graphene- ﬂake models are computationally less expensive compared than extended models while providing a faithful description of graphene. Finally, the graphene-ﬂake model allows us to carry out the large-scale screening studies at the hybrid functional DFT-level (see below for details), which is required to accurately capture the adsorption energies of the CO2RR intermediates on doped graphene catalysts (Vijay et al., 2020). 2.2 Density Functional Theory Methods 2.2 Density Functional Theory Methods All density functional theory (DFT) (Hohenberg and Kohn, 1964; Kohn and Sham, 1965) calculations were carried out using Gaussian 09 (RB.01) (Frisch et al., 2009) package with the help of the Gauss View 5 (Dennington et al., 2009) visualizer. Exchange-correlation effects were described using the B3PW91 (Perdew et al., 1991) functional with Grimme’s D3 (Grimme et al., 2010) dispersion corrections. The metallic atoms were treated within the effective core potential (ECP) formalism and the LANL2DZ (Hay and Wadt, 1985a; Hay and Wadt, 1985b; Wadt and Hay, 1985) basis set, while for all the other atoms (C, H, O, N, and B) the 6-31 + g (d,p) (Petersson et al., 1988; Petersson and Al-Laham, 1991) basis set was used. The atomic structures were relaxed until the maximum residual force reached below 0.02 eV Å−1. The satisfactory performance of the adopted computational approach has been demonstrated previously for graphene (Verma et al., 2018) and metallic clusters (Verma and Kishore, 2017; Verma and Kishore, 2018; Verma and Kishore, 2019). Pristine graphene sheets are chemically inert, but their reactivity can be tuned with dopants and defects. The single vacancy is arguably the simplest point defect in graphene; however, larger point defects such as di- and tri-vacancies are also observed in experiments (Warner et al., 2012; He et al., 2014) forming after the coalescence of two or three neighboring mono- vacancies (Trevethan et al., 2014). Different vacancy and doping February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 3 Verma et al. Modiﬁed Graphene for CO2 Reduction FIGURE 1 \| Various models of heteroatom doped graphene electrodes: (A) M_SV, (B) M2_SV, (C) MPt_SV, (D) M2_2SV, (E) M_3N, (F) M_DV, (G) M_4N, and (H) M2_TV. The elements carbon, nitrogen and hydrogen are in gray, blue and lavender colors, respectively. The dopant atom is a model atom in purple color. The anchoring Pt atom in M2_SV model is presented in teal color. FIGURE 1 \| Various models of heteroatom doped graphene electrodes: (A) M_SV, (B) M2_SV, (C) MPt_SV, (D) M2_2SV, (E) M_3N, (F) M_DV, (G) M_4N, and (H) M2_TV. The elements carbon, nitrogen and hydrogen are in gray, blue and lavender colors, respectively. The dopant atom is a model atom in purple color. The anchoring Pt atom in M2_SV model is presented in teal color. 2.2 Density Functional Theory Methods Both implicit and explicit solvent effects were mostly excluded from the present study. Considering explicit solvation would have been computationally intractable and the implicit solvent models do not markedly affect the adsorption properties of CO2RR intermediates in constant charge calculations as, e.g., explicit hydrogen bonding cannot be captured with such models (Heenen et al., 2020). The exclusion of solvation effects from modeling the electrocatalytic reactions is a common approximation that has been shown to successfully describe a variety of electrocatalytic systems (Frydendal et al., 2015; Busch et al., 2016; Calle-Vallejo et al., 2017; Valter et al., 2018). However, we have checked the effect of implicit solvation for some of the consider structures. In particular, we addressed implicit solvent effects using the polarizable continuum model (PCM) (Scalmani and Frisch, 2010) for the best performing electrodes identiﬁed from the vacuum calculations. For this, we computed the overpotentials using Equation 6 with and without implicit solvent for the best performing catalysts, and the differences are at largest 0.16 V vs. RHE corresponding to 0.16 eV difference in reaction energies as shown in Table 1. vibrational frequencies were computed to determine the vibrational entropy and zero-point energy contributions in the thermodynamics of CO2RR. The thermodynamic parameters were calculated at 298.15 K temperature and 1 atm pressure. In the Supplementary Material, we present the methodology utilized to compute the thermodynamic entropy, enthalpy, and free energy. The charge analysis was based on the Mulliken scheme (Mulliken, 1955), which is assumed to be accurate enough for the comparison of similar models and similar basis sets. Frontiers in Energy Research \| www.frontiersin.org 2.3 Computation of Catalyst Stability Therefore, we assume that the electrosorption valency is independent of the potential (c(U) ≈c), which is a good approximation at potentials close to the reference potential but breaks down as U ≫ Uref . This is seen as unphysically large equilibrium potential values for CO2 adsorption steps–the true upper bound for the CO2→CO2 −should be close to its thermodynamic reduction potential of −1.9 V vs. SHE (Bratsch, 1989). Here, we approximate the electrosorption valency using the ﬁtting approach (Schmickler and Guidelli, 2014), which gives: CO2aq + δe−+ Gr # CO2()δe− (3a) CO2()δe−+ (1 −δ)e−+ H2Oaq # COOH() + OH−aq (3b) COOH() + e−# CO() + OH−aq (3c) CO() + # COaq + Gr (3d) (3d) where Gr (also referred as “”) is the graphene catalyst and δ denotes the partial charge. When Equations 3a,b, are separated, the mechanism is sequential (ETPT) whereas if they are summed the PCET mechanism is followed. To compute the electrochemical CO2RR thermodynamics along both the ETPT and PCET pathways in alkaline conditions, the CHE method was extended to include partial charge transfer and ETPT steps. Note that we are interested in ET taking place during adsorption, i.e., electrosorption, not outer- sphere ET to a dissolved CO2 forming CO− 2(aq). The crucial difference is that the solution phase ET to CO2 can be considered as an outer-sphere ET reaction the thermodynamics of which are not affected by the electrode material. In that case, the tabulated (Bratsch, 1989) reduction potentials can simply be used for computing the thermodynamics of this outer-sphere ET reaction. For such outer-sphere reactions, the thermodynamics can also be accurately computed using standard approaches (Marenich et al., 2014). Here, we instead consider an inner- sphere ET taking place during adsorption, and in this case, the electrode cannot be neglected due to the strong interactions and hybridization between the electronic states of electrode and CO2. For such reactions, the potential-dependent thermodynamics require going beyond standard approaches and methods such as grand canonical DFT are required (Hörmann et al., 2020). Grand canonical DFT as implemented presently is, however, too costly for large-scale screening studies and we propose a computationally feasible extension to CHE to enable addressing (partial) ET. The extension is motivated by the recent decoupled hydrogen electrode method (Lindgren et al., 2020), which combines the CHE with grand canonical DFT. 2.3 Computation of Catalyst Stability By combining the electrosorption valency and CHE, the reaction thermodynamics for each elementary step can be calculated using: 2.4 Computation of Reaction Free Energies The CO2 adsorption free energy (GAds) over the doped graphene electrodes was calculated using: GAds GMGr+CO2 −GMGr + GCO2 (2) (2) where GMGr+CO2 is the free energy of CO2 adsorbed over modiﬁed- graphene, GMGr is the free energy of the bare modiﬁed-graphene, and GCO2 is the free energy of gas-phase CO2. ΔG1(RHE) GCOδe− 2 () −G(Gr) −GCO2aq + cU(RHE)(4a) ΔG2(RHE) ΔGw + 1 −cU(RHE) + G(COOH()) −1 2 GH2aq −GCOδe− 2 () (4b) ΔG1(RHE) GCOδe− 2 () −G(Gr) −GCO2aq + cU(RHE)(4a) ΔG2(RHE) ΔGw + 1 −cU(RHE) + G(COOH()) −1 2 GH2aq −GCOδe− 2 () (4b) ΔG3(RHE) ΔGw + U(RHE) + G(CO()) + GH2Oaq −1 2 GH2aq −G(COOH()) (4c) ΔG4 GCOaq + G(Gr) −G(CO()) (4d) To model alkaline conditions, we consider water as the hydrogen donor as commonly assumed for alkaline CO2RR (Yin et al., 2019). In general, the reaction may take place either via a sequential ETPT or a concerted PCET mechanism (Dunwell et al., 2018). In the ETPT pathway, the ﬁrst ET denotes partial charge transfer taking place during CO2 adsorption on the graphene catalyst, which is followed by further partial ET and PT. For both the sequential and concerted pathways, the elementary steps for CO2RR to CO are: (4b) ΔG3(RHE) ΔGw + U(RHE) + G(CO()) + GH2Oaq −1 2 GH2aq −G(COOH()) (4c) ΔG4 GCOaq + G(Gr) −G(CO()) (4d) (4c) (4d) A complete derivation of Equation 4 is presented in the Supplementary Material. Here ΔGw ~μH+ + ~μOH−−μH2O is the water dissociation free energy (0.83 eV at pH 14) (Bratsch, 1989) Equations 4a,b are written for the ETPT pathway and when added together, the reaction free energy for the concerted PCET step is obtained, i.e., ΔG1+2(RHE) ΔG1(RHE) + ΔG2(RHE). c (zΔGads(U)/zU) is the electrosorption valency (Schmickler and Guidelli, 2014), which changes the adsorption energy as ΔΔGads(U) cΔU. Note that c ≤δ and the partial charge transfer upon adsorption do not directly change the Gibbs free energy. As discussed in the SM, γ is a potential-dependent quantity and could be obtained from experiments or using potential- dependent grand canonical DFT, (Melander et al., 2019; Hörmann et al., 2020) but such data is not available for the catalysts studied here. 2.3 Computation of Catalyst Stability p y y Electrode stability is a prerequisite for maintaining catalytic activity for extended operation time(Krasheninnikov et al., 2009; Wang et al., 2013; Back et al., 2017). In doped graphene electrodes, the deactivation is thought to take place through dopant dissolution (Shao et al., 2019). The dissolution will result in the formation of a graphene vacancy, while the dissolved atoms will likely form metallic nanoparticles or other stable products. Hence, the thermodynamic stability of the studied catalysts was evaluated against pristine graphene and the most stable (bulk) phase of the dopant. The catalyst stability was ﬁrst referenced against gas-phase metal ions and subsequently against the dopant’s most stable bulk phase by utilizing experimental cohesive energies (Kittel, 2004). This way the formation energy is compared against the most stable bulk phase of a given dopant. Finally, the formation free energy (GFE) is computed as: Metal-modiﬁed graphene structures are subject to spin- inversion and, therefore, spin-unrestricted DFT was used and a careful investigation of spin multiplicities was necessary to locate the lowest energy spin states. To ensure the correct spin- multiplicity, single point energies at different spin multiplicities were calculated followed by re-optimizations of the atomic structures at the correct spin-multiplicity. The Mulliken charge (qM), magnetic moment (m), and the most stable spin multiplicity for each dopant in the considered graphene frameworks can be found in Supplementary Table S1. Note that all the formation and binding free energy calculations were performed using the most stable spin state structures and their energetics. (1) GFE 1/N[GMGr −(N × GM + N × Gvacant Gr) + N × Gcoh] (1) where N is the number of dopant atoms, GMGr is the free energy of the doped graphene electrode, GM is the free energy of the metal atom in the gas-phase, Gvacant Gr is the free energy of the vacant graphene, and Gcoh is the cohesive energy of the dopant(s) (Kittel, 2004). Vibrational frequency calculations were performed for all the optimized structures to conﬁrm that they are true minima. The February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 4 Modiﬁed Graphene for CO2 Reduction Verma et al. valency (Schmickler and Guidelli, 2014) to model the partial ET taking place during CO2 adsorption to capture the potential- dependency of this step. 2.3 Computation of Catalyst Stability We utilize a computationally more feasible approximation based on the electrosorption c ≈δ 0.84 + 0.16 × exp −3χGr −χCO2 2 (5) (5) where χs are Pauling electronegativities. χGr ≈2.5 for all metal- doped graphene electrodes. χCO2 ≈1.5 is computed using the experimental (NIST database, 2020) CO2 electron afﬁnity and ionization energies for computing its Mulliken electronegativity (Mulliken, 1934). The Mulliken electronegativity is then converted to the Pauling scale using the linear correlation (Herrick, 2005) between the two scales. To compare and analyze the electrochemical thermodynamics, we also deﬁne the thermodynamic equilibrium potential (Ueq), the thermodynamic limiting potential (UL), and overpotential (η) (Durand et al., 2011) as further discussed in the SM. However, our deﬁnition will take into account the ETPT steps with partial charge transfer in terms of the electrosorption valency. February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 5 Modiﬁed Graphene for CO2 Reduction Verma et al. FIGURE 2 \| Formation free energies (eV) of various heteroatom doped graphene models. Increasing negative energetics denote stronger formation stability of dopant(s) to the vacant site(s). FIGURE 2 \| Formation free energies (eV) of various heteroatom doped graphene models. Increasing negative energetics denote stronger formation stability of dopant(s) to the vacant site(s). FIGURE 2 \| Formation free energies (eV) of various heteroatom doped graphene models. Increasing negative energetics denote stronger formation stability of dopant(s) to the vacant site(s). UEq(RHE) ⎡⎣ i ΔGi(U 0 vs. RHE)⎤⎦ne,tot (6a) UL(ETPT) −maxΔG1(U 0)c, ΔG2(U 0)1 −c, ΔG3(U 0)(6b) UL(PCET) −max{ΔG1+2(U 0), ΔG3(U 0)} (6c) ηPCET/ETPT \|Ueq −UL(PCETETPT)\| (6d) −6.47 eV; however, doping with coinage metals and a few of base and platinum group metals is thermodynamically unfavorable (see Supplementary Table S5). The formation free energies of B- and N-doped graphene are above −6 eV, which demonstrates their extremely strong interaction with the graphene single vacancy. (6a) In the case of M2_2SV, the metal dopants are also elevated out of the graphene plane similar to M_SV. Almost all M2_2SV systems are symmetric with the dopants occupying two vacancies. The formation free energy of stable M2_2SV electrodes ranges from −0.42 eV to −5.70 eV. Doping by non- metallic B and N atoms is extremely exergonic (∼−5.5 eV). 2.3 Computation of Catalyst Stability We found that, upon structural optimization, the Ag, Al, and Cu dopants spontaneously drift away from the vacancy to form dimers, therefore, they are excluded from electrochemical thermodynamics studies. where ne,tot 2 is the total number of electrons transferred in the CO2RR to CO. 3.1 Structures and Stabilities Thermodynamic stability against dissolution is a key material property mandatory for maintaining electrocatalytic activity and, therefore, we ﬁrst addressed the stability of our graphene model electrodes against pristine graphene and the thermodynamically most stable (bulk) phase of the dopant. Figure 2 summarizes the Gibbs free energies for formation, computed according to Equation 1. Additional data on formation thermodynamics with numerical values, zero- point energy corrections, and formation enthalpies are reported in the Supplementary Tables S2–S5. We found that majority of the studied dimers (M2_SV and MPt_SV) in single vacancies are thermodynamically unstable and these results are therefore presented and discussed solely in the SM. For other systems, the ﬁndings of Figure 2 are discussed in more detail below. 3.1.2 Di- and Tri-vacant Graphene In the M_DV graphene, dopant atoms are bound to four nearest undercoordinated C atoms and remain in the basal plane due to the large size of the double vacancy compared to the dopant’s atomic radius. The considered non-metallic elements B and N feature benzene-like hexagons with two out of the four undercoordinated carbon atoms. According to the formation free energy analysis, all M_DV graphene structures are highly stable with formation energies ranging from −1.43 eV to −5.28 eV. The tri-vacancy of M2_TV models can accommodate the dopant pair in two different orientations: 1) both dopant atoms above the tri-vacancy (Supplementary Figure S4A) and 2) one atom above and one below the tri-vacancy (Supplementary Figure S4B). We found that the latter geometry is more stable (see Supplementary Figure S4C) and was selected for further studies. In the M2_TV structure, the dopant atoms are bound to the three nearest undercoordinated carbon atoms. Non-metallic B and N dimers break down during structural optimization to form benzene-like structures, and the resulting structures mimic mono-vacant graphenes. The computed formation free energies show that non-metallic dopants in the M2_TV structure are highly stable, while 3.1.3 N-Coordinated Graphene All the optimized M_3N and M_4N structures are highly symmetric and show no buckling behavior. We ﬁnd that the average M-N bond distances along with other geometrical features of both N-coordinated structures are similar to their carbon coordinated counterparts (M_SV and M_DV) (Zhao et al., 2019; Kattel et al., 2012). Our calculations suggest that the incorporation of three nitrogen atoms enhances the dopant’s stability as can be inferred from the formation free energies ranging from −1.22 eV to −3.28 eV, which are more exergonic than their M_SV counterparts. The formation free energies of M_4N structures vary from −1.84 eV to −2.99 eV and are comparable to the values computed for M_DV models. However, it is evident that the non-metallic dopants prefer 4C-coordination as their formation free energies (see Figure 2) are much more exergonic than in the 4N- coordination environment. Therefore, the 4N-coordination environment does not enhance the stability of metallic dopants compared to 4C-coordination in agreement with previous computational results (Zhao et al., 2019). Figure 3B illustrates the adsorption geometry of CO2 on Mn2_2SV. The adsorption geometry is similar on all the CO2- activating M2_2SV structures. CO2 adsorption is exergonic on B-, Co-, Mn-, and N-based 2SV structures with adsorption free energies varying from −0.18 eV to −0.68 eV. However, only Co2_2SV and Mn2_2SV are able to activate CO2 and facilitate modest electron transfer to the molecule. While Fe2_2SV and Ru2_2SV structures can also activate CO2, they exhibit thermoneutral adsorption. On the other hand, upon structural optimization, the coinage and a few base metal (Al, Ni, and Zn) dopants stem out of the vacancies and form CO2-dimer complex above graphene structure, and hence these systems were excluded from further studies. In general, the geometric and energetic stabilities of modiﬁed- graphene sheets depend strongly on the dopant and vacancy but some overall trends are observed. The Ag-, Au-, Cr-, and Mo- doped carbon-coordinated graphene structures are always unstable irrespective of the type of vacancy whereas all the considered dopants in 3N- and 4N-environments are highly stable, which is in line with previous studies (Tripkovic et al., 2013; Li et al., 2015; Ju et al., 2017; Pan et al., 2018; Varela et al., 2019; Wang et al., 2019; Zhang et al., 2019; Koshy et al., 2020). 3.1.1 Single and Two Mono-Vacant Graphene The M_SV graphenes are the simplest doped graphene structures (see Figure 1A). Calculations showed that all metal atoms exhibit slight protrusion out of the graphene plane upon bonding with the dangling carbon atoms, and the vertical distance between a metal atom and graphene plane varies from 0.73 Å to 1.19 Å in good agreement with previous computational studies (Li et al., 2015; Verma et al., 2018; Zhao et al., 2019). About half of the M_SV systems are thermodynamically stable as can be seen from Figure 2. The free energy of formation ranges from −0.19 to February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 6 Modiﬁed Graphene for CO2 Reduction Verma et al. metallic M2_TV electrodes are at best modestly stable, and Ag-, Au-, Cr-, Mo-, and Ru-doped graphenes are unstable. metallic M2_TV electrodes are at best modestly stable, and Ag-, Au-, Cr-, Mo-, and Ru-doped graphenes are unstable. Among the M_SV electrodes, CO2 adsorption is exergonic only on Fe_SV (−0.12 eV). In this case, CO2 remains linear upon adsorption (see Figure 3A) and the oxygen atom interacts weakly with the metal site having Fe–O distance 2.16 Å and small electron transfer from CO2 (0.22 e). A similar pattern is observed for the M_DV electrodes as well. The CO2 adsorption is exergonic only on Fe_DV (see Figure 3D) with similar binding conﬁguration and free energy as observed for Fe_SV; however, the charge transfer (0.47 e) from CO2 is rather high compared to Fe_SV. On the B- and N-doped SV sheets, CO2 merely physisorbs as seen from the long B-C and N-C distances, which are over 3.1 Å. This is similar to the CO2 adsorption on the pristine graphene (Rad and Foukolaei, 2015). The activation of CO2, i.e., non-linear CO2 structure, occurs only on Mn_SV and Cr- and Mo-based SV and DV structures but the CO2 adsorption is thermodynamically unfavorable. Note that the formation of Cr- and Mo-doped SV electrodes are thermodynamically unstable as well. 3.1.3 N-Coordinated Graphene Apart from Ag- and Cu-doped M_3N models, the majority of M_3N structures exhibit promising CO2 adsorption characteristics as CO2 both adsorbs and becomes activated with the O–C–O angle varying between 127.9o to 150.4o. The molecule interacts with the metal dopant via the carbon atom and one oxygen in a bidentate conformation shown in Figure 3C for the Mo_3N structure. In the case of non-metallic dopants, CO2 favors a monodentate adsorption conﬁguration. The free energies of adsorption range from −0.24 eV to −0.74 eV being comparable to the values found for the M2_2SV structures. CO2 adsorption on 4N-coordinated electrodes is surprisingly unfavorable: roughly half of the M_4N structures can activate CO2 but only Mo_4N (see Figure 3E) exhibits exergonic adsorption free energy (−0.89 eV) with exceptionally large charge transfer (−0.67 e). Similar to M_3N graphenes, CO2 favors a bidentate adsorption geometry on Mo_4N with O–C–O angle of 130.7o. Given the experimentally proven performance of 4N catalysts, it is unexpected to observe rather high endergonic adsorption energies on these electrodes. Possible reasons for the observed weak adsorption include, e.g., the neglected axial ligands and dipolar interactions as further elucidated in the Discussion section, or the deformation of the electrode structure upon adsorption. To test the latter, we computed the deformation energy for Fe_4N in the same way as in ref (Nykänen and Honkala, 2011). but the contribution from structural deformation is very small and therefore unfavourable 3.4 COOH and CO Binding: Scaling Relations CO and COOH adsorption energies typically exhibit scaling relations and are therefore often taken as CO2RR activity and selectivity descriptors (Peterson and Nørskov, 2012; Koper, 2013a; Kuhl et al., 2014; Nitopi et al., 2019). However, there are indications (Kim et al., 2014; Li et al., 2015; Back et al., 2017) that the CO/COOH scaling relations and descriptors are not valid for graphene-based electrodes. To study whether scaling relations can be established among the materials studied here, we attempt to ﬁnd correlations between the CO2, COOH, and CO adsorption free energies. This analysis is carried out for the M_3N and M2_2SV electrodes only as M_4N and M2_TV electrodes contain too few data points. The binding energies are presented in the Supplementary Table S11 and are utilized in the next section to study the electrochemical thermodynamics. To rank and identify prospective CO2RR electrocatalysts, we introduce a four-level selection criteria (see the ﬂow chart in Supplementary Figure S6). First, we screen all bare electrodes and select the ones with negative formation free energy but disregard the systems with geometric instability. The electrodes fulﬁlling the ﬁrst criterion are subjected to CO2 adsorption energy screening. Catalysts, over which the adsorption free energy is below or equal to 0.15 eV (i.e., more adsorbing) are selected. Third, after passing the screening of binding free energies, the next step is CO2 activation for which we set the threshold value of 150o for O–C–O angle. Fourth, after previous three criteria, we consider the net charge transfer due to the binding of CO2 to the electrode as another criteria. In this section, the systems demonstrating net charge transfer below or equal to 0.15 are selected. The upper range of 0.15 eV in the binding free energies of CO2 and in the net charge transfer is considered to account for any possible number sensitivity due to the DFT methods (Gauthier et al., 2020). This is an acceptable condition considering the fact that the electronic structure results often deviate by 0.1 eV due to different DFT functionals or basis sets. From Supplementary Figures S8, S9, it is evident that a poor correlation exists between the binding free energies of COOH and CO2 over M_3N (R2 0.07) and M2_2SV (R2 0.24) electrodes. Similar results are observed for COOH/CO scaling relationships on M_3N (R2 0.29) and M2_2SV (R2 0.08) electrodes as well. 3.3 Selection Criteria After analyzing the electrode stabilities and CO2 adsorption, we develop the selection criteria for identifying the most promising modiﬁed graphene CO2RR electrocatalysts from a large pool of materials. For instance, the doping of heteroatom to the graphene electrode should be highly stable to avoid any possible dissolution. The binding of CO2 to the electrode material should be strong enough to avoid being the limiting step at industrially relevant current densities/potentials. Furthermore, at high pH conditions, Tafel slopes approach 120 mV/dec (Gabardo et al., 2018), which is indicative of electron-transfer (ET) step being the limiting step (Dunwell et al., 2018). Therefore, CO2 should be activated upon adsorption, become negatively charged, and bind strongly: these features are ranked based on O-C-O angle, negative charge transfer, and negative GAds values. 3.2 CO2 Adsorption p As discussed in the Introduction, CO2 adsorption and activation are crucial steps for CO2RR in alkaline conditions as the adsorption is often thermodynamically unfavorable and limits CO2RR activity (Jones et al., 2014; Gauthier et al., 2019; Yu et al., 2019; Vijay et al., 2020). Furthermore (partial) charge transfer during CO2 adsorption may take place and this process is sensitive to the electrode potential but formally independent of solution pH. CO2 may also be activated upon adsorption, which is seen as the deviation of the O–C–O angle from 180o. Under alkaline conditions, where only weak proton donors (water) are present, the CO2 adsorption and an accompanied ET step are important contrary to pure PCET steps. Therefore, we closely monitor the CO2 adsorption geometry and energy over all of the considered electrodes, regardless of electrode stability. Figure 3 displays CO2 binding geometries on the stable M_SV, M2_2SV, M_3N, M_DV, and M_4N electrodes whereas binding geometries on stable M2_SV, MPt_SV, and M2_TV structures are illustrated in Supplementary Figure S5. Electronic structure information such as magnetic moments and stable spin multiplicities are supplied in Supplementary Table S6 for each system, and numerical values for adsorption energies, enthalpies, and free energies are provided in the Supplementary Tables S7–S10. February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 7 Verma et al. Modiﬁed Graphene for CO2 Reduction FIGURE 3 \| The optimized binding geometries of CO2 over considered electrodes: (A) Fe_SV, (B) Mn2_2SV, (C) Mo_3N, (D) Fe_DV, and (E) Mo_4N. The elements carbon, oxygen, and nitrogen are in gray, red, and blue colors, respectively. FIGURE 3 \| The optimized binding geometries of CO2 over considered electrodes: (A) Fe_SV, (B) Mn2_2SV, (C) Mo_3N, (D) Fe_DV, and (E) Mo_4N. The elements carbon, oxygen, and nitrogen are in gray, red, and blue colors, respectively. M_DV, M2_2SV, M2_TV, M_3N, and M_4N, is supplied in Supplementary Figure S7. Based on the selection criteria, the qualiﬁed systems to study the electrochemical thermodynamics are: M2_2SV (Co, Mn, Rh, Fe, and Ru), M2_TV (Al, Cu, and Zn), M_3N (Al, N, Rh, Cr, Fe, Mo, Pd, Pt, and Ru), and M_4N (Mo and Ru). Thus, a total of 19 electrodes are selected for further study. Note that none of M_SV and M_DV electrodes qualify the present selection criteria. adsorption energies cannot be attributed to structural deformations. Frontiers in Energy Research \| www.frontiersin.org 3.4 COOH and CO Binding: Scaling Relations These poor scaling relations are in agreement with previous studies on graphene-based electrodes (Li et al., 2015; Back et al., 2017; Li et al., 2019; Guo et al., 2020), which indicates that graphene electrodes may break scaling relationships making these materials an interesting class of CO2RR electrocatalysts. However, it needs to be recognized that the number of materials subjected to the scaling relation studies is small and too far-reaching conclusions on breaking scaling should be avoided. Yet, it has been shown that even slight modiﬁcations to N-doped graphene electrodes lead to different scaling relations (Guo et al., 2020). A combined plot of three key selection parameters, i.e., adsorption free energy (GAds), O–C–O angle, and charge transfer (q) for each of the electrode categories, namely, M_SV, February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 8 Modiﬁed Graphene for CO2 Reduction Verma et al. FIGURE 4 \| Electrochemical thermodynamics of CO2 reduction reaction using PCET and ETPT mechanisms over four selected electrodes: (A) Pt_3N, (B) Al_3N, (C) Ru2_2SV, and (D) Mo_4N. FIGURE 4 \| Electrochemical thermodynamics of CO2 reduction reaction using PCET and ETPT mechanisms over four selected electrodes: (A) Pt_3N, (B) Al_3N, (C) Ru2_2SV, and (D) Mo_4N. 5 Electrochemical Thermodynamics dd h l h i l f f h However, the CO adsorption free energy (∼−1 eV) February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 9 Modiﬁed Graphene for CO2 Reduction Verma et al. FIGURE 5 \| Calculated theoretical limiting ((A,C), 0 to −2.0 V vs. RHE) and overpotential ((B,D),0–1.5 V vs. RHE) of electrocatalytic CO2RR process via PCET (A,B) and ETPT (C,D) mechanisms. FIGURE 5 \| Calculated theoretical limiting ((A,C), 0 to −2.0 V vs. RHE) and overpotential ((B,D),0–1.5 V vs. RHE) of electrocatalytic CO2RR process via PCET (A,B) and ETPT (C,D) mechanisms. etical limiting ((A,C), 0 to −2.0 V vs. RHE) and overpotential ((B,D),0–1.5 V vs. RHE) of electrocatalytic CO2RR process via PCET (A,B) s on the electrode is too strong, which makes it susceptible to catalyzing further reduction or poisoning. they are absent from PCET due to the assumption of simultaneous electron and proton transfer. Overall, these differences modify the reaction thermodynamics, and we observed that the charge transfer during adsorption has a profound effect on the electrocatalytic thermodynamics as The last example, given in Figure 4D, features very different limiting potentials for the ETPT and PCET pathways on the Mo_4N electrode. In this case, CO2 adsorbs strongly, signiﬁcant charge transfer (∼−0.7 e) to the molecule takes place, and the COOH formation is quite endergonic at zero potential (red line). Together these factors result in a large negative limiting potential along the ETPT pathway, while the PCET exhibits a fairly modest overpotential. Similar behavior is seen for the Mo_3N electrode as well and we conclude that the N-Mo-modiﬁed graphene catalysts are unsuitable for a ETPT mechanism but could work for the PCET pathway. TABLE 1 \| Thermodynamic overpotentials in the presence of implicit solvation (ηaq) (V vs. RHE) for the three promising ETPT and PCET electrocatalysts. Also, the magnitude of deviation (Δη) from the vacuum phase (ηvac) is indicated. Electrode ηaq Δη = ηvac −ηaq ETPT PCET ETPT PCET Pt_3N 0.75 0.15 0.05 −0.04 Ru_3N 0.63 0.04 −0.02 0.05 Ru_4N 0.28 0.11 0.00 0.16 February 2021 \| Volume 8 \| Article 606742 0 TABLE 1 \| Thermodynamic overpotentials in the presence of implicit solvation (ηaq) (V vs. RHE) for the three promising ETPT and PCET electrocatalysts. Also, the magnitude of deviation (Δη) from the vacuum phase (ηvac) is indicated. 5 Electrochemical Thermodynamics dd h l h i l f f h from ideal and a large negative limiting potential is required due to the highly endergonic CO2→COOH step and the signiﬁcant partial charge transfer (∼0.7 e) in this step. 3.5 Electrochemical Thermodynamics To address the electrochemical performance of the qualiﬁed electrodes, electrochemical thermodynamics for CO2 reduction to CO were computed using Equations 4, 6. While Figure 4 displays the thermodynamic proﬁles for selected materials with varying behavior along the PCET and ETPT pathways, the limiting potentials and overpotentials are illustrated in Figure 5. For other materials, not shown here, the potential energy surfaces are displayed in Supplementary Figure S10 and the explicit potential values are collected in Supplementary Table S12. CO2RR thermodynamics on the Al_3N catalysts, shown in Figure 4B, provides an example where both the ETPT and PCET pathways attain same limiting potential. Herein, CO2 adsorption energy is modestly exergonic and features insigniﬁcant charge transfer. Furthermore, CO formation is highly unfavourable compared to COOH and is therefore identiﬁed as the potential- determining step. As the reaction free energy for CO formation (see Equation 4c) is same for both PCET and ETPT pathways, the resulting thermodynamic potentials are also equal. As displayed in Figure 4A, a large potential difference (0.69 V) between the ETPT and PCET pathways is observed for the Pt_3N catalyst. We attribute this difference to the inclusion of the CO2 binding in the former mechanism. The CO2 adsorption step itself is rather exergonic and associated with a modest charge transfer (∼−0.3 e) from the electrode. Along the PCET pathway, COOH and CO formation steps are mildly exergonic at zero potential resulting in positive equilibrium potential. This observation suggests that PCET pathway on Pt_3N is nearly thermodynamically ideal and requires zero limiting potential (blue PES). On the other hand, the ETPT mechanism is far In Figure 4C, the Ru2_2SV electrode is considered. CO2 adsorption is almost thermoneutral at zero potential (red line) associated with a small charge transfer (∼−0.2 e) from the electrode. The formation of COOH is identiﬁed as the limiting step owing to its high endergonicity. Altogether these factors make ETPT and PCET limiting potentials rather similar, which in turn implies that this material is a promising candidate for both ETPT and PCET pathways, i.e., for both alkaline and acid conditions. February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 4 DISCUSSION Our computational predictions show that the CO2RR activity and selectivity on a given graphene-based electrode depend on the mechanism, i.e., whether the PCET or ETPT pathway is followed. For conditions favoring the PCET mechanism, the N-coordinated graphene electrodes demonstrate the best performance among the studied models. The aforementioned holds true also when ETPT is operational but this typically leads to larger limiting potentials. Pd_3N is an exception to this rule and is the only electrode for which a lower overpotential along the ETPT than the PCET pathway is observed. This anomaly is caused by the slightly positive CO2 binding energy and minor charge transfer. In general, we ﬁnd that the coupled PCET is thermodynamically more favorable but for kinetic reasons the ETPT pathway may be preferred (Koper, 2013b). The thermodynamic potentials are extensively used for ranking or comparing the expected performance of different catalysts, and limiting potentials and overpotentials should be close to zero for the best performing catalysts. Figure 5 presents both potentials for the PCET and ETPT pathways. For all the qualiﬁed electrodes, the overpotentials are within 0.7 V (vs. RHE) for PCET pathways. The best-performing electrodes include Ru, Rh, and Mo-doped 3N graphene and Ru and Mo-doped 4N graphene for which the limiting potentials vary from 0 to −0.52 V vs. RHE and overpotentials range from 0.06 to 0.21 V vs. RHE. For the ETPT mechanism, the best-performing electrodes contain Pd_3N, Ru_3N, Pt_3N, Ru2_2SV, and Ru_4N with limiting potentials between −0.48 and −0.86 V vs. RHE and overpotentials varying from 0.28 to 0.8 V vs. RHE. The limiting potentials found for the studied structures are either comparable or even smaller than those previously computed for 4N and 3N catalysts (Guo et al., 2020; Pan et al., 2020), non-metallic defected graphene (Siahrostami et al., 2017), and single-atom doped metal catalysts (Lim et al., 2014). Overall, the ETPT mechanism leads to higher overpotentials and limiting potentials than the PCET pathway, which we take to indicate that pure ET steps are detrimental for the reaction thermodynamics as commonly suggested (Koper, 2013b). Interestingly, we ﬁnd that three of the studied materials–Ru_3N, Pt_3N, and Ru_4N–exhibit very promising electrochemical thermodynamics along both the ETPT and PCET pathways making these materials promising candidates for CO2RR under both acidic (PCET) and alkaline (ETPT) conditions. y y The obtained results show that the ETPT pathway is highly sensitive to the CO2 electrosorption energy and charge transfer. 4 DISCUSSION Ignoring these important features from mechanistic consideration and focusing solely on the PCET mechanism, often leads to a different potential-determining step, and underestimation of thermodynamic potentials. For example, the current and previous computational studies (Li et al., 2019; Vijay et al., 2020) on Fe_4N show that the CO2 binding itself introduces a thermodynamic barrier of ∼0.9 eV, whereas, the PCET pathway (Guo et al., 2020) presents a quite small (∼0.1 eV) thermodynamic barrier for producing COOH. These results demonstrate that CO2 adsorption directly affects the CO2RR elementary thermodynamics and should therefore be considered. The adsorption process and its potential- dependency are particularly important under alkaline conditions where the coupled PCET becomes less likely, and at high current densities where mass transfer and adsorption are the limiting processes. We identiﬁed that slightly exergonic CO2 adsorption associated with minor charge transfer is a desirable feature for promising catalysts. The conclusions from a large number of studies (see Introduction) neglecting the adsorption step and focusing only on PCET steps are limited to acidic conditions or to the electrodes that cannot catalyze the ETPT pathway. Next, we simulated the three promising ETPT and PCET electrocatalysts (Ru_3N, Pt_3N, and Ru_4N) using the PCM model (Scalmani and Frisch, 2010) for water to check the solvation effects on the overpotentials, and their deviations from the vacuum phase thermodynamics. The adsorption free energies of CO2 in the aqueous medium are lower (GAds(aq) Pt_3N −0.49 eV; GAds(aq) Ru_3N −0.08 eV; and GAds(aq) Ru_4N 0.02 eV) than those of computed for the vacuum phase, but charge transfer from electrodes to the CO2 in the solution phase is signiﬁcantly higher (−0.5 e to −0.6 e). The aqueous phase PCET and ETPT limiting potentials of Pt_3N and Ru_3N are slightly lower compared to vacuum phase, whereas, slightly higher limiting potentials are observed for Ru_4N. As shown in Table 1, including solvation introduces only small changes to The importance of the adsorption step can be further illustrated by the widely studied Fe_4N electrode. Our results suggest that the Fe_4N electrode (Zhang et al., 2018; Li et al., 2019; Vijay et al., 2020) does not belong to the best performing CO2RR materials. We attribute this to the thermodynamically unfavorable CO2 adsorption step, which agrees with the recent ﬁnding that the adsorption step on Fe_4N is thermodynamically uphill even under high electrode potentials and ﬁeld strengths (Vijay et al., 2020). 5 Electrochemical Thermodynamics dd h l h i l f f h p y Comparison of the ETPT potential energy proﬁles given in Figure 4 shows that CO2 adsorption and associated charge transfer are pivotal for this mechanism to be operational, but Frontiers in Energy Research \| www.frontiersin.org 10 Modiﬁed Graphene for CO2 Reduction Verma et al. manifested by the limiting potentials shown in Figures 4, 5. We found that the slightly exergnonic CO2 adsorption is beneﬁcial as it enables the formation of a stable CO2-catalyst complex and prevents excessively endergonic COOH formation. Furthermore, charge transfer associated with CO2 adsorption should be as low as possible: if the COOH formation requires only a very small degree of charge transfer, a very high reductive potential is needed to make this step thermodynamically favorable as shown in, e.g., Figure 4D. In a case where no charge transfer takes place during CO− 2 + H+ →COOH, the COOH formation step is fully chemical and cannot be controlled by the electrode potential. The very large limiting potentials along the ETPT pathways are likely an artifact of the simple charge transfer and electrosorption model adopted herein, but we believe that the present model enables qualitatively correct comparison of thermodynamics between ETPT and PCET processes. overpotentials. The largest change is observed for Ru_4N (Δη 0.16 V), which is still relatively small given typical DFT inaccuracies. We ascribe the observed difference to slightly more stable adsorption of COOH in the solution phase compared to vacuum phase. Finally, it can be concluded that the implicit solvation framework does not signiﬁcantly change the thermodynamics computed in gas-phase, and thus the conclusions remain largely unaffected by the implicit solvation. 5 CONCLUSION adsorption and mass transfer at industrially relevant current densities (Vijay et al., 2020). Even though our decoupled pathway and CO2RR analyses are based on a relatively simple and ideal electrosorption valency concept, the identiﬁcation of CO2 adsorption as a crucial step is expected to be rather general. In particular, our approach provides a fast approach to identify a catalyst, where ET during adsorption is an important feature. In this work, we identiﬁed promising modiﬁed graphene electrodes for CO2RR to CO under alkaline conditions. Robust selection criteria based on thermodynamic stability, CO2 adsorption thermodynamics, and potential-dependent reaction free energies were devised and applied. The computational hydrogen electrode concept was extended to treat decoupled PCET steps at electrode surfaces to account for the possible decoupled ETPT mechanism in alkaline conditions. We utilized this development to evaluate the effect of pure charge transfer during adsorption, i.e., electrosorption, and found that a high degree of partial charge transfer during CO2 is detrimental to electrocatalyst performance and that moderately strong CO2 adsorption energy without charge transfer leads to promising electrode materials. We identiﬁed metal sites coordinated to three nitrogen atoms (M_3N) and two single vacancy metal (M2_2SV) electrodes as highly promising materials for CO2RR following either the coupled or decoupled pathways. N-coordinated Ru and Pt electrodes exhibit promising characteristics for both coupled and decoupled pathways making these materials interesting candidates as pH-universal CO2RR electrodes and call for further experimental and computational studies. y g p p In addition to accounting for ETPT and electrosorption, future studies should also consider the role of “innocent” ligands introduced by the pH or the supporting electrolyte. A recent joint experimental and computational work (Li et al., 2019) has shown that axial ligands are important in determining the CO2RR to CO performance on M_4N-type materials. While the experimental results exhibit a high current density and Faradaic efﬁciency toward CO at applied potentials of −0.5 and −0.6 V vs. RHE, the calculations predict a signiﬁcant thermodynamic barrier of ∼0.9 eV for CO2 adsorption in agreement with our results. As the initial computational results were not in line with the experimental ﬁndings, the effect of axial adsorption of H2O and OH was considered. While the adsorption energy was found to be further destabilized with axial OH adsorption, the axially adsorbed H2O stabilized CO2 adsorption on Fe_4N by ∼0.3 eV. DATA AVAILABILITY STATEMENT The datasets presented in this study can be accessed at https:// gitlab.com/ 1341 compcatjyu/co2rr-on-graphene-electrodes. 5 CONCLUSION Including the axial H2O ligand in the computational Fe_4N model resulted in a better agreement with experiments as the CO2 adsorption posed only a small thermodynamic barrier (0.06 eV) at −0.6 V vs. RHE. Furthermore, the simulated Pourbaix diagrams (Li et al., 2019) conﬁrmed that Fe_4N electrode may bind axial ligands on both sides. While the kinetic role of the axial ligands was not considered for CO2RR, the oxygen reduction reaction kinetics were shown to be very sensitive to the presence of “innocent” axial ligands (Rebarchik et al., 2020). ACKNOWLEDGMENTS The computational resources were provided by the CSC–IT Center for Science, Espoo, Finland (https://www.csc.ﬁ/en/) and the FGCI–Finnish Grid and Cloud Infrastructure. FUNDING MM was supported by the Academy of Finland grants 307853 and 317739. AV and KH were supported by the Academy of Finland grant 317739. AUTHOR CONTRIBUTIONS AV performed the calculations and analysed the data. MM developed the thermodynamic model and oversaw the analysis. KH oversaw the calculations and analysis. All authors contributed to conceptualising the work and writing the manuscript. ( ) Previous studies (Peterson and Nørskov, 2012; Kuhl et al., 2014; Nitopi et al., 2019) suggested that CO binding strength is a descriptor that determines the CO2RR product distribution. Materials with strong CO adsorption are either poisoned or produce C1- or C2-species or form hydrogen via the competing HER. On the contrary, electrodes binding CO weakly yield CO as the major product due to favourable CO desorption kinetics. Apart from Al-, Cr-, and Pd-dopants, the majority of 3N-coordinated electrodes bind CO quite strongly suggesting that they might be poisoned by CO. The CO adsorption on Al_3N (ηETPT/PCET 0.69 V) is weak and leads to facile CO desorption. Intermediate CO adsorption energies on Cr_3N and Pd_3N suggest that these two materials are promising electrocatalysts for producing C1- or C2- molecules; however, the former suffers from rather high limiting potential making Pd_3N the only 3N-graphene electrode suitable for further CO reduction. With the exception of Ru2_2SV, all the other 2SV electrodes exhibit CO binding strength within ∼−0.7 eV indicating that they will primarily produce CO. The M2_2SV electrodes show low overpotentials but feature high negative limiting potentials (above −1V) except for Ru2_2SV, which is identiﬁed as the most promising M2_2SV structure. Both M_4N and M2_TV electrodes suffer from either strong CO binding or high limiting potentials, and are therefore unsuitable for CO production. 4 DISCUSSION Therefore, CO2RR on Fe_4N is limited by CO2 February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 11 Modiﬁed Graphene for CO2 Reduction Verma et al. REFERENCES 9, 7894–7899. doi:10.1021/acscatal.9b01899 Bohra, D., Chaudhry, J. H., Burdyny, T., Pidko, E. A., and Smith, W. A. (2019). Modeling the electrical double layer to understand the reaction environment in a CO electrocatalytic system. Energy Environ. Sci. 12, 3380–3389. doi:10.1039/ C9EE02485A Guo, S.-X., Li, F., Chen, L., MacFarlane, D. R., and Zhang, J. (2018). Polyoxometalate-promoted electrocatalytic CO reduction at nanostructured silver in dimethylformamide. ACS Appl. Mater. Interfaces 10, 12690–12697. doi:10.1021/acsami.8b01042 Hansen, H. A., Varley, J. B., Peterson, A. A., and Nørskov, J. K. (2013). Understanding trends in the electrocatalytic activity of metals and enzymes for CO reduction to CO. J. Phys. Chem. Lett. 4, 388–392. doi:10.1021/jz3021155 Bratsch, S. G. (1989). Standard electrode potentials and temperature coefﬁcients in water at 298.15 K. J. Phys. Chem. Ref. Data 18, 1–21. doi:10.1063/1.555839 Brownson, D. A. C., Kampouris, D. K., and Banks, C. E. (2012). Graphene electrochemistry: fundamental concepts through to prominent applications. Chem. Soc. Rev. 41, 6944–6976. doi:10.1039/C2CS35105F Hay, P. J., and Wadt, W. R. (1985a). Ab initio effective core potentials for molecular calculations. potentials for K to Au including the outermost core orbitals. J. Chem. Phys. 82, 299–310. doi:10.1063/1.448975 Busch, M., Halck, N. B., Kramm, U. I., Siahrostami, S., Krtil, P., and Rossmeisl, J. (2016). Beyond the top of the volcano? – a uniﬁed approach to electrocatalytic oxygen reduction and oxygen evolution. Nano. Energy 29, 126–135. doi:10. 1016/j.nanoen.2016.04.011 Hay, P. J., and Wadt, W. R. (1985b). Ab initio effective core potentials for molecular calculations. potentials for the transition metal atoms Sc to Hg. J. Chem. Phys. 82, 270–283. doi:10.1063/1.448799 Calle-Vallejo, F., Krabbe, A., and García-Lastra, J. M. (2017). How covalence breaks adsorption-energy scaling relations and solvation restores them. Chem. Sci. 8, 124–130. doi:10.1039/C6SC02123A He, Z., He, K., Robertson, A. W., Kirkland, A. I., Kim, D., Ihm, J., et al. (2014). Atomic structure and dynamics of metal dopant pairs in graphene. Nano Lett. 14, 3766–3772. doi:10.1021/nl500682j Chen, Y., Li, C. W., and Kanan, M. W. (2012). Aqueous CO reduction at very low overpotential on oxide-derived Au nanoparticles. J. Am. Chem. Soc. 134, 19969–19972. doi:10.1021/ja309317u Heenen, H. H., Gauthier, J. A., Kristoffersen, H. H., Ludwig, T., and Chan, K. (2020). Solvation at metal/water interfaces: an ab initio molecular dynamics benchmark of common computational approaches. J. Chem. Phys. 152, 144703. doi:10.1063/1.5144912 Cheng, Y., Yang, S., Jiang, S. P., and Wang, S. (2019). Supported single atoms as new class of catalysts for electrochemical reduction of carbon dioxide. REFERENCES efﬁcient CO electroreduction to CO. Energy Environ. Sci. 11, 2531–2539. doi:10.1039/C8EE01684D efﬁcient CO electroreduction to CO. Energy Environ. Sci. 11, 2531–2539. doi:10.1039/C8EE01684D Gao, D., McCrum, I. T., Deo, S., Choi, Y.-W., Scholten, F., Wan, W., et al. (2018). Activity and selectivity control in CO electroreduction to multicarbon products over CuOx catalysts via electrolyte design. ACS Catal. 8, 10012–10020. doi:10. 1021/acscatal.8b0258 Akhade, S. A., Luo, W., Nie, X., Bernstein, N. J., Asthagiri, A., and Janik, M. J. (2014). Poisoning effect of adsorbed CO during CO electroreduction on late transition metals. Phys. Chem. Chem. Phys. 16, 20429–20435. doi:10.1039/ C4CP03340J Gattrell, M., Gupta, N., and Co, A. (2006). A review of the aqueous electrochemical reduction of CO to hydrocarbons at copper. J. Electroanal. Chem. 594, 1–19. doi:10.1016/j.jelechem.2006.05.013 Aljabour, A., Coskun, H., Apaydin, D. H., Ozel, F., Hassel, A. W., Stadler, P., et al. (2018). Nanoﬁbrous cobalt oxide for electrocatalysis of CO reduction to carbon monoxide and formate in an acetonitrile-water electrolyte solution. Appl. Catal. B. 229, 163–170. doi:10.1016/j.apcatb.2018.02.017 Gauthier, J. A., Chen, L. D., Bajdich, M., and Chan, K. (2020). Implications of the fractional charge of hydroxide at the electrochemical interface. Phys. Chem. Chem. Phys. 22, 6964–6969. doi:10.1039/C9CP05952K Back, S., Kim, H., and Jung, Y. (2015a). Selective heterogeneous CO electroreduction to methanol. ACS Catal. 5, 965–971. doi:10.1021/cs501600x Gauthier, J. A., Fields, M., Bajdich, M., Chen, L. D., Sandberg, R. B., Chan, K., et al. (2019). Facile electron transfer to CO during adsorption at the metal— solution interface. J. Phys. Chem. C. 123, 29278–29283. doi:10.1021/acs.jpcc.9b10205 Back, S., Yeom, M. S., and Jung, Y. (2015b). Active sites of Au and Ag nanoparticle catalysts for CO electroreduction to CO. ACS Catal. 5, 5089–5096. doi:10.1021/ acscatal.5b00462 Grimme, S., Antony, J., Ehrlich, S., and Krieg, H. (2010). A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, 154104. doi:10.1063/1.3382344 Back, S., Lim, J., Kim, N.-Y., Kim, Y.-H., and Jung, Y. (2017). Single-atom catalysts for CO electroreduction with signiﬁcant activity and selectivity improvements. Chem. Sci. 8, 1090–1096. doi:10.1039/C6SC03911A Guo, C., Zhang, T., Liang, X., Deng, X., Guo, W., Wang, Z., et al. (2020). Single transition metal atoms on nitrogen-doped carbon for CO electrocatalytic reduction: CO production or further CO reduction? Appl. Surf. Sci. 533, 147466. doi:10.1016/j.apsusc.2020.147466 Bagger, A., Ju, W., Varela, A. S., Strasser, P., and Rossmeisl, J. (2019). Electrochemical CO reduction: classifying Cu facets. ACS Catal. REFERENCES Small Methods 3, 1800440. doi:10.1002/smtd.201800440 Herrick, D. R. (2005). Connecting pauling and mulliken electronegativities. J. Chem. Theor. Comput. 1, 255–260. doi:10.1021/ct049942a Dennington, R., Keith, T. A., and Millam, J. M. (2009). Gaussview Version 5. Semichem Inc. Shawnee Mission KS. Hohenberg, P., and Kohn, W. (1964). Inhomogeneous electron gas. Phys. Rev. 136, B864. doi:10.1103/PhysRev.136.B864 DeWulf, D. W., Jin, T., and Bard, A. J. (1989). Electrochemical and surface studies of carbon dioxide reduction to methane and ethylene at copper electrodes in aqueous solutions. J. Electrochem. Soc. 136, 1686. doi:10.1149/1.2096993 Hori, Y. i. (2008). “Electrochemical CO reduction on metal electrodes,” in Modern aspects of electrochemistry (New York, NY: Springer), 89–189. Hori, Y., Kikuchi, K., Murata, A., and Suzuki, S. (1986). Production of methane and ethylene in electrochemical reduction of carbon dioxide at copper electrode in aqueous hydrogencarbonate solution. Chem. Lett. 15, 897–898. doi:10.1246/cl. 1986.897 Dunwell, M., Luc, W., Yan, Y., Jiao, F., and Xu, B. (2018). Understanding surface- mediated electrochemical reactions: CO reduction and beyond. ACS Catal. 8, 8121–8129. doi:10.1021/acscatal.8b02181 Durand, W. J., Peterson, A. A., Studt, F., Abild-Pedersen, F., and Nørskov, J. K. (2011). Structure effects on the energetics of the electrochemical reduction of CO by copper surfaces. Surf. Sci. 605, 1354–1359. doi:10.1016/j.susc.2011.04.028 Hori, Y., Murata, A., and Takahashi, R. (1989). Formation of hydrocarbons in the electrochemical reduction of carbon dioxide at a copper electrode in aqueous solution. J. Chem. Soc. Faraday Trans. 1 85, 2309–2326. doi:10.1039/ F19898502309 Ferrante, F., Prestianni, A., Cortese, R., Schimmenti, R., and Duca, D. (2016). Density functional theory investigation on the nucleation of homo-and heteronuclear metal clusters on defective graphene. J. Phys. Chem. C. 120, 12022–12031. doi:10.1021/acs.jpcc.6b02833 Hörmann, N. G., Marzari, N., and Reuter, K. (2020). Electrosorption at metal surfaces from ﬁrst principles. Npj Comput. Mater. 6, 136. doi:10.1038/s41524- 020-00394-4 Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., et al. (2009). GaussianËœ09 revision B.01. Wallingford CT: Gaussian Inc. Jia, C., Dastafkan, K., Ren, W., Yang, W., and Zhao, C. (2019). Carbon-based catalysts for electrochemical CO reduction. Sust. Energy Fuels 3, 2890–2906. doi:10.1039/C9SE00527G Frydendal, R., Busch, M., Halck, N. B., Paoli, E. A., Krtil, P., Chorkendorff, I., et al. (2015). Enhancing activity for the oxygen evolution reaction: the beneﬁcial interaction of gold with manganese and cobalt oxides. ChemCatChem. 7, 149–154. doi:10.1002/cctc.201402756 Jia, M., Fan, Q., Liu, S., Qiu, J., and Sun, Z. (2019). SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenrg.2020.606742/ full#supplementary-material. February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 12 Modiﬁed Graphene for CO2 Reduction Verma et al. REFERENCES doi:10.1021/ja403162r Nykänen, L., and Honkala, K. (2011). Density functional theory study on propane and propene adsorption on Pt(111) and PtSn alloy surfaces. J. Phys. Chem. C. 115, 9578–9586. doi:10.1021/jp1121799 Pan, F., Li, B., Sarnello, E., Fei, Y., Feng, X., Gang, Y., et al. (2020). Pore-edge tailoring of single atomic iron-nitrogen sites on graphene for enhanced CO reduction. ACS Catal. 10, 10803–10811. doi:10.1021/acscatal.0c02499 Lee, M.-Y., Ringe, S., Kim, H., Kang, S., and Kwon, Y. (2020). Electric ﬁeld mediated selectivity switching of electrochemical CO reduction from formate to CO on carbon supported Sn. ACS Energy Lett. 5, 2987–2994. doi:10.1021/ acsenergylett.0c01387 Pan, F., Zhang, H., Liu, K., Cullen, D., More, K., Wang, M., et al. (2018). Unveiling active sites of CO reduction on nitrogen-coordinated and atomically dispersed iron and cobalt catalysts. ACS Catal. 8, 3116–3122. doi:10.1021/acscatal.8b00398 Li, J., Prˇslja, P., Shinagawa, T., Martín Fernández, A. J., Krumeich, F., Artyushkova, K., et al. (2019). Volcano trend in electrocatalytic CO reduction activity over atomically dispersed metal sites on nitrogen-doped carbon. ACS Catal. 9, 10426–10439. doi:10.1021/acscatal.9b02594 Perdew, J. P., Ziesche, P., and Eschrig, H. (1991). Electronic structure of solids’91. Berlin: Akademie Verlag. Li, X., Bi, W., Chen, M., Sun, Y., Ju, H., Yan, W., et al. (2017). Exclusive Ni–N sites realize near-unity CO selectivity for electrochemical CO reduction. J. Am. Chem. Soc. 139, 14889–14892. doi:10.1021/jacs.7b09074 Pérez-Gallent, E., Marcandalli, G., Figueiredo, M. C., Calle-Vallejo, F., and Koper, M. T. (2017). Structure-and potential-dependent cation effects on CO reduction at copper single-crystal electrodes. J. Am. Chem. Soc. 139, 16412–16419. doi:10.1021/jacs.7b10142 Li, Y., Su, H., Chan, S. H., and Sun, Q. (2015). CO electroreduction performance of transition metal dimers supported on graphene: a theoretical study. ACS Catal. 5, 6658–6664. doi:10.1021/acscatal.5b01165 Peterson, A. A., Abild-Pedersen, F., Studt, F., Rossmeisl, J., and Nørskov, J. K. (2010). How copper catalyzes the electroreduction of carbon dioxide into hydrocarbon fuels. Energy Environ. Sci. 3, 1311–1315. doi:10.1039/C0EE00071J Lim, H.-K., Shin, H., Goddard, W. A., Hwang, Y. J., Min, B. K., and Kim, H. (2014). Embedding covalency into metal catalysts for efﬁcient electrochemical conversion of CO. J. Am. Chem. Soc. 136, 11355–11361. doi:10.1021/ja503782w Peterson, A. A., and Nørskov, J. K. (2012). Activity descriptors for CO electroreduction to methane on transition-metal catalysts. J. Phys. Chem. Lett. 3, 251–258. doi:10.1021/jz201461p Lindgren, P., Kastlunger, G., and Peterson, A. A. (2020). A challenge to the G 0 interpretation of hydrogen evolution. ACS Catal. 10, 121–128. REFERENCES Chem. Interfacial Electrochem. 245, 223–244. doi:10.1016/0022-0728(88)80071-8 Kittel, C. (2004). Introduction to solid state physics. New York, NY: Wiley. Mermin, N. D. (1965). Thermal properties of the inhomogeneous electron gas. Phys. Rev. 137, A1441–A1443. doi:10.1103/PhysRev.137.A1441 Kohn, W., and Sham, L. J. (1965). Self-consistent equations including exchange and correlation effects. Phys. Rev. 140, A1133. doi:10.1103/PhysRev.140.A1133 Koper, M. T. M. (2013a). Theory of multiple proton-electron transfer reactions and its implications for electrocatalysis. Chem. Sci. 4, 2710–2723. doi:10.1039/ C3SC50205H Mistry, H., Reske, R., Zeng, Z., Zhao, Z.-J., Greeley, J., Strasser, P., et al. (2014). Exceptional size-dependent activity enhancement in the electroreduction of CO over Au nanoparticles. J. Am. Chem. Soc. 136, 16473–16476. doi:10.1021/ja309317u Mulliken, R. S. (1934). A new electroafﬁnity scale; Together with data on valence states and on valence ionization potentials and electron afﬁnities. J. Chem. Phys. 2, 782–793. doi:10.1063/1.1749394 Koper, M. T. M. (2013b). Theory of the transition from sequential to concerted electrochemical proton–electron transfer. Phys. Chem. Chem. Phys. 15, 1399–1407. doi:10.1039/C2CP42369C Mulliken, R. S. (1955). Electronic population analysis on LCAO-MO molecular wave functions. I. J. Chem. Phys. 23, 1833–1840. doi:10.1063/1.1740588 Koshy, D. M., Chen, S., Lee, D. U., Stevens, M. B., Abdellah, A. M., Dull, S. M., et al. (2020). Understanding the origin of highly selective CO electroreduction to CO on Ni, N-doped carbon catalysts. Angew. Chem. Int. Ed. 132, 4072–4079. doi:10. 1002/anie.201912857 NIST database (2020). Nist webbook. Available at: https://webbook.nist.gov/cgi/ cbook.cgi?IDC124389. Krasheninnikov, A., Lehtinen, P., Foster, A. S., Pyykkö, P., and Nieminen, R. M. (2009). Embedding transition-metal atoms in graphene: structure, bonding, and magnetism. Phys. Rev. Lett. 102, 126807. doi:10.1103/physrevlett.102.126807 Nitopi, S., Bertheussen, E., Scott, S. B., Liu, X., Engstfeld, A. K., Horch, S., et al. (2019). Progress and perspectives of electrochemical CO reduction on copper in aqueous electrolyte. Chem. Rev. 119, 7610–7672. doi:10.1021/acs.chemrev. 8b00705 Kuhl, K. P., Hatsukade, T., Cave, E. R., Abram, D. N., Kibsgaard, J., and Jaramillo, T. F. (2014). Electrocatalytic conversion of carbon dioxide to methane and methanol on transition metal surfaces. J. Am. Chem. Soc. 136, 14107–14113. doi:10.1021/ja505791r Nørskov, J. K., Rossmeisl, J., Logadottir, A., Lindqvist, L., Kitchin, J. R., Bligaard, T., et al. (2004). Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B. 108, 17886–17892. doi:10.1021/jp047349j Lazar, P., Karlickỳ, F., Jureˇcka, P., Kocman, M., Otyepková, E., ˇSafáˇrová, K., et al. (2013). Adsorption of small organic molecules on graphene. J. Am. Chem. Soc. 135, 6372–6377. REFERENCES Single-atom catalysis for electrochemical CO reduction. Curr. Opin. Green Sustain. Chem. 16, 1–6. doi:10.1016/j.cogsc.2018.11.002 Jones, J.-P., Prakash, G. S., and Olah, G. A. (2014). Electrochemical CO reduction: recent advances and current trends. Isr. J. Chem. 54, 1451–1466. doi:10.1002/ ijch.201400081 Gabardo, C. M., Seiﬁtokaldani, A., Edwards, J. P., Dinh, C.-T., Burdyny, T., Kibria, M. G., et al. (2018). Combined high alkalinity and pressurization enable February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 13 Verma et al. Modiﬁed Graphene for CO2 Reduction Ju, W., Bagger, A., Hao, G.-P., Varela, A. S., Sinev, I., Bon, V., et al. (2017). Understanding activity and selectivity of metal-nitrogen-doped carbon catalysts for electrochemical reduction of CO. Nat. Commun. 8, 1–9. doi:10. 1038/s41467-017-01035-z in alkaline electrolyte. J. Am. Chem. Soc. 142, 15438–15444. doi:10.1021/jacs. 0c06779 MacDowell, N., Florin, N., Buchard, A., Hallett, J., Galindo, A., Jackson, G., et al. (2010). An overview of CO capture technologies. Energy Environ. Sci. 3, 1645–1669. doi:10.1039/C004106H Kattel, S., Atanassov, P., and Kiefer, B. (2012). Stability, electronic and magnetic properties of in-plane defects in graphene: a ﬁrst-principles study. J. Phys. Chem. C. 116, 8161–8166. doi:10.1021/jp2121609 Malkhandi, S., and Yeo, B. S. (2019). Electrochemical conversion of carbon dioxide to high value chemicals using gas-diffusion electrodes. Curr. Opin. Chem. Eng. 26, 112–121. doi:10.1016/j.coche.2019.09.008 Kibria, M. G., Edwards, J. P., Gabardo, C. M., Dinh, C.-T., Seiﬁtokaldani, A., Sinton, D., et al. (2019). Electrochemical CO reduction into chemical feedstocks: from mechanistic electrocatalysis models to system design. Adv. Mater. 31, 1807166. doi:10.1002/adma.201807166 Marenich, A. V., Ho, J., Coote, M. L., Cramer, C. J., and Truhlar, D. G. (2014). Computational electrochemistry: prediction of liquid-phase reduction potentials. Phys. Chem. Chem. Phys. 16, 15068–15106. doi:10.1039/C4CP01572J Melander, M. M. (2020). Grand canonical rate theory for electrochemical and electrocatalytic systems I: general formulation and proton-coupled electron transfer reactions. J. Electrochem. Soc. 167, 116518. doi:10.1149/1945-7111/aba54b Kim, D., Resasco, J., Yu, Y., Asiri, A. M., and Yang, P. (2014). Synergistic geometric and electronic effects for electrochemical reduction of carbon dioxide using gold–copper bimetallic nanoparticles. Nat. Commun. 5, 1–8. doi:10.1038/ncomms5948 Melander, M. M., Kuisma, M. J., Christensen, T. E. K., and Honkala, K. (2019). Grand-canonical approach to density functional theory of electrocatalytic systems: thermodynamics of solid-liquid interfaces at constant ion and electrode potentials. J. Chem. Phys. 150, 041706. doi:10.1063/1.5047829 Kim, J., Summers, D., and Frese, K., Jr (1988). Reduction of CO and CO to methane on Cu foil electrodes. J. Electroanal. REFERENCES Dislocation-driven deformations in graphene. Science 337, 209–212. doi:10.1126/science.1217529 Sun, Z., Ma, T., Tao, H., Fan, Q., and Han, B. (2017). Fundamentals and challenges of electrochemical CO reduction using two-dimensional materials. Inside Chem. 3, 560–587. doi:10.1016/j.chempr.2017.09.009 Wu, J., Shariﬁ, T., Gao, Y., Zhang, T., and Ajayan, P. M. (2019). Emerging carbon- based heterogeneous catalysts for electrochemical reduction of carbon dioxide into value-added chemicals. Adv. Mater. 31, 1804257. doi:10.1002/adma. 201804257 Tachikawa, H., and Kawabata, H. (2011). Ground and low-lying excited electronic states of graphene ﬂakes: a density functional theory study. J. Phys. B Atom. Mol. Opt. Phys. 44, 205105. doi:10.1088/0953-4075/44/20/205105 Yin, Z., Peng, H., Wei, X., Zhou, H., Gong, J., Huai, M., et al. (2019). An alkaline polymer electrolyte CO electrolyzer operated with pure water. Energy Environ. Sci. 12, 2455–2462. doi:10.1039/C9EE01204D Tian, Z., Priest, C., and Chen, L. (2018). Recent progress in the theoretical investigation of electrocatalytic reduction of CO. Adv. Theory Simul. 1, 1800004. doi:10.1002/adts.201800004 Trevethan, T., Latham, C. D., Heggie, M. I., Briddon, P. R., and Rayson, M. J. (2014). Vacancy diffusion and coalescence in graphene directed by defect strain ﬁelds. Nanoscale 6, 2978–2986. doi:10.1039/c3nr06222h Yu, F., Wei, P., Yang, Y., Chen, Y., Guo, L., and Peng, Z. (2019). Material design at nano and atomic scale for electrocatalytic CO reduction. Nano Mater. Sci. 1, 60–69. doi:10.1016/j.nanoms.2019.03.006 Tripkovic, V., Vanin, M., Karamad, M., Björketun, M. E., Jacobsen, K. W., Thygesen, K. S., et al. (2013). Electrochemical CO and CO reduction on metal-functionalized porphyrin-like graphene. J. Phys. Chem. C. 117, 9187–9195. doi:10.1021/jp306172k Zhang, B., Zhang, J., Shi, J., Tan, D., Liu, L., Zhang, F., et al. (2019). Manganese acting as a high-performance heterogeneous electrocatalyst in carbon dioxide reduction. Nat. Commun. 10, 1–8. doi:10.1038/s41467-019-10854-1 Zhang, C., Yang, S., Wu, J., Liu, M., Yazdi, S., Ren, M., et al. (2018). Electrochemical CO reduction with atomic iron-dispersed on nitrogen-doped graphene. Adv. Energy Mater. 8, 1703487. doi:10.1002/aenm.201703487 Valter, M., Busch, M., Wickman, B., Grönbeck, H., Baltrusaitis, J., and Hellman, A. (2018). Electrooxidation of glycerol on gold in acidic medium: a combined experimental and DFT study. J. Phys. Chem. C. 122, 10489–10494. doi:10.1021/ acs.jpcc.8b02685 Zhao, T., Tian, Y., Wang, Y., Yan, L., and Su, Z. (2019). Mechanistic insight into electroreduction of carbon dioxide on FeNx (x 0–4) embedded graphene. Phys. Chem. Chem. Phys. 21, 23638–23644. doi:10.1039/C9CP03370J Varela, A. S., Ju, W., and Strasser, P. (2018). REFERENCES H., and Chan, K. (2020). Dipole-ﬁeld interactions determine the CO reduction activity of 2D Fe-N-C single-atom catalysts. ACS Catal. 10, 7826–7835. doi:10.1021/ acscatal.0c01375 Scalmani, G., and Frisch, M. J. (2010). Continuous surface charge polarizable continuum models of solvation. i. general formalism. J. Chem. Phys. 132, 114110. doi:10.1063/1.3359469 Schmickler, W., and Guidelli, R. (2014). The partial charge transfer. Electrochim. Acta. 127, 489–505. doi:10.1016/j.electacta.2014.02.057 Wadt, W. R., and Hay, P. J. (1985). Ab initio effective core potentials for molecular calculations. potentials for main group elements Na to Bi. J. Chem. Phys. 82, 284–298. doi:10.1063/1.448800 Schneider, J., Jia, H., Muckerman, J. T., and Fujita, E. (2012). Thermodynamics and kinetics of CO, CO, and H binding to the metal centre of CO reduction catalysts. Chem. Soc. Rev. 41, 2036–2051. doi:10.1039/C1CS15278E Wang, H., Wang, Q., Cheng, Y., Li, K., Yao, Y., Zhang, Q., et al. (2012). Doping monolayer graphene with single atom substitutions. Nano Lett. 12, 141–144. doi:10.1021/nl2031629 Shang, R., Steinmann, S. N., Xu, B.-Q., and Sautet, P. (2020). Mononuclear Fe in N-doped carbon: computational elucidation of active sites for electrochemical oxygen reduction and oxygen evolution reactions. Cat. Sci. Tech. 10, 1006–1014. doi:10.1039/C9CY01935A Wang, H., Feng, Q., Cheng, Y., Yao, Y., Wang, Q., Li, K., et al. (2013). Atomic bonding between metal and graphene. J. Phys. Chem. C. 117, 4632–4638. doi:10. 1021/jp311658m Shao, Y., Dodelet, J.-P., Wu, G., and Zelenay, P. (2019). PGM-free cathode catalysts for PEM fuel cells: a mini-review on stability challenges. Adv. Mater. 31, 1807615. doi:10.1002/adma.201807615 Wang, W.-H., Himeda, Y., Muckerman, J. T., Manbeck, G. F., and Fujita, E. (2015). CO hydrogenation to formate and methanol as an alternative to photo-and electrochemical CO reduction. Chem. Rev. 115, 12936–12973. doi:10.1021/acs. chemrev.5b00197 Shi, C., Hansen, H. A., Lausche, A. C., and Nørskov, J. K. (2014). Trends in electrochemical CO reduction activity for open and close-packed metal surfaces. Phys. Chem. Chem. Phys. 16, 4720–4727. doi:10.1039/C3CP54822H Wang, X., Zhao, Q., Yang, B., Li, Z., Bo, Z., Lam, K. H., et al. (2019). Emerging nanostructured carbon-based non-precious metal electrocatalysts for selective electrochemical CO reduction to CO. J. Math. Chem. A. 7, 25191–25202. doi:10. 1039/C9TA09681G Siahrostami, S., Jiang, K., Karamad, M., Chan, K., Wang, H., and Nørskov, J. (2017). Theoretical investigations into defected graphene for electrochemical reduction of CO. ACS Sustain. Chem. Eng. 5, 11080–11085. doi:10.1021/ acssuschemeng.7b03031 Warner, J. H., Margine, E. R., Mukai, M., Robertson, A. W., Giustino, F., and Kirkland, A. I. (2012). REFERENCES doi:10.1021/ acscatal.9b02799 Petersson, a., Bennett, A., Tensfeldt, T. G., Al-Laham, M. A., Shirley, W. A., and Mantzaris, J. (1988). A complete basis set model chemistry. I. the total energies of closed-shell atoms and hydrides of the ﬁrst-row elements. J. Chem. Phys. 89, 2193–2218. doi:10.1063/1.455064 Lu, Q., and Jiao, F. (2016). Electrochemical CO reduction: electrocatalyst, reaction mechanism, and process engineering. Nano. Energy 29, 439–456. doi:10.1016/j. nanoen.2016.04.009 Petersson, G., and Al-Laham, M. A. (1991). A complete basis set model chemistry. II. open-shell systems and the total energies of the ﬁrst-row atoms. J. Chem. Phys. 94, 6081–6090. doi:10.1063/1.460447 Lu, Q., Rosen, J., Zhou, Y., Hutchings, G. S., Kimmel, Y. C., Chen, J. G., et al. (2014). A selective and efﬁcient electrocatalyst for carbon dioxide reduction. Nat. Commun. 5, 1–6. doi:10.1038/ncomms4242 Qu, L., Liu, Y., Baek, J.-B., and Dai, L. (2010). Nitrogen-doped graphene as efﬁcient metal-free electrocatalyst for oxygen reduction in fuel cells. ACS Nano. 4, 1321–1326. doi:10.1021/nn901850u. PMID: 20155972 Lu, X., Zhu, C., Wu, Z., Xuan, J., Francisco, J. S., and Wang, H. (2020). In situ observation of the pH gradient near the gas diffusion electrode of CO reduction February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 14 Verma et al. Modiﬁed Graphene for CO2 Reduction Rad, A. S., and Foukolaei, V. P. (2015). Density functional study of Al-doped graphene nanostructure towards adsorption of CO, CO and HO. Synth. Met. 210, 171–178. doi:10.1016/j.synthmet.2015.09.026 Verma, A. M., and Kishore, N. (2019). First-principles study on the gas-phase decomposition of bio-oil oxygenated compounds over the palladium catalyst surface. Phys. Chem. Chem. Phys. 21, 22320–22330. doi:10.1039/C9CP04858H Verma, A. M., and Kishore, N. (2018). Molecular modeling approach to elucidate gas phase hydrodeoxygenation of guaiacol over a Pd (111) catalyst within DFT framework. J. Mol. Model 24, 254. doi:10.1007/s00894-018-3803-8 Rebarchik, M., Bhandari, S., Kropp, T., and Mavrikakis, M. (2020). How noninnocent spectator species improve the oxygen reduction activity of single-atom catalysts: microkinetic models from ﬁrst-principles calculations. ACS Catal. 10, 9129–9135. doi:10.1021/acscatal.0c01642 Verma, A. M., and Kishore, N. (2017). Molecular simulations of palladium catalysed hydrodeoxygenation of 2-hydroxybenzaldehyde using density functional theory. Phys. Chem. Chem. Phys. 19, 25582–25597. doi:10.1039/ C7CP05113A Ringe, S., Clark, E. L., Resasco, J., Walton, A., Seger, B., Bell, A. T., et al. (2019). Understanding cation effects in electrochemical CO reduction. Energy Environ. Sci. 12, 3001–3014. doi:10.1039/C9EE01341E Vijay, S., Gauthier, J. A., Heenen, H. H., Bukas, V. J., Kristoffersen, H. REFERENCES Molecular nitrogen-carbon catalysts, solid metal organic framework catalysts, and solid metal/nitrogen-doped carbon (MNC) catalysts for the electrochemical CO reduction. Adv. Energy Mater. 8, 1703614. doi:10.1002/aenm.201703614 Zhu, W., Zhang, Y.-J., Zhang, H., Lv, H., Li, Q., Michalsky, R., et al. (2014). Active and selective conversion of CO to CO on ultrathin Au nanowires. J. Am. Chem. Soc. 136, 16132–16135. doi:10.1021/ja5095099 Varela, A. S., Ju, W., Bagger, A., Franco, P., Rossmeisl, J., and Strasser, P. (2019). Electrochemical reduction of CO on metal-nitrogen-doped carbon catalysts. ACS Catal. 9, 7270–7284. doi:10.1021/acscatal.9b01405 Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Varela, A. S. (2020). The importance of pH in controlling the selectivity of the electrochemical CO reduction. Curr. Opin. Green Sustain. Chem. 26, 100371. doi:10.1016/j.cogsc.2020.100371 Copyright © 2021 Verma, Honkala and Melander. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Vasileff, A., Zheng, Y., and Qiao, S. Z. (2017). Carbon solving carbon’s problems: recent progress of nanostructured carbon-based catalysts for the electrochemical reduction of CO. Adv. Energy Mater. 7, 1700759. doi:10.1002/aenm.201700759 Verma, A. M., Agrawal, K., and Kishore, N. (2018). Binding of phenolic model compounds with noble metal doped graphene sheets. Comp. Theo. Chem. 1134, 37–46. doi:10.1016/j.comptc.2018.05.001 February 2021 \| Volume 8 \| Article 606742 Frontiers in Energy Research \| www.frontiersin.org 15
https://openalex.org/W3136345358	http://entalpipendidikan.ppj.unp.ac.id/index.php/epk/article/download/133/pdf	Indonesian	null	Pengembangan Modul Senyawa Hidrokarbon Berbasis Pendekatan Saintifik Dengan Pertanyaan Probing Promting Untuk Siswa Kelas XI SMA/MA	Entalpi Pendidikan Kimia	2,021	cc-by	3,228	Amalia Firdaus1 and Ellizar1 1Pendidikan Kimia, Fakultas Matematika dan IPA, Universitas Negeri Padang, Jl. Prof. Dr. Hamka, Air Tawar Barat, Padang Utara, Sumatera Barat, Indonesia. 25171. 1Pendidikan Kimia, Fakultas Matematika dan IPA, Universitas Negeri Padang, Jl. Prof. Dr. Hamka, Air Tawar Barat, Padang Utara, Sumatera Barat, Indonesia. 25171. *non_jalius@yahoo.com Keywords: Modules, Hydrocarbon Compounds, Scientific Approach, Probing Prompting Technique, Plomp Model. Entalpi Pendidikan Kimia e-issn: 2774-5171 Development of Hidrocarbon Compount Module Based on Scientific Approach with Probing and Prompting Questions for Class XI SMA/MA Amalia Firdaus1 and Ellizar1 ABSTRACT One of the variations of teaching materials that can make students active is a chemistry module based on a scientific approach with probing promting questions. The purpose of this research is to develop a module of hydrocarbon compounds based on a scientific approach with probing promting questions and to test the validity and practicality of the products produced. This type of research is research and development (R&D) using the plomp development model. The plomp development model has 3 stages, preliminary research stage, the development stage (prototype phase) and the assessment phase. Validation was carried out by 5 validators and small group tests by 6 students and field tests were carried out on 13 students at Sman 1 Pasaman. The research instrument used a questionnaire consisting of a validity questionnaire and a practicality questionnaire, the data collection technique was through filling out a questionnaire. The validation questionnaire data was analyzed using the Aikens'v formula with an average value of Vof 0.83 with the valid category. The practicality test was analyzed using a descriptive statistical formula with a practicality value for a small group of 88% with a very practical category. The practicality test results in the field test on the response questionnaire of the teacher were 88% and the student response questionnaire was 90% with the very practical category. 88 Amalia Firdaus & Ellizar Kata Kunci : Modul, Senyawa Hidrokarbon, Pendekatan Saintifik, Teknik Probing Prompting, Model Plomp Tahapan pembelajaran saintifik diantaranya mengamati, menanya, mengumpulkan informasi, mengasosiasi dan mengkomunikasikan (Daryanto, 2013). Berdasarkan tahapan-tahapan ilmiah tersebut, peserta didik dituntut aktif dan berfikirkritis dalam proses pembelajaran. Penggunaan pendektan saintifik dalam proses belajar mengajar mampu menambah motivasi (Roslina, 2014) dan membuat nilai yang diperoleh peserta didik semakin baik (Ayu dkk., 2015). ABSTRAK Variasi bahan ajar yang bisa membuat peserta didik aktif salah satunya ialah modul kimia berbasis pendekatan saintifik dengan pertanyaan probing prompting. Penelitian ini bertujuan untuk mengembangkan modul senyawa hidrokarbon berbasis pendekatan saintifik dengan pertanyaan probing prompting serta melakukan uji validitas dan praktikalitas produk yang dihasilkan. Jenis penelitian ini ialahResearch and Development (R&D) dengan menggunakan model pengembangan plomp. Model pengembangan plompmemiliki 3 tahap yaitu tahap penelitian awal (preliminary research), tahap pengembangan (prototype phase) dan tahap penilaian (assesment phase). Validasi dilakukan oleh 5 orang validator dan uji small groupdilakukan oleh 6 orang peserta didik sertauji lapangan dilakukan oleh 13 orang peserta didik dan 3 orang guru kimia di SMAN 1 Pasaman. Instrumen penelitian ini menggunakan angket yang terdiri dari angket validitas dan angket praktikalitas. Teknik pengumpulan data melalui pengisian angket. Data angket validasi dianalisis menggunakan formula Aiken’s V dengan nilai rata-rata V sebesar 0,83 dengan kategori valid. Uji praktikalitas dianalisis menggunakan formula statistik deskriptif dengan nilai praktikalitas pada kelompok kecil diperoleh hasil 88% dengan kategori sangat praktis. Hasilpraktikalitas pada uji lapangan pada angket respon guru ialah 88% serta angket respon siswaialah 90% dengan kategori sangat praktis. Kata Kunci : Modul, Senyawa Hidrokarbon, Pendekatan Saintifik, Teknik Probing Prompting, Model Plomp 1. PENDAHULUAN Kurikulum adalah rencana dalam pembelajaran yang membantu peserta didik agar dapat mengembangkan kemampuan dirinya sehingga mempunyai karakter yang diharapkan oleh masyarakat (Daryanto, 2014). Proses belajar mengajarpada kurikulum 2013 menggunakan pendekatan saintifik untuk semua jenjang pendidikan (Kemendikbud, 2013). Pembelajaran berbasis saintifik adalah pembelajaran yang menerapkan tahapan- tahapan saintis dalam proses pembelajaran untuk menciptakan pemahaman melalui tahapan saintifik. Proses belajar mengajar berbasis tahapan saintis bertujuan untuk membimbing peserta didik agar dapat berperan dalam proses pemahaman konsep, hukum atau prinsip melalui tahapan- tahapan ilmiah (Handayani & Legi, 2016). Selain menggunakan pendekatan saintifik yang dapat memotivasi dan membuat nilai siswa menjadi lebih baik, pemilihan teknik dalam bertanya juga dapat membantu proses belajar menjadi lebih hidup. Pertanyaan yang diajukandalam prosespembelajaran akan meningkatkan partisipasi peserta didik baik dalam 89 Entalpi Pendidikan Kimia Salah satu jenis media pembelajaranyang dapat digunakan yaitu modul. Modul adalah variasi media pembelajaran secara tertulis yang dapat mempermudah peserta didik dalam mengolah informasi (Marnesya & Ellizar, 2020). bertanya maupun menjawab pertanyaan (Ellizar, 2009). Sementara itu, dalam proses pembelajarantidak jarang guru kurang melontarkan pertanyaan kepada peserta didik. Keaktifan peserta didik dalam pembelajaran dapat ditingkatkan melalui teknik bertanya probing dan prompting (Jalius, 2012). Penggunaan modul sebagai salah satu media pembelajaran mampu meningkatkan minat dan hasil belajar peserta didik (Ellizar, dkk., 2013). Selain itu, modul juga dapat dijadikan sebagai acuan dalam mengajar oleh guru dandapat digunakan tanpa adanya pengawasan dari guru (Wulansari, 2011). Peserta didik juga dapat menggunakan modul sesuai dengan kecepatan belajar masing-masing (Rahmi & Ilham, 2014). Modul berbasis saintifik adalah modul yang memiliki tahapan- tahapan saintifik (Alfionita & Gazali, 2014). Teknik probing prompting dapat meningkatkan kemampuan berfikir dan komunikasi siswa (Mayasari, 2014), mempermudah peserta didik untuk menyimpulkan pembelajaran (Mutmainnah, Ali, & Napitupulu, n.d.) dan memperbaiki perolehan nilai peserta didik (Bagus dkk., 2016), serta meningkatkan efektivitas dalam menjawab dan kualitas jawaban (Guspatni & Andromeda, 2018), dapat memudahkan peserta didik ketika melakukan pemecahan masalah (Mustika & Buana, 2017). Materi senyawa hidrokarbon merupakan salah satu bagian dari ilmu kimia yang bersifat abstrak karena banyaknya reaksi dan konsep mikroskopis yang tidak dapat diamati secara langsung (Kurniawati, 2011). Penelitian sebelumnya yangberjudul “Pengembangan Modul Reaksi Reduksi dan Oksidasi Berbasis Pendekatan Saintifik dengan Menerapkan Teknik Probing dan Prompting untuk Pembelajaran Kimia Kelas XI SMA/MA”. Berdasarkan penelitian tersebut didapatkan hasil bahwa modul yang dikembangkan sudah valid dan praktis dengan perolehan nilai momen kappa untuk validitas sebesar 0,87dan untuk praktikalitas sebesar 0,86dengan kategori tinggi (Ellizar, 2018). 1. PENDAHULUAN Berdasarkan hasil wawancara dengan beberapa responden dari dua sekolah yaitu di SMAN 13 Padang dan SMAN 3 Bukittinggi didapatkan hasil bahwa minat dan semangat peserta didikketika mempelajari materi tentang senyawa hidrokarbon masih rendah karena peserta didik merasa kesulitan dalam memahami konsep penting dalam materi tersebut. Media pembelajaran yang digunakan di sekolah yaitu LKS, ppt dan buku paket.Berdasarkan permasalahan di atas, peneliti tertarik untuk mengembangkan modul dengan judul “Pengembangan Modul Senyawa Hidrokarbon Berbasis Pendekatan Saintifik dengan Pertanyaan Probing dan Prompting untuk Siswa Kelas XI SMA/MA”. Selanjurnya, penelitian yang dilakukan oleh Ulandari dengan judul “Pengembangan Modul Berbasis Saintifik untuk Melatih Kemampuan Berfikir Kritis pada Materi Gerak Harmonis di SMAN Balung” menyatakan bahwa penggunaan modul berbasis saintifik dapat meningkatkan berfikir kritis peserta didik dengan skor N-Gain sebesar 0,53 dengan kategori sedang (Di & Balung, 2018). 90 Amalia Firdaus & Ellizar kuesioner validasi dan kuesioner praktikalitas. Cara pengambilan informasi dilakukan dengan cara menyebarkan kuesioner. Teknis analisis data validitas didasarkan kepada categorical judgments menggunakan skala Aiken’s V (Heri, 2016). Hasil yang diperoleh dari angket validitas dianalisis menggunakan formula Aiken’s V. Formula Aiken’s V sebagai berikut (Aiken, 1980). 2. METODE 91 Entalpi Pendidikan Kimia Entalpi Pendidikan Kimia Tabel 2. Kriteria Penilaian Praktikalitas (Riduwan, 2009) Interval Kategori 80% - 100% Sangat Praktis 60% - 79% Praktis 40% - 59% Cukup Praktis 20% - 39% Kurang Praktis x ≤ 19% Tidak Praktis Tabel 2. Kriteria Penilaian Praktikalitas (Riduwan, 2009) Interval Kategori 80% - 100% Sangat Praktis 60% - 79% Praktis 40% - 59% Cukup Praktis 20% - 39% Kurang Praktis x ≤ 19% Tidak Praktis 2. Materi senyawa h merujuk pada text book, 3. Tahapan saintifik mer jurnal dan sumber lainny 4. Teknik bertanya probin merujuk pada buku, sumber lainnya. 2. Materi senyawa hidrokarbon merujuk pada text book, 2. Materi senyawa hidrokarbon merujuk pada text book, 3. Tahapan saintifik merujuk pada jurnal dan sumber lainnya, 4. Teknik bertanya probing promting merujuk pada buku, jurnal dan sumber lainnya. 3.1.1. Analisis Kebutuhan 3.1.1. Analisis Kebutuhan Langkah analisis kebutuhan dilakukan melalui tanya jawab bersama 2 orang guru kimia di SMAN 3 Bukittinggi dan SMAN 13 Padang serta penyebaran angket kepada siswa kelas XI dari masing-masingsekolah. Kegiatan yang dilakukan fokus pada masalah yang terjadi dalam proses pembelajaran pada materi senyawa hidrokarbon terutama yang berkaitan dengan bahan ajar yang digunakan. Hasil analisis kebutuhan menunjukkan bahwa belum tersedianya bahan ajar pada materi senyawa hidrokarbon yang dapat meningkatkan keaktifan peserta didik untuk mengkonstruk pengetahuan berdasarkan tuntutan Kurikulum2013. 3.1.4. Pengembangan Kerangka Konseptual 3. HASIL DAN DISKUSI 3. HASIL DAN DISKUSI Konsep-konsep yang penting pada materi senyawa hidrokarbon antara lain, hidrokarbon siklik dan alifatik, hidrokarbon jenuh dan tak jenuh, isomer pada senyawa hidrokarbon serta reaksi pada senyawa hidrokarbon. 2. METODE Jenis penelitian yang digunakan adalah penelitian dan pengembangan atau Research and Development (R&D). Metode penelitian dan pengembangan digunakan untuk menghasilkan dan menguji keefektifan produk tertentu (Sugiyono, 2013). Penelitian dibatasi sampai tahap uji validitas dan praktikaitas produk yang dihasilkan. Model pengembangan yang digunakan pada penelitian ini adalah model pengembangan Plomp yang terdiri dari 3 tahapan yaitu (1) penelitian awal (preliminary research), (2) tahap pembentukan prototipe (prototypestage), dan (3) tahap penilaian (assessment phase) (Plomp, 2007). s = r – lo Keterangan: g lo = Skor terendah dalam kategori (penskoran) (dalam hal ini = 1) c = Banyaknya kategori yang dipilih penilai (dalam hal ini = 5) r = Skor yang diberikan oleh penilai n = Banyak penilai lo = Skor terendah dalam kategori (penskoran) (dalam hal ini = 1) c = Banyaknya kategori yang dipilih penilai (dalam hal ini = 5) c = Banyaknya kategori yang dipilih penilai (dalam hal ini = 5) Tahapan penelitian berupapenelitian awal yang terdiri atas empat tahapan yaitu (a) analisis kebutuhan, (b) analisis konteks (c) analisis literatur dan (d) pembentukan kerangka konseptual. Tahap selanjutnya berupa pembentukan prototipe yang terdiri atas 4 tahapan yaitu (a) prototipe I, (b) prototipe II, (c) prototipe III dan (d) prototipe IV. Tahap terakhir yaitu tahap penilaian atau assesment phase yaitu melakukan uji praktikalitas pada skala yang lebih besar. r = Skor yang diberikan oleh penilai n = Banyak penilai r = Skor yang diberikan oleh penilai n = Banyak penilai Pedoman penilaian validitas berdasarkan skala Aiken’s V sebagai berikut. Tabel 1. Kategori Keputusan berdasarkan Skala Aiken’s V Skala Aiken’s V Kategori V≤0.4 Kurang 0.4 < V ≤ 0.8 Sedang 0.8 < V Valid Tabel 1. Kategori Keputusan berdasarkan Sk l Aik ’ V Peneliti memilih 2 orang dosen kimia FMIPA UNP dan 3 orang guru kimia di SMAN 1 Pasaman sebagai validator. Penetapan validator dilakukan berdasarkan pendapat (Sugiyono, 2013) bahwa validasi dapat dilakukan oleh orang yag ahli dalam bidang kimia.Uji praktikalitas melibatkan 13 orang peserta didik kelas XI dan 3 orang guru kimia di SMAN 1 Pasaman. Penelitian dilaksanakan di SMAN 1 Pasaman dikarenakan saat penelitian dilakukan sedang terjadi pandemi sehingga peneliti memilih sekolah yang masih melaksanakan sekolah tatap muka sebagai tempat penelitian. Instrumen pada penelitian ini menggunakan 2 buah kuesioner yaitu Hasil yang diperoleh dari angket praktikalitas dianalisis menggunakan statistik deskriptif dengan rumus: Nilai yang didapatkan kemudian diinterpretasikan sesuai dengan kriteria seperti Tabel 2. 3.2 Tahap Pembentukan Prototipe. 3.2.1 Prototype 1 Hasil dari perancangan pada penelitian awal (preliminary research) direalisasikan ke dalam bentuk prototipe 1 yang berupa modul. Perancangan modul dilakukan menggunakan aplikasi Microsoft Word, adapun jenis tulisan yang digunakan yaitu Hobo Std, Bradley Hand ITC dan Times New Roman. Komponen pada modul merujuk pada Depdiknas (2008) yaitu cover, kata pengantar, daftarisi, peta konsep, petunjuk penggunaan modul, KI, KD, IPK, tujuan pembelajaran, lembaran kegiatan, lembaran kerja, lembar evaluasi, kunci lembaran kerja dan evaluasi, dan kepustakaan. Desain cover modul dapat dilihat pada Gambar 1. 3.1.2. Analisis Konteks 3.1.2. Analisis Konteks Analisis konteks dilakukan dengan menganalisis kompetensi dasar (KD) 3.1 serta 4.1 menjadi beberapa indikator pencapaian kompetensi (IPK) sertatarget pembelajaran yang akan dicapai oleh peserta didik. Gambar 1. Cover Modul 3.1.3. Studi Literatur Hasil dari tahapan studi literatur sebagai berikut 3.2.3 Prototype III 3.2.3 Prototype III Hasil dari prototipe II dievaluasi secara formatif yaitu penilaian ahli (expert review) dan uji coba satu-satu (one to one evaluation). Berdasarkan wawancara uji satu-satu diperoleh hasil bahwa modul yang dibuat sudah menarik dengan warna yang cerah sehingga membuat siswa tertarik Gambar 3. Hasil Validasi Modul 3.1.3. Studi Literatur 1. Komponen yang terdapat pada modul merujuk pada Depdiknas (Depdiknas, 2008). Gambar 1. Cover Modul 92 Amalia Firdaus & Ellizar Cover pada modul didesain semenarik mungkin dengan menampilkan judul, identitas penulis serta contoh senyawa hidrokarbon pada kehidupan sehari-hari. Pada modul juga terdapat lembar kegiatan yang memiliki tahapan saintifik dengan pertanyaan probing promting. Contoh lembar kegiatan dapat dilihat di Gambar 2. untuk menggunakan modul. Selain itu, tampilan pada modul menurut siswa juga sudah menarik, begitu juga dengan pertanyaan-pertanyaan yang disajikan. Penggunaan bahasa pada modul dapat dengan mudah dimengerti oleh peserta didik. Prototipe II yang dihasilkan selanjutnya dilakukan validasi oleh 5 orang ahli. Penetapan validator berdasarkan pendapat (Sugiyono, 2013) yang menyatakan bahwa validasi dapat dilakukan oleh orang yang ahli dibidang kimia. Pengolahan data validasi menggunakan rumus Aiken’s V. Validasi ini menilai beberapa aspek yaituisi modul, unsur dalam penyajian, penggunaan bahasa dan segi kegrafisannya. Gambar 2. Lembar kegiatan Dari penilaian oleh para ahli didapati hasil rata-rata nilai V semua aspek ialah 0,83 dengan kategori valid. Tahapan selanjutnya yaitu melakukan revisi sesuai saran dari para ahli sehingga diperoleh Protipe III. Hasil validasi disajikan pada Gambar 3. Gambar 2. Lembar kegiatan 3.2.2 Prototype II 3.2.2 Prototype II Prototipe I yang dihasilkan akan dilakukan evaluasi formatif yaitu evaluasi diri sendiri (self evaluation). Self evaluation dilakukan melaluichecklistterhadap spesifikasi dari desain modul. Berdasarkan hasil evaluasi diperoleh bahwa modul yang dirancang sudah sesuai dengan komponen modul secara umum. Hanya saja ada beberapa list yang tidak tersedia di dalam modul. Maka dilakukan revisi terhadap modul yang dirancang. Gambar 3. Hasil Validasi Modul 3.2.4 Prototype IV 3.2.4 Prototype IV Hasil dari prototipe III yang telah valid dilakukan uji praktikalitas pada komunitas kecil (small group) yang beranggotakan 6 orang peserta didik kelas XI SMAN 1 Pasaman.Angket praktikalitas ini berisi tentang tingkat kemudahan pelaksanaan 93 Entalpi Pendidikan Kimia dan penggunaan produk berupa waktu pelaksanaan, biaya, pengelolaan dan penafsiran hasilnya (Mudjijo, 1995). Berdasarkan uji coba kelompok kecil diperoleh hasil bahwa modul memiliki tingkat praktikalitas sangat tinggi. Terdapat beberapa komponen pada modul yang harus direvisi sehingga menghasilka prototipe IV. Gambar 5. Hasil Praktikalitas Modul Gambar 5. Hasil Praktikalitas Modul 3.1. Tahap Penilaian Prototipe IV yang dihasilkan dilakukan uji lapangan pada kelompok siswa yang lebih besar. Pada tahap fild test dilakukan uji praktikalitas kepada 13 orang peserta didik kelas XI dan 3 orang guru kimia SMAN 1 Pasaman dengan menggunakan lembar penilaian berupa angket. Tujuan dilakukannya uji praktikalitas adalah untuk mengetahui sejauh mana pemahaman dan tangggapan siswa serta tenaga pendidik mengenai modul yang dihasilkan (Sukardi, 2011). 4. SIMPULAN Bahan ajar berupa modul berbasis pendekatan saintifik dengan pertanyaan probing dan prompting pada materi senyawa hidrokarbon kelas XI SMA/MA dapat dikembangkan dengan menggunakan model pengembangan Plomp. Modul yang dikembangkan memiliki tingkat validitas 0,83 dengan kategori valid dan tingkat praktikalitas pada angket respon guru 88% dan angket respon peserta didik 90% dengan kategori sangat tinggi. Berdasarkan pengolahan data pada lembar angket praktikalitas untuk respon guru diperoleh hasil sebesar 88% dengan kategori sangat tinggi dan untuk respon pesertadidikdiperoleh hasil sebesar 90% dengan kategori sangat tinggi. Saran untuk peneliti selanjutnya, diharapkan agar dilakukan pengujian efektivitas pada modul senyawa hidrokarbon berbasis pendekatan saintifik dengan pertanyaan probing prompting yang dihasilkan. Hal ini menunjukkan bahwa modul senyawa hidrokarbon berbasis pendekatan saintifik dengan pertanyaan probing prompting yang dihasilkan sudah valid dan praktis untuk dapat digunakan dalam proses pembelajaran. Hasil dari tahap penilaian ini akan menghasilkan prototype final yang telah valid dan praktis. REFERENSI Aiken, L. R. (1980). valid-invalid) validity- moderate validity)., (V), 955–959. Alfionita, T., & Gazali, F. (2014). Deskripsi Modul Hukum-Hukum Dasar Berbasis Pendekatan Saintifik Terhadap Hasil Belajar Peserta Didik. Ranah Research Journal of Multidicsiplinary Researh and Development, 3(2), 32–38. Ayu, K., Astiti, T., Suardika, I. W. R., & Ardana, I. K. (2015). Pengaruh Pendekatan Saintifik Terhadap Hasil Belajar Pengetahuan IPA Tema Tempat Tinggalku Pada Siswa Kelas IV Ditinjau dari Karakteristik Pertanyaan Guru di SDN Gugus Budi Utomo. E-Journal PGSD Universitas Pendidikan Ganesha, 3. Bagus, I. M., Putra, S., Garminah, N. N., Bagi pendidik diharapkan agar membimbing peserta didik mengenai penggunaan modul dan pembelajaran berbasis pendekatan saintifik dengan pertanyaan probing promtingterlebih dahulu sebelum peserta didik dapat belajar secara mandiri. 94 Amalia Firdaus & Ellizar dengan Pendekatan Saintifik. Jakarta: Menteri Pendidikan dan Kebudayaan PSMA. Wibawa, I. M. C., Guru, P., & Dasar, S. (2016). PENGARUH PROBING- PROMPTING TERHADAP HASIL BELAJAR PADA SISWA KELAS IV. E- Journal PGSD Universitas Pendidikan Ganesha, 4(1). Kurniawati, I. L. (2011). Pengembangan Modul Pembelajaran Hibrid Learning pada Mata Pelajaran Kimia SMA Kelas X Dalam Materi Hidrokarbon. Bimafika, 3, 284–291. Daryanto. (2013). Strategi dan Tahapan Mengajar. Bandung: Yama Widya. Marnesya, C. A., & Ellizar. (2020). EFEKTIVITAS MODUL SISTEM KOLOID BERBASIS PENDEKATAN SAINTIFIK DENGAN PERTANYAAN PROBING-PROMPTING. Ranah Research Journal of Multidicsiplinary Researh and Development, 2(4), 80–85. Daryanto. (2014). Pendekatan Pembelajaran Sintifik Kurikulum 2013. (J. Media, Ed.). Jakarta. Depdiknas. (2008). Pengembangan Bahan Ajar. Jakarta: Departemen Pendidikan Nasional. Di, H., & Balung, S. (2018). Pengembangan modul berbasis saintifik untuk melatih kemampuan berpikir kritis pada materi gerak harmonis di sman balung 1). Jurnal Pembelajaran Fisika, 7, 15–21. Mayasari, Y. (2014). Penerapan Teknik Probing Prompting dalam Pembelajaran Matematika Peserta didik Kelas VIII MTsN Lubuk Buaya. Jurnal Pendidikan Matematika, 3(1), 56–61. Ellizar. (2009). Models Of Teaching By Constructivism Approach with Module. Jurnal Kependidikan Triadik, 12(1). Mudjijo. (1995). Tes Hasil Belajar. Jakarta: Bumi Aksara. Ellizar., Bayharti., & Andromeda. (2013). Pengaruh Motivasi dan Pembelajaran Kimia Menggunakan Modul dan Tanpa Modul Terhadap Hasil Belajar Kimia di. Prosiding Simarata FMIPA Universitas Lampung, 117–124. Mustika, H., & Buana, L. (2017). Penerapan model pembelajaran probing prompting terhadap kemampuan pemecahan masalah matematika siswa. MES: Journal of Mathematics Education and Science, 2(2). Mutmainnah, S., Ali, M., & Napitupulu, D. (n.d.). Penerapan Teknik Pembelajaran Probing -Prompting Untuk Meningkatkan Hasil Belajar Fisika pada Siswa Kelas VIII A SMP Negeri I Banawa Tengah. Jurnal Pendidikan Fisika Tadulako, 2(1), 38–43. Ellizar, S. (2018). Entalpi Pendidikan Kimia Sukardi. (2011). Evaluasi Pendidikan, Prinsip, Dan Operasionalnya. Yogyakarta: Bumi Aksara. Wulansari, W. (2011). PENGARUH PENGGUNAAN MODUL TERHADAP PRESTASI BELAJAR SISWA PADA MATA PELAJARAN AKUNTANSI KELAS XI IPS Wahyu Wulansari, 1–11. Entalpi Pendidikan Kimia REFERENSI Pengembangan Modul Reaksi Reduksi Dan Oksidasi Berbasis Pendekatan Saintifik Dengan Menerapkan Teknik Probing Dan Prompting Untuk Pembelajaran Kimia Kelas XI SMA/ MA. Menara Ilmu, 12(12), 91–100. Plomp, T. (2007). Educational Design Research:An Introduction, An Introduction to Educational Research Enschede. Netherland: NationalInstitute For Curriculum Development. Guspatni,. Andromeda., & Bayharti. (2018). Peningkatan Aktivitas Menjawab dan Kualitas Jawaban Mahasiswa dengan Pertanyaan Prompting pada Mata Kuliah Strategi Pembelajaran Kimia Jurusan Kimia , FMIPA , Universitas Negeri Padang. Jurnal Eksakta Pendidikan, 2, 101–107. Rahmi, A., & Ilham, Y. (2014). Pengembangan Bahan Ajar Modul pada Materi Hidrokarbon di SMA Negeri 11 Banda Aceh. Jurnal Pendidikan Sains Indonesia, 02(01), 12–26. Handayani, F., & Legi, W. F. (2016). PENGEMBANGAN MODUL KESETIMBANGAN KIMIA BERBASIS PENDEKATAN SAINTIFIK UNTUK KELAS XI SMA / MA. Journal of Sainstek, 8(1), 85–97. Roslina. (2014). Pengaruh Penggunaan Modul Berbasis Pendekatan Saintifik Terhadap Motivasi dan Hasil Belajar Peserta Didik Influence of Learning Interest , Learning Environment and Parent Participation on Biology Learning Outcomes of State High School Students in Wajo Reg. Prosding Seminar Nasional Biologi, (2011), 2013– 2017. Heri, R. (2016). Analisis Kuantitatif Instrumen Penelitian. Yogyakarta: Parama Publishing. Jalius, E. (2012). Pengembagan Program Pembelajaran. Padang: Universitas Negeri Padang. Sugiyono. (2013). Metode Penelitian Kuantitatif, Kualitatif, dan R&D. Bandung: Alfabeta. Kemendikbud. (2013). Pembelajaran Berbasis Kompetensi Mata Pelajaran Kimia 95 Sukardi. (2011). Evaluasi Pendidikan, Prinsip, Dan Operasionalnya. Yogyakarta: Bumi Aksara. Wulansari, W. (2011). PENGARUH Entalpi Pendidikan Kimia PENGGUNAAN MODUL TERHADAP PRESTASI BELAJAR SISWA PADA MATA PELAJARAN AKUNTANSI KELAS XI IPS Wahyu Wulansari, 1–11. 96
https://openalex.org/W4320732201	https://www.nature.com/articles/s41467-023-36480-6.pdf	English	null	A flexible artificial chemosensory neuronal synapse based on chemoreceptive ionogel-gated electrochemical transistor	Nature communications	2,023	cc-by	13,870	Article https://doi.org/10.1038/s41467-023-36480-6 A ﬂexible artiﬁcial chemosensory neuronal synapse based on chemoreceptive ionogel- gated electrochemical transistor Hamna Haq Chouhdry 1,2, Dong Hyun Lee3, Atanu Bag 3,4 & Nae-Eung Lee 1,2,3,4,5,6,7 The human olfactory system comprises olfactory receptor neurons, projection neurons, and interneurons that perform remarkably sophisticated functions, including sensing, ﬁltration, memorization, and forgetting of chemical stimuli for perception. Developing an artiﬁcial olfactory system that can mimic these functions has proved to be challenging. Herein, inspired by the neuronal network inside the glomerulus of the olfactory bulb, we present an artiﬁcial chemosensory neuronal synapse that can sense chemical stimuli and mimic the functions of excitatory and inhibitory neurotransmitter release in the synapses between olfactory receptor neurons, projection neurons, and inter- neurons. The proposed device is based on a ﬂexible organic electrochemical transistor gated by the potential generated by the interaction of gas molecules with ions in a chemoreceptive ionogel. The combined use of a chemoreceptive ionogel and an organic semiconductor channel allows for a long retentive memory in response to chemical stimuli. Long-term memorization of the excitatory chemical stimulus can be also erased by applying an inhibitory electrical stimulus due to ion dynamics in the chemoresponsive ionogel gate electrolyte. Applying a simple device design, we were able to mimic the exci- tatory and inhibitory synaptic functions of chemical synapses in the olfactory system, which can further advance the development of artiﬁcial neuronal systems for biomimetic chemosensory applications. Signal transduction and perception in the sensory nervous system (SNS) are carried out by a series of synaptic events beginning at the sensory organ and ending in the brain1,2. Sensory organs with receptors convert a speciﬁc type of stimulus into action potentials and transmit pre-processed sensory signals to the central nervous system. Sensory organs, including those in tactile, auditory, visionary, gustatory, and olfactory systems, are often composed of sensory neurons with receptors at their own terminal or receptor cells innervated with afferent neurons, with the generated signals communicated by neu- rotransmitters across synapses to other neurons closer to the brain. Chemoreceptors, mechanoreceptors, photoreceptors, nociceptors, and thermoreceptors are the ﬁve major receptors in sensory organs 1SKKU Advanced Institute of Nano Technology (SAINT), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 2Department of Nano Science and Technology, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 3School of Advanced Materials Science & Engineering, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 4Research Centre for Advanced Materials Technology, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 1SKKU Advanced Institute of Nano Technology (SAINT), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 2Department of Nano Science and Technology, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 3School of Advanced Materials Science & Engineering, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 4Research Centre for Advanced Materials Technology, Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 5Samsung Advanced Institute for Health Sciences & Technology (SAIHST), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 6Institute of Quantum Biophysics (IQB), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 7Biomedical Institute for Convergence at SKKU (BICS), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. e-mail: abag@skku.edu; nelee@skku.edu A ﬂexible artiﬁcial chemosensory neuronal synapse based on chemoreceptive ionogel- gated electrochemical transistor 5Samsung Advanced Institute for Health Sciences & Technology (SAIHST), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 6Institute of Quantum Biophysics (IQB), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. 7Biomedical Institute for Convergence at SKKU (BICS), Sungkyunkwan University, Suwon, Gyeonggi-do 16419, Republic of Korea. e-mail: abag@skku.edu; nelee@skku.edu Nature Communications\| (2023) 14:821 Nature Communications\| (2023) 14:821 1 1 Article https://doi.org/10.1038/s41467-023-36480-6 with cations within the chemoresponsive ionogel. The potentiation effect of chemical gas pulsing can be impeded by applying inhibitory electrical pulses, changing the overall PSC of the postsynaptic mem- brane. This report provides a description of an artiﬁcial chemosensory synapse for artiﬁcial olfactory systems in which the IG layer responds to chemical and electrical stimuli and, in turn, contributes to the modulation of the PSC, exhibiting both potentiation and inhibition. The ACNS demonstrated an ability to mimic biological synapses between ORNs and PNs as wellas interneurons, and to retain ions in the channel for an extended period, resulting in long-term memorization of the stimuli. that respond to stimuli and carry out transduction3–10. Energy-efﬁcient and intelligent signal processing by the SNS has sparked extensive research efforts to mimic biological sensory neuronal systems, including receptors11–13. Emulating the synaptic connection between sensory receptors and afferent neurons has been of great interest for researchers involved with neuromorphic sensor engineering14–19. p g g Artiﬁcial sensory systems have advanced towards intelligent sen- sor systems that sense, ﬁlter, compute, and memorize at the device level20–22. Most relevant research to date has been limited to mechanoreceptors15 or photoreceptors22,23. Although a few attempts have been made to artiﬁcially mimic chemoreceptive sensory systems, the development of artiﬁcial chemoreceptors has been challenging due to their complexity24–28. Mimicking a chemoreceptive sensory system often requires connecting chemical sensors to separate synaptic devices, making for complex structures and fabrication pro- cesses, and involves applying the concept of synaptic plasticity mod- ulation through electrical pulsing under exposure to chemical gas, rather than chemical pulsing29,30. Chemoreceptors are found exclu- sively in olfactory and gustatory systems. Stimuli in the form of odors or tastants binding to receptor cells generate action potentials and release neurotransmitters from the receptor cells to afferent neurons. The olfactory system consists of a network of chemical synapses in the glomerulus of the olfactory bulb31. Synaptic events in these synapses can be excitatory or inhibitory. Results Biological synapses in the olfactory system form between the axons of olfactory receptor neurons (ORNs) and the dendrites of two projection neurons (PNs), i.e., mitral and tufted cells, in a spherical structure called the glomerulus, which is located in the olfactory bulb. The ORNs articulated with the same odor-binding proteins release excitatory neurotransmitters (glutamate) in the same synaptic zone in the glomerulus31,43. Two inhibitory interneurons of periglomerular cells (PGCs) and granule cells (GCs) form synapses inside the glomerulus with PNs (Fig. 1a). The PGCs release the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) in the glomerulus, forming inhibi- tory synapses with PNs32,34. By modulating the presynaptic signals, these synapses can produce excitatory or inhibitory postsynaptic sig- nals and convey complex details to the postsynaptic receptors acti- vated by respective neurotransmitters, resulting in overalllearning and forgetting or modulation of signals in a single device. By simply emulating the functions of these neurons, an artiﬁcial chemoreceptive system can be built and utilized as a component of bio-inspired sensor signal processors that exhibit sensing, learning, forgetting, and mod- ulating signals for intrinsically intelligent olfactory systems44. Figure 1b depicts a schematic of the proposed ACNS, based on an OECT that can be gated by both chemical and electrical stimuli and function with excitatory and inhibitory action potentials, respectively, for the ACNS. Details of the OECT fabrication process for the ACNS are described in the Materials and Methods section. OECTs have been widely studied for possible use in bio-interfaces, chemical and biolo- gical sensing, and synaptic devices18,19,21,45. Stimuli (either electrical or chemical) in an OECT can cause an IG to inject or extract cations from the p-type PEDOT:PSS channel, inducing de-doping or doping in the channel, respectively46. Subsequent modulations in PSC (ΔPSC), which translate to potentiation and inhibition, can be realized by applying chemical and electrical stimuli, respectively (Fig. 1c). Due to the nature of the gating effect in the ionogel under chemical stimulation, potentiation of the OECT synapse can be obtained by chemical pulsing without accompanying electrical gate pulsing. In this study, the pulse duration time (Ton) and the number of pulses (Pn) of the applied stimuli were varied. This phenomenon resembles neurotransmitter transport across a biological synapse when an action potential is transported through the cleft of the synapse. Results Here, we present a ﬂexible artiﬁcial chemosensory neuronal synapse (ACNS) with synaptic functions that emulate the inhibition and potentiation of biological synapses in the glomerulus of the olfactory bulb. Our ACNS is based on an organic electrochemical transistor (OECT) with a poly(3,4-ethylenedioxythiophene) poly- styrene sulfonate (PEDOT:PSS) channel and an ionogel (IG) based on 1- ethyl-3-methylimidazolium bis(triﬂuoromethylsulfonyl)imide ([EMIM] [TFSI]) ionic liquid (IL), poly(ethylene glycol) diacrylate (PEGDA) monomer, and a 1-hydroxycyclohexyl phenyl ketone photoinitiator as a chemoreceptive material as well as gate electrolyte layer. PEDOT:PSS has been widely described as an ion-permeable p-type conducting polymer in OECTs due to its tunable conductivity, ease of deposition, and use in ﬂexible and stretchable devices37–39. Because ILs, including [EMIM][TFSI], possess high thermal and chemical stability, non- volatility, and high carrier-inducing and gas-solvating abilities, they can be used as gas-sensing layers in electrochemical or potentiometric gas sensors40–42. An ACNS based on an OECT can emulate neurotransmitter release (in the form of ions) from the presynaptic membrane (IG electrolyte), resulting in excitation or inhibition of the postsynaptic membrane (PEDOT:PSS channel) in response to an electrical or che- mical external stimulus. Due to ion permeation of the channel layer, high volumetric capacitance allows OECTs to operate at relatively low voltages suitable for emulating biological synaptic functions. Nitrogen dioxide (NO2) gas was used as a potentiating chemical stimulus, and electrical gate pulsing was used as an inhibiting stimulus to drive the ions between the presynaptic and postsynaptic membrane. Upon exposure to NO2 gas, the ACNS showed a stable response toward dif- ferent concentrations of NO2 and extended retention of the changes in an excitatory postsynaptic current (PSC), i.e., synaptic weight (SW), which demonstrates long-term memorization. The ACNS is gated by the potential generated by the interaction between NO2 molecules Figure 1d provides a comparative breakdown of the components of the biological chemosensory synapse in the olfactory system and ACNS for artiﬁcial olfactory systems. The channel forms the synaptic cleft of the artiﬁcial chemical synapse, similar to the synaptic cleft between the presynaptic ORNs and postsynaptic PNs. The IG corresponds to the presynaptic membrane in ORNs, receiving the stimulus, and the drain electrode in the OECT acts as a postsynaptic membrane in the mitral/ tufted cells (PNs) and collects the PSC (i.e., drain current, Id) for further signal processing. A ﬂexible artiﬁcial chemosensory neuronal synapse based on chemoreceptive ionogel- gated electrochemical transistor The olfactory bulb in particular has more inhibitory synapses compared with the other areas of the brain, which are believed to be helpful in the adaptation, perception, dis- crimination, regulation, and attenuation of odor signals, and odor coding (e.g., shaping the overall output of the olfactory bulb)32–34. The adaptation, potentiation, or inhibition of signals in chemosensory systems can occur locally by neuromodulation of excitatory and inhi- bitory synapses, altering conducting states and potentials31,35. How- ever, no previous reports address the functions of adaptation, potentiation, or inhibition of signal in artiﬁcial olfactory systems24,25,28,36. Nature Communications\| (2023) 14:821 Results S1a), and established fro the transfer curves that an IG with 90 wt.% IL achieved a comparativ (b) Anion NO2 PSS- Cation Inhibitory electrical stimulus Excitatory chemical stimulus PSC Channel Ionogel PEDOT+ OECT S G D (a) Olfactory tract Mitral cell (PN) Periglomerular cell (PGC) Glomerulus Granule cell (GC) Olfactory receptor neurons (ORNs) Chemical stimulus Olfactory bulb (d) Biological chemosensory synapse in olfactory system Presynaptic membrane in ORNs Chemical synapse Postsynaptic membrane in PNs Postsynaptic current (PSC) Presynaptic membrane in PGCs Glutamate (Excitatory neurotransmitter) GABA (Inhibitory neurotransmitter) Ionogel (IG) Channel Cations Anions Gate electrode S-D electrodes Drain Current Id G S D Artificial chemosensory synapse for artificial olfactory systems (c) 0 200 Electrical stimulus Ton = 0.5 s Pn = 30 Vd = -0.2 V NO2 ) m p p ( 0 1 Vg (V) 0 500 1000 1500 -2.32 -2.36 -2.40 Inhibition Chemical stimulus Ton = 30 s Pn = 5 Id (mA) Time (s) Potentiation (b) Anion NO2 PSS- Cation Inhibitory electrical stimulus Excitatory chemical stimulus PSC Channel Ionogel PEDOT+ OECT S G D (a) Olfactory tract Mitral cell (PN) Periglomerular cell (PGC) Glomerulus Granule cell (GC) Olfactory receptor neurons (ORNs) Chemical stimulus Olfactory bulb (c) 0 200 Electrical stimulus Ton = 0.5 s Pn = 30 Vd = -0.2 V NO2 ) m p p ( 0 1 Vg (V) 0 500 1000 1500 -2.32 -2.36 -2.40 Inhibition Chemical stimulus Ton = 30 s Pn = 5 Id (mA) Time (s) Potentiation (a) (b) Anion NO2 PSS- Cation Inhib electr stimu Excitatory chemical stimulus PSC Channel Ionogel PEDOT+ OECT S G D O bitory trical mulus (c) 0 200 Electrical stimulus Ton = 0.5 s Pn = 30 Vd = -0.2 V NO2 ) m p p ( 0 1 Vg (V) 0 500 1000 1500 -2.32 -2.36 -2.40 Inhibition Chemical stimulus Ton = 30 s Pn = 5 Id (mA) Time (s) Potentiation (b) (c) (d) Biological chemosensory synapse in olfactory system Presynaptic membrane in ORNs Chemical synapse Postsynaptic membrane in PNs Postsynaptic current (PSC) Presynaptic membrane in PGCs Glutamate (Excitatory neurotransmitter) GABA (Inhibitory neurotransmitter) Ionogel (IG) Channel Cations Anions Gate electrode S-D electrodes Drain Current Id G S D Artificial chemosensory synapse for artificial olfactory systems (d) Biological chemosensory synapse in olfactory system Artificial chemosensory synapse for artificial olfactory systems (d) describe an electrically regulated artiﬁcial chemical synapse (Fig. 2b). Results When a chemical stimulus in the form of an NO2 gas pulse is applied to the IG, anions, behaving as an excitatory neuro- transmitter (glutamate) diffuse into the channel layer, increase the PSC of the device (Fig. 1c). The gate electrode forms an inhibitory synapse with the channel layer, acting as an inhibitory presynaptic membrane. The electrical stimulus applied to the gate electrode controls the ﬂow of Nature Communications\| (2023) 14:821 2 Article https://doi.org/10.1038/s41467-023-36480-6 ons to the channel, playing a role in controlling an inhibitory neuro- transmitter (GABA) and impeding the ΔPSC induced by excitatory chemical stimuli. Chemical synapses release neurotransmitters in the synaptic cleft by opening ion channels when an action potential applied to a pre- synaptic membrane transmits signals from the presynaptic to the describe an electrically regulated artiﬁcial chemical synapse (Fig. 2 To characterize the device under an inhibitory electrical stimulus, assessed the comparative roles of IL and IG electrolytes in the OECT sensing and their synaptic properties. We fabricated and tested d ferent IG compositions (Supplementary Fig. Nature Communications\| (2023) 14:821 https://doi.org/10.1038/s41467-023-36480-6 To demon- strate similarproperties in anACNS, weinvestigated the SADP function of both IL- and IG-gated devices by applying a Vg pulse with an amplitude of 1 V (pulse width of 0 5 s) (Fig 2d) It was evident that the The conductance state of the device after gate biasing and the rate of the recovery process of the cations in the channel are the essential parameters that control and directly inﬂuence the synaptic plasticity of the ACNS. Synaptic plasticity implies activity-dependent changes in synaptic transmission between presynaptic and postsynaptic mem- branes, whereas SW is the strength of the connection between the two membranes due to synaptic activity. To understand the synaptic properties of our device under gate biasing, we investigated the synaptic functions, such as paired-pulse depression (PPD), spike amplitude-dependent plasticity (SADP), spike duration-dependent plasticity (SDDP), and spike number-dependent plasticity (SNDP), which are the basic properties of artiﬁcial synaptic systems. To mea- sure the changes in source-drain current (Id), which is the equivalent of PSC, the source-drain voltage (Vd) at the postsynaptic membrane was ﬁxed at −0.2 V throughout the investigation, which is low enough not to obstruct the ion dynamics. Figure 2c shows the PSC variation of the ACNS when two consecutive Vg pulses of 1 V (pulse width of 0.5 s) were applied on the IG (90 wt.% IL)-gated OECT, emulating the PPD. The SW calculated by \|(PSCSW – PSCin)/PSCin\| after 30 s of termination of the applied gate pulses was 6.8%, where PSCin and PSCSW are the PSC values before and after applying the gate pulses where the SW is cal- culated, respectively. The higher retention time and the slow recovery of the PSC to its initial state also contributed toward higher SW values. To demonstrate the retention of PSC state after electrical stimulation, the decay time constant (τ) of the device was analyzed by employing the exponential decay function. Supplementary Fig. S5 shows the analyzed decay time constants for both IL- and IG (90 wt.% IL)-gated devices at different magnitudes of the electrical stimulus. It is established from the decay constant estimation that IG (90 wt.% IL)-gated device shows longer retention times compared to IL-gated device (up to 19 × 103 s at sti- mulation amplitude, Vg = 1.5 V) and slow decay of PSC to its initial state, generating higher SW values. https://doi.org/10.1038/s41467-023-36480-6 When Vg was removed, the cations slowly diffused back to the electrolyte, returning the neutral PEDOT0 to its original doped state (PEDOT+) (Fig. 2b, schematic III). This shows that the application of a positive Vg will drive the device into a low-conductance state from an initially high-conductance state. The conductance state of the device after gate biasing and the rate of the recovery process of the cations in the channel are the essential parameters that control and directly inﬂuence the synaptic plasticity of the ACNS. Synaptic plasticity implies activity-dependent changes in synaptic transmission between presynaptic and postsynaptic mem- branes, whereas SW is the strength of the connection between the two membranes due to synaptic activity. To understand the synaptic properties of our device under gate biasing, we investigated the synaptic functions, such as paired-pulse depression (PPD), spike amplitude-dependent plasticity (SADP), spike duration-dependent plasticity (SDDP), and spike number-dependent plasticity (SNDP), which are the basic properties of artiﬁcial synaptic systems. To mea- sure the changes in source-drain current (Id), which is the equivalent of PSC, the source-drain voltage (Vd) at the postsynaptic membrane was ﬁxed at −0.2 V throughout the investigation, which is low enough not to obstruct the ion dynamics. Figure 2c shows the PSC variation of the ACNS when two consecutive Vg pulses of 1 V (pulse width of 0.5 s) were applied on the IG (90 wt.% IL)-gated OECT, emulating the PPD. The SW calculated by \|(PSCSW – PSCin)/PSCin\| after 30 s of termination of the applied gate pulses was 6.8%, where PSCin and PSCSW are the PSC values before and after applying the gate pulses where the SW is cal- culated, respectively. After applying a stimulation, PSC decays to its initial state rapidly or slowly, depending on the retention behavior of the ions in the channel, which is usually referred to as short- or long-term plasticity (STP or LTP, respectively). The number of neurotransmitters released into the synaptic cleft can vary depending on the presynaptic mem- brane potential. Application of repetitive, prolonged, and large action potentials can convert STP to LTP in chemical synapses. https://doi.org/10.1038/s41467-023-36480-6 resulting in de-doping or doping phenomena. c Chemical (NO2) and electrical (Vg) stimuli are applied as presynaptic excitatory and inhibitory action potentials, respectively, which can result in ion exchange between the channel and electrolyte, modulating the Id i.e., the postsynaptic current (PSC). The graph shows the PSC variation with different stimuli, where Ton is the pulse duration time and Pn is the number of pulses of the applied chemical and electrical stimuli. d A comparative breakdown of biological and artiﬁcial chemosensory neuronal synapse components. Fig. 1 \| A biological olfactory system and the concept of an artiﬁcial chemo- sensory neuronal synapse (ACNS). a Synaptic connections between the olfactory receptor neurons (ORNs), mitral cells (PNs), and interneurons form inside the glomerulus in the olfactory bulb. The synapses can be excitatory or inhibitory. The subsequent postsynaptic current (PSC) is conveyed to the olfactory tract for fur- ther processing. b An ACNS based on an organic electrochemical transistor (OECT) with an ionogel as the sensing layer as well as the gate electrolyte. Ions in the ionogel are injected or extracted from the organic semiconductor channel, with a pulse duration of 5 s was applied to both IL- and IG (90 wt.% IL)- gated devices to estimate ΔPSC. The SNDP characteristics of the devices were also determined by applying repetitive Vg pulses with a ﬁxed amplitude and pulse width (1 V and 0.5 s, respectively) (Fig. 2f). The SW values of the devices were determined for SADP, SDDP, and SNDP functions. The high retention of the ΔPSC state after applying the electrical stimuli affected the SW values of the devices, which was conﬁrmed by increasing the Vg pulse amplitude (0.5, 0.8, 1, 1.2, and 1.5 V), while keeping the pulse width constant at 0.5 s to estimate the dependence of PSC on pulse amplitude, i.e., SADP (Fig. 2g). The detailed time-dependent PSC of IL- and IG (90 wt.% IL)-gated devices showing SADP can be found in Supplementary Figs. S3a and S4a, respectively. The SW recorded after 30 s of termination of Vg pulsing gradually increased from 0.5% to 5.4% for IL-gated devices and 0.7–10.9% for IG-gated devices. The overall increase in the SW value after increasing the amplitude of the Vg pulses indicates that a higher- magnitude Vg can drive more cations into the channel, resulting in longer retention times and higher SW values, and converting STP to LTP. Results To characterize the device under an inhibitory electrical stimulus, we assessed the comparative roles of IL and IG electrolytes in the OECT in sensing and their synaptic properties. We fabricated and tested dif- ferent IG compositions (Supplementary Fig. S1a), and established from the transfer curves that an IG with 90 wt.% IL achieved a comparatively lower threshold voltage (Vth) and higher on-off current ratio (Ion/Ioff) ions to the channel, playing a role in controlling an inhibitory neuro- transmitter (GABA) and impeding the ΔPSC induced by excitatory chemical stimuli. Chemical synapses release neurotransmitters in the synaptic cleft by opening ion channels when an action potential applied to a pre- synaptic membrane transmits signals from the presynaptic to the postsynaptic membrane (Fig. 2a)47. A similar approach can be used to Nature Communications\| (2023) 14:821 3 Article https://doi.org/10.1038/s41467-023-36480-6 https://doi.org/10.1038/s41467-023-36480-6 These results indicate that IG-gated devices can achieve a higher SW compared with IL-gated devices. Similarly, the PSC recorded by increasing the pulse width from 0.1 to 5 s to emulate SDDP function (Supplementary Figs. S3b and S4b for IL- and IG-gated devices, respectively) resulted in an overall increase in the SW for the pulse duration in both IL- and IG-gated devices (Fig. 2h). Due to the repetitive stimulation of the device (SNDP), the PSC increases with the number of electrical stimuli (Vg pulses with 1 V of amplitude and 0.5 s of width), as shown in Supplementary Figs. S3c and S4c for both the IL- and IG (90 wt.% IL)-gated devices, respectively. The estimated SW values increased up to 70% in the IG-gated device and 20% in the IL-gated device (Fig. 2i), respectively, demonstrating LTP at a greater number of applied Vg pulses when the number of electrical stimuli increased from 1 to 50. (Supplementary Fig. S1b). The IGs with lower IL content (40 and 60 wt.% IL) resulted in a harder solid structure whereas higher IL content in the IG (90 wt.% IL) can help form a soft gel-like structure of IG (Supplementary Fig. S1c). The fundamental transfer and output char- acteristics of IL- and IG (with 90 wt.% IL)-gated OECTs (Supplementary Fig. S2a–d) demonstrated device operation in depletion mode. The transfer curves of six IG (90 wt.% IL)-gated devices fabricated on the same substrate show good reproducibility in their transfer curves (Supplementary Fig. S2e) and also their estimated Vth and Ion/Ioff values (Supplementary Fig. S2f, g). When the gate pulsing was applied in the form of positive gate voltage (Vg), the cations in the electrolyte were pushed into the PEDOT:PSS channel, lowering the channel con- ductance from its original doped state (PEDOT+) (Fig. 2b, schematic I). This depleted the number of holes in the channel, resulting in de- doping of the original PEDOT+ in the polymer chain and creating neutral PEDOT0 (Fig. 2b, schematic II). When Vg was removed, the cations slowly diffused back to the electrolyte, returning the neutral PEDOT0 to its original doped state (PEDOT+) (Fig. 2b, schematic III). This shows that the application of a positive Vg will drive the device into a low-conductance state from an initially high-conductance state. (Supplementary Fig. S1b). https://doi.org/10.1038/s41467-023-36480-6 The IGs with lower IL content (40 and 60 wt.% IL) resulted in a harder solid structure whereas higher IL content in the IG (90 wt.% IL) can help form a soft gel-like structure of IG (Supplementary Fig. S1c). The fundamental transfer and output char- acteristics of IL- and IG (with 90 wt.% IL)-gated OECTs (Supplementary Fig. S2a–d) demonstrated device operation in depletion mode. The transfer curves of six IG (90 wt.% IL)-gated devices fabricated on the same substrate show good reproducibility in their transfer curves (Supplementary Fig. S2e) and also their estimated Vth and Ion/Ioff values (Supplementary Fig. S2f, g). When the gate pulsing was applied in the form of positive gate voltage (Vg), the cations in the electrolyte were pushed into the PEDOT:PSS channel, lowering the channel con- ductance from its original doped state (PEDOT+) (Fig. 2b, schematic I). This depleted the number of holes in the channel, resulting in de- doping of the original PEDOT+ in the polymer chain and creating neutral PEDOT0 (Fig. 2b, schematic II). When Vg was removed, the cations slowly diffused back to the electrolyte, returning the neutral PEDOT0 to its original doped state (PEDOT+) (Fig. 2b, schematic III). This shows that the application of a positive Vg will drive the device into a low-conductance state from an initially high-conductance state. (Supplementary Fig. S1b). The IGs with lower IL content (40 and 60 wt.% IL) resulted in a harder solid structure whereas higher IL content in the IG (90 wt.% IL) can help form a soft gel-like structure of IG (Supplementary Fig. S1c). The fundamental transfer and output char- acteristics of IL- and IG (with 90 wt.% IL)-gated OECTs (Supplementary Fig. S2a–d) demonstrated device operation in depletion mode. The transfer curves of six IG (90 wt.% IL)-gated devices fabricated on the same substrate show good reproducibility in their transfer curves (Supplementary Fig. S2e) and also their estimated Vth and Ion/Ioff values (Supplementary Fig. S2f, g). When the gate pulsing was applied in the form of positive gate voltage (Vg), the cations in the electrolyte were pushed into the PEDOT:PSS channel, lowering the channel con- ductance from its original doped state (PEDOT+) (Fig. 2b, schematic I). This depleted the number of holes in the channel, resulting in de- doping of the original PEDOT+ in the polymer chain and creating neutral PEDOT0 (Fig. 2b, schematic II). https://doi.org/10.1038/s41467-023-36480-6 S7a, b) The time-dependent ΔPSC of the IL-gated device upon bending was not signiﬁcant under no gas exposure, but resulted in a larger abrupt PSC (b) (d) (g) (e) (h) (f) (i) 0.5 0.8 1 1.2 1.5 0 3 6 9 12 IL IG ,t h gie W cit p a n y S SW (%) Pulse amplitude (V) Ton = 0.5 s Pn = 1 0.1 0.3 0.5 1 3 5 0 3 6 9 Vg = 1 V Pn = 1 ,t h gie W cit p a n y S SW (%) Pulse duration, Ton (s) IL IG 1 2 5 10 20 50 0 20 40 60 Ton = 0.5 s Vg = 1 V IL IG ,t h gie W cit p a n y S SW (%) Number of pulses, Pn 0 10 20 30 40 50 900 600 300 0 0 1 Ton = 5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( 0 10 20 30 40 50 900 600 300 0 0 1 IL IG ( C S P A) Time (s) Ton = 0.5 s Pn = 5 Vg ) V ( 0 10 20 30 40 50 600 300 0 0 1 Ton = 0.5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( 0 20 40 -0.6 -0.8 -1.0 -1.2 -1.4 , t n e rr u C nia r D Id (mA) Time (s) SW PSCin PSCSW Vg = 1 V 2 pulses (0.5 s) Vd = -0.2 V I III II Vg PEDOT+ PSS- Cation Hole PEDOT0 Anion S D G Vg > 0 Vg = 0 III II (b) (d) (e) (f) 0 10 20 30 40 50 900 600 300 0 0 1 Ton = 5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( 0 10 20 30 40 50 900 600 300 0 0 1 IL IG ( C S P A) Time (s) Ton = 0.5 s Pn = 5 Vg ) V ( 0 10 20 30 40 50 600 300 0 0 1 Ton = 0.5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( (f) (d) (e) (g) ( 0.5 0.8 1 1.2 1.5 0 3 6 9 12 IL IG ,t h gie W cit p a n y S SW (%) Pulse amplitude (V) Ton = 0.5 s Pn = 1 (h) (i) (i) 1 2 5 10 20 50 0 20 40 60 Ton = 0.5 s Vg = 1 V IL IG ,t h gie W cit p a n y S SW (%) Number of pulses, Pn (g) (h) 0.1 0.3 0.5 1 3 5 0 3 6 9 Vg = 1 V Pn = 1 ,t h gie W cit p a n y S SW (%) Pulse duration, Ton (s) IL IG device operations. https://doi.org/10.1038/s41467-023-36480-6 These results suggest that the liquid nature of ILs imits their use in solid-state devices. In contrast, IGs possess a solid- state nature and ionic conductivity, which is beneﬁcial for stable device operations. Because our device is fabricated on a ﬂexible polyimide substrate, we also tested the electrical properties of IL- and IG-gated devices under mechanical bending, in which the measured transfer curves showed evidence of instability in IL-gated devices upon bending compared with IG-gated devices (Supplementary Fig. https://doi.org/10.1038/s41467-023-36480-6 Nature Communications\| (2023) 14:821 4 https://doi.org/10.1038/s41467-023-36480-6 Article (a) (c) Action potential PSC (b) (d) (g) (e) (h) (f) (i) 0.5 0.8 1 1.2 1.5 0 3 6 9 12 IL IG ,t h gie W cit p a n y S SW (%) Pulse amplitude (V) Ton = 0.5 s Pn = 1 0.1 0.3 0.5 1 3 5 0 3 6 9 Vg = 1 V Pn = 1 ,t h gie W cit p a n y S SW (%) Pulse duration, Ton (s) IL IG 1 2 5 10 20 50 0 20 40 60 Ton = 0.5 s Vg = 1 V IL IG ,t h gie W cit p a n y S SW (%) Number of pulses, Pn 0 10 20 30 40 50 900 600 300 0 0 1 Ton = 5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( 0 10 20 30 40 50 900 600 300 0 0 1 IL IG ( C S P A) Time (s) Ton = 0.5 s Pn = 5 Vg ) V ( 0 10 20 30 40 50 600 300 0 0 1 Ton = 0.5 s Pn = 1 IL IG ( C S P A) Time (s) Vg ) V ( 0 20 40 -0.6 -0.8 -1.0 -1.2 -1.4 , t n e rr u C nia r D Id (mA) Time (s) SW PSCin PSCSW SW = PSCSW - PSCin / PSCin Vg = 1 V 2 pulses (0.5 s) Vd = -0.2 V I III II Vg PEDOT+ PSS- Cation Hole PEDOT0 Anion S D G Vg > 0 Vg = 0 III II I Vg = 0 (a) Action potential PSC (b) Vg PEDOT+ PSS- Cation Hole PEDOT0 Anion S D G Vg > 0 Vg = 0 III II I Vg = 0 Vg > 0 Vg = 0 III II I Vg = 0 SC Vg DOT+ PSS- tion Anion D G Vg > 0 Vg = 0 III II I Vg = 0 (a) (c) Action potential PSC (b) 0 20 40 -0.6 -0.8 -1.0 -1.2 -1.4 , t n e rr u C nia r D Id (mA) Time (s) SW PSCin PSCSW SW = PSCSW - PSCin / PSCin Vg = 1 V 2 pulses (0.5 s) Vd = -0.2 V I III II Vg PEDOT+ PSS- Cation Hole PEDOT0 Anion S D G Vg > 0 Vg = 0 III II I Vg = 0 (c) 0 20 40 -0.6 -0.8 -1.0 -1.2 -1.4 , t n e rr u C nia r D Id (mA) Time (s) SW PSCin PSCSW SW = PSCSW - PSCin / PSCin Vg = 1 V 2 pulses (0.5 s) Vd = -0.2 V I III II (c) (a) The longer retention time of the PSC and higher SW in the IG-gated devices is presumably the result of the blocking effect created by the highly networked porous structure in the polymer matrix of IG, which s believed to slow the redistribution of ions in an IG upon removal of the electric ﬁeld. https://doi.org/10.1038/s41467-023-36480-6 Increased SW and τ values along with shift from STP to LTP was observed, by increasing the magnitude of electrical stimuli. After applying a stimulation, PSC decays to its initial state rapidly or slowly, depending on the retention behavior of the ions in the channel, which is usually referred to as short- or long-term plasticity (STP or LTP, respectively). The number of neurotransmitters released into the synaptic cleft can vary depending on the presynaptic mem- brane potential. Application of repetitive, prolonged, and large action potentials can convert STP to LTP in chemical synapses. To demon- strate similarproperties in anACNS, weinvestigated the SADP function of both IL- and IG-gated devices by applying a Vg pulse with an amplitude of 1 V (pulse width of 0.5 s) (Fig. 2d). It was evident that the ΔPSC in the IG (90 wt.% IL)-gated device was larger than in the IL-gated device. To further investigate the synaptic behavior, we characterized SDDP characteristics of the devices (Fig. 2e). A single Vg pulse of 1 V The results provided in Fig. 2 indicate that, with repetitive, pro- longed, and higher magnitude of stimulation, we could observe LTP in ACNS, akin to chemical synapses in the SNS. The IG-gated device achieved a higher SW in all cases. This can be attributed to the slow redistribution of ions in networked IL channels inside the polymer matrix of an IG after the electric ﬁeld across it is removed. Optical microscopy conﬁrmed a uniformly deposited IG (~150 μm thick) (Supplementary Fig. S6). The dropped IL exhibited a stability problem, presumably due to hydrophobicity, instability due to the ﬂow of IL on the device surface, and difﬁculty establishing a uniformly thin layer. https://doi.org/10.1038/s41467-023-36480-6 Also, to study the dependency of ΔPSC on the pulse duration, the ΔPSC was estimated for 10 and 30 ppm of NO2 with varying pulse duration (20, 30, and 60 s), as shown in Supplementary Fig. S10e, S10f, respectively. Both studies showed a gradual increase in the ΔPSC with increasing the concentration and duration of NO2 pulses for a ﬁxed duration and concentration, respectively. Transfer curves were measured to determine the basic char- acteristics of the device upon gas exposure. They showed an increase in the on-state current and a large positive shift in the Vth of both IG (90 wt.% IL)- (ΔVth = 1.6 V) and IL-gated devices (ΔVth = 0.75 V) when exposed to 400 ppm of NO2 (Fig. 3c). Changes in Vth of the IG-gated device at varying concentrations of NO2 are shown in Supplementary Fig. S9a. The ΔVth was also observed under chemical exposure without electrical gate biasing. A positive shift in Vth after gas exposure may be attributable to the increasing number of anions in the channel. The presence of NO2 can consequently induce a negative gating effect at the IG, resulting in an increase of PSC. The transfer curves show a larger Vth shift in the IG-gated device, which indicates that more anions penetrated the channel layer compared with the IL-gated device when both devices were exposed to the same concentration of NO2 for the same duration of time. This is presumably due to the faster movement of ions in the thin layer of the IG when an electric ﬁeld is present across it. The spin-coated and cured IG forms a thin layer on the device compared with the same amount of IL (Supplementary Fig. S6). We also compared the responses of an IG (90 wt.% IL)-gated OECT and an OECT without IG (Supplementary Fig. S9b). Compared to the IG (90 wt.% IL)-gated OECT, the device without IG showed a small, recover- able response to 400 ppm NO2. The results indicate that the IG (90 wt.% IL)-gated OECT structure can be utilized to get a high response with high retention of the induced changes in PSC via the chemical- gating effect. In a conventional gas sensor, the device recovers to its original state once it is no longer exposed to the gas. Time-dependent PSC measurements with different concentrations, durations, and the number of chemical pulses (Fig. https://doi.org/10.1038/s41467-023-36480-6 y g In the SNS, multiple chemical synapses formed between the ORNs and PNs are controlled by varying the electrochemical excitation (known as an action potential), which plays a large part in perception. When a chemical stimulus (in the form of gas molecules) is applied to the ORNs, the ORNs sense and sequentially transport a corresponding action potential through the neuron, which releases neurotransmitters via exocytosis into the synaptic cleft, and the resulting changes in PSC are transported by PNs to higher olfactory centers for further pro- cessing (Fig. 3a). The schematic in Fig. 3b illustrates the ACNS concept, in which the device mimics a synaptic connection between the ORN and PN in an olfactory system. To emulate the concept, our ACNS establishes a chemoreceptive function for chemical stimulation using an IG that can respond to the chemical properties of a gas. We checked the device responses under four different analytes (NO2, NH3, SO2, and acetic acid) as shown in Supplementary Fig. S8 and continued the further experiments with NO2. Upon exposure to NO2 gas, the gas molecules interact with cations in the IG, which drives ions into the channel in the manner of a neurotransmitter, resulting in the mod- ulation of PSC. In all cases, the IG-gated device achieved a higher ΔPSC compared with the IL-gated device, which can be attributed to the more effective diffusion of ions from the thin IG to the channel due to NO2 absorption. Chemical pulses with a longer exposure time (width of pulse), when applied continuously (with a short resting time between the pulses), may saturate the Id of the device, indicating a limited absorption capacity of the IG under long-term gas exposure (Supplementary Fig. S10a). Similarly, the PSC variation of the device indicates a satu- rated state when the concentration of NO2 pulse increases in a step-like manner with time (Supplementary Fig. S10b). Concentration- dependent changes in PSC (i.e., ΔPSC) of ACNS for lower concentra- tion NO2 (10, 30, 50, and 100 ppm) with a ﬁxed pulse duration of 30 s (Supplementary Fig. S10c) were obtained and the limit of detection of the device was calculated as low as 2.66 ppm from the data (Supple- mentary Fig. S10d)48. https://doi.org/10.1038/s41467-023-36480-6 diffuse to their original distribution once the action potential is removed. c PSC variation and an explanation of SW in an artiﬁcial chemical synapse when two consecutive Vg pulses are applied, corresponding to the operation mechanism described in (b). d–f Time-dependent measurement of ΔPSC for SADP, SDDP and SNDP properties. Changes in SW were measured for the devices with ionogel (IG with 90 wt% IL) and ionic liquid (IL) electrolytes by varying the (g) amplitude, (h) duration, and (i) number of electrical stimuli (i.e., gate pulses). The source-drain voltage (Vd) i.e., the reading voltage, is maintained at −0.2 V for all cases. Fig. 2 \| Electrical regulation of the ACNS. a Schematic of a biological chemical synapse. When an action potential is generated, neurotransmitters are released from the presynaptic membrane to the postsynaptic membrane. b Schematic dia- gram of the artiﬁcial chemical synapse. The inset shows the operating mechanism of the artiﬁcial synapse, (I) before an action potential (Vg pulse) is applied, and (II) when an action potential (Vg > 0) is applied to the presynaptic gate electrode, the ions diffuse from the electrolyte to the channel which results in a change in ΔPSC, which can be observed between source and drain electrodes. (III) Ions slowly change in the ΔPSC upon gas exposure, in comparison with the ΔPSC of the IG-gated device (Supplementary Fig. S7c) presumably due to the instability of IL in the device under a bended condition (Supplemen- tary Fig. S7d). concentration of the NO2 pulse (with a duration of 30 s) was increased from 100 to 600 ppm (Fig. 3d). The higher the concentration, the stronger the response. Similarly, the dependence of ΔPSC on the chemical pulse duration (Fig. 3e) and the pulse number (Fig. 3f) indi- cates increases in the response, with increasing the duration and number, respectively. This was conﬁrmed by the estimated SW after 300 s for the dependence of ΔPSC on the pulse concentration, dura- tion and number. The SW value increased from 0.04 to 3% in the case of the IL-gated device and 1.2 to 16% for the IG-gated device when the NO2 concentration increased from 200 to 600 ppm (Fig. 3g). Similarly, a gradual increase in SW was observed upon increases of the pulse duration (Fig. 3h) and number (Fig. 3i). Nature Communications\| (2023) 14:821 https://doi.org/10.1038/s41467-023-36480-6 Because our device is fabricated on a ﬂexible polyimide substrate, we also tested the electrical properties of IL- and IG-gated devices under mechanical bending, in which the measured transfer curves showed evidence of instability in IL-gated devices upon bending compared with IG-gated devices (Supplementary Fig. S7a, b). The time-dependent ΔPSC of the IL-gated device upon bending was not signiﬁcant under no gas exposure, but resulted in a larger abrupt The longer retention time of the PSC and higher SW in the IG-gated devices is presumably the result of the blocking effect created by the highly networked porous structure in the polymer matrix of IG, which is believed to slow the redistribution of ions in an IG upon removal of the electric ﬁeld. These results suggest that the liquid nature of ILs limits their use in solid-state devices. In contrast, IGs possess a solid- state nature and ionic conductivity, which is beneﬁcial for stable Nature Communications\| (2023) 14:821 5 Article https://doi.org/10.1038/s41467-023-36480-6 https://doi.org/10.1038/s41467-023-36480-6 The potentiation mechanism by chemical stimulatio therefore be explained by the chemically driven gating effect, w induced by the potential in the IG generated by interactions amon (a) (c) (d) (e) (f) (g) ORN PN Chemical stimulus Synapse formation between ORN and PN Presynaptic action potential PSC (b) TFSI- EMIM+ N2O4 PEDOT+ NO2 PEDOT+ PSS- Cation Anion -1 0 1 2 3 4 5 0 -2 -4 -6 -8 Drain current, Id (mA) Gate Voltage, Vg (V) IL IG N2 NO2 Vth Shift 0 200 400 600 NO2 (ppm) 0 500 0 -100 -200 -300 ΔPSC (μA) ΔPSC (μA) ΔPSC (μA) 0 500 Time (s) 0 500 IG IL Ton = 30 s 0 200 NO2 (ppm) Ton = 60 s Ton = 40 s Ton = 20 s 0 500 0 -100 0 500 Time (s) 0 500 IG IL (h) (i) 0 200 Ton = 30 s NO2 (ppm) 0 400 0 -50 -100 IG IL 0 500 Time (s) 0 500 1000 1 2 5 0 1 2 3 4 5 Ton = 30 s NO2 200 ppm Synaptic Weight, SW (%) Number of pulses, Pn IL IG 200 400 600 0 4 8 12 16 0.0 0.1 0.2 0.3 Ton = 30 s Synaptic Weight, SW (%) NO2 concentration (ppm) IL IG 20 40 60 0 1 2 3 4 5 NO2 200 ppm Synaptic Weight, SW (%) Pulse duration, Ton (s) IL IG (a) ORN PN Chemical stimulus Synapse formation between ORN and PN Presynaptic action potential PSC NO PEDOT+ PSS- (a) ( (d) (e) ORN Ch st (b) TFSI- EMIM+ N2O4 PEDOT+ NO2 PEDOT+ PSS- Cation Anion (a) (c) (d) (e) (f) ORN PN Chemical stimulus Synapse formation between ORN and PN Presynaptic action potential PSC (b) TFSI- EMIM+ N2O4 PEDOT+ NO2 PEDOT+ PSS- Cation Anion -1 0 1 2 3 4 5 0 -2 -4 -6 -8 Drain current, Id (mA) Gate Voltage, Vg (V) IL IG N2 NO2 Vth Shift (a) (c) -1 0 1 2 3 4 5 0 -2 -4 -6 -8 Drain current, Id (mA) Gate Voltage, Vg (V) IL IG N2 NO2 Vth Shift (b) (c) (d) (e) (f) (g) 0 200 400 600 NO2 (ppm) 0 500 0 -100 -200 -300 ΔPSC (μA) ΔPSC (μA) ΔPSC (μA) 0 500 Time (s) 0 500 IG IL Ton = 30 s 0 200 NO2 (ppm) Ton = 60 s Ton = 40 s Ton = 20 s 0 500 0 -100 0 500 Time (s) 0 500 IG IL (h) (i) 0 200 Ton = 30 s NO2 (ppm) 0 400 0 -50 -100 IG IL 0 T (f) g ΔPSC (μA) (i) 0 200 Ton = 30 s NO2 (ppm) 0 400 0 -50 -100 IG IL 0 500 Time (s) 0 500 1000 (e) (e) ΔPSC (μA) 0 200 NO2 (ppm) Ton = 60 s Ton = 40 s Ton = 20 s 0 500 0 -100 0 500 Time (s) 0 500 IG IL (h) (f) (d) (i) Time (s) 1 2 5 0 1 2 3 4 5 Ton = 30 s NO2 200 ppm Synaptic Weight, SW (%) Number of pulses, Pn IL IG (i) (h) (g) 20 40 60 0 1 2 3 4 5 NO2 200 ppm Synaptic Weight, SW (%) Pulse duration, Ton (s) IL IG 200 400 600 0 4 8 12 16 0.0 0.1 0.2 0.3 Ton = 30 s Synaptic Weight, SW (%) NO2 concentration (ppm) IL IG hindering the diffusion of the anions in the channel to their initial positions. https://doi.org/10.1038/s41467-023-36480-6 3d–f) indicate that the increase in the PSC is preserved after termination of gas exposure and does not recover to its original state for hundreds of seconds. We observed long-term retention of the ΔPSC, resulting in long-term memorization of the input stimuli, even after the termination of gas pulsing. This unique behavior has rarely been reported and may be attributable to the solvation of NO2 in [EMIM][TFSI]49,50 (Fig. 3b). The Lewis structure of NO2 is consistent with the presence of an unpaired electron, which makes it susceptible to reactions. In a stable state, NO2 readily forms a dimer (N2O4) in the IL/IG, lowering the energy of the entire system51,52. The mechanism of NO2 absorption in [EMIM][TFSI] has been reported by several groups42,49, indicating a strong tendency for NO2 to be dimerized in the stable state. Dimerized N2O4 tends to form π–π interactions with the imidazole cation, [EMIM]+ in the IL/IG (Fig. 3b). Because the NO2 molecules interact with the cations in the IL/IG, the anions presumably diffuse into the channel layer due to the negative gating effect created in the IG and interact readily with PEDOT+ for stable charge redistribution, making room for more NO2 molecules in the electrolyte. This may result in an overall increase in PSC due to the doping effect in the channel. The phenomenon of penetration of To demonstrate the emulation of the chemoreceptive and synaptic functions in our ACNS and understand the operating mechanism, time-dependent ΔPSC data under chemical gating were measured. The dynamic ΔPSC data of the IL- and IG (90 wt.% IL)-gated devices was measured at Vd = −0.2 V with different concentrations, durations, and successive periods of exposure to the chemical (NO2 gas) pulses. The results showed an increase in ΔPSC when the Nature Communications\| (2023) 14:821 6 Article https://doi.org/10.1038/s41467-023-36480-6 into the channel and an increase in the ΔPSC can be also ed by applying a negative electrical pulsing at the gate (Sup- ntary Fig. S11). The results indicate that the anions indeed diffuse e channel upon chemical stimulation of the device. Unlike the hindering the diffusion of the anions in the channel to their positions. https://doi.org/10.1038/s41467-023-36480-6 3 \| Chemical regulation of the ACNS. a A biological chemical synapse with chemical stimulus. A chemical stimulus (in the form of gas molecules) is sensed by the ORN, and a corresponding action potential is then transported through the neuron, releasing neurotransmitters via exocytosis into the synaptic cleft. The resulting changes in ΔPSC are transported by PN to higher olfactory centers for further processing. b An artiﬁcial chemical synapse with chemical stimulus. NO2 dimerizes to form N2O4 in the stable state, facilitating π–π interactions between the [EMIM]+ cations and N2O4. Cations in the electrolyte are solvated by gas molecules, resulting in diffusion of [TFSI]– anions into the channel to interact with PEDOT+, increasing the ΔPSC due to gating inducement of the doping effect. c Transfer Measurement of the activity-dependent changes in synaptic plasticity in the synapses of the SNS can help predict postsynaptic signal enhancement or decrement under dynamic stimuli. The order of and the interval between the excitatory and inhibitory action poten- tials in the synapses can inﬂuence the overall postsynaptic signals. To predict the changes in the SW due to the time interval between two different stimulations, the effect of excitatory chemical pulsing fol- lowed by inhibitory electrical pulsing on the change in the ΔPSC, with different time intervals between the two events, is shown in Fig. 4d. In contrast, Fig. 4e displays the reverse situation, i.e., inhibitory electrical pulsing followed by excitatory chemical pulsing. Anions and cations diffuse into the channel during the excitatory and inhibitory events, respectively, mimicking excitatory and inhibitory neurotransmitter release in the synaptic cleft. A short interval between the two events in the biological synapses indicates the presence of neurotransmitters from the ﬁrst synaptic event in the synaptic cleft when the neuro- transmitters from the second event are released, affecting overall synaptic plasticity. Similarly, in the case of ACNS, anions, and cations can be present concurrently in the channel if the excitatory and inhi- bitory synaptic events occur at sufﬁciently short time intervals. As a parameter describing the different time intervals between the two events, we deﬁned the difference in stimulation time between the excitatory (chemical pulsing) and inhibitory (electrical pulsing) events as Δt (Δt = t2 −t1, where t1 is the ending time of the ﬁrst stimulation, and t2 is the starting time of the second stimulation). An increase in Δt between the excitatory and inhibitory events, as demonstrated in Fig. https://doi.org/10.1038/s41467-023-36480-6 The excitatory chemical stimulus drives the anions from the electrolyte into the channel layer, resulting in an increase in the PSC due to the doping effect in the p-type organic semiconductor channel. An electrical sti- mulus followed by a chemical stimulus will force these anions to dif- fuse out of the channel back into the electrolyte layer and instead drive the cations into the channel layer, resulting in a decrease of PSC due to de-doping in the channel. This diffusion phenomenon of cations and anions into and out of the channel layer under different inducements mimics the behavior of neurotransmitters in olfactory inhibitory and excitatory chemical synapses. y y To mimic the functions of excitatory and inhibitory synaptic events in the glomerulus, we ﬁrst tested the response of the ACNS to excitatory chemical stimuli followed by inhibitory electrical stimuli (Fig. 4c). The IG (90 wt.% IL)-gated device was selected for further measurements (for comparison of IL- and IG-gated devices, see Sup- plementary Fig. S12). The ΔPSC is shown with excitatory chemical pulses resulting from varying the NO2 concentration (100, 200, 400, and 600 ppm, with a chemical pulse duration of 30 s), followed by inhibitory electrical pulses (a Vg amplitude of 1.3 V with an electrical pulse duration of 0.5 s). As the concentration of the excitatory che- mical stimulus, i.e., NO2 concentration, increased, the overall ΔPSC value was enhanced. Because the device exhibited a long-term mem- ory effect when exposed to the gas and did not recover to its initial state for an extended period, this property can be used to depict several memory states of the device after each successive chemical pulsing. The ACNS did not show signiﬁcant degradation in the memory state after the termination of gas pulses and maintained the PSC level even after a few hundred seconds. The ΔPSC level changed when fol- lowed by 200 continuous inhibitory electrical pulses of positive Vg (with a duration time of 0.5 s). The inhibitory electrical pulsing of positive Vg decreased the ΔPSC level, and at lower NO2 concentrations of 200 and 400 ppm was able to bring the ΔPSC level back to the initial state. https://doi.org/10.1038/s41467-023-36480-6 The potentiation mechanism by chemical stimulation can therefore be explained by the chemically driven gating effect, which is induced by the potential in the IG generated by interactions among NO2 molecules and cations. Because NO2 molecules form stable bonds with cations in the IL/IG, it is difﬁcult to desorb the gas from the IL/IG without external aid, resulting in long-term retention of the charge carriers in the anions into the channel and an increase in the ΔPSC can be also observed by applying a negative electrical pulsing at the gate (Sup- plementary Fig. S11). The results indicate that the anions indeed diffuse into the channel upon chemical stimulation of the device. Unlike the electrical stimulation, the anions are retained in the channel for an extended period upon chemical stimulation, indicating a strong inter- action between the solvated complex of cations and NO2 in the IL/IG, Nature Communications\| (2023) 14:821 7 7 Article https://doi.org/10.1038/s41467-023-36480-6 curves show a positive shift in threshold voltage (Vth) upon exposure to NO2 gas (400 ppm). The IG (90 wt.% IL)-gated OECT shows a larger positive shift and greater gas absorption. d Time-dependent measurement of the ΔPSC with different con- centrations (200, 400, and 600 ppm) of NO2 with Ton = 30 s. The corresponding SW is presented in (g). e Time-dependent measurement of ΔPSC with different dura- tion times, i.e., Ton of NO2 (200 ppm) exposure. The corresponding SW is quantiﬁed in (h) for both IL- and IG-gated devices. f Time-dependent measurement of ΔPSC with different numbers of NO2 (200 ppm) pulses i.e., Pn (1, 2, and 5 pulses) indicates a gradual increase in SW, which is depicted in (i). curves show a positive shift in threshold voltage (Vth) upon exposure to NO2 gas (400 ppm). The IG (90 wt.% IL)-gated OECT shows a larger positive shift and greater gas absorption. d Time-dependent measurement of the ΔPSC with different con- centrations (200, 400, and 600 ppm) of NO2 with Ton = 30 s. The corresponding SW is presented in (g). e Time-dependent measurement of ΔPSC with different dura- tion times, i.e., Ton of NO2 (200 ppm) exposure. The corresponding SW is quantiﬁed in (h) for both IL- and IG-gated devices. f Time-dependent measurement of ΔPSC with different numbers of NO2 (200 ppm) pulses i.e., Pn (1, 2, and 5 pulses) indicates a gradual increase in SW, which is depicted in (i). Fig. https://doi.org/10.1038/s41467-023-36480-6 4d, e (from bottom to top), caused changes in ΔPSC and then SW. The overall normalized change in SW (calculated at 4τ interval after the two stimulations in all cases) vs. Δt (the difference in the stimulation time between the excitatory and inhibitory events) was calculated and is presented in Fig. 4f (for quantitative values, see the data in Supple- mentary Fig. S14). When an excitatory chemical stimulus was applied ahead of the inhibitory electrical stimuli, a gradual decrease in SW was observed with increasing Δt, and the opposite trend was seen when inhibitory electrical stimuli were applied ahead of the excitatory sti- mulus. This behavior can be attributed to the fact that, when the inhibitory pulses are applied immediately after an excitatory pulse, i.e., Δt = 5 s (red in Fig. 4d), the ions do not have enough time to achieve a stable equilibrium state in the device. The inhibitory electrical pulses would not be able to expel a large number of anions from the channel layer as the anions from the IG are still inside the channel due to the gradual interaction of gas molecules and cations.In comparison, if Δt is large (160 s) (green in Fig. 4d), the ion distribution can converge on a stable equilibrium state after the gas absorption, and ions can be extracted from the channel more efﬁciently by applying inhibiting electrical pulses. At very high Δt (260 s), the inﬂuence of ﬁrst stimu- lation on the second stimulation would be almost negligible. However, when inhibitory electrical stimuli were applied ahead of an excitatory chemical stimulation (Fig. 4e), a gradual increase in the SW with increasing Δt were observed. When Δt is small (red in Fig. 4e), the decrease in ΔPSC due to inhibitory electrical pulsing will be compen- sated for immediately by the excitatory chemical pulsing, making the channel, which implies long-term memory. Inhibition of PSC can be achieved by applying an external electrical ﬁeld, which can mitigate the chemically driven gating effect, resulting in the extraction of anions in the channel and injection of cations freed from the solvated complex of NO2 gas molecules and cations. The schematic representation of our ACNS mimicking of the functions of a chemical synapse with excitatory or inhibitory pre- synaptic membranes in the biological olfactory system (Fig. 4a), is shown in Fig. 4b, describing the working principle when chemically exciting and electrically inhibiting stimuli are applied. https://doi.org/10.1038/s41467-023-36480-6 C i i \| (2023) 82 9 (d) ( C S P A) 40 0 -40 40 0 -40 40 0 -40 0 200 400 40 0 -40 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( ) Time (s) (e) (d) 15 0 -15 15 0 -15 15 0 -15 0 200 400 15 0 -15 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P A) Time (s) 40 0 -40 40 0 -40 40 0 -40 0 200 400 40 0 -40 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P ) A Time (s) (c) (f) Ton = 0.5 s Pn = 200 NO2 On Off Ton = 30 s Pn = 5 0.0 1.3 Vg (V) 100ppm 200ppm 400ppm 600ppm 0 500 1000 1500 100 0 -100 PSC ( A) Time (s) ) A (e) (c) (d) (f) (g) (f) Time (s) ( C S P ) A 0 50 100 150 200 250 300 0.0 0.2 0.4 0.6 0.8 1.0 SWnorm t (s) Exci Inhi Inhi Exci (g) (g) 0 400 30 60 30 30 30 20 Ton = 0.5 s NO2 (ppm) Pn 30 0 1 30 100 60 30 30 30 30 20 30 Ton (s) Pn = 1 Vg (V) 0 1000 2000 3000 4000 5000 6000 100 0 -100 PSC ( A) Time (s) 20 30 0 400 20 20 Pn 0 1 60 Ton (s) 3800 3900 4000 100 0 -100 Time (s) 20 Fig. 4 \| Excitatory and inhibitory functions of the ACNS. a Chemical synapses inside the glomerulus acquire different action potentials from ORN and PGC and transfer the modulated postsynaptic signals to the olfactory bulb. b Schematic of the ACNS, mimicking the biological olfactory system and showing the excitatory and inhibitory functions. The chemical stimulus contributes to excitatory PSC by diffusion of anions, whereas the electrical stimulus results in PSC inhibition by driving cations into the channel. The net modulated PSC is collected at the source- drain electrodes. https://doi.org/10.1038/s41467-023-36480-6 Article https://doi.org/10.1038/s41467-023-36480-6 Inhibitory action potential (Vg) Excitatory action potential (NO2) Postsynaptic current (Id) PEDOT+ PSS- Cation Anion Hole PEDOT0 NO2 S D G (b) (g) (c) (f) (e) (d) 0 400 30 60 30 30 30 20 Ton = 0.5 s NO2 (ppm) Pn 30 0 1 30 100 60 30 30 30 30 20 30 Ton (s) Pn = 1 Vg (V) 100 0 -100 PSC ( A) 20 30 0 400 20 20 Pn 0 1 60 Ton (s) 100 0 -100 20 Ton = 0.5 s Pn = 200 NO2 On Off Ton = 30 s Pn = 5 0.0 1.3 Vg (V) 100ppm 200ppm 400ppm 600ppm 0 500 1000 1500 100 0 -100 PSC ( A) Time (s) 15 0 -15 15 0 -15 15 0 -15 0 200 400 15 0 -15 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P A) Time (s) 40 0 -40 40 0 -40 40 0 -40 0 200 400 40 0 -40 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P ) A Time (s) 0 50 100 150 200 250 300 0.0 0.2 0.4 0.6 0.8 1.0 SWnorm t (s) Exci Inhi Inhi Exci (a) PSC (Mitral cell) Inhibitory action potential (PGC) Excitatory action potential (ORN) rticle https://doi.org/10.1038/s41467 023 36480 6 Inhibitory action potential (Vg) Excitatory action potential (NO2) Postsynaptic current (Id) PEDOT+ PSS- Cation Anion Hole PEDOT0 NO2 S D G (b) (a) PSC (Mitral cell) Inhibitory action potential (PGC) Excitatory action potential (ORN) (b) (a) Postsynaptic current (Id) S D (g) (c) (f) (e) (d) 0 400 30 60 30 30 30 20 Ton = 0.5 s NO2 (ppm) Pn 30 0 1 30 100 60 30 30 30 30 20 30 Ton (s) Pn = 1 Vg (V) 0 1000 2000 3000 4000 5000 6000 100 0 -100 PSC ( A) Time (s) 20 30 0 400 20 20 Pn 0 1 60 Ton (s) 3800 3900 4000 100 0 -100 Time (s) 20 Ton = 0.5 s Pn = 200 NO2 On Off Ton = 30 s Pn = 5 0.0 1.3 Vg (V) 100ppm 200ppm 400ppm 600ppm 0 500 1000 1500 100 0 -100 PSC ( A) Time (s) 15 0 -15 15 0 -15 15 0 -15 0 200 400 15 0 -15 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P A) Time (s) 40 0 -40 40 0 -40 40 0 -40 0 200 400 40 0 -40 t2 t1 NO2 400 ppm Pn = 1 Ton = 30 s Vg = 1 V Pn = 30 Ton = 0.5 s t ( C S P ) A Time (s) 0 50 100 150 200 250 300 0.0 0.2 0.4 0.6 0.8 1.0 SWnorm t (s) Exci Inhi Inhi Exci ( ) PSC (Mitral cell) Inhibitory action potential (PGC) Excitatory action potential (ORN) ig. https://doi.org/10.1038/s41467-023-36480-6 4 \| Excitatory and inhibitory functions of the ACNS. a Chemical synapses nside the glomerulus acquire different action potentials from ORN and PGC and ransfer the modulated postsynaptic signals to the olfactory bulb. b Schematic of he ACNS, mimicking the biological olfactory system and showing the excitatory nd inhibitory functions. The chemical stimulus contributes to excitatory PSC by iffusion of anions, whereas the electrical stimulus results in PSC inhibition by riving cations into the channel. The net modulated PSC is collected at the source- rain electrodes. c The ACNS with pulsed chemical stimuli at different concentrations followed by an equal number of electrical pulses, showing poten- tiation ﬁrst and then inhibition functions. The changes in PSC with (d) excitatory followed by inhibitory and (e) inhibitory stimuli followed by an excitatory stimulus with different time intervals (Δt) between the two types of stimuli. The consequent normalized SW is calculated and presented in (f). g Under random excitatory and inhibitory stimuli, the ACNS shows gradual modulation in the ΔPSC according to the time and number of applied pulses. https://doi.org/10.1038/s41467-023-36480-6 For a higher concentration of excitatory NO2 pulses, a higher voltage of inhibitory electrical pulse amplitude (Vg) pulses would be required for the ΔPSC to recover to its initial state, which means a higher electrical potential would be required to deplete the cations from the channel layer and turn off the device (consequently extract- ing the anions). Further demonstration of the time-dependent ΔPSC of the IG-gated device, with a ﬁxed excitatory chemical stimulus followed by inhibitory electrical stimuli of different pulse amplitude (Vg), pulse duration (Ton), and pulse number (Pn), is shown in Supplementary Fig. S13. Nature Communications\| (2023) 14:821 8 Methods Materials To demonstrate the capability of the ACNS for an application to the data-efﬁcient bio-inspired perception of input stimuli, the SW values from the time-dependent response (ΔPSC) of IG (90 wt.% IL)- gated OECT under chemical stimulation (NO2 with Ton = 30 s) with varying the concentration from 100 to 400 ppm (Supplementary Fig. S15a) were used for training and prediction of the concentration using a machine learning (ML) algorithm. Here, for the purpose, we measured the performance of the ACNS eight times at each con- centration (total of 32 samples) for the classiﬁcation of different con- centrations of NO2 (ranging from 100 to 400 ppm) utilizing the support vector machine (SVM) in Classiﬁcation Learner application in the ML toolbox available in the MATLAB software. Instead of utilizing a full-time-dependent variation of ΔPSC of the device for ML, as an example, we considered the SW values at 3 instances (at 90, 210, and 330 s at equally spaced time interval of ~120 s) after the chemical sti- mulation (NO2 with Ton = 30 s) for generating the dataset for ML. The results were shown to have test accuracy of 81.2% (Supplementary Fig. S15b). From the results, it can be concluded that four different concentrations of NO2 can be classiﬁed successfully after training, even utilizing signiﬁcantly fewer training and test data from the SW values, which enhances the data efﬁciency of the proposed ACNS through sensing and memorization. The accuracy of the prediction is expected to be improved with more reproducibility of devices and the use of arrayed devices. The proposed ACNS can also enhance energy efﬁciency by reducing the amount of data. Furthermore, it is expected that the SW values from the device can be encoded and fed into a neuromorphic hardware for practical neuromorphic signal processing. For the OECT channel layer, PEDOT:PSS (poly(3,4-ethylenediox- ythiophene) polystyrene sulfonate) solution was purchased from Heraeus (Clevios PH 1000). 1% volume Capstone FS-30, 4% volume ethylene glycol (EG), 0.5% volume glycidyloxypropyl)trimethoxysilane (GOPS) and 0.04 g/ml xylitol were added in PEDOT:PSS and stirred overnight before their use. Adding EG, GOPS, and xylitol to the PED- OT:PSS channel improves electrical and mechanical stability53. Cap- stone was purchased from Chemours, and EG, GOPS, and xylitol were purchased from Sigma Aldrich. Discussion Our study successfully demonstrated thatthe excitatoryand inhibitory synaptic functions of the ACNS are similar to excitatory and inhibitory neurotransmitter release in biological synapses. An ACNS based on an OECT with chemoresponsive ionogel acting as a gate electrolyte can operate simultaneously at a low voltage and in a range of conductance states, depending on the duration, amplitude/concentration, and repetition of the applied chemical and electrical stimuli. Various synaptic properties were emulated in the ACNS by varying the elec- trical stimulation conditions, with high SW in the case of IG-gated devices. Interestingly, the chemical pulsing resulted in the chemically driven gating effect, which effectively modulated PSC, and unusual long-term retention and memory behavior due to the unique combi- nation of organic semiconductor channel and chemoresponsive IG material. In our device, chemical stimulation increased the potential in pre-synapse membrane and, in turn, modulated postsynaptic current, which closely emulates the ORNs connected with projection neurons. https://doi.org/10.1038/s41467-023-36480-6 4e), cations that penetrate the channel due to the electrical stimuli will have enough time to diffuse back to the IG, and the excitatory chemical stimulus will further increase the conductance state, increasing the overall ΔPSC and SW. This can help predict changes in the SW due to the time interval between the different stimulation events. The ACNS was tested over an extended period of 6000 s to determine the operational stability and responses of the device under random stimulations (Fig. 4g). Random excitatory and inhibitory pul- ses of varying pulse durations and numbers were applied, respectively, with different intervals between the successive stimuli. The device indicated a change in the ΔPSC with different excitatory chemical and inhibitory electrical stimuli of 400 ppm NO2 and Vg of 1 V, respectively. Continuous measurement over 6000 s revealed the ability of the device to repeatedly carry out the functions of long-term memoriza- tion through excitatory chemical pulses and depression through inhibitory electrical pulses. We successfully emulated the excitation and inhibition functions of chemical synapses in an artiﬁcial olfactory system, leading to sensing, memorization, forgetting, and overall modulation of the synaptic signals. OECT fabrication Th i f The transistor was fabricated on a 50-μm-thick polyimide ﬁlm, soni- cated in acetone, ethanol, and deionized water. 75 nm-thick source, drain, and gate electrodes were deposited by thermal evaporation of gold through a shadow mask. The prepared PEDOT:PSS solution was then spin-coated as the channel layer between the source-drain elec- trodes (channel area = 5 × 1 mm) followed by annealing at 150 °C for 2 h in an N2 environment has been reported to result in acceptable elec- trical and mechanical stability in electrolytes53. The source-drain elec- trodes were encapsulated using polydimethylsiloxane, and the devices were treated with O2 plasma before the electrolyte was dropped on the channel and gate area to reduce the surface tension due to the hydrophobic nature of IL. 1 μL of IL was directly dropped onto the IL- gated OECTs. For IG-gated OECTS, the prepared IG solution was spin- coated on the channel and gate area and cured for 20 s under ultra- violet radiation (365 nm wavelength, 5 mW/m2 power). The schematics of the detailed fabrication steps are shown in Supplementary Fig. S16a. The device design is shown in Supplementary Fig. S16b. Methods Materials The IG was prepared by mixing the IL 1-Ethyl-3-methylimidazolium bis(triﬂuoromethylsulfonyl)imide ([EMIM][TFSI]), poly(ethylene gly- col) diacrylate (PEGDA) monomer, and 1-hydroxycyclohexyl phenyl ketone photoinitiator (Photoiniator 184) in the ratio of 90:8:2. All three components of the IG were purchased from Sigma Aldrich. Different concentrations (90 wt.%, 60 wt.%, and 40 wt.%) of the IGwere prepared and tested, out of which 90 wt.% was selected for the IG-gated device. The comparison of different concentrations of the IG is illustrated in the Supplementary information. https://doi.org/10.1038/s41467-023-36480-6 c The ACNS with pulsed chemical stimuli at different concentrations followed by an equal number of electrical pulses, showing poten- tiation ﬁrst and then inhibition functions. The changes in PSC with (d) excitatory followed by inhibitory and (e) inhibitory stimuli followed by an excitatory stimulus with different time intervals (Δt) between the two types of stimuli. The consequent normalized SW is calculated and presented in (f). g Under random excitatory and inhibitory stimuli, the ACNS shows gradual modulation in the ΔPSC according to the time and number of applied pulses. Nature Communications\| (2023) 14:821 9 9 Article https://doi.org/10.1038/s41467-023-36480-6 The simple design of the ACNS incorporates both chemical and electrical gating phenomena and can achieve a prolonged stable response even after random electrical and chemical stimulation. Pre- vious studies have seldom reported sensing, memorization, forgetting, and overall modulation of sensory signals under simultaneous stimu- lation of electrical and chemical stimuli in a single device. Although we used only one type of gas as a chemical stimulus, the same approach can be used to incorporate more sensing layers and arrayed devices that sense and differentiate among multiple chemosensory signals. Modulation of the SW value, which can be varied by the different patterns of stimuli, can be used to classify different stimuli using machine learning. The SW values can also be encoded as spiking sig- nals and supplied for bio-inspired neuromorphic signal processing, such as those involved in a spiking neural network, with signiﬁcant reductions in the amount of data required. In conclusion, the biomi- metic chemosensory neuronal system described here exhibits various functions similar to chemosensory neurons in the biological olfactory system. This work can further broaden the research toward a chemo- sensory system for use in the artiﬁcial olfactory systems. net change in ΔPSC and SW small. At larger Δt (purple in Fig. 4e), cations that penetrate the channel due to the electrical stimuli will have enough time to diffuse back to the IG, and the excitatory chemical stimulus will further increase the conductance state, increasing the overall ΔPSC and SW. This can help predict changes in the SW due to the time interval between the different stimulation events. net change in ΔPSC and SW small. At larger Δt (purple in Fig. References 1. Torre, V., Ashmore, J., Lamb, T. & Menini, A. Transduction and adaptation in sensory receptor cells. J. Neurosci. 15, 7757–7768 (1995). 26. Qian, C. et al. Oxygen‐detecting synaptic device for realization of artiﬁcial autonomic nervous system for maintaining oxygen home- ostasis. Adv. Mater. 32, 2002653 (2020). 2. Jessell, T. M. & Kandel, E. R. Synaptic transmission: a bidirectional and self-modiﬁable form of cell-cell communication. Cell 72, 1–30 (1993). 27. Qian, C. et al. Risk‐perceptional and feedback‐controlled response system based on NO2‐detecting artiﬁcial sensory synapse. Adv. Funct. Mater. 32, 2112490 (2022). 3. Hao, J., Bonnet, C., Amsalem, M., Ruel, J. & Delmas, P. Transduction and encoding sensory information by skin mechanoreceptors. Pﬂug. Arch. 467, 109–119 (2015). 28. Chu, Y., Tan, H., Zhao, C., Wu, X. & Ding, S.-J. Power-efﬁcient gas- sensing and synaptic diodes based on lateral pentacene/a-IGZO PN junctions. ACS Appl. Mater. Interfaces 14, 9368–9376 (2022). 4. Genovese, F., Reisert, J. & Kefalov, V. J. Sensory transduction in photoreceptors and olfactory sensory neurons: common features and distinct characteristics. Front. Cell Neurosci. 15, 414 (2021). 29. Han, J. et al. Artiﬁcial olfactory neuron for an in‐sensor neuro- morphic nose. Adv. Sci. 9, 2106017 (2022). 30. Wang, T. et al. A bio‐inspired neuromorphic sensory system. Adv. Intell. Sys. 4, 2200047 (2022). 5. Lumpkin, E. A. & Caterina, M. J. Mechanisms of sensory transduction in the skin. Nature 445, 858–865 (2007). 6. Van De Water, T. R. Determinants of neuron-sensory receptor cell interaction during development of the inner ear. Hear Res. 22, 265–277 (1986). 31. Harvey, J. & Heinbockel, T. Neuromodulation of synaptic transmis- sion in the main olfactory bulb. Int. J. Environ. Res. Public Health 15, 2194 (2018). 7. Spielman, A. I. Chemosensory function and dysfunction. Crit. Rev. Oral. Biol. Med. 9, 267–291 (1998). 32. Urban, N. N. Lateral inhibition in the olfactory bulb and in olfaction. Physiol. Behav. 77, 607–612 (2002). 8. Breslin, P. A. S. & Spector, A. C. Mammalian taste perception. Curr. Biol. 18, R148–R155 (2008). 33. Zhou, F.-W., Shao, Z.-Y., Shipley, M. T. & Puche, A. C. Short-term plasticity in glomerular inhibitory circuits shapes olfactory bulb output. J. Neurophysiol. 123, 1120–1132 (2020). 9. Schepers, R. J. & Ringkamp, M. Thermoreceptors and thermo- sensitive afferents. Neurosci. Biobehav. Rev. 34, 177–184 (2010). 34. Arruda, D., Publio, R. & Roque, A. C. The periglomerular cell of the olfactory bulb and its role in controlling mitral cell spiking: a computational model. References PLoS ONE 8, e56148 (2013). 10. Sneddon, L. U. Comparative physiology of nociception and pain. Physiology 33, 63–73 (2018). 35. Wu, A., Yu, B. & Komiyama, T. Plasticity in olfactory bulb circuits. Curr. Opin. Neurobiol. 64, 17–23 (2020). 11. Jung, Y. H., Park, B., Kim, J. U. & Kim, T. Bioinspired electronics for artiﬁcial sensory systems. Adv. Mater. 31, 1803637 (2019). 36. Song, Z. et al. A ﬂexible conformable artiﬁcial organ-damage memory system towards hazardous gas leakage based on a single organic transistor. Mater. Horiz. 6, 717–726 (2019). 12. Choi, S., Yang, J. & Wang, G. Emerging memristive artiﬁcial synapses and neurons for energy‐efﬁcient neuromorphic comput- ing. Adv. Mater. 32, 2004659 (2020). 13. He, Y. et al. Recent progress on emerging transistor‐based neuro- morphic devices. Adv. Intell. Syst. 3, 2000210 (2021). 37. Fan, X. et al. PEDOT:PSS for ﬂexible and stretchable electronics: modiﬁcations, strategies, and applications. Adv. Sci. 6, 1900813 (2019). 14. Wan, C. et al. An artiﬁcial sensory neuron with visual-haptic fusion. Nat. Commun. 11, 4602 (2020). 38. Kayser, L. V. & Lipomi, D. J. Stretchable conductive polymers and composites based on PEDOT and PEDOT:PSS. Adv. Mater. 31, 1806133 (2019). 15. Lee, Y. R., Trung, T. Q., Hwang, B.-U. & Lee, N.-E. A ﬂexible artiﬁcial intrinsic-synaptic tactile sensory organ. Nat. Commun. 11, 2753 (2020). 39. Zhang, S. et al. Tuning the electromechanical properties of PED- OT:PSS ﬁlms for stretchable transistors and pressure sensors. Adv. Electron Mater. 5, 1900191 (2019). 16. Kim, Y. et al. A bioinspired ﬂexible organic artiﬁcial afferent nerve. Science 360, 998–1003 (2018). 17. Xu, W., Min, S.-Y., Hwang, H. & Lee, T.-W. Organic core-sheath nanowire artiﬁcial synapses with femtojoule energy consumption. Sci. Adv. 2, e1501326 (2016). 40. Rehman, A. & Zeng, X. Methods and approaches of utilizing ionic liquids as gas sensing materials. RSC Adv. 5, 58371–58392 (2015). 41. Lei, Z., Dai, C. & Chen, B. Gas solubility in ionic liquids. Chem. Rev. 114, 1289–1326 (2014). 18. van de Burgt, Y. et al. A non-volatile organic electrochemical device as a low-voltage artiﬁcial synapse for neuromorphic computing. Nat. Mater. 16, 414–418 (2017). 42. Jin, M. L. et al. Scalable superior chemical sensing performance of stretchable ionotronic skin via a π‐hole receptor effect. Adv. Mater. 33, 2007605 (2021). 19. Kim, K., Sung, M., Park, H. & Lee, T. Organic synaptic transistors for bio‐hybrid neuromorphic electronics. Adv. Electron Mater. 8, 2100935 (2022). 43. Nagayama, S., Homma, R. Measurements The responsive properties of the device toward a target gas, NO2, were examined using a custom-built gas-sensing chamber connected to gas source by mass ﬂow controllers (KRO-4000S, KMB Tech) to control parameters such as gas ﬂow rate and humidity. All the experiments Nature Communications\| (2023) 14:821 10 Article https://doi.org/10.1038/s41467-023-36480-6 21. Ling, H. et al. Electrolyte-gated transistors for synaptic electronics, neuromorphic computing, and adaptable biointerfacing. Appl. Phys. Rev. 7, 011307 (2020). were performed at room temperature (20 to 25 °C) and relative humidity of 20%, where N2 was used as a reference and balancing gas. All electrical and time-dependent measurements of PSC were per- formed using a semiconductor parameter analyzer (Keithley, 4200 SCS). The channel layer thickness was examined by ﬁeld-emission scanning electron microscopy (JEO JSM-6500F) (Supplementary Fig. S16c). The thickness of the IG layer was veriﬁed by a custom digital optical microscope connected to a digital camera. 22. Lee, T. et al. Realization of an artiﬁcial visual nervous system using an integrated optoelectronic device array. Adv. Mater. 33, 2105485 (2021). 23. Yu, J. et al. Bioinspired mechano-photonic artiﬁcial synapse based on graphene/MoS 2 heterostructure. Sci. Adv. 7, 9117–9134 (2021). Data availability 24. Gao, Z. et al. An artiﬁcial olfactory system with sensing, memory and self-protection capabilities. Nano Energy 86, 106078 (2021). y Source data are provided with this paper. 25. Ban, C. et al. An artiﬁcial olfactory memory system for monitoring and recording of volatile organic compounds. Adv. Mater. Technol. 6, 2100366 (2021). Reprints and permissions information is available at http://www.nature.com/reprints 52. Liu, W.-G. & Goddard, W. A. First-principles study of the role of interconversion between NO2, N2O4, cis- ONO-NO2, and trans -ONO-NO2 in chemical processes. J. Am. Chem. Soc. 134, 12970–12978 (2012). Publisher’s note Springer Nature remains neutral with regard to jur- isdictional claims in published maps and institutional afﬁliations. 53. Nguyen, T. D., Trung, T. Q., Lee, Y. & Lee, N.-E. Stretchable and stable electrolyte‐gated organic electrochemical transistor synapse with a naﬁon membrane for enhanced synaptic properties. Adv. Eng. Mater. 24, 2100918 (2022). Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. https://doi.org/10.1038/s41467-023-36480-6 https://doi.org/10.1038/s41467-023-36480-6 manuscript. N.-E.L. planned and supervised the whole work and reviewed and edited the manuscript. 45. van de Burgt, Y., Melianas, A., Keene, S. T., Malliaras, G. & Salleo, A. Organic electronics for neuromorphic computing. Nat. Electron 1, 386–397 (2018). References & Imamura, F. Neuronal organization of olfactory bulb circuits. Front. Neural Circuits 8, 98 (2014). 20. Seo, S. et al. Artiﬁcial van der Waals hybrid synapse and its appli- cation to acoustic pattern recognition. Nat. Commun. 11, 3936 (2020). 44. Manzini, I., Schild, D. & di Natale, C. Principles of odor coding in vertebrates and artiﬁcial chemosensory systems. Physiol. Rev. 102, 61–154 (2022). Nature Communications\| (2023) 14:821 11 Article Acknowledgements This research was supported by the Basic Science Research Program (Grant Nos. 2019R1A6A1A03033215 and 2020R1A2C3013480) through the National Research Foundation of Korea (NRF), funded by the Ministry of Science and Technology and the Ministry of Education. Supplementary information The online version contains supplementary material available at 48. Bag, A. et al. A room-temperature operable and stretchable NO2 gas sensor composed of reduced graphene oxide anchored with MOF- derived ZnFe2O4 hollow octahedron. Sens Actuators B Chem. 346, 130463 (2021). Correspondence and requests for materials should be addressed to Atanu Bag or Nae-Eung Lee. Correspondence and requests for materials should be addressed to Atanu Bag or Nae-Eung Lee. 49. Yuan, G., Zhang, F., Geng, J. & Wu, Y.-T. Highly efﬁcient reversible adsorption of NO2 in imidazole sulfonate room temperature ionic liquids. RSC Adv. 4, 39572–39575 (2014). Peer review information Nature Communications thanks the anon- ymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available. 50. Liu, J. & Xu, Y. NOx absorption and conversion by ionic liquids. J. Hazard Mater. 409, 124503 (2021). 51. Yu, Q. & Gao, H. A simple determination of the NO2 dimerization equilibrium constant. J. Chem. Educ. 74, 233 (1997). Competing interests The authors declare no competing interests. 46. Tu, D. & Fabiano, S. Mixed ion-electron transport in organic elec- trochemical transistors. Appl. Phys. Lett. 117, 080501 (2020). 47. Faber, D. S. & Pereda, A. E. Two forms of electrical transmission between neurons. Front. Mol. Neurosci. 11, 427 (2018). Additional information Supplementary information The online version contains supplementary material available at h //d i /10 1038/ 41467 023 36480 6 Nature Communications\| (2023) 14:821 Author contributions H.H.C. carried out the experiments, analyzed the data, and wrote the manuscript. D.H.L. developed the initial devices and fabrication pro- cesses and developed the measurement protocol. A.B. conceptualized some experiments, analyzed the data, discussed the ideas, carried out computation for machine learning, and reviewed and edited the © The Author(s) 2023 © The Author(s) 2023 Nature Communications\| (2023) 14:821 Nature Communications\| (2023) 14:821 Nature Communications\| (2023) 14:821 12
https://openalex.org/W2440482312	https://europepmc.org/articles/pmc4912801?pdf=render	English	null	SCM: a practical tool to implement hospital-based syndromic surveillance	BMC research notes	2,016	cc-by	5,864	Abstract Background: Syndromic surveillance has been widely used for the early warning of infectious disease outbreaks, especially in mass gatherings, but the collection of electronic data on symptoms in hospitals is one of the funda- mental challenges that must be overcome during operating a syndromic surveillance system. The objective of our study is to describe and evaluate the implementation of a symptom-clicking-module (SCM) as a part of the enhanced hospital-based syndromic surveillance during the 41st World Exposition in Shanghai, China, 2010. Methods: The SCM, including 25 targeted symptoms, was embedded in the sentinels’ Hospital Information Systems (HIS). The clinicians used SCM to record these information of all the visiting patients, and data were collated and transmitted automatically in daily batches. The symptoms were categorized into seven targeted syndromes using pre- defined criteria, and statistical algorithms were applied to detect temporal aberrations in the data series. Results: SCM was deployed successfully in each sentinel hospital and was operated during the 184-day surveillance period. A total of 1,730,797 patient encounters were recorded by SCM, and 6.1 % (105,352 visits) met the criteria of the seven targeted syndromes. Acute respiratory and gastrointestinal syndromes were reported most frequently, accounted for 92.1 % of reports in all syndromes, and the aggregated time-series presented an obvious day-of-week variation over the study period. In total, 191 aberration signals were triggered, and none of them were identified as outbreaks after verification and field investigation. Conclusions: SCM has acted as a practical tool for recording symptoms in the hospital-based enhanced syndromic surveillance system during the 41st World Exposition in Shanghai, in the context of without a preexisting electronic tool to collect syndromic data in the HIS of the sentinel hospitals. Keywords: Syndromic surveillance, Infectious disease, Mass gatherings, Early warning, Outbreak detection augment traditional surveillance for nearly two decades [1–3]. Syndromic surveillance systems usually employ statistical algorithms to inspect unexpected changes in prodromic or pre-diagnostic data captured in electronic systems from a variety of sources, with the premise that aberrations in these data may provide an earlier indi- cation of a disease outbreak before the change can be observed in confirmed diagnoses [4, 5]. Among the mul- tiple sources of pre-diagnostic data (such as emergency department visits, ambulance trip logs, pharmacy sales, and work or school absentee rates), pre-diagnostic clini- cal syndromes among patients visiting hospitals are fre- quently used as a data source in syndromic surveillance systems [6–8]. © 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Correspondence: qsun@pdcdc.sh.cn; ywz126@vip.sina.com †ChuchuYe and Zhongjie Li contributed equally to this work 1 Research Base of Key Laboratory of Surveillance and Early‑warning on Infectious Disease in China CDC, Shanghai Pudong New Area Center for Disease Control and Prevention, Shanghai, China 2 Key Laboratory of Surveillance and Early‑warning on Infectious Disease, Chinese Center for Disease Control and Prevention, Beijing, China Full list of author information is available at the end of the article Abstract As syndromic surveillance is expected to SCM: a practical tool to implement hospital‑based syndromic surveillance Chuchu Ye1†, Zhongjie Li2†, Yifei Fu1, Yajia Lan3, Weiping Zhu1, Dinglun Zhou3, Honglong Zhang2, Shengjie Lai2,4, David L. Buckeridge5, Qiao Sun1 and Weizhong Yang2* Ye et al. BMC Res Notes (2016) 9:315 DOI 10.1186/s13104-016-2098-z Ye et al. BMC Res Notes (2016) 9:315 DOI 10.1186/s13104-016-2098-z BMC Research Notes Open Access Background Surveillance systems play a fundamental role on the mon- itoring and detecting outbreaks of infectious diseases. Syndromic surveillance, or the use of near “real-time” pre-diagnostic data and automated tools to detect and characterize unusual activity for further public health investigation, has been adopted in many countries to Ye et al. BMC Res Notes (2016) 9:315 Page 2 of 9 detect an epidemic at a very early stage, the timely col- lection, collation and analysis of a large amount of syn- drome data are key system components. we describe how SCM operated, and present the main results of targeted syndromes, and compare the patterns of data among different levels of hospitals, as well as the day-of-week effect. In most existing syndromic surveillance systems, a common strategy of obtaining data from hospitals is to automatically collect the chief complaints of patients and to classify the data into syndromic categories using standard codes or by natural language processing [9–11]. Syndromic surveillance, however, faces huge challenge while the chief complaints have not been documented electronically in the hospitals, which had been encoun- tered in Pudong New Area, Shanghai. The most of the hospitals in Pudong have an electric hospital informa- tion system (HIS) to record medical information. When a patient visits a hospital, a social insurance card or a provisional electronic card must be presented at the reception counter. Basic demographic information about the patient is recorded in the HIS, such as gender, date of birth, address and historical medical record. The HIS then generates a patient ID number, which is unique for the visit, and the patient is subsequently seen by a doc- tor. Information on the patient can be tracked by the ID number, which can be used to access the HIS record, including lab test orders, test results, prescriptions, and diagnoses. However, the fundamental information for syndromic surveillance-the chief complaints of the patient, are not routinely collected in HIS, as it is practi- cally recorded on paper at the stage of triage.h Surveillance sentinels and targeted syndromes Surveillance sentinels and targeted syndromes Twenty-one hospitals were selected as sentinel sites for PD-SEWS, including two tertiary hospitals, five second- ary hospitals and fourteen primary hospitals. Hospitals were chosen according to their location, catchment area, and patient volume [12]. In Pudong, a primary hospital is the smallest category of healthcare facility, provid- ing primary medical services such as medical treatment, prevention, healthcare, and rehabilitation for a commu- nity with a population of <100,000 persons. The second- ary hospitals provide general medical services, including medical treatment, prevention, healthcare, and reha- bilitation for larger communities (population >100,000 persons). The tertiary hospitals are regional healthcare centers providing specialized, high-complexity health- care services for several districts. Most of the sentinel sites were closed to the Exposition venue, where was also the part of Pudong with highest population density (Fig. 1).h The concerning diseases during mass gatherings and their typical symptoms were listed according to litera- ture review, and Delphi method was employed to con- sult 18 domestic epidemiologic and clinical professionals to score the disease severity, risk probabilities. Then a disease-risk matrix was draw and 40 diseases were pri- oritized in the surveillance, including local common diseases and some highly concerned diseases of importa- tion. The 25 most common and typical symptoms of the The 41st World Exposition was held in Shanghai city, China, from May 1st to October 31st, 2010. It has been the largest exposition in the history, with more than 73 million visitors from approximately 240 countries and organizations. A large flow of international visitors might pose a potential public health risk of disease importation, as infectious disease transmission can be promoted by an increasing population flow and density or by the acci- dental or deliberate introduction of unusual pathogens. Pudong New Area is the largest district of Shanghai city, containing 5.4 million residents and approximately 60 % areas of the World Exposition sites. A routine notifiable infectious disease reporting system has been used, which only collected information of specific kinds of infectious disease diagnosed by the clinicians. To enhance the sen- sitivity and timeliness of disease outbreaks detection dur- ing the Expo, a syndromic surveillance and early warning system, Pudong Syndromic surveillance and Early Warn- ing System (PD-SEWS), was implemented in this district. Fig. 1 The geographic location of sentinel hospitals at three different levels in Pudong New Area, Shanghai, China, 2010 Fig. Surveillance sentinels and targeted syndromes 1 The geographic location of sentinel hospitals at three different levels in Pudong New Area, Shanghai, China, 2010 To allow automated electronic access to chief com- plaints for PD-SEWS in the 41st World Exposition held in Shanghai in 2010, we have piloted to introduce a novel approach, the symptom-clicking-module (SCM), to gather syndrome data from hospitals, thereby ena- bling implementation of syndromic surveillance. Here, Fig. 1 The geographic location of sentinel hospitals at three different levels in Pudong New Area, Shanghai, China, 2010 Fig. 1 The geographic location of sentinel hospitals at three different levels in Pudong New Area, Shanghai, China, 2010 Ye et al. BMC Res Notes (2016) 9:315 Page 3 of 9 targeted diseases were enrolled and classified to seven targeted syndromes: acute respiratory, acute gastroin- testinal, rash with fever, neurological syndrome, hem- orrhagic fever, botulism-like syndrome and acute viral hepatitis (Table 1). could therefore check each row of the table quickly. For example, if a patient described the chief complaint as ‘fever and cough’, the clinician should click ‘fever’ in the first row and ‘cough’ in the second, as well as ‘none of the left’ for the other three rows. Clinicians were pre- vented from moving to the next step of treatment (such as prescribing a medication) until they finished recording symptoms by clicking the ‘save’ button (Fig. 3). Data collection and transmissionh The SCM was developed and embedded in the HIS for each sentinel hospital, with the 25 symptoms presented on the SCM interface as a 5 × 5 table which could be selected by single click on the screen (Fig. 2). In addition, we included an extra column to facilitate data entry and improve data quality. If no symptoms in a row were pre- sented, the ‘none of the left’ could be clicked. A clinician A record of symptoms in SCM was generated with a unique identity (ID) number and the basic demographic information registered in HIS. All data were stored in real time and transmitted automatically to the Pudong Public Health Database Center each day. A virtual pri- vate network (VPN) was used to connect securely with sentinel hospitals, and all data were supplied and ana- lyzed in an anonymous format, without access to per- sonal identifying information. For improving the data quality, duplicated records were rejected by checking the unique ID number, and if the data were not received from a sentinel site, the Public Health Database Center would send a notice to that hospital at 08:00 a.m. the next day. Once the database center received the surveillance data, all of the patient records were automatically grouped and aggregated into the seven syndromes according to the criteria listed in Table 1. If one patient’s symptoms referred to more than one syndrome, then the encounter would be counted separately for each syndrome. Table 1 The seven syndromes under surveillance in the 41st exposition, Pudong New Area, Shanghai City, China, 2010 Syndrome Typical symptoms Acute respiratory syndrome Fever with at least one of the following: cough, sputum, hemoptysis, chest pain, breathing difficulties Acute gastrointestinal syndrome Fever with at least one of the following: vomiting, diarrhea, pus/mucus in stool Rash with fever Fever with at least one of the following: herpes, maculopapular rash Neurological syndrome Fever with at least one of the following: headache, projectile vomiting, shock, altered consciousness, sudden body pain Hemorrhagic fever Fever with at least one of the following: skin or mucous congestion, petechiae, bleeding, bloody stool Botulism-like syndrome At least one of the following: sudden blurred vision, dysphagia Acute viral hepatitis At least one of the following: hepatosple- nomegaly, acute jaundice, lymphad- enopathy Table 1 The seven syndromes under surveillance in the 41st exposition, Pudong New Area, Shanghai City, China, 2010 Data analysis and aberration detection Data analysis and aberration detection We developed an interface connecting to the Pudong Public Health Database Center. The interface allowed authorized users to generate customized time series of total visits for each syndrome stratified by gender, age, syndrome, hospital, start dates and end dates. The cumu- lative sum (CUSUM) method, a control chart method commonly used in syndromic surveillance [4, 13], was applied daily to analyze the aggregated data of all sentinel Fig. 2 User interface of the symptom-clicking-module of PD-SEWS, Shanghai, China, 2010. All English words in the SCM were translated from Chinese words Fig. 2 User interface of the symptom-clicking-module of PD-SEWS, Shanghai, China, 2010. All English words in the SCM were translated from Chinese words Ye et al. BMC Res Notes (2016) 9:315 Page 4 of 9 Fig. 3 Framework of syndromic surveillance in Pudong New Area, Shanghai, China, 2010 ig. 3 Framework of syndromic surveillance in Pudong New Area, Shanghai, China, 2010 patients. If the signal indicated a potential outbreak, field investigation would be performed to obtain more detailed epidemiological information and necessary control measures would be conducted to prevent fur- ther spread of the disease. Outbreak is defined as the occurrence of cases of disease in excess of what would normally be expected in a defined geographical area and period, which was further quantitatively defined for each kind of disease in this study, so as to facilitate the outbreak confirmation and report. For example, an out- break of hand, foot and mouth (HFM) disease is defined as “within 1 week, at least five HFM disease cases occur in the same setting-e.g. kindergarten or school-or at least three cases of the disease occur in the same village or community” [15]. hospitals for detecting abnormal temporal increases [14]. For each targeted syndrome, CUSUM compared the pro- portion of syndrome counts in total visits in the current day (day 0) with the corresponding mean proportion and standard deviation of the past 7 days (day-7 to day-1). A signal was generated if the value of comparison exceeded two standard deviations. The surveillance and response team of Pudong Center for Disease Control and prevention (Pudong CDC) would monitor the warning signals routinely. When a signal was triggered, the verification would be con- ducted immediately by analyzing and reviewing the data to identify any unusual cluster of gender, age, occupa- tion or hospital. Data analysis and aberration detection Some signals would be verified and compared with the data in the routine notifiable dis- ease reporting system, which recorded the confirmed patients’ detailed information. Potential epidemic asso- ciation would be explored by calling the clinicians or In this study, the number of outpatients, the reporting frequency and the proportion of each syndrome were calculated. The average reporting frequencies of different Ye et al. BMC Res Notes (2016) 9:315 Page 5 of 9 Page 5 of 9 syndromes between weekdays and weekends were com- pared using a Chi square test, and a P value <0.05 was considered to be statistically significant. were higher on the weekdays than that in weekends, but the variation was only significant for acute gastrointesti- nal syndrome and acute viral hepatitis (Fig. 4). Addition- ally, in terms of the average reporting frequency of each syndrome across the sentinel hospitals, we found that the tertiary and secondary hospitals overall contributed much more than the primary hospitals, and the differences var- ied with syndromes. For rash with fever syndrome, the average reporting number in the tertiary hospitals was 151 times of that in the primary ones. However, for acute viral hepatitis syndrome, the average reporting numbers in the three levels of hospital were closer (Table 3). Results During the surveillance period (184 days) from May 1, 2010 to October 31, 2010, totally 1,730,797 patient encounters were collected through the SCM. Of these encounters, 105,352 (6.1 %) were classified into the seven targeted syndromes under surveillance. The most frequent syndromes were acute respiratory syndrome (59,793 encounters) and acute gastrointestinal syndrome (45,634 encounters), which together accounted for 92.1 % of total visitors in the seven targeted syndromes. Botu- lism-like syndrome and hemorrhagic fever syndrome had the lowest frequencies, with only 187 and 131 encoun- ters, respectively. Encounters classified as acute respira- tory, acute gastrointestinal, and neurological syndromes were reported every day, whereas the other syndromes (including rash with fever, acute viral hepatitis, botulism- like syndrome and hemorrhagic fever) were not found from 7 to 122 days during surveillance period (Table 2). Preparation for system operation A series of training sessions on the system were con- ducted for the clinicians in the sentinel hospitals before the practical operation. This approach helped to ensure that the clinicians would be familiar with SCM in HIS. The estimated average time cost of recording a patient’s syndrome by SCM was 10 s, according to our field inves- tigation in the last week of the training. A pilot surveil- lance operation was conducted 2 weeks before the formal running of this syndromic surveillance system, and the staff from Pudong CDC went to all the 21 sentinel hos- pitals during this pilot period to provide help if needed. During the surveillance period, the CUSUM triggered 191 signals, including 44 signals for acute respiratory syndrome, 41 signals for neurological syndrome, and 30 signals for acute gastrointestinal syndrome, and rash with fever syndrome, botulism-like syndrome, hemorrhagic fever and acute viral hepatitis had 22, 22, 17 and 15 sig- nals respectively. All of the signals had been verified, and field investigation and control measures had been con- ducted if a signal seemed to lead to an outbreak. Finally, none of the signals were confirmed as outbreaks. Mean- while, there was no infectious disease outbreak or public health emergency was confirmed and reported by local CDC through the other routine surveillance system dur- ing the same period. Discussionh The SCM was implemented successfully and operated for a surveillance period lasting 184 days in Expo 2010, which made it feasible to collect syndromic data and classify syndromes which was lacking in the HIS of sen- tinel hospitals. Approximately 6 % of patient encounters were classified into the seven surveillance syndromes, with the number of patients ranged from 131 in hemor- rhagic fever syndrome to 59,793 in acute respiratory syn- drome. Frequency of reporting varied by days of week for some syndromes, and the secondary and tertiary The reporting frequencies for different syndromes var- ied with the day of the week. Except for hemorrhagic fever, the frequency of all other six surveillance syndromes Table 2 Descriptive statistics of visit counts for the seven syndromes in PD-SEWS, Shanghai, China, May 1st to Oct 31st, 2010 Q1 first quartile value; Q3 third quartile value Syndrome Overall visits Proportion of the total visits (%) Mean visits per day Minimum visits per day Q1 visits per day Median visits per day Q3 visits per day Maximum visits per day Days with zero reporting Acute respiratory syndrome 59,793 3.45 325 143 273 310.5 356.5 642 0 Acute gastrointestinal syndrome 45,634 2.64 248 123 196.8 249 295.8 398 0 Neurological syndrome 6055 0.35 32.9 9 20 29 41 112 0 Rash with fever 1910 0.11 10.4 0 3.8 7.5 16 40 7 Acute viral hepatitis 790 0.05 4.3 0 1 3 6 26 25 Botulism-like syndrome 187 0.01 1 0 0 0 2 13 110 Hemorrhagic fever 131 0.01 0.7 0 0 0 1 7 122 atistics of visit counts for the seven syndromes in PD-SEWS, Shanghai, China, May 1st to Oct 31st, Table 2 Descriptive statistics of visit counts for the seven syndromes in PD-SEWS, Shanghai, China Table 2 Descriptive statistics of visit counts for the seven syndromes in PD-SEWS, Sh 2010 Ye et al. BMC Res Notes (2016) 9:315 Page 6 of 9 Fig. 4 Box plot of average reporting number of each day of week by syndrome detected by PD-SEWS, Shanghai, China, 2010 ig. 4 Box plot of average reporting number of each day of week by syndrome detected by PD-SEWS, Shanghai, China, 2010 hospitals reported the majority of encounters in syn- drome surveillance. data collection to make the establishment of syndromic surveillance more easily and cost-effectively [17, 18]. Discussionh The SCM in this study collated and analyzed syndromic data in a standard format across reporting hospitals, and allowed direct and automatic classification of symptoms into syndromes, avoiding to develop additional syndrome In recent years, enhanced surveillance is commonly implemented during mass gatherings [16], and syndro- mic surveillance system has been established globally. In some countries, more efforts are needed to improve Ye et al. BMC Res Notes (2016) 9:315 Page 7 of 9 Table 3 Average reporting amount of each syndrome per hospital by levels of hospital in Pudong New Area, Shanghai, China, 2010 Syndromes Average reporting amount by hospital’s level Ratio of primary to sec- ondary and tertiary hospital (a:b:c) Primary (a) Secondary (b) Tertiary (c) Acute respiratory syndrome 707.4 8544.0 3585.0 1:12:5 Acute gastrointestinal syndrome 756.0 5508.8 3753.0 1:7:5 Neurological syndrome 63.7 920.6 280.0 1:14:4 Rash with fever 4.4 106.4 658.5 1:24:151 Acute viral hepatitis 28.4 56.8 54.5 1:2:2 Botulism-like syndrome 3.3 15.0 33.0 1:5:10 Hemorrhagic fever 3.0 9.2 21.5 1:3:7 Table 3 Average reporting amount of each syndrome per hospital by levels of hospital in Pudong New Area, Shanghai, China, 2010 ting amount of each syndrome per hospital by levels of hospital in Pudong New Area, Shanghai, classification software [19, 20]. Instead of directly collect- ing the syndrome data subjectively decided by the clini- cians, SCM allowed clinicians to record the raw data of symptoms on a patient, and the system automatically grouped the encounters into corresponding surveillance syndromes using pre-defined syndrome definitions. Additionally, SCM was embedded into the HIS so that all the basic information of the patients could be extracted from the registration data collected automatically. The design of SCM saves time during data collection and helps to avoid errors due to manual syndrome classifica- tion, which is important and effective because the most existing HIS in China without recording chief complaints of patients. patients sought medical services at the nearest primary hospital. Additionally, the SCM is a new data collection module embedded into the routine HIS at each hospital, and the clinicians might need some time to accustom his change in their routine workflow. The strategy of con- ducting pilot surveillance is widely adopted by syndro- mic surveillance practice during mass gathering, which is key to improve the system’s acceptability and ensure the data quality, which has been also conducted in this study [22–24]. Discussionh In this study, no one outbreak was confirmed among the 191 signals generated by PD-SEWS during the Expo. Each signal, which represented an aberration on the surveillance data, was required to be timely verified and investigated by the local staff of Center for Disease Control and Prevention, and control measures should be conducted rapidly once a signal seemed to lead to an out- break. That’s the possible reason why no one signal was identified as an outbreak finally. To enhance the routine surveillance system by increasing the sensitive and timely detection of outbreaks was the important objective of syndromic surveillance, as well as to deal with each signal before it led to a real outbreak. The system would create a lot of extra work for Public Health Utility, that’s why we only suggested adopting the enhanced syndromic surveil- lance system during the specific period, such as holding mass gatherings, instead of conducting it as routine work after the Expo. We found that acute respiratory and acute gastrointes- tinal syndromes were reported most frequently, which is consistent with other syndromic surveillance systems [21–23]. These syndromes cover the symptoms of some common diseases, such as influenza and viral diarrhea, which are also the main infectious diseases in Shang- hai during the study period. Moreover, the day-of-week effect in our data suggests the residents in study area avoided to seek healthcare during weekend, possibly because hospitals could provide more medical services on weekdays. Therefore, the aberration detection algo- rithm on the syndromic surveillance system should take account of the day-of-week effect [22].hf f The variation in reporting across the different lev- els of hospital was also observed for each syndrome. We found that most of the fever with rash cases were reported by one tertiary hospital, which was the big- gest children’s hospital in Pudong. Fever with rash syn- drome usually presented among the diseases such as measles, rubella or chicken pox with high incidences in children. The primary hospitals reported the least cases in all of the syndromes, but the difference seemed much smaller for the acute viral hepatitis syndrome. This dif- ference is probably because most of the chronic hepatitis There were some limitations in our study. First, the cost-effectiveness of the SCM application was not evalu- ated systematically, but we performed a pilot investiga- tion of the SCM’s usability and acceptability in several hospitals at the beginning of the surveillance. Funding This study was supported by the Grants from the National Science and Technology Key Projects (No. 2009ZX10004-201, No. 2012ZX10004-201) and Ministry of Health, China (No. 201202006). These funding offices had no involvement in the design, data collection analysis, write up and decision for the results to be published. Consent for publication p Not applicable as the manuscript does not contain any individual persons data. Abbreviations SCM t SCM: symptom-clicking-module; HIS: hospital information system; PD-SEWS: pudong syndromic surveillance and early warning system; FEP: front end pro- cessor; PDCDC: Pudong New Area Center for disease control and prevention; VPN: virtual private network; CUSUM: cumulative sum. 5. Khan K, McNabb SJ, Memish ZA, Eckhardt R, Hu W, Kossowsky D, et al. Infectious disease surveillance and modelling across geographic frontiers and scientific specialties. Lancet Infect Dis. 2012;12(3):222–30. 5. Khan K, McNabb SJ, Memish ZA, Eckhardt R, Hu W, Kossowsky D, et al. Infectious disease surveillance and modelling across geographic frontiers and scientific specialties. Lancet Infect Dis. 2012;12(3):222–30. Author details 1 1 Research Base of Key Laboratory of Surveillance and Early‑warning on Infec- tious Disease in China CDC, Shanghai Pudong New Area Center for Disease Control and Prevention, Shanghai, China. 2 Key Laboratory of Surveillance and Early‑warning on Infectious Disease, Chinese Center for Disease Control and Prevention, Beijing, China. 3 Research Base of Key Laboratory of Surveil- lance and Early‑warning on Infectious Disease in China CDC, West China School of Public Health, Sichuan University, Chengdu, China. 4 Department of Geography and Environment, University of Southampton, Southampton, UK. 5 McGill University, Montreal, Canada. 9. Chapman WW, Christensen LM, Wagner MM, Haug PJ, Ivanov O, Dowl- ing JN, et al. Classifying free-text triage chief complaints into syndro- mic categories with natural language processing. Artif Intell Med. 2005;33(1):31–40. 10. South BR, Chapman WW, Delisle S, Shen S, Kalp E, Perl T, et al. Optimiz- ing A syndromic surveillance text classifier for influenza-like illness: does document source matter? AMIA Annu Symp Proc. 2008;692–6. 11. May L, Chretien JP, Pavlin JA. Beyond traditional surveillance: applying syndromic surveillance to developing settings–opportunities and chal- lenges. BMC Pub Health. 2009;9:242. Authors’ contributions 6. May LS, Griffin BA, Bauers NM, Jain A, Mitchum M, Sikka N, et al. Emergency department chief complaint and diagnosis data to detect influenza-like illness with an electronic medical record. West J Emerg Med. 2010;11(1):1–9. QS, WY, YL and ZL made substantial contributions to the design of this study. CY, ZL, WZ and YF made contributions to acquisition of data, data collection supervision. CY HZ and SL were involved in data analysis, interpretation of data and preparation of the manuscript. CY, ZL, HZ, DZ and DB were involved in drafting the manuscript or revising it. All authors read and approved the final manuscript. 7. Moore KM, Edgar BL, McGuinness D. Implementation of an auto- mated, real-time public health surveillance system linking emergency departments and health units: rationale and methodology. CJEM. 2008;10(2):114–9. 8. Khan K, Freifeld CC, Wang J, Mekaru SR, Kossowsky D, Sonricker AL, et al. Preparing for infectious disease threats at mass gatherings: the case of the Vancouver 2010 Olympic Winter Games. CMAJ. 2010;182(6):579–83. Conclusions 1. Chretien JP, Burkom HS, Sedyaningsih ER, Larasati RP, Lescano AG, Mun- daca CC, et al. Syndromic surveillance: adapting innovations to develop- ing settings. PLoS Med. 2008;5(3):e72. 1. Chretien JP, Burkom HS, Sedyaningsih ER, Larasati RP, Lescano AG, Mun- daca CC, et al. Syndromic surveillance: adapting innovations to develop- ing settings. PLoS Med. 2008;5(3):e72. 1. Chretien JP, Burkom HS, Sedyaningsih ER, Larasati RP, Lescano AG, Mun- daca CC, et al. Syndromic surveillance: adapting innovations to develop- ing settings. PLoS Med. 2008;5(3):e72. SCM has acted as a practical tool for recording symp- toms in the hospital-based enhanced syndromic sur- veillance system during the 41st World Exposition in Shanghai, in the context of without a preexisting elec- tronic tool to collect syndromic data in the HIS of the sentinel hospitals. 2. Wu TS, Shih FY, Yen MY, Wu JS, Lu SW, Chang KC, et al. Establishing a nationwide emergency department-based syndromic surveillance system for better public health responses in Taiwan. Bmc Pub Health. 2008;8:18. 2. Wu TS, Shih FY, Yen MY, Wu JS, Lu SW, Chang KC, et al. Establishing a nationwide emergency department-based syndromic surveillance system for better public health responses in Taiwan. Bmc Pub Health. 2008;8:18. 3. Abubakar I, Gautret P, Brunette GW, Blumberg L, Johnson D, Poumerol G, et al. Global perspectives for prevention of infectious diseases associated with mass gatherings. Lancet Infect Dis. 2012;12(1):66–74. 3. Abubakar I, Gautret P, Brunette GW, Blumberg L, Johnson D, Poumerol G, et al. Global perspectives for prevention of infectious diseases associated with mass gatherings. Lancet Infect Dis. 2012;12(1):66–74. g g 4. Yan WR, Nie SF, Xu B, Dong HJ, Palm L, Diwan VK. Establishing a web- based integrated surveillance system for early detection of infectious disease epidemic in rural China: a field experimental study. BMC Med Inform Decis Mak. 2012;12:4. Ethics approval and consent to participate The surveillance was approved by Shanghai Pudong New Area Center for Disease Control and Prevention Review Board. The patient information was anonymized and de-identified prior to analysis. Discussionh In addition, we have not systematically collected and evaluation the feedbacks from the clinicians during the study period. Nonetheless, according to the data collected by SCM and its operating results, it sounds acceptable for physicians to use during the special situation of a mass gathering. Ye et al. BMC Res Notes (2016) 9:315 Ye et al. BMC Res Notes (2016) 9:315 Page 8 of 9 Availability of data and materials Availability of data and materials Another limitation is that spatial cluster detection was not conducted in our study. Some algorithms such as space–time scan statistic [25] have been applied to detect spatial aberration, but SCM without the sufficient geo- graphic information of each patient to conduct this anal- ysis. Additionally, as the CUSUM method was applied to the aggregated data of all hospitals instead of each single sentinel site, this limited the ability to detect small out- breaks. However, as the surveillance data for each senti- nel hospital (including three different levels of hospitals) were not stable enough on the time series, which would lead to large amount of false alerts if performing the algo- rithms on single hospital level. Therefore, taking account of the balance between sensitivity and false alarm rate, we applied CUSUM method to the total number of each syndrome, and adopted a relatively lower threshold to ensure the sensitivity. In addition, the system could auto- matically demonstrate daily time series figure of surveil- lance data for each single hospital, which could assist in the local epidemiologist to detect the potential smaller outbreak with manual manner. Original data are available upon request from the first author. Competing interests We confirm that none of the authors have any competing interests. Acknowledgements We thank the clinicians in the sentinel hospitals for their assistance in data collection. We thank the clinicians in the sentinel hospitals for their assistance in data collection. Ye et al. BMC Res Notes (2016) 9:315 Page 9 of 9 12. Wang JF, Reis BY, Hu MG, Christakos G, Yang WZ, Sun Q, et al. Area disease estimation based on sentinel hospital records. PLoS One. 2011;6(8):e23428. 19. Lu HM, Chen H, Zeng D, King CC, Shih FY, Wu TS, et al. Multilingual chief complaint classification for syndromic surveillance: an experiment with Chinese chief complaints. Int J Med Inform. 2009;78(5):308–20. 20. Chapman WW, Dowling JN, Wagner MM. Classification of emergency department chief complaints into 7 syndromes: a retrospective analysis of 527,228 patients. Ann Emerg Med. 2005;46(5):445–55. 13. Fricker RD Jr, Hegler BL, Dunfee DA. Comparing syndromic surveillance detection methods: EARS’ versus a CUSUM-based methodology. Stat Med. 2008;27(17):3407–29. 21. Dafni UG, Tsiodras S, Panagiotakos D, Gkolfinopoulou K, Kouvatseas G, Tsourti Z, et al. Algorithm for statistical detection of peaks–syndromic surveillance system for the Athens 2004 Olympic Games. MMWR Morb Mortal Wkly Rep. 2004;53(Suppl):86–94. 14. Rogerson PA, Yamada I. Approaches to syndromic surveillance when data consist of small regional counts. MMWR Morb Mortal Wkly Rep. 2004;53(Suppl):79–85. 15. National health and family planning commission of China. Guideline for hand, foot and mouth disease outbreaks disposal (2012 edition). 2012. http://www.nhfpc.gov.cn/zhuzhan/wsbmgz/201304/2455757fe843447c 8289e1431b20a1a9.shtml. Accessed 18 May 2016. y p pp 22. Meyer N, McMenamin J, Robertson C, Donaghy M, Allardice G, Cooper DA. Multi-data source surveillance system to detect a bioterrorism attack during the G8 summit in Scotland. Epidemiol Infect. 2008;136(7):876–85. 16. Thackway S, Churches T, Fizzell J, Muscatello D, Armstrong P. Should cities hosting mass gatherings invest in public health surveillance and plan- ning? reflections from a decade of mass gatherings in Sydney, Australia. BMC Pub Health. 2009;9:324. 23. Yan W, Palm L, Lu X, Nie S, Xu B, Zhao Q, et al. ISS—an electronic syndro- mic surveillance system for infectious disease in rural China. PLoS One. 2013;8(4):e62749. 24. Harcourt SE, Fletcher J, Loveridge P, Bains A, Morbey R, Yeates A, et al. Developing a new syndromic surveillance system for the London 2012 Olympic and Paralympic Games. Epidemiol Infect. 2012;140(12):2152–6. 17. Ding Y, Fei Y, Xu B, Yang J, Yan W, Diwan VK, et al. Acknowledgements Measuring costs of data collection at village clinics by village doctors for a syndromic surveil- lance system—a cross sectional survey from China. Bmc Health Serv Res. 2015;15:287. 25. Nordin JD, Goodman MJ, Kulldorff M, Ritzwoller DP, Abrams AM, Klein- man K, et al. Simulated anthrax attacks and syndromic surveillance. Emerg Infect Dis. 2005;11(9):1394–8. 18. Fan Y, Wang Y, Jiang H, Yang W, Yu M, Yan W, et al. Evaluation of outbreak detection performance using multi-stream syndromic surveillance for influenza-like illness in rural Hubei Province, China: a temporal simulation model based on healthcare-seeking behaviors. PLoS One. 2014;9(11):e112255. 25. Nordin JD, Goodman MJ, Kulldorff M, Ritzwoller DP, Abrams AM, Klein- man K, et al. Simulated anthrax attacks and syndromic surveillance. Emerg Infect Dis. 2005;11(9):1394–8. y 22. Meyer N, McMenamin J, Robertson C, Donaghy M, Allardice G, Cooper DA. Multi-data source surveillance system to detect a bioterrorism attack during the G8 summit in Scotland. Epidemiol Infect. 2008;136(7):876–85. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and we will help you at every step:
https://openalex.org/W4252687917	https://www.researchsquare.com/article/rs-75406/latest.pdf	English	null	Intramedullary Nailing Combined With Adjunct Plate or Blocking Screw for Proximal Tibial Fractures	Research Square (Research Square)	2,020	cc-by	5,758	Intramedullary Nailing Combined With Adjunct Plate or Blocking Screw for Proximal Tibial Fractures Jie li Xi'an Honghui Hospital Qian Wang Xi'an Honghui Hospital Zhong Li Xi'an Honghui Hospital Kun Zhang (  zk612730@163.com ) Xi'an Honghui Hospital Research Article Keywords: proximal tibial fracture, interlocking intramedullary nail, adjunct plate, blocking screw Posted Date: July 14th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-75406/v2 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Intramedullary Nailing Combined With Adjunct Plate or Blocking Screw for Proximal Tibial Fractures Jie li Xi'an Honghui Hospital Qian Wang Xi'an Honghui Hospital Zhong Li Xi'an Honghui Hospital Kun Zhang (  zk612730@163.com ) Xi'an Honghui Hospital Research Article Keywords: proximal tibial fracture, interlocking intramedullary nail, adjunct plate, blocking screw Posted Date: July 14th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-75406/v2 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Research Article Posted Date: July 14th, 2021 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. R d F ll Li License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/16 Abstract Objective: To compare the traditional approach of intramedullary nail with an extra plate versus the nail combined with blocking screws for proximal tibia fractures without the knee involved. Objective: To compare the traditional approach of intramedullary nail with an extra plate versus the nail combined with blocking screws for proximal tibia fractures without the knee involved. Methods: From January 2013 to January 2017, a total of 36 patients who suffered from proximal tibial fractures unaffecting the knee were enrolled into this prospective study, and divided into two groups by random number table method. Of them, 19 patients received an interlocking intramedullary nail combined with an extra plate for internal fixation of the fractures (the plate group), while the remaining 17 patients had fractures fixed with the nail combined with blocking screws (the screw group). The perioperative, follow-up and radiographic data were compared between the two groups. Results: All the 36 patients underwent operation smoothly without iatrogenic neurovascular injuries. The plate group proved superior the screw group regarding to operation time and intraoperative X-ray exposure (P＜0.05) , nevertheless the former was inferior to the latter in implant cost and hospital stay (P＜ 0.05) . The follow-up period lasted for 12～24 months with a mean of (15.62±4.71) months. There were no statistically significant differences in the time to return ambulation and the time to full weight-bearing activity between the two groups (P＞0.05) . At the latest follow up, no statistically significant differences were found between the two groups regarding knee range of motion and Johner-Wruhs grades for clinical consequences (P＞0.05) . In terms of anterior knee pain, the difference between the two groups was not statistically significant (P＞0.05) . In respect of radiographic assessment, the plate group had significantly less residual malalignment than the screw group, including anteroposterior and lateral displacements, as well as angulations in coronal and sagittal planes (P＜0.05) . To the latest follow up, all patients in both group got bony healing of the fractures without a statistical difference in fracture healing time between them (P＞0.05) , and no loosening or breaking of the implants were showed on images in anyone of them. Conclusion: Both the nail plus plate and nail plus blocking screw do achieve satisfactory clinical outcomes for proximal tibial fractures unaffecting the knee. By comparison, the nail combined with plate facilitates to regain and maintain better alignment of the leg regardless of higher implant cost. 2.1 Patient selection Inclusion criteria: 1) Proximal tibial fracture; 2) Can tolerate surgery, and knee joint movement function was normal before this injury; 3) The fracture occurred within 1 week of surgery; 4) Agreed to participate in the study and sign an informed consent form; and 5) Patients with follow-up time ≥ 12 months. Exclusion criteria: 1) Pathological fracture; 2) Severe osteoporosis; 3) Patients with a history of mental illness; 4) Obvious surgical contraindications; 5) Patients with compartment syndrome; 6) Open fractures of Gustilo–Anderson class II and above; 7) Vascular injury patients; 8) Patients with narrow medullary cavity (intramedullary nails cannot be used); 9) Patients with open osteophytes; and 10) Patients with tibial deformity and those whose tibia cannot be fixed with intramedullary nails. Backgroud Proximal tibial fractures account for 5–11% of all tibial shaft fractures. This type of fracture is often caused by high-energy injury. In addition to the fracture itself, the fracture is often accompanied by different degrees of soft tissue injury[1]. Conservative treatment of fractures at this site often leads to complications, such as fracture malunion, nonunion, rotational displacement, and adjacent joint stiffness[2]. In order to avoid these complications, clinicians currently use surgical methods to treat fractures in this area. At present, there are many methods for internal fixation of this type of fracture, such as intramedullary nails, bone plates, and external fixators. However, it is still unclear which type of internal fixation is best[3]. In recent years, with the understanding of clinicians on the importance of skin soft tissue and bone blood transport, it has become more and more common to use the intramedullary nail Page 2/16 Page 2/16 technique to treat this type of fracture[4]. Due to the special anatomy of the proximal tibia, it is easy to use the traditional approach of the intramedullary nail; however, this can cause poor resetting and angular deformity. For this reason, the attachment of a small plate to the nail is advisable. In addition, with the maturity of the barrier nail technology and its application in proximal humerus fractures, new ideas and technology to solve the above problems have been presented that have achieved good clinical results. Consequently, two types of surgical intervention are now available for the treatment of proximal tibial fractures: intramedullary nailing with adjunct plate and intramedullary nailing with blocking screw. However, the evidence is currently insufficient to determine which of these two surgical approaches is better. For this reason, based on the literature that has been published so far and the evidence for the lack of treatment for such fractures, we prospectively analyzed 36 cases of proximal tibial dry fractures admitted to our hospital from January 2013 to January 2017 to compare the efficacy and related factors of each treatment. 2.3.1 Subgingival approach, intramedullary nail, adjunct plate Plate group: Under general anesthesia, the patient is placed in a supine position, and the affected side is raised. Make an incision of about 6 cm in length at the fracture site of the proximal and posterior side of the calf, and make a layer-by-layer incision to expose the fractured end. Be careful not to strip the periosteum in a large area. The anatomical relationship of the fracture reduction can be distinguished. Use after reduction under direct vision After shaping, the 5 to 6-hole 3.5 mm system reconstruction locking bone plate is placed on the posterior medial side of the tibia, and the single cortical locking screws are screwed in in turn. After fluoroscopy confirms the reduction and fixation position is satisfied, close the incision. Then use the anterior inferior patella approach, split the patellar ligament in the middle, clear the subpatellar fat pad, and expose the tibial plateau slope. Before the opening of the medullary cavity, the correct nail entry point should be determined under fluoroscopy. This operation is particularly important. The correct entry point should be located at the anterior midline of the cartilage, slightly higher than the tibial tubercle[5]. Use an opener to make a hole in the direction of the guide pin, insert the reamed long guide pin, select the appropriate intramedullary nail and place it along the guide pin after stepwise reaming, and try to make the tip of the tibial intramedullary nail close to the articular surface of the distal tibia. Perform the interlocking nails at the far and near ends. After the far and proximal ends are locked, the tail cap is installed, the saline is flushed, the drainage tube is left, and the incision is sutured layer by layer. 2.3.2 Blocking nail technology Blocking nail group: The patient is placed in a supine position, and t patellar ligament in the middle, clear the subpatellar fat pad, and ex Before the opening of the medullary cavity, the correct nail insertion fluoroscopy. Use an opener to make a hole in the direction of the gu fluoroscopy. Correct the lateral position, remove the finger reducer ac outside of the broken end and the front and back angles, and drill a position as a temporary stop nail, insert the finger reducer again, and the position of the blocking nail when the inside and outside of the b and back angles. The following steps are the same as the steel plate Blocking nail group: The patient is placed in a supine position, and the affected side is raised. Split the patellar ligament in the middle, clear the subpatellar fat pad, and expose the slope of the tibial plateau. Before the opening of the medullary cavity, the correct nail insertion point should be determined under fluoroscopy. Use an opener to make a hole in the direction of the guide pin, insert a finger reducer, and fluoroscopy. Correct the lateral position, remove the finger reducer according to the angled inside and outside of the broken end and the front and back angles, and drill a 2.5 mm Kirschner wire in the proper position as a temporary stop nail, insert the finger reducer again, and see through the broken end. Adjust the position of the blocking nail when the inside and outside of the broken end are angled and the front and back angles. The following steps are the same as the steel plate group. 2.2 General information This study prospectively analyzed 36 cases of proximal tibial dry fractures admitted to the Department of Trauma and Orthopaedics, Xi'an Honghui Hospital from January 2013 to January 2017. Sealed envelopes encoded according to computer randomization order were randomly grouped for the patients under study. The envelopes were opened by a roving nurse after the patient entered the operating room, to inform the surgeon which internal fixation method to adopt. The patients who received the traditional approach of intramedullary nail and small plate internal fixation was classified as the “steel plate” group, and the internal fixation with the barrier nail technique was classified as the “barrier nail” group. The preoperative general information of the two groups of patients is shown in Table 1. There was no significant difference in age, gender, injury factors, and fracture types between the two groups (P>0.05)。 This study was approved by the ethics committee of this unit, and all patients provided signed, informed consent. Page 3/16 2.3 Surgical procedure All patients were given cephalosporin antibiotics to prevent infection and limb elevation All patients were given cephalosporin antibiotics to prevent infection and limb elevation for 24 to 48 hours. X-ray films were reviewed on the second day after surgery, and patients were encouraged to perform primary and passive functional exercises on the knees, ankles, and toes. CPM (Joint function training machine) machines were used to assist exercise if necessary. Anterior and posterior radiographs of the affected tibia were taken after the drainage tube was removed the day Page 4/16 following surgery. Wound dressings were removed after two weeks. The patients were followed up once a month for the first three months after the operation, every 3–6 months after 3 months, and every 6–12 months after 1 year. The X-ray film showed that the affected limbs were partially loaded after the formation of the callus, and were completely loaded after the fracture end was healed. 2.5 Observation index and efficacy evaluation The operation time, the number of intraoperative fluoroscopy, the cost of surgical consumables, the length of hospital stay and early complications were recorded in the two groups of patients; the clinical results were evaluated by the time of ground movement, the time of full weight bearing, the range of motion of the knee joint and the Johner-Wruhs rating. Perform imaging examinations and use Freedmanand Johnson's method to measure tibia coronal and sagittal angulation deformities, namely: varus and valgus deformity on the coronal plane and anteroposterior angulation deformity on the sagittal plane, and angulation on any plane >5° is regarded as poor reset [6]. Record the fracture healing time. At the last follow-up, the therapeutic effect was evaluated according to the Johner-Wruhs tibial shaft fracture postoperative evaluation standard[7]. as follows: Excellent: good fracture healing, no neurological and vascular complications, no internal and external valgus deformity, anterior and posterior angle <5°, rotational shift angle <5°, shortening <5 mm, normal knee and ankle joint activity, subtalar joint activity degree >75%, no pain, normal gait, unlimited exercise; Good: fractures healed well, mild vascular nerve stimulation, internal and external turning angle <5°, anterior and posterior angle <10°, rotational shift angle <10°, shortening <10 mm, knee joint mobility >80%, ankle joint mobility >75%, subtalar joint mobility >50%, occasional pain, normal gait, limited exercise; Fair: fractures healed well, moderate blood vessels, nerve stimulation, internal and external angles <10°, anterior and posterior angles <20°, rotational displacement <20°, shortening <20 mm, knee mobility >75%, ankle joint mobility >50%, subtalar joint mobility <50%, moderate pain, mild lameness, severe exercise is severely restricted; Poor: fracture not healed or bone infection, amputation, severe vascular nerve stimulation, internal and external angulation angle >10°, anterior and posterior angle >20°, rotational displacement >20°, shortening >20 mm, knee joint activity degree <75%, ankle joint activity <50%, severe pain, obvious claudication, cannot do vigorous activity. 2.6 Statistical analysis Statistical analysis was performed using IBM SPSS 19.0 statistical software (SPSS, USA). The homogeneity of variance is expressed by`x ± s. Two independent sample t-tests were used to compare the two groups. The rate was compared using the χ2 test, and P < 0.05 was considered to be statistically significant. Results 3.1. Perioperative condition 3.1. Perioperative condition 3.1. Perioperative condition Page 5/16 Page 5/16 All 36 patients successfully completed the operation without any vascular or nerve injury during the operation. The perioperative data of the two groups of patients are shown in Table 2. The operation time and the number of intraoperative fluoroscopy in the plate group were significantly better than those in the blocking nail group (P＜0.05); but the incision length and intraoperative blood loss in the steel plate group were significantly greater than those in the blocking nail group(P＜0.05); The plate group was significantly lower than the blocking nail group in terms of surgical consumables cost and average hospital stay (P＜ 0.05). Of the 19 patients in the steel plate group, 15 wounds healed well as scheduled, 3 wounds healed delayed, 1 wound developed pus, and healed after debridement; 17 wounds in the blocking nail group had 15 wounds healed well as scheduled, and 2 wounds delayed Healed, no pus phenomenon appeared. Of the 19 patients in the steel plate group, there were 5 thrombosis, 1 was proximal thrombosis, and 4 were distal thrombosis; of the 17 patients in the blocking nail group, 3 patients had thrombus, all of which were distal thrombosis. 3.2 Follow-up results Patients in both groups were followed up for 12-24 months after operation, with an average of (15.62±4.71) months. During the follow-up, none of the patients suffered another serious trauma, and there was no revision surgery. The follow-up data of the two groups of patients are shown in Table 3. There was no significant difference between the two groups in walking time and full weight bearing time (P>0.05). At the last follow-up, there was no significant difference between the two groups in terms of knee range of motion, Johner-Wruhs postoperative functional rating and anterior knee pain (P>0.05). At the last follow-up, of the 19 cases in the plate group, 14 cases were completely painless, 4 cases had mild pain while walking, and 1 case had obvious pain; 16 cases walked normally without claudication, 3 cases had mild claudication; 17 cases had normal squatting activities. 2 cases were slightly restricted in squatting activities; 18 cases recovered pre-injury exercise and work ability, and 1 case did not return to the level of pre-injury exercise and work ability. Among the 17 cases in the blocking nail group, 14 cases were completely painless, 2 cases had mild pain while walking, and 1 case had obvious pain; 15 cases walked normally without claudication, 2 cases had mild claudication, 16 cases had normal squatting activities, and 1 case had normal squatting activities. Squatting activities were slightly restricted; 15 cases recovered their pre-injury exercise and work ability, and 2 cases did not return to the level of pre- injury exercise and work ability.A typical case picture is shown in figure (steel plate group 3.3.1, barrier nail group 3.3.2). Discussion Although intramedullary nailing is standard in the fixation of most femoral and tibial shaft fractures, it is still challenging for proximal tibial fractures[8]. Early literature reports that the use of iliac bone marrow nails for the treatment of proximal tibial fractures has a rate of malformation of up to 84%, especially for forward angular and valgus deformities[9]. Through the study of the anatomical features of the proximal tibia, this is thought to be due to the knee joint in the high flexion position of the anterior aspect of the tibia and the inner goose foot traction[10]. In order to be able to use intramedullary nails to treat proximal tibial fractures and reduce or avoid the above-mentioned malformation, try to achieve the purpose of anatomical reset. Dunbar[11]et al first proposed, before using the intramedullary nail, that 3.5mm system compression plate single-layer cortical fixation should be used for direct reduction of the proximal tibial fracture. After direct reduction, the traditional approach of intramedullary nail should be used to fix the proximal tibial fracture. In the 33 patients he studied, the effect of anatomical reduction after surgery was achieved.In our 19 cases of small inferior nail attachment to the intramedullary nail, we found that the probability of angular displacement and poor force line of the fracture ends was significantly reduced compared to the blocking nail technique. Later retrospective studies showed that the infection rate was not related to the surgical procedure. Yoon et al reached the same conclusion. We also reached this conclusion after comparing the postoperative infection rate between the two groups, that is, there was no difference in the infection rate between the two groups. In the study of Yoon et al, it was first proposed that the inferior approach of intramedullary nail plus small plate surgery can reduce the operation time. The same conclusion was obtained through our research. One of the biggest reasons for this reduction in operating time is that the fracture was opened and fixed under direct vision before fixation with intramedullary nailing. We also consider this factor to result in lower intraoperative fluoroscopy times of this surgical method compared to the blocking nail technique, an important factor for the better overall postoperative reduction. Kubiak et al[12]. reported that the use of percutaneously assisted intramedullary nailing can solve the above problems, but the precise reduction of the fracture is insufficient. 3.3 Image evaluation The postoperative residual fracture displacement and imaging bone healing time of the two groups of patients are shown in Table 4. Postoperative images showed that the residual lateral displacement, anteroposterior displacement, coronal and sagittal angulation of the plate group were significantly smaller than those of the blocking nail group, and the differences were statistically significant (P<0.05). By the time of the last follow-up, the two groups of patients had achieved imaging bone union. There was no significant difference in fracture healing time between the two groups (P>0.05). There was no Page 6/16 loosening or displacement of internal fixation in all patients. Images of two typical cases are shown in Figure 1 and Figure 2. loosening or displacement of internal fixation in all patients. Images of two typical cases are shown in Figure 1 and Figure 2. Discussion Our pursuit of perfection for resetting is another reason for not adopting this method. In order to solve the valgus of the anterior aspect of the proximal humerus and the valgus forward angulation caused by the inner goose foot traction in the high flexion position, Krettek et al[13]. proposed a blocking nail based on the proximal anatomical features of the humerus. The concept, carried out with biomechanical experiments, proved that the use of barrier nails can significantly reduce the probability of the tibia forward angle. Through the comparative study of its literature, the reduction rate can reach 25%. Since frequent intraoperative fluoroscopy is required in the use of the stab nail technique in order to find an appropriate blocking position, this does not intend to increase the number of surgical fluoroscopy and the amount of radiation, which is also a reason for prolonging the overall operation time. We demonstrate Page 7/16 Page 7/16 this by statistical analysis of the number of intraoperative radiation and the time of surgery in both groups. However, this procedure has the unparalleled advantages of soft tissue disturbance[14]. The soft tissue can be repaired quickly after bony structural force line and parasitic recovery.We compared the average hospitalization days between the two groups, We used a comparative analysis of the average hospitalization days between the two groups, considering that this is also a reason for patients with shorter treatment time in the hospital after using the technique of blocking nails. We analyzed the statistical analysis of the fracture healing time, nonunion, and knee resection in the treatment of proximal tibial fractures by adding small plate and occlusion nail technique to the submedullary nail. The results were not statistically significant (P >0.05). Although the Steel plate group was open-reset, we fully protected the periosteum during open reduction. In addition, we used the shape- reconstruction to lock the bone plate, which acted as an inner stent to reduce the damage to bone blood. In addition, the auxiliary small plate incision is closed before the insertion of the intramedullary nail, which causes the bone debris and growth factors generated during the reaming to be retained, and the internal factors of bone healing are not destroyed, Therefore, there is no statistically significant difference in the time of fracture healing and nonunion. Discussion Therefore, both procedures can be regarded as minimally invasive operations, and the iatrogenic soft tissue injury is small, so there is no significant difference in the statistical analysis of the range of motion at the final review of the knee joint.In the selection of internal fixation consumables, the steel plate group used extra fixed consumables, resulting in a lower average cost of surgical consumables between the two groups. The difference was statistically significant (P <0.05). The most serious complication in this study was anterior knee pain. The incidence of knee pain in the plate group was 26.3%, and this figure was 17.6% in the block group. 31.9% of the report, the author's analysis may be related to the study of patients with open sacral ligament. Djahangiri et al. [15] performed an 18-month follow-up of 96 cases of tibial shaft fractures and found that 52% of patients had pain in the patellofemoral ligaments ,14% of patients who do not have sacral ligaments have knee pain. The difference was statistically significant. The incidence of knee pain in the plate group was higher than that in the block group (P <0.05), which may be related to the additional anterior and lateral incisions of the knee joint. It has been reported in the literature in China that the occurrence of anterior knee pain is not related to the sacral ligament, and it is related to the internal fixation in the body for too long. In this study, five cases of anterior knee pain occurred in the plate group, four of which were improved after internal fixation, and three cases of anterior knee pain were occurred in the nail group, and improved after removal of the internal fixation. Toivanen et al. [16, 17] reported that the patellofemoral ligament did not reduce the incidence of chronic anterior knee pain; 67% to 71% of the anterior knee pain remained after removal of the intramedullary nail, and the difference after 8 years of follow-up was not statistically significant. However, in this study, there were cases of pain in the sacral ligaments, so the results of the study are biased Therefore the causes of knee pain need to be further explored Funding Shaanxi Provincial Social Development Scienceand Technology Research Project (2016SF-340) Availability of data and materials Shaanxi Provincial Social Development Scienceand Technology Research Project (2016SF-340) Availability of data and materials The detailed data and materials of this study were available from the corresponding author through emails on reasonable request. Acknowledgements he authors would like to thank all the staff of the participating departments Conclusions Page 8/16 In summary, for proximal tibial fractures, the treatment of fracture healing can be achieved by the intramedullary nail with small plate or block nail technique. Both techniques can achieve satisfactory Page 8/16 results in terms of therapeutic effect. However, the former requires additional surgical operation of the incision and internal fixation materials, while the latter requires special surgical operation techniques. Although the former has a lower overall operation time, intraoperative fluoroscopy times, and postoperative horn deformity, the average cost of surgical supplies for this technique is higher and the hospital stay is longer. The latter is exactly the opposite. Both groups patients had complications of anterior knee pain, and this was slightly higher in the plate group. Although this complication was improved after internal fixation and removal, it is worthy of the attention of clinicians. Ethics approval and consent to participate Our study had been verified and approved by the ethics committee of the Department of Orthopaedic Surgery, Hong Hui Hospital, Xi'an Jiao tong University and followed the Declaration of Helsinki. all the included patients had signed the written informed consents. All methods were performed in accordance with the relevant guidelines and regulations. Authors’ contributions J, Q performed the experimental work. Q, Z, and K evaluated the data. J wrote the manuscript. All authors read and approved the final manuscript. Consent for publication Not applicable Page 9/16 The authors declare that they have no competing interests. The authors declare that they have no competing interests. References 1. Court-Brown CM, McBirnie J. The epidemiology of tibial fractures. J Bone Joint Surg Br. 1995;77(3):417-421. 2. Milner SA, Davis TR, Muir KR, Greenwood DC, Doherty M. Long-term outcome after tibial shaft fracture: is malunion important?. J Bone Joint Surg Am. 2002;84(6):971-980. 3. Nork SE, Barei DP, Schildhauer TA, et al. Intramedullary nailing of proximal quarter tibial fractures. J Orthop Trauma. 2006;20(8):523-528. 4. Naik MA, Arora G, Tripathy SK, Sujir P, Rao SK. Clinical and radiological outcome of percutaneous plating in extra-articular proximal tibia fractures: a prospective study. Injury. 2013;44(8):1081-1086. 5. Freedman EL, Johnson EE. Radiographic analysis of tibial fracture malalignment following intramedullary nailing. Clin Orthop Relat Res. 1995;(315):25-33. 6. SPRINT Investigators, Bhandari M, Guyatt G, et al. Study to prospectively evaluate reamed intramedually nails in patients with tibial fractures (S.P.R.I.N.T.): study rationale and design. BMC Musculoskelet Disord. 2008;9:91. 7. Johner R, Wruhs O. Classification of tibial shaft fractures and correlation with results after rigid internal fixation. Clin Orthop Relat Res. 1983;(178):7-25. 8. Jeffcoach DR, Sams VG, Lawson CM, et al. Nonsteroidal anti-inflammatory drugs' impact on nonunion and infection rates in long-bone fractures. J Trauma Acute Care Surg. 2014;76(3):779-783. 9. Zelle BA. Intramedullary nailing of tibial shaft fractures in the semi-extended position using a suprapatellar portal technique. Int Orthop. 2017;41(9):1909-1914. 10. Dunbar RP, Nork SE, Barei DP, Mills WJ. Provisional plating of Type III open tibia fractures prior to intramedullary nailing. J Orthop Trauma. 2005;19(6):412-414. 11. Yoon RS, Gage MJ, Donegan DJ, Liporace FA. Intramedullary Nailing and Adjunct Permanent Plate Fixation in Complex Tibia Fractures. J Orthop Trauma. 2015;29(8):e277-e279. 12. Kubiak EN, Camuso MR, Barei DP, Nork SE. Operative treatment of ipsilateral noncontiguous unicondylar tibial plateau and shaft fractures: combining plates and nails. J Orthop Trauma. 2008;22(8):560-565. Page 10/16 13. Krettek C, Miclau T, Schandelmaier P, Stephan C, Möhlmann U, Tscherne H. The mechanical effect of blocking screws ("Poller screws") in stabilizing tibia fractures with short proximal or distal fragments Page 10/16 Page 10/16 after insertion of small-diameter intramedullary nails. J Orthop Trauma. 1999;13(8):550-553. 14. Court-Brown CM. Reamed intramedullary tibial nailing: an overview and analysis of 1106 cases. J Orthop Trauma. 2004;18(2):96-101. 15. Djahangiri A, Garofalo R, Chevalley F, et al. Closed and open grade I and II tibial shaft fractures treated by reamed intramedullary nailing. Med Princ Pract. 2006;15(4):293-298. 16. Väistö O, Toivanen J, Kannus P, Järvinen M. References Anterior knee pain after intramedullary nailing of fractures of the tibial shaft: an eight-year follow-up of a prospective, randomized study comparing two different nail-insertion techniques. J Trauma. 2008;64(6):1511-1516. 17. Toivanen JA, Väistö O, Kannus P, Latvala K, Honkonen SE, Järvinen MJ. Anterior knee pain after intramedullary nailing of fractures of the tibial shaft. A prospective, randomized study comparing two different nail-insertion techniques. J Bone Joint Surg Am. 2002;84(4):580-585. 17. Toivanen JA, Väistö O, Kannus P, Latvala K, Honkonen SE, Järvinen MJ. Anterior knee pain after intramedullary nailing of fractures of the tibial shaft. A prospective, randomized study comparing two different nail-insertion techniques. J Bone Joint Surg Am. 2002;84(4):580-585. Tables Table 1 Preoperative baseline data between the two groups Item Steel plate group(n=19) Blocking nail group(n=17) P- Value Age 34.10±15.30 36.90±12.80 0.558 Male/Female 12/7 9/8 0.736 Mechanism of injury Road traffic accident 13 15 0.236 Fall from height 6 2 AO classification AO-41A2 10 14 0.083 AO-41A3 9 3 Table 1 Preoperative baseline data between the two groups 3 Page 11/16 Table 2 Comparison of perioperative indicators between the two groups Item Steel plate group(n=19) Blocking nailing group(n=17) P- Value Hospital stay（d） 9.20±3.50 7.10±2.60 0.047 operation time（min） 82.90±20.40 125.80±35.60 <0.001 Intraoperative perspective（times） 28.60±9.30 110.30±24.80 <0.001 Incision length(cm) 7.34±4.12 4.52±2.39 0.024 Intraoperative blood loss (ml) 434.21±78.74 161.76±12.52 0.003 Consumables cost（RMB） 13498.30±983.60 8947.60±698.40 <0.001 Table 3 Follow-up results and comparison between the two groups Item Steel plate group(n=19) Blocking nailing group(n=17) P- Value Knee activity（°） 115.20±12.50 119.70±15.80 0.347 Walking time (weeks) 8.62±3.17 7.53±2.36 0.684 Full weight-bearing time(weeks) 14.52±2.31 13.94±3.65 0.235 Knee pain(cases） yes 5 3 0.695 no 14 14 Johner-Wruhs score(cases) excellent 14 14 0.809 good 3 2 ok 2 1 poor 0 0 Table 2 Comparison of perioperative indicators between the two groups Item Steel plate group(n=19) Blocking nailing group(n=17) P- Value Hospital stay（d） 9.20±3.50 7.10±2.60 0.047 operation time（min） 82.90±20.40 125.80±35.60 <0.001 Intraoperative perspective（times） 28.60±9.30 110.30±24.80 <0.001 Incision length(cm) 7.34±4.12 4.52±2.39 0.024 Intraoperative blood loss (ml) 434.21±78.74 161.76±12.52 0.003 Consumables cost（RMB） 13498.30±983.60 8947.60±698.40 <0.001 Table 2 Comparison of perioperative indicators between the two groups Table 3 Follow-up results and comparison between the two groups Item Steel plate group(n=19) Blocking nailing group(n=17) P- Value Knee activity（°） 115.20±12.50 119.70±15.80 0.347 Walking time (weeks) 8.62±3.17 7.53±2.36 0.684 Full weight-bearing time(weeks) 14.52±2.31 13.94±3.65 0.235 Knee pain(cases） yes 5 3 0.695 no 14 14 Johner-Wruhs score(cases) excellent 14 14 0.809 good 3 2 ok 2 1 poor 0 0 Table 3 Follow-up results and comparison between the two groups Page 12/16 Page 12/16 Page 12/16 Table 4 Comparison of postoperative residual fracture displacement and imaging bone healing time between the two groups Item Steel plate group(n=19) Blocking nail group(n=17) P- Value Coronal angle (°) 2.20±0.60 2.90±0.90 0.009 Sagittal angulation（°） 1.30±0.30 2.70±0.40 <0.001 Lateral displacement（mm） 1.21±0.17 1.94±0.09 0.012 Anterior and posterior displacement (mm) 2.30±0.11 3.23±0.45 <0.001 Bone healing time (weeks) 16.20±3.90 14.80±3.10 0.245 Table 4 Comparison of postoperative residual fracture displacement and imaging bone healing time between the two groups Table 4 Comparison of postoperative residual fracture displacement and imaging bone healing time between the two groups Figures Page 13/16 Page 14/16 igure 1 A typical case from the steel plate group A 46-year-old male patient was treated fo AO41-A3) following a traffic injury on the fifth day after the injury. The traditional x the intramedullary nail with small plate technique. 1&2:Preoperative and lateral roximal tibial fracture and displacement was obvious; 3:Immediately after the op &5:On the second day after surgery, the X-ray films showed that the tibial line was Figure 1 A typical case from the steel plate group A 46-year-old male patient was treated for a left humeral fracture (AO41-A3) following a traffic injury on the fifth day after the injury. The traditional approach was used to fix the intramedullary nail with small plate technique. 1&2:Preoperative and lateral X-ray films showed proximal tibial fracture and displacement was obvious; 3:Immediately after the operation was completed; 4&5:On the second day after surgery, the X-ray films showed that the tibial line was restored and the Page 14/16 Page 14/16 ternal fixation was satisfactory; 6&7: After 3 months, the fracture is almost healed; 8–11:One ye rgery, the fracture was completely healed and appearance and function were restored. gure 2 internal fixation was satisfactory; 6&7: After 3 months, the fracture is almost healed; 8–11:One year after surgery, the fracture was completely healed and appearance and function were restored. internal fixation was satisfactory; 6&7: After 3 months, the fracture is almost healed; 8–11:One year after surgery, the fracture was completely healed and appearance and function were restored. rgery, the fracture was completely healed and appearance and function were re g y, p Page 15/16 Figure 2 A typical case from the blocking nail group A 53-year-old female patient was treated for a proximal racture of the left tibia (AO41-A2) following a traffic injury. The patient was treated with the nail echnique. 1&2: Preoperative and lateral X-ray films showed tibial fracture with obvious displacement; 3 Figure 2 A typical case from the blocking nail group A 53-year-old female patient was treated for a proximal fracture of the left tibia (AO41-A2) following a traffic injury. The patient was treated with the nail technique. 1&2: Preoperative and lateral X-ray films showed tibial fracture with obvious displacement; 3: Page 15/16 The finger resetter was inserted into the rear side image to show the forward angle; 4: By inserting the blocking nail on the back of the proximal end, the forward angle is eliminated; 5: The front and rear images show that there is still obvious shift inside and outside; 6: Transcutaneous clamp reset is invalid; 7&8: The double plane blocking nail improved the angular displacement and restored the tibial line; 9: Appearance after surgery; 10 & 11: On the second day after surgery, the X-ray films of the anterior and lateral position show that the tibial line was restored and the internal fixation was satisfactory; 12–15: Postoperative fracture healing and functional appearance in June. Page 16/16
https://openalex.org/W4302399515	https://orbi.uliege.be/bitstream/2268/212285/1/ncomms15264.pdf	English	null	Cis-perturbation of cancer drivers by the HTLV-1/BLV proviruses is an early determinant of leukemogenesis.	HAL (Le Centre pour la Communication Scientifique Directe)	2,017	cc-by	18,575	ARTICLE Received 4 Dec 2016 \| Accepted 14 Mar 2017 \| Published 23 May 2017 1 Unit of Animal Genomics, GIGA-R, Universite´ de Lie`ge (ULg), Avenue de l’Hoˆpital 11, B34, Lie`ge 4000, Belgium. 2 Laboratory of Experimental Hematology, Institut Jules Bordet, Universite´ Libre de Bruxelles (ULB), Boulevard de Waterloo 121, Brussels 1000, Belgium. 3 Service d’he´matologie, Hoˆpital Universitaire Necker, Universite´ Rene´ Descartes, Assistance publique hoˆpitaux de Paris, 149-161 rue de Se`vres, Paris 75010, France. 4 Vaccine and Infectious Disease Organization, VIDO-Intervac, University of Saskatchewan, 120 Veterinary Road, Saskatoon, Canada S7N 5E3. 5 Laboratoire de Virologie, AP-HP, Hoˆpital Necker-Enfants Malades, Universite´ Paris Descartes, Sorbonne Paris Cite´, EA7327, 149 rue de Se`vres, Paris 75010, France. 6 INSERM U1163-ERL8254, Institut Imagine, 24 B Boulevard du Montparnasse, Paris 75010, France. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed to A.Vd.B. (email: anne.vandenbroeke@bordet.be). ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 A k A n estimated 10–20 million people are infected with the Human T-cell leukaemia virus type-1 (HTLV-1) worldwide1. In B5% of infected individuals the virus provokes adult T-cell leukaemia/lymphoma (ATL), an aggressive T-cell malignancy with a poor prognosis2. Bovine leukaemia virus (BLV) is closely related to HTLV-1 and causes a very similar B-cell leukaemia in cattle and sheep3,4. The virus infects B50 million dairy cattle worldwide inducing substantial economic costs in infected herds5. Like HTLV-1 in humans, following a long period of asymptomatic infection (several years in cattle, several decades in humans) B5% of BLV-infected animals develop leukaemia/lymphoma. In addition to infecting bovines, it is possible to experimentally infect sheep with BLV, providing a powerful model for studying deltaretrovirus-induced tumours6,7. In contrast to cattle, infected sheep systematically develop leukaemia/lymphoma and in a shorter time frame (B20 months). A further advantage of this model is that it is possible to monitor animals from before infection to terminal leukaemia/lymphoma, recapitulating many of the stages observed in HTLV-1-associated human malignancy. unknown. Like HTLV-1, BLV expresses antisense transcripts, AS1 and AS2, driven by 30LTR-dependent promoter activity and constitutively produced in leukaemic cells24. In addition, BLV strongly expresses RNA polymerase III-dependent microRNAs overlapping AS1 and contributing to B40% of microRNAs in the tumour cell25. In the fraction of infected individuals who do progress, many years separate the initial infection from the development of leukaemia/lymphoma. This indicates that infection with BLV/HTLV-1 is not sufﬁcient to provoke tumour development and that secondary events are required to make the transition to a neoplasm. A recent study examined the landscape of mutations in ATLs and found frequent alterations enriched in T-cell related pathways and immunosurveillance11. As regards BLV-induced tumours, beyond limited studies that reported frequent genome instability and mutation of p53 (refs 26,27), the occurrence of secondary events in BLV malignancies remains largely unexplored. As BLV and HTLV-1 vary little in sequence both within and between hosts, and as wide variations exist in clone abundance between infected individuals and over time, it is hypothesized that the proviral integration site is the principal attribute that distinguishes one infected cell clone from the other, thus is a key element on the road to malignancy. ARTICLE Using a quantitative high-throughput sequencing (HTS) approach to characterize the genomic environment of the provirus, previous studies have addressed the role of the genomic integration site in determining clonal expansion and the potential for malignant transformation of cells carrying integrated HTLV-1 or BLV8–10,28,29. Although HTLV-1 and BLV preferentially integrate in transcriptionally active genomic regions, near transcriptional start sites and transcription factor-binding sites, there was no reported evidence of recurrent proviral integration. In tumours, there were no clear hotspots of HTLV-1/BLV integration associated with leukaemic clones, although the ontology of the nearest downstream gene was associated with malignant clones in 6% of the ATL cases10,12. Despite a long history of study in both the HTLV-1 and BLV models, key steps on the road from initial infection to leukaemia development remain poorly elucidated. In chronic stages of infection, HTLV-1 and BLV propagate primarily through clonal expansion of infected T- or B-cells, respectively, resulting in the presence of multiple clones of varying abundance each uniquely identiﬁed by their proviral integration site in the host genome. Following a protracted incubation period, one of these clones expands, leading to the accumulation of malignant cells in the peripheral blood (leukaemia) and/or diverse tissues (lymphoma)4,8–10. Tumour cells consist of a predominant malignant T- or B-cell clone and chieﬂy harbour a single integrated provirus, yet integration sites are very variable10–12. As a consequence, it has been widely believed that virus-encoded products drive clonal proliferation and inﬂuence oncogenic progression. Historically the focus of research in BLV/HTLV-1 has been on the oncogenic potential of the viral TAX protein. TAX can immortalize rodent cells in vitro and induces tumours in transgenic mice, supporting the hypothesis that it is an essential contributor to oncogenesis13,14. TAX activates transcription of the provirus and of many host genes, promotes cell-cycle progression and interacts with DNA repair mechanisms15,16. However, the lack of TAX (and other viral sense transcript) expression in the majority of BLV/HTLV-1- induced malignancies and the high frequency of proviruses containing alterations inactivating TAX points to a more complicated picture17–20. While TAX expression provides a proliferative advantage to the infected clone, it also makes this clone a target for cytotoxic immune response21. Therefore, it may be advantageous for the virus to evade the strong immune response to TAX by silencing expression from the positive strand. ARTICLE It is thus widely accepted that TAX fulﬁls an essential role at early stages of the oncogenic process, yet is not required for late-stage precipitation to monoclonal malignancy. Despite the variability of integration sites in fully transformed cells, a role for proviral integration and cis-perturbation of host genes in HTLV-1/BLV-induced clonal expansion cannot be excluded. Here, to explore this hypothesis, we carried out RNA-seq of primary tumours in both the human disease and the animal model in combination with HTS mapping of proviral integration sites. We show that HTLV-1/BLV proviruses are integrated in the vicinity of cancer drivers, which they perturb either by provirus-dependent transcription termination or as a result of viral antisense RNA-dependent cis-perturbation. The same pattern is observed at asymptomatic stages of the disease, indicating that provirus-dependent host gene perturba- tion triggers initial ampliﬁcation of the corresponding clones, requiring additional alterations to develop full-blown leukaemia/lymphoma. p p g y Over the last years it has become increasingly apparent that another viral product, HTLV-1 basic leucine zipper (bZIP) factor (HBZ) encoded from the minus-strand, plays an important role in the life cycle and oncogenic potential of the virus. HBZ downregulates HTLV-1 transcription, promotes T-cell prolifera- tion and displays oncogenic properties in transgenic mice, suggesting a critical role in HTLV-1-mediated leukemogen- esis16,22. In contrast to TAX, HBZ is consistently expressed in infected cells in vivo, regardless of their transformation status. HBZ appears to exert its effects through both the transcript and the protein23; however, the precise mechanisms by which HBZ contributes to the oncogenic process remain largely Cis-perturbation of cancer drivers by the HTLV-1/BLV proviruses is an early determinant of leukemogenesis Nicolas Rosewick1,2,, Keith Durkin1,, Maria Artesi1,, Ambroise Marc¸ais3, Vincent Hahaut1, Philip Griebel4, Natasa Arsic4, Ve´ronique Avettand-Fenoel5, Arse`ne Burny2, Carole Charlier1, Olivier Hermine3,6, Michel Georges1 & Anne Van den Broeke1,2 Human T-cell leukaemia virus type-1 (HTLV-1) and bovine leukaemia virus (BLV) infect T- and B-lymphocytes, respectively, provoking a polyclonal expansion that will evolve into an aggressive monoclonal leukaemia in B5% of individuals following a protracted latency period. It is generally assumed that early oncogenic changes are largely dependent on virus- encoded products, especially TAX and HBZ, while progression to acute leukaemia/lymphoma involves somatic mutations, yet that both are independent of proviral integration site that has been found to be very variable between tumours. Here, we show that HTLV-1/BLV proviruses are integrated near cancer drivers which they affect either by provirus-dependent transcription termination or as a result of viral antisense RNA-dependent cis-perturbation. The same pattern is observed at polyclonal non-malignant stages, indicating that provirus- dependent host gene perturbation contributes to the initial selection of the multiple clones characterizing the asymptomatic stage, requiring additional alterations in the clone that will evolve into full-blown leukaemia/lymphoma. 1 Unit of Animal Genomics, GIGA-R, Universite´ de Lie`ge (ULg), Avenue de l’Hoˆpital 11, B34, Lie`ge 4000, Belgium. 2 Laboratory of Experimental Hematology, Institut Jules Bordet, Universite´ Libre de Bruxelles (ULB), Boulevard de Waterloo 121, Brussels 1000, Belgium. 3 Service d’he´matologie, Hoˆpital Universitaire Necker, Universite´ Rene´ Descartes, Assistance publique hoˆpitaux de Paris, 149-161 rue de Se`vres, Paris 75010, France. 4 Vaccine and Infectious Disease Organization, VIDO-Intervac, University of Saskatchewan, 120 Veterinary Road, Saskatoon, Canada S7N 5E3. 5 Laboratoire de Virologie, AP-HP, Hoˆpital Necker-Enfants Malades, Universite´ Paris Descartes, Sorbonne Paris Cite´, EA7327, 149 rue de Se`vres, Paris 75010, France. 6 INSERM U1163-ERL8254, Institut Imagine, 24 B Boulevard du Montparnasse, Paris 75010, France. These authors contributed equally to this work. Correspondence and requests for materials should be addressed to A.Vd.B. (email: anne.vandenbroeke@bordet.be). 1 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications TURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| w NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications Results b Of the examined tumour samples, 87.8% were characterized by a single predominant malignant clone deﬁned by a single integration site and clonal abundance ranging from 79.4 to 99.9%. For the remaining tumour clones B6.75% showed evidence of two integrations, while B4.05% carried three or four integration sites, with each integration displaying equivalent abundance characteristic of multiple proviruses in a single clone rather than the co-occurrence of multiple tumour clones. Finally, one tumour showed evidence of seven integrations (Supplementary Table 1 and Supplementary Data 2). These observations are consistent with previous ﬁndings of multiple proviruses in 9–15% of ATL cases10,11,32. The integration sites were deﬁned 86% of the time by both 50 long-terminal-repeat (LTR)-ﬂanking and 30LTR-ﬂanking host sequences and 14% of the time by 30LTR-ﬂanking sequences only (29% of ATLs), in agreement with earlier reports describing 50LTR-defective (but never 30LTR- defective) proviruses in both ATLs and B-cell tumours11,17,32. Provirus-dependent transcription interruption of host genes. In 27 of the analysed tumours (11 in human and 16 in ruminants), the provirus was integrated in the intron of a gene with same transcriptional orientation (assuming dominant 50LTR to 30LTR viral transcription). We refer to this group as ‘genic insertion, concordant’. In 21 of these, the RNA-seq data revealed premature termination of transcription and polyadenylation of the interrupted host gene at the viral poly-(A) signal (441 and 603 bp within the BLV and HTLV-1 50LTR, respectively). This was accompanied by severe reduction of downstream exon reads (P ¼ 5.9e-08, Mann–Whitney U-test), strongly suggesting cis-allele truncation of the affected host genes (Fig. 1). Expression of downstream exons was halved on average, suggesting that provirus- dependent transcriptional termination operates in all infected cells. In two of these tumours read numbers of downstream exons were even signiﬁcantly o50% of controls, supporting additional perturbation of the trans-allele, as expected for tumour suppressor genes (tumours T1345 and M2532, 88% and 96% downregulation of MSH2 (ref. 33) and STARD7 (ref. 34), respectively; Fig. 1c). The list of truncated genes includes established anti-oncogenes such as MSH2 and BRCC3 (refs 33,35). Preferential integration of HTLV-1/BLV near cancer drivers. A naı¨ve examination of the proviral integration sites revealed a number of striking instances of repeated integration into the same genomic vicinity. Three B-cell leukaemias had BLV inte- gration sites falling within an B80 kb region upstream of FOXR2 (M251, M138 and M21, OAR3.1 chrX: 47,545,987, 47,600,711 and 47,618,373, respectively). Results b We obtained samples from 44 adult T-cell leukaemias/lymphoma (ATLs) (from 35 patients, HTLV-1, human disease) and 47 B-cell leukaemias (from 43 individuals, BLV, animal model). The animal sample set included 15 bovine tumours (natural disease) and 32 tumours from BLV-infected sheep, a well-established experi- mental model for BLV/HTLV-1 (refs 6,7) (Supplementary Data 1). We utilized stranded RNA-seq data (91 tumours) in combination with an improved version of DNA-seq based high-throughput mapping of integration sites (56 tumours) to simultaneously proﬁle proviral integrations, measure clonal abundance and identify virus– host transcriptional interactions in tumours9,24,29–31. We identiﬁed NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 2 ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Table 1 \| HTLV-1/BLV interacting host genes are enriched in cancer driver genes. Gene number Enrichment (P-values) Random* Random Paraw Exprz Expr Paraw Tumour samplesy Genic proviruses 41 0.0029 0.0016 0.0219 0.0111 All proviruses excluding genic concordant poly-(A) 65 0.0004 o1e-05 0.0006 o1e-05 Intergenic proviruses excluding exon capture 30 0.0015 0.0012 0.0006 0.0004 All proviruses 82 o1e-05 o1e-05 0.0004 o1e-05 Asymptomatic samplesy 723 o1e-05 o1e-05 o1e-05 0.0002 Random simulated gene sets. wSimulated gene sets that include information about paralogs. zExpression-matched simulated gene sets. yGene subsets as deﬁned in Supplementary Data 3 and Supplementary Table 4. Table 1 \| HTLV-1/BLV interacting host genes are enriched in cancer driver genes. integration at speciﬁc sites promotes tumorigenesis, i.e., selection. To distinguish between these hypotheses, we veriﬁed whether the 54 genic (versus 38 intergenic) proviral integration sites identiﬁed across tumours were enriched in known cancer drivers using seven publicly available cancer driver lists (Supplementary Fig. 2). Indeed, under the selection model, genes interrupted by the provirus are expected to be enriched in cancer drivers (the affected gene, if any, is a priori more difﬁcult to pinpoint for intergenic insertions). The enrichment was signiﬁcant compared to random sets of genes matched for expression level in lymphocytes (0.0016oPo0.0219), strongly suggesting that—in contrast to the prevailing view—HTLV-1 and BLV integration sites are overrepresented in the vicinity of genes connected to cancer, and that the resulting perturbation is an essential driver of tumour formation (Table 1, Supplementary Data 3 and Supplementary Table 2). 92 distinct proviral integration sites considering all available tumours (54 genic and 38 intergenic, Supplementary Table 1). NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications Results b (a) Normalized RNA-seq read counts of upstream (left) and downstream (right) exons relative to the proviral integration site in the tumour set characterized by genic-concordant proviruses (red plot, N ¼ 21) and control tumours without integration in that gene (blue plot). *P ¼ 5.878e-08 (Mann–Whitney U-test). (b) Transcription patterns of human leukaemia ATL66 shown as RNA-seq sense (red) and antisense (blue) coverage mapped to the proviral (top) or host (bottom) genomes visualized in Integrative Genomic Viewer (IGV)59. Top panel: HTLV-1 proviral genome ﬂanked by 50LTR/30LTR redundant regions (U3, R, U5) that contain regulatory elements, transcriptional start sites (TSS) and poly-(A) signal. Positive-strand transcripts (red) encode structural and regulatory (TAX/REX) proteins; spliced HBZ antisense transcripts (blue) expressed from negative-strand. ATL66 RNA-seq coverage of HTLV-1: HBZ antisense transcripts and upstream coverage exposing hybrid transcripts; positive coverage of 50LTR reveals read-through transcription and provirus- dependent premature polyadenylation of host gene OSBP. Absence of 50LTR-driven viral transcription. Bottom panel: mapping to host genome (hg19). Small box: HTLV-1 integration in OSBP introns 9–10 (opposite orientation). OSBP exons 10–14 show decreased coverage (ATL66/control ATLs (N ¼ 39): 52% decrease, ATL66 OSBP downstream/upstream exons, fold-change ¼ 0.52). Sense coverage: 30LTR-dependent chimeric transcript in antisense overlap with OSBP. (c) Transcription patterns of bovine T1345/ovine M2532 B-cell tumours shown as RNA-seq sense (red) and antisense (blue) coverage mapped to the proviral (top) or host (bottom) genomes. Top: BLV genome, annotation and T1345 RNA-seq coverage representative of both tumours: AS1 antisense transcription; positive coverage of 50LTR reveals host gene transcription (read-through) and provirus-dependent premature polyadenylation. Bottom panels: mapping to host genomes (UMD3.1 and OAR3.1). Small box: BLV integration in MSH2 (ref. 33) intron 6 or STARD7 (ref. 34) intron 5. Decrease of MSH2 and STARD7 downstream exon coverage (MSH2 T1345/control tumours (N ¼ 14), 88% decrease; T1345 MSH2 downstream/upstream exons, fold-change ¼ 0.12 and STARD7: 96% decrease (control tumours, N ¼ 31), downstream/upstream exon fold-change ¼ 0.08). Antisense coverage: 30AS-dependent chimeric transcript in antisense overlap with MSH2/STARD7. See also Supplementary Fig. 3 for RNA-seq coverage assignment to 50LTR/30LTR. Host genome Figure 1 \| Provirus-dependent host gene interruption in HTLV-1/BLV primary tumours with genic-concordant proviral integration. (a) Normalized RNA-seq read counts of upstream (left) and downstream (right) exons relative to the proviral integration site in the tumour set characterized by genic-concordant proviruses (red plot, N ¼ 21) and control tumours without integration in that gene (blue plot). Results b Mapping the RNA-seq reads to the proviral and host reference genomes revealed in all tumours the complete absence of The observed non-random distribution can either indicate that some sites are more prone to proviral integration, possibly reﬂecting speciﬁc chromatin features28, or/and reﬂect the fact that NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 2 ** 5′LTR HTLV-1 9023 bp U3 R U5 3′LTR HBZ usHBZ Provirus integrated in host genome minus strand HTLV-1 IS:hg19-chr11:59,349,193 chr11:59,385,000 Host - hg19 45 kb OSBP Viral poly-A signal Viral poly-A signal 3′LTR 5′LTR 5′ poly-A signal 5′LTR T1345 M2532 chr3:104,053,000 [0–20] [0–10] [0–500] [0–500] M2531 Ovine control tumour B3171 Bovine control tumour chr11:29,648,305 BLV Host - bosTau6/UMD3.1 85 kb Host - OAR3.1 12 kb MSH2 STARD7 chr11.29,730,537 chr3:104,066,000 * * * * * * * * * * * * 8720 bp BLV IS: bosTau6/UMD3.1 chr11:29,666,537 5′LTR 3′LTR Viral poly-A signal BLV 5′LTR 3′LTR BLV miRNAs 3′LTR AS1 AS2 U3 R U5 U3 R U5 GAG-POL ENV TAX/REX R3 G4 5′ poly-A signal Provirus genome ATL66 ATL30 human control tumour chr11:59,340,000 [0–70] [0–60] [0–10] [0–10] [0–80] [0–30] [0–10] [0–180] [0–60] [0–60] * Host genome Provirus genome Host genome Host genome * * * * U3 R U5 GAG-PRO-POL ENV P13 P21 P30 TAX/REX P12.1 P12.2 [0–90] [0–40] 1 0 –1 log2(fold change) –2 –3 Upstream exons Downstream exons a b c 2 * 1 0 –1 log2(fold change) –2 –3 Upstream exons Downstream exons a b * 9023 bp U3 R U5 HBZ usHBZ Provirus integrated in host genome minus strand HTLV-1 IS:hg19-chr11:59,349,193 chr11:59,385,000 Host - hg19 45 kb OSBP Viral poly-A signal Viral poly-A signal 3′LTR 5′LTR 5′LTR T1345 M2532 chr3:104,053,000 [0–20] [0–10] [0–500] [0–500] M2531 Ovine control tumour B3171 Bovine control tumour chr11:29,648,305 BLV Host - bosTau6/UMD3.1 85 kb Host - OAR3.1 12 kb MSH2 STARD7 chr11.29,730,537 chr3:104,066,000 * * * * * * * * * * * * 8720 bp BLV IS: bosTau6/UMD3.1 chr11:29,666,537 5′LTR 3′LTR Viral poly-A signal BLV 5′LTR 3′LTR BLV miRNAs 3′LTR AS1 AS2 U3 R U5 U3 R U5 GAG-POL ENV TAX/REX R3 G4 5′ poly-A signal Provirus genome ATL66 ATL30 human control tumour chr11:59,340,000 [0–70] [0–60] [0–10] [0–10] [0–80] [0–30] [0–10] [0–180] [0–60] [0–60] * Host genome Provirus genome Host genome Host genome * * * * U3 R U5 GAG-PRO-POL ENV P13 P21 P30 TAX/REX P12.1 P12.2 [0–90] [0–40] 1 0 –1 log2(fold change) –2 –3 Upstream exons Downstream exons c Figure 1 \| Provirus-dependent host gene interruption in HTLV-1/BLV primary tumours with genic-concordant proviral integration. Results b Additionally, two pairs of ATLs had integration sites in the same genomic region (ATL1_ly/ATL2_7 within 300 kb of RBFOX1, ATL1_ly/ATL58_23 within 600 kb of CA10). Finally, a bovine B-cell/ATL tumour pair had integration sites 17 and 13 kb from the same host gene, respectively (T15, ATL4_2, TMEM67), while an ovine B-cell/ATL tumour pair had their proviruses integrated within and 430 kb down- stream of the same host gene, respectively (M395/ATL25_10, ELF2). These observations contrast with previous studies that did show preferential integration in transcriptionally active regions, but failed to observe hotspots of HTLV-1 integration in ATLs9–11,28. We showed by simulation that the observed degree of recurrent integration into narrow genomic regions signiﬁcantly exceeded expectation assuming random proviral integration (P ¼ 0.0024; Supplementary Fig. 1). Consistent with this conclusion, one of the ovine B-cell tumours showed evidence of BLV integration into UBASH3B, which was identiﬁed as the target gene of HTLV-1 integration in one of the ATLs analysed by WGS in the recent study of Kataoka et al.11. Viral antisense RNA-dependent cis-perturbation of host genes. Demonstrating 50LTR-dependent transcriptional termination of the interrupted gene in 23% (21/92) of the tumours leaves the question open of what alternative molecular mechanisms perturb the presumed cancer drivers in the remaining 77% of tumours. To gain some insights into what these mechanisms might be, we carefully mined the RNA-seq data from the 71 (i.e., 92–21) remaining tumours. NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 (a) Schematics of RNA-seq coverage mapped to HTLV-1/BLV in ATLs/B-cell tumours and antisense-predominant viral transcription. Top: proviral genome and simpliﬁed HTLV-1/BLV common annotation: positive-strand transcripts (red), spliced HBZ/AS antisense transcripts (blue). (i–iv): RNA-seq coverage. In all tumours, (i) absence of 50LTR-dependent positive-strand coverage (structural proteins and TAX, Supplementary Fig. 3), (ii) 30LTR-dependent HBZ/AS antisense transcripts (dark-blue) and upstream coverage (light-blue), (iii) hybrid antisense reads that span HBZ/AS exon 1-host and 50LTR-host boundaries, supporting 30LTR-dependent chimeric transcripts. In 21 proviruses (genic-concordant), (iv) positive coverage of host-50LTR-U3/R boundary (viral poly-(A)- dependent host transcript truncation). See also Supplementary Fig. 3: coverage for 24 representative tumours and Supplementary Fig. 4: secondary types of virus-host transcriptional interactions. (b) Four main patterns of viral antisense RNA-dependent transcriptional interactions with the tumour genome: upper left: genic integration, concordant gene-provirus transcriptional orientation. Viral 50LTR poly-(A)-dependent gene interruption, downstream exon decreased expression (blue boxes, see also Fig. 1) and 30AS-dependent hybrid transcript in antisense overlap with upstream sequences. Splicing to host cryptic SA (10/21 proviruses). Upper right: genic integration, discordant gene-provirus transcriptional orientation. 30AS-dependent virus–host hybrid transcript and HBZ/AS exon 1 SD sequestration by upstream host exon(s) (i.e., tumour M160/ICA1; Fig. 3a). Capture of viral HBZ/AS exon 2 by downstream host gene exon (10/27 proviruses). Bottom left: intergenic integration, concordant gene-provirus transcriptional orientation. 30AS-dependent virus–host chimeric transcript in antisense overlap with host gene (i.e., ATL1_Ly/RGCC; Supplementary Fig. 5a). Bottom right: intergenic integration, discordant gene–provirus transcriptional orientation. 30AS-dependent chimeric transcript in sense overlap with upstream host gene(s) (i.e., tumours M138, M251 and M21/FOXR2 and RRAGB; Fig. 3b,c), and capture of exon 2 or novel exon that creates non-canonical isoforms (i.e., tumour LB120/SEPT11; Fig. 3d). Six proviruses were integrated in gene deserts (Supplementary Table 1 and Supplementary Data 2). See Supplementary Figs 4 and 6: comprehensive characterization of dominant and secondary types of virus-host interactions in tumours. RNA-seq splice junctions/breakpoints were validated by RT–PCR for representative tumours of each group (Supplementary Fig. 7). and 32 non-coding genes (Supplementary Data 2 and 4 and Supplementary Table 4). The distance between the perturbed host gene and the interacting provirus averaged 172 kb, ranging from zero in the cases of genic insertions to 1,300 kb in a case of intergenic insertion. Interactions were observed with a single (45 integrations) or multiple host genes per provirus (15 with 2, 3 with 3, and 2 with 4 genes). NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Host genome U3 (i) (ii) All tumours 21 proviruses (iii) SA (iv) RNA-Seq sense coverage RNA-Seq antisense coverage R U5 Host genome U3 R U5 HTLV-1 HBZ BLV AS SD Genomic 5′LTR 3′LTR HTLV-1 / BLV provirus Concordant gene-provirus orientation Discordant gene-provirus orientation Host gene Host gene Host gene Host gene HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV 5’ viral poly-A signal SA SA SA Provirus-mapping Virus-host hybrid transcript Captured exons Host-mapping Virus to host splicing Host to host splicing SA SD SA SA SA SD SD SD SD SD SD Genic integration Intergenic integration 5′ 5′ 5′ 3′ 3′ 3′ 3′ N=27/92 N=27/92 N=24/92 Novel exon N=8/92 * 5′ 5′LTR 5′LTR 5′LTR 5′LTR 3′LTR 3′LTR 3′LTR 3′LTR 5′ poly-A signal env tax/rex Splice acceptor Splice donor a b Figure 2 \| Viral antisense RNA-dependent transcriptional interactions with the host genome in HTLV-1/BLV primary tumours. (a) Schematics of RNA-seq coverage mapped to HTLV-1/BLV in ATLs/B-cell tumours and antisense-predominant viral transcription. Top: proviral genome and simpliﬁed HTLV-1/BLV common annotation: positive-strand transcripts (red), spliced HBZ/AS antisense transcripts (blue). (i–iv): RNA-seq coverage. In all tumours (i) absence of 50LTR-dependent positive-strand coverage (structural proteins and TAX, Supplementary Fig. 3), (ii) 30LTR-dependent HBZ/AS antisense transcripts (dark-blue) and upstream coverage (light-blue), (iii) hybrid antisense reads that span HBZ/AS exon 1-host and 50LTR-host boundaries, supporting 30LTR-dependent chimeric transcripts. In 21 proviruses (genic-concordant), (iv) positive coverage of host-50LTR-U3/R boundary (viral poly-(A) dependent host transcript truncation). See also Supplementary Fig. 3: coverage for 24 representative tumours and Supplementary Fig. 4: secondary types o virus-host transcriptional interactions. (b) Four main patterns of viral antisense RNA-dependent transcriptional interactions with the tumour genome: uppe left: genic integration, concordant gene-provirus transcriptional orientation. Viral 50LTR poly-(A)-dependent gene interruption, downstream exon decreased expression (blue boxes, see also Fig. 1) and 30AS-dependent hybrid transcript in antisense overlap with upstream sequences. Splicing to host cryptic SA (10/21 proviruses). Upper right: genic integration, discordant gene-provirus transcriptional orientation. 30AS-dependent virus–host hybrid transcript and HBZ/AS exon 1 SD sequestration by upstream host exon(s) (i.e., tumour M160/ICA1; Fig. 3a). Capture of viral HBZ/AS exon 2 by downstream host gene exon (10/27 proviruses). Bottom left: intergenic integration, concordant gene-provirus transcriptional orientation. 30AS-dependent virus–host chimeric transcript in antisense overlap with host gene (i.e., ATL1_Ly/RGCC; Supplementary Fig. 5a). Bottom right: intergenic integration, discordant gene–proviru transcriptional orientation. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 30AS-dependent chimeric transcript in sense overlap with upstream host gene(s) (i.e., tumours M138, M251 and M21/FOXR2 and RRAGB; Fig. 3b,c), and capture of exon 2 or novel exon that creates non-canonical isoforms (i.e., tumour LB120/SEPT11; Fig. 3d). Six proviruses were integrated in gene deserts (Supplementary Table 1 and Supplementary Data 2). See Supplementary Figs 4 and 6: comprehensive characterization of dominant and secondary types of virus-host interactions in tumours. RNA-seq splice junctions/breakpoints were validated by RT–PCR for representative tumours of each group (Supplementary Fig 7) NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 ARTICLE Host genome U3 (i) (ii) All tumours 21 proviruses (iii) SA (iv) RNA-Seq sense coverage RNA-Seq antisense coverage R U5 Host genome U3 R U5 HTLV-1 HBZ BLV AS SD Genomic 5′LTR 3′LTR HTLV-1 / BLV provirus 5′ poly-A signal env tax/rex a a (i) (ii) All tumours 21 proviruses (iii) SA (iv) RNA-Seq sense coverage RNA-Seq antisense coverage HTLV-1 HBZ BLV AS SD Concordant gene-provirus orientation Discordant gene-provirus orientation Host gene Host gene Host gene Host gene HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV 5’ viral poly-A signal SA SA SA Provirus-mapping Virus-host hybrid transcript Captured exons Host-mapping Virus to host splicing Host to host splicing SA SD SA SA SA SD SD SD SD SD SD Genic integration Intergenic integration 5′ 5′ 5′ 3′ 3′ 3′ 3′ N=27/92 N=27/92 N=24/92 Novel exon N=8/92 * 5′ 5′LTR 5′LTR 5′LTR 5′LTR 3′LTR 3′LTR 3′LTR 3′LTR tax/rex Splice acceptor Splice donor b p Concordant gene-provirus orientation Discordant gene-provirus orientation Host gene Host gene Host gene Host gene HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV HTLV-1 / BLV 5’ viral poly-A signal SA SA SA Provirus-mapping Virus-host hybrid transcript Captured exons Host-mapping Virus to host splicing Host to host splicing SA SD SA SA SA SD SD SD SD SD SD Genic integration Intergenic integration 5′ 5′ 5′ 3′ 3′ 3′ 3′ N=27/92 N=27/92 N=24/92 Novel exon N=8/92 * * 5′ 5′LTR 5′LTR 5′LTR 5′LTR 3′LTR 3′LTR 3′LTR 3′LTR Splice acceptor Splice donor b b Discordant gene-provirus orientation Discordant gene-provirus orientation pendent transcriptional interactions with the host genome in HTLV-1/BLV primary tumours. (a) Schematics of ure 2 \| Viral antisense RNA-dependent transcriptional interactions with the host genome in HTLV-1/BLV primary t Figure 2 \| Viral antisense RNA-dependent transcriptional interactions with the host genome in HTLV-1/BLV primary tumours. Results b **P ¼ 5.878e-08 (Mann–Whitney U-test). (b) Transcription patterns of human leukaemia ATL66 shown as RNA-seq sense (red) and antisense (blue) coverage mapped to the proviral (top) or host (bottom) genomes visualized in Integrative Genomic Viewer (IGV)59. Top panel: HTLV-1 proviral genome ﬂanked by 50LTR/30LTR redundant regions (U3, R, U5) that contain regulatory elements, transcriptional start sites (TSS) and poly-(A) signal. Positive-strand transcripts (red) encode structural and regulatory (TAX/REX) proteins; spliced HBZ antisense transcripts (blue) expressed from negative-strand. ATL66 RNA-seq coverage of HTLV-1: HBZ antisense transcripts and upstream coverage exposing hybrid transcripts; positive coverage of 50LTR reveals read-through transcription and provirus- dependent premature polyadenylation of host gene OSBP. Absence of 50LTR-driven viral transcription. Bottom panel: mapping to host genome (hg19). Smal box: HTLV-1 integration in OSBP introns 9–10 (opposite orientation). OSBP exons 10–14 show decreased coverage (ATL66/control ATLs (N ¼ 39): 52% decrease, ATL66 OSBP downstream/upstream exons, fold-change ¼ 0.52). Sense coverage: 30LTR-dependent chimeric transcript in antisense overlap with OSBP. (c) Transcription patterns of bovine T1345/ovine M2532 B-cell tumours shown as RNA-seq sense (red) and antisense (blue) coverage mapped to the proviral (top) or host (bottom) genomes. Top: BLV genome, annotation and T1345 RNA-seq coverage representative of both tumours: AS1 antisense transcription; positive coverage of 50LTR reveals host gene transcription (read-through) and provirus-dependent premature polyadenylation. Bottom panels: mapping to host genomes (UMD3.1 and OAR3.1). Small box: BLV integration in MSH2 (ref. 33) intron 6 or STARD7 (ref. 34) intron 5. Decrease of MSH2 and STARD7 downstream exon coverage (MSH2 T1345/control tumours (N ¼ 14), 88% decrease; T1345 MSH2 downstream/upstream exons, fold-change ¼ 0.12 and STARD7: 96% decrease (control tumours, N ¼ 31), downstream/upstream exon fold-change ¼ 0.08). Antisense coverage: 30AS-dependent chimeric transcript in antisense overlap with MSH2/STARD7. See also Supplementary Fig. 3 for RNA-seq coverage assignment to 50LTR/30LTR. Most importantly, it revealed the systematic interactions between the antisense transcripts (30AS) and host genes located upstream of the provirus. (Fig. 2b, Supplementary Fig. 4 and Supplementary Table 3). These interactions conformed to four major schemes: capture of host exons located upstream of the provirus by the ﬁrst viral 50LTR-dependent sense transcripts (50S) (corresponding to the GAG, POL, ENV structural genes and the regulatory genes including TAX), yet abundant 30LTR-dependent antisense tran- scripts corresponding to the previously described HTLV-1 HBZ and BLV AS1/2 RNAs11,22,24 (Fig. 2a and Supplementary Fig. 3). 4 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 In 21 cases, the interacting host gene was not the gene most closely located to the provirus. Like these 71 proviruses, the 21 genic-concordant proviruses described above all showed evidence of 30LTR dependent transcription with antisense overlap of upstream sequences in addition to the poly-(A) dependent interruption type of perturbation (Figs 1 and 2b, top left panel). In 5 of these 21 tumours, we observed additional interactions of this 30LTR-dependent antisense transcript with the adjacent upstream HBZ/AS exon (genic (27/71 tumours) and intergenic (12/71 tumours), discordant), antisense overlap of genes located upstream of the provirus by long 30LTR-dependent antisense transcripts (genic (6/71 tumours) and intergenic (8/71 tumours) concordant), sense overlap of genes located upstream of the provirus by long 30LTR-dependent antisense transcripts (23/71 tumours, intergenic discordant) and the capture of viral HBZ/AS exon 2 by host gene transcripts (10/27 tumours, genic- discordant) (Figs 2b and 3 and Supplementary Fig. 5). The genuine nature of the chimeric transcripts detected by RNA-seq was tested by RT–PCR (7 transcripts) or 30 modiﬁed RACE (2 transcripts) and conﬁrmed for all of these (Fig. 3c and Supplementary Fig. 7). g pp y g At least one such antisense RNA-dependent interaction was observed in every one of the 71 tumours. Across all 71 tumours, a total of 92 genes were involved, including 60 coding 5 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 SEPT11 BLV AS FOXR2 SEPT11 RRAGB RRAGB FOXR2 ICA1 BLV AS ICA1 ICA1 d c b a IS: chr4:16,124,876 chr4:16,150,000 OAR3.1 * [0–300] [0–160] [0–80] [0–100] [0–75] [0–70] 50 kb exon 13 OAR3.1 220 kb chrX:47,630,000 IS:chrX:47,600,711 IS: chrX:47,618,373 BLV BLV 5′ 3′ 5′ 3′ IS: chrX:47,562,215 chr6:93,525,000 IS: chr6:93,525,293 Bovine tumour LB120 Bovine control tumour B3191 RNA-Seq mapping to provirus RNA-Seq mapping to host Splice junction Unspliced fusion transcript exon1 bosTau6 60 kb chr6:93,585,000 RNA-Seq 3′LTR 5′LTR SD SD SA SA BLV 3′ 5′ Novel exon Novel exon exon 1 exon 2 exon 2 [0–200] [0–230] BLV 5′ 3′ Ovine tumour M160 chrX: 47,410,000 Ovine control tumour M126 Ovine tumour M21 Ovine tumour M138 Ovine tumour M251 Ovine control tumour M210 Ovine tumour M138 Ovine tumour M251 Ovine control tumour L267 200 kb BLV RNA-Seq 5′ 5′LTR 3′LTR 3′ chr4:16,100,000 e 3 \| Viral antisense RNA-dependent cis-perturbation of host genes in representative tumours with discordant proviruses. RNA-seq an and sense (red) coverages of tumours with genic (a) and intergenic (b–d) discordant proviruses. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 RNA-seq antisense (blue) and sense (red) coverages of tumours with genic (a) and intergenic (b–d) discordant proviruses. Upper (yellow) and lower (black) IGV tracks: tumour of interest and control tumour, respectively. (a) Genic-discordant provirus and host gene cis-perturbation by 30AS-dependent capture of upstream exons. Ovine tumour M160: capture and increased coverage of ICA1 exons 13–15; read counts M160 ICA1 upstream/downstream exons fold- change ¼ 11.54; ICA1 upstream exons M160/control tumours (N ¼ 31) fold-change ¼ 6.89. Box: hybrid RNA-seq split reads spanning BLV AS exon 1 and ICA1 exon 13. (b) Intergenic-discordant provirus and interaction with multiple host genes: RNA-seq of three independent ovine tumours M251, M138, M21 with BLV integration upstream of FOXR2 (ref. 42) (80 kb-window) reveals sense overlap of FOXR2 and RRAGB (ref. 66) by 30AS-dependent hybrid transcripts (160, 220 and 240 kb in length respectively). (c) 30AS-capture RNA-seq reveals AS exon 1–FOXR2/RRAGB splice junctions and ectopic expression of FOXR2. Mapping and Sashimi plots of reads obtained from 30AS-capture RNA libraries (STAR splice-aware aligner) prepared from modiﬁed 30-RACE products of tumours M138 and M251 reveal FOXR2 coverage as well as chimeric split reads exposing splicing events between BLV AS exon 1 and upstream genomic sequences including FOXR2 and RRAGB (white tracks). BLV integration causes ectopic expression of FOXR2 a gene that is not expressed in normal ovine B-cells (L267: control tumour), consistent with a gain-of-function of this well-established oncogene. (d) Intergenic-discordant proviruses generate novel transcript isoforms: host gene cis-perturbation by sense overlap of upstream gene by 30AS-dependent hybrid transcript with capture of both exon 2 and a novel exon upstream of canonical exon 1, creating two novel isoforms (bovine tumour LB120: BLV integration upstream of SEPT11 (ref. 67), SEPT11 exon 1 skipping). See also Supplementary Fig. 5 for representative tumours with concordant proviruses. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 b SEPT11 BLV AS FOXR2 SEPT11 RRAGB d c chr6:93,525,000 IS: chr6:93,525,293 Bovine tumour LB120 Bovine control tumour B3191 RNA-Seq mapping to provirus RNA-Seq mapping to host Splice junction Unspliced fusion transcript exon1 bosTau6 60 kb chr6:93,585,000 RNA-Seq 3′LTR 5′LTR SD SD SA SA BLV 3′ 5′ Novel exon Novel exon exon 1 exon 2 exon 2 [0–200] [0–230] Ovine tumour M21 Ovine tumour M138 Ovine tumour M251 Ovine control tumour M210 Ovine tumour M138 Ovine tumour M251 Ovine control tumour L267 200 kb c d Bovine tumour LB120 Figure 3 \| Viral antisense RNA-dependent cis-perturbation of host genes in representative tumours with discordant proviruses. RNA-seq antisense (blue) and sense (red) coverages of tumours with genic (a) and intergenic (b–d) discordant proviruses. Upper (yellow) and lower (black) IGV tracks: tumour of interest and control tumour, respectively. (a) Genic-discordant provirus and host gene cis-perturbation by 30AS-dependent capture of upstream exons. Ovine tumour M160: capture and increased coverage of ICA1 exons 13–15; read counts M160 ICA1 upstream/downstream exons fold- change ¼ 11.54; ICA1 upstream exons M160/control tumours (N ¼ 31) fold-change ¼ 6.89. Box: hybrid RNA-seq split reads spanning BLV AS exon 1 and ICA1 exon 13. (b) Intergenic-discordant provirus and interaction with multiple host genes: RNA-seq of three independent ovine tumours M251, M138, M21 with BLV integration upstream of FOXR2 (ref. 42) (80 kb-window) reveals sense overlap of FOXR2 and RRAGB (ref. 66) by 30AS-dependent hybrid transcripts (160, 220 and 240 kb in length respectively). (c) 30AS-capture RNA-seq reveals AS exon 1–FOXR2/RRAGB splice junctions and ectopic expression of FOXR2. Mapping and Sashimi plots of reads obtained from 30AS-capture RNA libraries (STAR splice-aware aligner) prepared from modiﬁed 30-RACE products of tumours M138 and M251 reveal FOXR2 coverage as well as chimeric split reads exposing splicing events between BLV AS exon 1 and upstream genomic sequences including FOXR2 and RRAGB (white tracks). BLV integration causes ectopic expression of FOXR2 a gene that is not expressed in normal ovine B-cells (L267: control tumour), consistent with a gain-of-function of this well-established oncogene. (d) Intergenic-discordant proviruses generate novel transcript isoforms: host gene cis-perturbation by sense overlap of upstream gene by 30AS-dependent hybrid transcript with capture of both exon 2 and a novel exon upstream of canonical exon 1, creating two novel isoforms (bovine tumour LB120: BLV integration upstream of SEPT11 (ref. 67), SEPT11 exon 1 skipping). See also Supplementary Fig. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Upper (yellow) and lower (black) IGV t ur of interest and control tumour, respectively. (a) Genic-discordant provirus and host gene cis-perturbation by 30AS-dependent capture of up . Ovine tumour M160: capture and increased coverage of ICA1 exons 13–15; read counts M160 ICA1 upstream/downstream exons fold- e ¼ 11.54; ICA1 upstream exons M160/control tumours (N ¼ 31) fold-change ¼ 6.89. Box: hybrid RNA-seq split reads spanning BLV AS exon xon 13. (b) Intergenic-discordant provirus and interaction with multiple host genes: RNA-seq of three independent ovine tumours M251, M13 BLV integration upstream of FOXR2 (ref. 42) (80 kb-window) reveals sense overlap of FOXR2 and RRAGB (ref. 66) by 30AS-dependent hybr ripts (160, 220 and 240 kb in length respectively). (c) 30AS-capture RNA-seq reveals AS exon 1–FOXR2/RRAGB splice junctions and ectop ssion of FOXR2. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Mapping and Sashimi plots of reads obtained from 30AS-capture RNA libraries (STAR splice-aware aligner) prepared from m CE products of tumours M138 and M251 reveal FOXR2 coverage as well as chimeric split reads exposing splicing events between BLV AS exo eam genomic sequences including FOXR2 and RRAGB (white tracks) BLV integration causes ectopic expression of FOXR2 a gene that is not ex SEPT11 BLV AS FOXR2 SEPT11 RRAGB RRAGB FOXR2 ICA1 BLV AS ICA1 ICA1 d c b a IS: chr4:16,124,876 chr4:16,150,000 OAR3.1 * [0–300] [0–160] [0–80] [0–100] [0–75] [0–70] 50 kb exon 13 OAR3.1 220 kb chrX:47,630,000 IS:chrX:47,600,711 IS: chrX:47,618,373 BLV BLV 5′ 3′ 5′ 3′ IS: chrX:47,562,215 chr6:93,525,000 IS: chr6:93,525,293 Bovine tumour LB120 Bovine control tumour B3191 RNA-Seq mapping to provirus RNA-Seq mapping to host Splice junction Unspliced fusion transcript exon1 bosTau6 60 kb chr6:93,585,000 RNA-Seq 3′LTR 5′LTR SD SD SA SA BLV 3′ 5′ Novel exon Novel exon exon 1 exon 2 exon 2 [0–200] [0–230] BLV 5′ 3′ Ovine tumour M160 chrX: 47,410,000 Ovine control tumour M126 Ovine tumour M21 Ovine tumour M138 Ovine tumour M251 Ovine control tumour M210 Ovine tumour M138 Ovine tumour M251 Ovine control tumour L267 200 kb BLV RNA-Seq 5′ 5′LTR 3′LTR 3′ chr4:16,100,000 RRAGB RRAGB FOXR2 ICA1 BLV AS ICA1 ICA1 c b a IS: chr4:16,124,876 chr4:16,150,000 OAR3.1 * * [0–300] [0–160] [0–80] [0–100] [0–75] [0–70] 50 kb exon 13 OAR3.1 220 kb chrX:47,630,000 IS:chrX:47,600,711 IS: chrX:47,618,373 BLV BLV 5′ 3′ 5′ 3′ IS: chrX:47,562,215 BLV 5′ 3′ Ovine tumour M160 chrX: 47,410,000 Ovine control tumour M126 Ovine tumour M21 Ovine tumour M138 Ovine tumour M251 Ovine control tumour M210 BLV RNA-Seq 5′ 5′LTR 3′LTR 3′ chr4:16,100,000 a SEPT11 BLV AS FOXR2 SEPT11 RRAGB RRAGB FOXR2 ICA1 BLV AS ICA1 d c b [0–160] [0–80] [0–100] [0–75] [0–70] exon 13 OAR3.1 220 kb chrX:47,630,000 IS:chrX:47,600,711 IS: chrX:47,618,373 BLV BLV 5′ 3′ 5′ 3′ IS: chrX:47,562,215 chr6:93,525,000 IS: chr6:93,525,293 Bovine tumour LB120 Bovine control tumour B3191 RNA-Seq mapping to provirus RNA-Seq mapping to host Splice junction Unspliced fusion transcript exon1 bosTau6 60 kb chr6:93,585,000 RNA-Seq 3′LTR 5′LTR SD SD SA SA BLV 3′ 5′ Novel exon Novel exon exon 1 exon 2 exon 2 [0–200] [0–230] BLV 5′ 3′ Ovine tumour M160 chrX: 47,410,000 Ovine control tumour M126 Ovine tumour M21 Ovine tumour M138 Ovine tumour M251 Ovine control tumour M210 Ovine tumour M138 Ovine tumour M251 Ovine control tumour L267 200 kb RNA-Seq 5′LTR 3′LTR Figure 3 \| Viral antisense RNA-dependent cis-perturbation of host genes in representative tumours with discordant proviruses. NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 STARD7 b a 3 2 1 0 Occurence in cancer driver gene lists HTLV-1 human BLV ovine BLV bovine Genic concordant: Viral poly-(A)-dependent premature interruption No evidence of viral poly-(A)-dependent premature interruption Genic discordant Intergenic concordant Intergenic discordant 5 4 3 2 Pvalue = 0.05 Random TCGA generanker AllOnco–G AllOnco Tamborero Tamborero–H Cancer5000 Cancer gene census Meta–analysis Downstream nearest gene –log10(Pvalue) 0 0 1 2 1 2 Random Para Expr Expr Para P value = 0.05 1 0 Occurence in cancer driver gene lists Upstream nearest gene –log10(Pvalue) MSH2 TRNAU1AP SNUPN RAPGEF6 OSBPL8 OSBP NAPG LYRM4 KPNA3 WDR82 NUMB MAST4 BRCC3 RASA3 MYCBP2 KSR1 DNMT3A NF1 SMAD2 DDX10 MTCP1 PDGFRB DSG2 GAS6 HUWE1 IRS2 MCC ARHGEF4 CA10 CENPQ DSG1 DSG4 FARS2 GNGT1 ICA1 IL12A KCND2 LHPP MUT N4BP2 PPP1R12C PRPSAP2 SEPT11 SNIP1 SPOCK1 STK17A STK31 STX5 TCF4 TLE4 TLR9 TMSB15B UBE2D3 CCRN4L CDX1 CGGBP1 COL2A1 DSG3 ELF2 FAM168B FOXR2 HECTD1 KLHL14 LCORL NXF1 RAB9B RBFOX1 RGCC RRAGB SCAF8 SEC11C SNX33 SPSB1 TMEM67 TNKS TOR1AIP2 UBASH3B VPS39 Figure 4 \| HTLV-1/BLV interacting host genes are connected to cancer. (a) Transcript-interacting host genes were ranked according to their occurren seven cancer driver lists used for enrichment (Table 1 and Supplementary Table 2). Top panel: viral poly-(A) truncated genes disrupted by genic-conco proviruses. Bottom panel: all genes interacting with 30AS-dependent transcripts. Cancer drivers are equally represented between provirus types (genic, intergenic) and across species. Underlined genes: genes for which transcriptional patterns in tumours are shown in Figs 1 and 3 and Suppleme Fig. 5. Recurrent genes between tumours: green symbols. Genes absent from cancer driver lists for which literature screen supported connection to ca This list comprises FOXR2, RRAGB, ELF2 and SPSB1 (refs 40,42,66,68), genes that exhibit undeniable oncogenic properties. UBASH3B (BLV, sheep) identiﬁed as the target gene of HTLV-1 integration in one of the ATLs analysed in a recent WGS study11. The remaining protein-coding genes and intera non-coding RNAs not previously reported in cancer (Supplementary Data 2 and Supplementary Table 4, antisense transcript-interacting lncRNAs) represent a potential resource of novel candidate cancer drivers of both the coding and noncoding class of genes. (b) Host genes upstream (y-axi proviral integration sites in ATLs and B-cell tumours (92 sites) show signiﬁcant enrichment in cancer drivers in contrast to the corresponding downst host genes (x-axis), supporting antisense-dependent cancer driver perturbation by HTLV-1/BLV proviruses. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 The direct target genes of genic proviruses excluded from the analysis. The signiﬁcance of the enrichment was computed for seven publicly available cancer driver lists and for all list combine means of a meta analysis (Supplementary Data 3 and Supplementary Table 2) Observed scores were compared to simulated scores obtained from STARD7 a 3 2 1 0 Occurence in cancer driver gene lists HTLV-1 human BLV ovine BLV bovine Genic concordant: Viral poly-(A)-dependent premature interruption No evidence of viral poly-(A)-dependent premature interruption Genic discordant Intergenic concordant Intergenic discordant 5 4 3 2 1 0 Occurence in cancer driver gene lists MSH2 TRNAU1AP SNUPN RAPGEF6 OSBPL8 OSBP NAPG LYRM4 KPNA3 WDR82 NUMB MAST4 BRCC3 RASA3 MYCBP2 KSR1 DNMT3A NF1 SMAD2 DDX10 MTCP1 PDGFRB DSG2 GAS6 HUWE1 IRS2 MCC ARHGEF4 CA10 CENPQ DSG1 DSG4 FARS2 GNGT1 ICA1 IL12A KCND2 LHPP MUT N4BP2 PPP1R12C PRPSAP2 SEPT11 SNIP1 SPOCK1 STK17A STK31 STX5 TCF4 TLE4 TLR9 TMSB15B UBE2D3 CCRN4L CDX1 CGGBP1 COL2A1 DSG3 ELF2 FAM168B FOXR2 HECTD1 KLHL14 LCORL NXF1 RAB9B RBFOX1 RGCC RRAGB SCAF8 SEC11C SNX33 SPSB1 TMEM67 TNKS TOR1AIP2 UBASH3B VPS39 a HTLV-1 human BLV ovine BLV bovine Genic concordant: Viral poly-(A)-dependent premature interruption No evidence of viral poly-(A)-dependent premature interruption Genic discordant Intergenic concordant Intergenic discordant b Pvalue = 0.05 Random TCGA generanker AllOnco–G AllOnco Tamborero Tamborero–H Cancer5000 Cancer gene census Meta–analysis Downstream nearest gene –log10(Pvalue) 0 0 1 2 1 2 Random Para Expr Expr Para P value = 0.05 Upstream nearest gene –log10(Pvalue) P b Downstream nearest gene –log10(Pvalue) Figure 4 \| HTLV-1/BLV interacting host genes are connected to cancer. (a) Transcript-interacting host genes were ranked according to their occurrence in seven cancer driver lists used for enrichment (Table 1 and Supplementary Table 2). Top panel: viral poly-(A) truncated genes disrupted by genic-concordant proviruses. Bottom panel: all genes interacting with 30AS-dependent transcripts. Cancer drivers are equally represented between provirus types (genic, intergenic) and across species. Underlined genes: genes for which transcriptional patterns in tumours are shown in Figs 1 and 3 and Supplementary Fig. 5. Recurrent genes between tumours: green symbols. Genes absent from cancer driver lists for which literature screen supported connection to cancer. This list comprises FOXR2, RRAGB, ELF2 and SPSB1 (refs 40,42,66,68), genes that exhibit undeniable oncogenic properties. UBASH3B (BLV, sheep): identiﬁed as the target gene of HTLV-1 integration in one of the ATLs analysed in a recent WGS study11. NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 5 for representative tumours with concordant proviruses. Figure 3 \| Viral antisense RNA-dependent cis-perturbation of host genes in representative tumours with discorda (bl ) d ( d) h ( ) d (b d) d d ( ll dent cis-perturbation of host genes in representative tumours with discordant proviruses. RNA-seq antisense NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 6 ARTICLE ARTICLE The list of genic hotspots includes established cancer drivers such as ARID1A, ARID1B, CBL-B, PIK3CA and PTEN45–48 (Fig. 5c). Most interestingly, we observed a very strong signature of non-randomness of proviral orientation within hotspots (Fig. 5d–f). Of the 674 (76%) genic integration hotspots, 517 showed a signiﬁcant bias (FDRo0.1) towards either concordant (350) or discordant (167) orientation (with regard to the orientation of the host gene as deﬁned above; concordant: provirus and gene in the same orientation; discordant: provirus and gene in opposite orientation). The genic hotspots with predominant concordant integration most probably point towards events of insertional inactivation by transcriptional termination, while hotspots with predominant discordant integration may point towards viral antisense-dependent exon captures. This ﬁnding certainly weakens the alternative (versus selection) hypothesis of chromatin-feature-dependent integration (which is unlikely to be orientation-dependent). Equally interesting, we observed the same orientation bias for non-genic integrations. Of the 48, 38 (79%) non-genic integration hotspots showed an orientation bias with FDRo0.1 (see also Fig. 5e,f). This observation is in agreement with our hypothesis of BLV antisense RNA-dependent perturbation of cancer driver genes. haematological system development and function), in only 6% of the ATL cases examined (11/197) and only when restricting the analyses to genes located downstream but not upstream of the provirus (contrary to our ﬁndings which point upstream genes, consistent with antisense-dependent interaction). The reasons for these apparent discrepancies between their and our ﬁndings are not known. Proviral antisense-dependent exon capture has the potential to cause expression of non-canonical isoforms of expressed genes, or ectopic expression of genes that are normally silent in the lymphoid lineage. Intriguing examples include N-terminal truncated isoforms of ELF2 and TCF4 (members of the Ets family of transcription factors and DNA binding transcrip- tional regulators of the Wnt pathway, respectively)40,41, and ectopic expression of the well-established oncogene FOXR2 (refs 42,43) in three independent ovine tumours (Fig. 3b,c). The capture of viral HBZ/AS exon 2 by host gene transcripts (10/27 genic-discordant cases) may cause premature transcription termination of the host gene at the HBZ/AS exon 2 poly (A) site (Fig. 2b, top right panel and Supplementary Fig. 4, scheme v). This suggests that genic-discordant insertions have the potential to affect the same gene by two distinct mechanisms—transcript interruption and downstream exon capture—consistent with observations in Sleeping Beauty transposon induced tumours in mice44. ARTICLE The genic hotspots with predominant concordant integration most probably point towards events of insertional inactivation by transcriptional termination, while hotspots with predominant discordant integration may point towards viral antisense-dependent exon captures. This ﬁnding certainly weakens the alternative (versus selection) hypothesis of chromatin-feature-dependent integration (which is unlikely to be orientation-dependent). Equally interesting, we observed the same orientation bias for non-genic integrations. Of the 48, 38 (79%) non-genic integration hotspots showed an orientation bias with FDRo0.1 (see also Fig. 5e,f). This observation is in agreement with our hypothesis of BLV antisense RNA-dependent perturbation of cancer driver genes. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 in promoting early-stage polyclonal expansion rather than late-stage precipitation to full-blown monoclonal tumour. To discriminate between these two hypotheses, we utilized the BLV experimental model in sheep. A considerable advantage of this model is that—contrary to the natural diseases in human and cattle—all infected sheep develop leukaemia/lymphoma, tumour onset occurs within a much shorter time frame (20 months on average) and it is possible to monitor infected animals at the very early stages of infection. We ﬁrst comprehensively analysed proviral integration sites at early nonmalignant stages (characterized by the presence of multiple clones of low abundance). This was achieved by very deep, high-throughput DNA-seq-based mapping of BLV integration sites for 10 infected but still asymptomatic sheep (proviral load range: 0.02–34%, clone abundance range: 0.002–9.524%; Supplementary Data 1). It uncovered 66,557 unique integration events. Examining their chromosomal distribution revealed extreme non-randomness, deﬁning 674 genic and 48 intergenic hotspots of integration (genome-wide corrected Po0.05) (Fig. 5). We showed by simulation that the majority of genic hotspots could not be explained by expression level and gene size alone (false discovery rate (FDR)o0.1 for 468/674 genic hotspots; proportion of true alternative (i.e., not explainable by expression level and size alone) hypotheses (p1) ¼ 0.67). The average number of integration sites per hotspot was 33 (range: 12–322) for genic hotspots and 37 (range: 12–202) for intergenic hotspots. The average size was 67,180 bp (range: 11,180–302,000) and 80,570 bp (range: 23,040–504,200) for genic and intergenic hotspots, respectively. Genes involved in genic hotspots showed a highly signiﬁcant enrichment in cancer drivers (Po1e-5; Supplementary Data 3) and a robust overlap with the 74 30AS-interacting host genes identiﬁed in the HTLV-1/BLV tumour set (P ¼ 0.00073; Fig. 5b and Supplementary Table 5). ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 in promoting early-stage polyclonal expansion rather than late-stage precipitation to full-blown monoclonal tumour. To discriminate between these two hypotheses, we utilized the BLV experimental model in sheep. A considerable advantage of this model is that—contrary to the natural diseases in human and cattle—all infected sheep develop leukaemia/lymphoma, tumour onset occurs within a much shorter time frame (20 months on average) and it is possible to monitor infected animals at the very early stages of infection. We ﬁrst comprehensively analysed proviral integration sites at early nonmalignant stages (characterized by the presence of multiple clones of low abundance). This was achieved by very deep, high-throughput DNA-seq-based mapping of BLV integration sites for 10 infected but still asymptomatic sheep (proviral load range: 0.02–34%, clone abundance range: 0.002–9.524%; Supplementary Data 1). It uncovered 66,557 unique integration events. Examining their chromosomal distribution revealed extreme non-randomness, deﬁning 674 genic and 48 intergenic hotspots of integration (genome-wide corrected Po0.05) (Fig. 5). We showed by simulation that the majority of genic hotspots could not be explained by expression level and gene size alone (false discovery rate (FDR)o0.1 for 468/674 genic hotspots; proportion of true alternative (i.e., not explainable by expression level and size alone) hypotheses (p1) ¼ 0.67). The average number of integration sites per hotspot was 33 (range: 12–322) for genic hotspots and 37 (range: 12–202) for intergenic hotspots. The average size was 67,180 bp (range: 11,180–302,000) and 80,570 bp (range: 23,040–504,200) for genic and intergenic hotspots, respectively. Genes involved in genic hotspots showed a highly signiﬁcant enrichment in cancer drivers (Po1e-5; Supplementary Data 3) and a robust overlap with the 74 30AS-interacting host genes identiﬁed in the HTLV-1/BLV tumour set (P ¼ 0.00073; Fig. 5b and Supplementary Table 5). The list of genic hotspots includes established cancer drivers such as ARID1A, ARID1B, CBL-B, PIK3CA and PTEN45–48 (Fig. 5c). Most interestingly, we observed a very strong signature of non-randomness of proviral orientation within hotspots (Fig. 5d–f). Of the 674 (76%) genic integration hotspots, 517 showed a signiﬁcant bias (FDRo0.1) towards either concordant (350) or discordant (167) orientation (with regard to the orientation of the host gene as deﬁned above; concordant: provirus and gene in the same orientation; discordant: provirus and gene in opposite orientation). ARTICLE For the majority of intergenic proviruses (26/38), standard RNA-seq revealed sense or antisense gene overlap by provirus-dependent hybrid transcripts, yet without direct evidence of exon capture. Nevertheless, enrichment in cancer drivers for the 30 corresponding protein-coding genes was robust and highly signiﬁcant (0.0015oPo0.0004, Table 1 and Supplementary Data 3), suggesting that these antisense-driven transcripts have a functional impact on the corresponding genes despite the lack of obvious transcriptional effects. The strong bias for upstream interaction is in agreement with the absence of 50LTR dependent mRNA transcription from the proviral positive-strand in all tumours examined, consistent with previous reports that showed antisense-predominant HTLV-1/BLV transcription and 50LTR epigenetic silencing or TAX mutations in tumours11,18–20,24 (Fig. 2a and Supplementary Fig. 3). It is noteworthy that we observed 30LTR-driven antisense-dependent chimeric transcripts involving well-established cancer drivers located upstream of the provirus in the 11 ATLs with 50LTR- deleted defective HTLV-1 proviruses (i.e., SPSB1; Supplementary Fig. 8). This strongly suggests that in fully malignant clones, positive-strand—and paradoxically TAX—silencing accompanies HTLV-1/BLV antisense-dependent host gene cis-perturbation presumably allowing the malignant clone to proliferate under strong host immune control21. g Note that Kataoka et al. also reported HTLV-1-dependent read-through transcripts in ATLs. From 53 integration sites, the authors identiﬁed 12 aberrantly spliced fusions with the host genome, which all were produced from genic proviruses (N ¼ 23). They did not report interactions with host genes in the case of intergenic proviruses (N ¼ 30). While the authors conclude that the relevance of aberrant transcripts observed in ATL is unknown, our ﬁndings show that they tend to occur with genes that are enriched in known cancer drivers and hence play an important role in tumorigenesis. Virus–host hybrid-RNA signatures at asymptomatic stages. To further test this hypothesis, we followed up on this DNA-seq- based study and performed RNA-seq of BLV antisense-enriched RNA from the same asymptomatic animals (30AS-capture RNA-seq). This revealed BLV 30AS-host chimeric transcripts involving multiple genes per sample (range of 4–276, mean of 84; Supplementary Table 3), supporting a total of 723 interacting host genes. Genes involved showed a highly signiﬁcant level of Altogether, our ﬁndings support the notion that cis-perturba- tion of cancer drivers by the HTLV-1/BLV proviruses is an essential component of leukemogenesis. Non-random provirus distribution at asymptomatic stages. The results reported thus far were obtained on late-stage tumours. NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 The remaining protein-coding genes and interacting non-coding RNAs not previously reported in cancer (Supplementary Data 2 and Supplementary Table 4, antisense transcript-interacting lncRNAs) represent a potential resource of novel candidate cancer drivers of both the coding and noncoding class of genes. (b) Host genes upstream (y-axis) of proviral integration sites in ATLs and B-cell tumours (92 sites) show signiﬁcant enrichment in cancer drivers in contrast to the corresponding downstream host genes (x-axis), supporting antisense-dependent cancer driver perturbation by HTLV-1/BLV proviruses. The direct target genes of genic proviruses were excluded from the analysis. The signiﬁcance of the enrichment was computed for seven publicly available cancer driver lists and for all list combined by means of a meta-analysis (Supplementary Data 3 and Supplementary Table 2). Observed scores were compared to simulated scores obtained from N ¼ 100,000 size-matched random or expression-matched gene sets, including information about paralogs (Random Para, Expr Para) or not (Random, Expr) (Supplementary Fig. 2). Symbol code: simulated gene sets, colour code: cancer driver lists. host gene, either by antisense overlap (2 tumours) or by exon capture (3 tumours; Supplementary Fig. 5b). species (1e-05oPo0.0006; Table 1 and Supplementary Data 3). Interacting genes included well-established cancer-connected genes such as DNMT3A, SMAD2, MTCP1 and TLE4 (refs 36–39) (Fig. 4a). Further supporting a role for viral antisense RNA-dependent interaction in tumorigenesis, we observed a signiﬁcant enrichment in cancer drivers for the 43 nearest protein-coding genes located in a 1 Mb-window upstream of intergenic proviral integration sites while the matched list of 48 genes located downstream of these integration sites was not enriched (Fig. 4b). Note that Cook et al. (2014) presented evi- dence for loose association of HTLV-1 integration sites with Gene Ontology terms (cell morphology, immune cell trafﬁcking, Provirus-interacting genes are enriched in cancer drivers. To test the biological relevance of the observed antisense RNA-dependent interactions, we examined the enrichment of the 65 (60 þ 5) antisense interacting protein-coding genes in known cancer drivers (gene subsets and criteria for gene inclusion; Supplementary Data 3 and Supplementary Table 4). The enrichment was highly signiﬁcant and robust, and dependent on both the 33 genic and 38 intergenic insertion sites of the three 7 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 recurrent capture of the same host genes between asymptomatic individuals (Po1e-5), a signiﬁcant overlap with transcript- interacting host genes previously identiﬁed in malignant clones of the three species (P ¼ 0.00085; Fig. 5b and Supplementary Table 6), and a highly signiﬁcant enrichment in cancer drivers (Po1e-05; Table 1 and Supplementary Data 3). As expected, we observed a very strong overlap between the genes exposed by 30AS-capture RNA-seq hybrid reads and the genic hotspots identiﬁed by DNA-seq mapping of integration sites in the same asymptomatic individuals (Po1e-5; Fig. 5b). ARTICLE They do not discriminate between a role for proviral integration NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 8 ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 hotspots deﬁned by DNA-seq (Fig. 5g). It also uncovered typical splice junctions between the BLV AS transcript and the corresponding upstream gene. This revealed cases of potential activation of important oncogenes (i.e., ID3 (ref. 49), Fig. 5h), and included other examples (in addition to FOXR2) that may lead to ectopic expression of otherwise silent genes (i.e., CCDC80; Fig. 5i). These observations are in perfect agreement with our hypothesis of viral antisense RNA-dependent perturbation of cancer drivers. hotspots deﬁned in sheep, we performed simulations assuming that a fraction w of x proviral integration sites are sampled from clones that are undergoing expansion due to perturbation of one of y cancer drivers (out of a total of 20,000 genes), of which a fraction z is reported in cancer driver lists. The remaining fraction (1 w) of proviral integration sites are assumed to be sampled from infected but non-expanded leukocyte clones. w was varied from 0.01 to 1, x from 1,000 to 100,000, y from 500 to 3,000 and z from 0.25 to 1. It was obvious that the statistical test for enrichment in cancer drivers was considerably more powerful than that for the detection of integration hotspots for a substantial proportion of parameter space (Supplementary Fig. 10). This is mostly due to the requirement to adjust the hotspot test but not the enrichment test for multiple testing. Particularly noteworthy was the effect of a decreasing proportion of integration sites sampled from expanded clones (w). This fundamental power difference between the two tests may thus very well explain the apparent discrepancy observed with the human data, and the fact that integration hotspots were not reported before. The model used in the simulations, characterized by two distinct populations of infected lymphocytes (one expanding with proviral insertions affecting cancer drivers, and one not expanding with random proviral insertions), predicts that the degree of integration in hotspots and of enrichment in cancer drivers should be correlated with corresponding clonal abundance (as expanding clones are by deﬁnition more abundant). We tested this prediction using the large number of insertion sites available in asymptomatic sheep and found that it was indeed the case (Supplementary Table 9). Assessing integration hotspots and cancer driver enrichment. Discussion T k Taken together, our results strongly support the notion that cis-perturbation of cancer drivers by the provirus is a major determinant of early clonal expansion in both BLV and HTLV-1 induced leukaemia. We provide circumstantiated evidence that the absence of easily detectable integration hotspots yet enrich- ment in cancer drivers in the human natural host (contrary to the ﬂagrant hotspots detected in the ovine model) may reﬂect species-speciﬁc dynamics of infected-expanding versus infected- nonexpanding lymphocyte populations and hence proportions. This may point towards disparities in the antiviral immune response—a major driving force underlying clone abundance in HTLV-1 individuals21,28—between the experimental model and the human disease. It is tempting to speculate that this may also Figure 5 \| Hotspots of proviral integration at polyclonal nonmalignant stages of infection. (a) Genome-wide distribution of BLV integration sites in asymptomatic sheep samples. Y-axis: number of integration sites per genomic bin (100 kb overlapping genomic windows sliding by steps of 50 kb). Hotspots of proviral integration were identiﬁed by simulation (Methods), deﬁning 674 genic and 48 intergenic hotspots (Po0.05). (b) Signiﬁcant recurrence (P-values) between genes revealed by BLV genic integration hotspots (674 genes), antisense-RNA interacting genes identiﬁed in tumours (74 genes, HTLV-1/BLV) and genes identiﬁed by 30AS-capture RNA-seq of asymptomatic samples (723 genes). All gene subsets showed robust cancer driver enrichment (Po1e-05 and 7e-04 for asymptomatic and tumour samples, respectively, Supplementary Data 3). (c) Top 50 genic integration hotspots. Comprise gene classes like chromatin modiﬁers, E3-ubiquitin ligases and tumour suppressors (ARID1B, CBL-B, PTEN). Genes in bold: also identiﬁed in tumour RNA-seq data set (TLE4 and STK17A: ATLs, TCF4: bovine B-cell tumour). (d) Genic integration hotspot in tumour suppressor TLE4 (ref. 39): arrows represent proviruses (50–30 orientation). Of 163 sites, 159 show identical orientation (ratio same/opposite: 0.97) consistent with genic-concordant proviruses predicted to cause TLE4 loss-of-function. TLE4 also affected in tumour data set (ATL62_2). (e) Intergenic integration hotspot upstream of RTN4 (ref. 69): 121/124 sites show identical orientation (ratio same/opposite: 0.97) consistent with intergenic-discordant proviruses predicted to cause RTN4 activation (gain-of-function). Mixed genic–intergenic hotspot type shown in Supplementary Fig. 9 (ATF7IP). (f) Provirus pairs from genic (red) and intergenic (blue) IS data sets were scored for relative orientation and same/opposite ratios computed for all combinations of pairs (Methods). Bias towards same orientation is associated with provirus proximity, consistent with non-randomness of proviral orientation in hotspots. ARTICLE When infected animals develop full-blown leukaemia following the asymptomatic polyclonal phase, the tumour clone is assumed to expand and dominate, yet coexist with multiple infected clones that remain in the background. Further mining the data from the DNA-seq-based integration mapping of 22/32 ovine tumours indeed revealed 8,015 minor proviral integration sites in addition to the 22 dominant ones described above. As expected, this novel collection of integration sites was characterized by highly signiﬁcant non-randomness, deﬁning 65 hotspots (61 genic and 4 intergenic) that all overlapped with the hotspots deﬁned in the asymptomatic samples. Forty-ﬁve (42 genic and 3 intergenic, 70%) of the hotspots were characterized by signiﬁcant (FDRr0.10) orientation bias (33 concordant, 9 discordant for genic hotspots). Also as expected the genes corresponding to the genic hotspots were highly enriched in cancer drivers (Po1e-5). The same approach was then applied to 31/44 human ATL cases. This allowed for the identiﬁcation of 4,628 minor proviral inte- gration sites in addition to the 32 tumour-speciﬁc ones described above. Surprisingly, examining their chromosomal distribution did not reveal any evidence for integration hotspots or orientation bias. We expanded our collection of HTLV-1 proviral integration sites with a recently published data set of 11,279 sites10,50 to reach a total of 15,939, but this did not alter the outcome regarding integration hotspots or orientation bias. These ﬁndings are in agreement with results published by others10,11,28. Paradoxically, the 7,155 human genic integration sites were very signiﬁcantly biased towards known cancer driver genes (Po1e-5), genic hotspots deﬁned in sheep (P ¼ 2.88e-13) and cis-perturbed host genes previously identiﬁed in malignant clones of the three species (P ¼ 6.68e-05) (Supplementary Tables 7 and 8). To understand the apparently discrepant results in the human samples, i.e., lack of evidence for integration hotspots yet strong enrichment of genic integration sites in cancer drivers and genic NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Most inter- estingly, the 30AS-capture RNA-seq data revealed mapping of viral antisense RNA-host chimeric reads upstream of intergenic ARID1B PTPRJ CBLC NCOA7 UBAC2 TLE4 STIM2 ARID1A ARHGAP24 STK17A FAM65B NCOA1 SWAP70 KLF12 ZDHHC14 KMT2C ADAM19 ATF7IP PLEKHA2 ADD3 INPP5D SPTLC2 KCNQ5 CTSH ZNF217 FLI1 ST3GAL1 KAT6A MARCH1 KDM6A MEF2C PAPD5 RASGRP1 CCDC6 SVIL TCF4 PTEN CCDC88C DIAPH1 NABP1 DOCK11 CNTRL COL19A1 RASA1 SMIM14 FOXP1 RASGRP3 TTC7A PPP3CC PIK3CA BLV AS CCDC80 ID3 BLV-AS host hybrid transcript BLV-AS host hybrid transcript BLV AS CCDC80 ID3 h f d a 1 150 100 IS in 100 kb sliding window 50 0 BLV asymptomatic RNA AS-capture (723) 571 7 7 60 0.00073 100 50 400 800 5′–3′ provirus 3′–5′ provirus 3′ 5′ 5′ 3′ RNA-Seq sense coverage RNA-Seq antisense coverage RNA-Seq sense AS-capture coverage RNA-Seq antisense AS-capture coverage RNA-Seq RNA-Seq AS-capture Standard Standard AS-capture 0 50 kb 120 kb chr2:242,140,000 chr1:175,730,000 chr1:175,850,000 chr2:242,190,000 10 RNA-Seq RNA-Seq 5 0 10 20 g 1.00 0.75 0.50 0.25 0.00 100 Genic Same/opposite provirus orientation ratio between integration sites Intergenic 1,000 10,000 Distance between two integration sites 100,000 1,000,000 1,000 10 Upstream Downstream AS-capture RNA-Seq virus-host chimeric read count RTN4 TLE4 e chr2:56,637,000 162 kb 170 kb chr2:56,799,000 chr3:68,780,000 chr3:68,950,000 150 100 50 0 RNA-Seq 600 400 200 0 RNA-Seq HTLV-1/BLV tumour RNA-seq fusion (74) BLV asymtomatic genic hotspots (674) 519 145 <1e-5 0.00085 300 200 100 0 IS number 2 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 99 percentile Genes present in cancer-driver lists 95 percentile Median 21 22 23 24 25 26 X i b c URE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications a 1 150 100 IS in 100 kb sliding window 50 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 99 percentile 95 percentile Median 21 22 23 24 25 26 X a a BLV asymptomatic RNA AS-capture (723) 571 7 7 60 0.00073 HTLV-1/BLV tumour RNA-seq fusion (74) BLV asymtomatic genic hotspots (674) 519 145 <1e-5 0.00085 2 b ARID1B PTPRJ CBLC NCOA7 UBAC2 TLE4 STIM2 ARID1A ARHGAP24 STK17A FAM65B NCOA1 SWAP70 KLF12 ZDHHC14 KMT2C ADAM19 ATF7IP PLEKHA2 ADD3 INPP5D SPTLC2 KCNQ5 CTSH ZNF217 FLI1 ST3GAL1 KAT6A MARCH1 KDM6A MEF2C PAPD5 RASGRP1 CCDC6 SVIL TCF4 PTEN CCDC88C DIAPH1 NABP1 DOCK11 CNTRL COL19A1 RASA1 SMIM14 FOXP1 RASGRP3 TTC7A PPP3CC PIK3CA 300 200 100 0 IS number Genes present in cancer-driver lists c b c d TLE4 chr2:56,637,000 162 kb chr2:56,799,000 150 100 50 0 RNA-Seq RTN4 e 170 kb chr3:68,780,000 chr3:68,950,000 600 400 200 0 RNA-Seq d e BLV AS CCDC80 ID3 BLV-AS host hybrid transcript BLV-AS host hybrid transcript BLV AS CCDC80 ID3 h f 100 50 400 800 5′–3′ provirus 3′–5′ provirus 3′ 5′ 5′ 3′ RNA-Seq sense coverage RNA-Seq antisense coverage RNA-Seq sense AS-capture coverage RNA-Seq antisense AS-capture coverage RNA-Seq RNA-Seq AS-capture Standard Standard AS-capture 0 50 kb 120 kb chr2:242,140,000 chr1:175,730,000 chr1:175,850,000 chr2:242,190,000 10 RNA-Seq RNA-Seq 5 0 10 20 g 1.00 0.75 0.50 0.25 0.00 100 Genic Same/opposite provirus orientation ratio between integration sites Intergenic 1,000 10,000 Distance between two integration sites 100,000 1,000,000 1,000 10 Upstream Downstream AS-capture RNA-Seq virus-host chimeric read count i RE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications f g t h i NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 9 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Somatic mutations Proviral integration HBZ/AS expression TAX expression Infection CTL growth suppression Asymptomatic polyclonal phase Leukaemia Infected but not expanded/persistent lymphocyte clones (proviral integration NOT perturbing cancer driver) Infected expanded/persistent lymphocyte clones (proviral integration perturbing cancer driver) M li l ( i l i i bi d i i i ) Uninfected lymphocyte Proviral integration CTL growth suppression Infected but not expanded/persistent lymphocyte clones (proviral integration NOT perturbing cancer driver) Infected expanded/persistent lymphocyte clones (proviral integration perturbing cancer driver) Malignant clone (proviral integration perturbing cancer driver + somatic mutations) Figure 6 \| Model of leukemogenesis by HTLV-1/BLV. After infection by HTLV-1/BLV the fate of a given infected T-cell/B-cell clone depends on the proviral integration site within the host genome, the expression of TAX and HTLV-1-HBZ/BLV-AS, the host CTL response to HTLV-1/BLV antigens and somatic mutations in the host genome. Asymptomatic polyclonal stage: the integration of HTLV-1/BLV proviruses in the vicinity of cancer drivers causes their perturbation and hence favours the persistence/survival/expansion of the corresponding infected clone (green clones). Clones in which proviral insertions do not affect cancer drivers show a modest survival (purple clones). The relative contribution of each infected clone to the polyclonal population of infected cells results from the balance between cancer driver perturbation, expression of TAX and HBZ/AS that both promote cell growth, and negative selection by the host CTL response. The prolonged life-span of clones in which cancer drivers are perturbed favours the acquisition of further somatic alterations in the host genome. Malignant stage: the accumulation of somatic changes ultimately precipitates the progression of one of the clones to full-blown malignancy (green clone—red integration and orange leukaemic clone). The tumour clone originates from an expanded/persistent clone yet not necessarily the most abundant one. The absence of TAX expression in the tumour clone confers a survival advantage through escape from the strong CTL immune response. the acquisition of somatic alterations in the host genome play a critical role in tumour development, we herein uncover an additional previously unrecognized yet complementary mechan- ism that contributes to leukemogenesis. We demonstrate that in tumour clones the HTLV1/BLV proviruses are integrated in the vicinity of cancer driver genes which they affect by either premature transcription interruption or antisense dependent cis-perturbation. We show that the same pattern already exists at early asymptomatic stages of infection. NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Thus, cis-perturbation of key host genes may contribute to malignant progression by providing a polyclonal background of infected cells with increased survival or proliferation. This extended half-life will promote the accumulation of further secondary mutations in the rest of the genome, ultimately precipitating the progression of one of these clones to full-blown ATL/B-cell malignancy (Fig. 6). Our results suggest that pharmacological repression of 30LTR-dependent transcription may lessen polyclonal expansion during the asymptomatic/chronic stage of the disease, thereby delaying the emergence of the tumour. underlie the observed species-speciﬁc latency time. For example, if the half-life of infected-expanding lymphocytes was more drastically increased in sheep than in human and bovine, it might explain their higher proportion among infected lymphocytes (and hence increased power to detect integration hotpots) and reduced latency time (due to the increased probability to acquire the somatic mutations needed for progression to full-blown cancer). p g Combining DNA-seq and RNA-seq data, yields a list of 674 putative leukaemia driver genes of which 370 were reported in one or several cancer-related gene lists used for enrichment analysis. Of the 50 most signiﬁcant ovine genic hotspots (Fig. 5c), 40 were reported in these lists. Six of the 10 remaining genes, despite not being listed, were genes for which literature search supported undeniable oncogenic properties (STIM2, ARHGAP24, KLF12, KMT2C, CCDC88C and DOCK11), and 4 genes were—to our knowledge—not previously reported in cancer, yet may include new candidate drivers. MARCH1, a gene that exhibits E3 ubiquitin ligase activity and was recently reported to regulate MHC class II turnover51, is an example of such a candidate. Thus, the catalogue of genes revealed by proviral integration hotspot identiﬁcation represents a potential resource of novel cancer drivers. It may be particularly attractive for the discovery of new cancer-related noncoding RNAs considering that several genes uncovered by this work belong to that class. Discussion T k (g) Virus–host chimeric transcripts uncovered by 30AS-capture RNA-seq map to genomic region upstream of intergenic hotspots consistent with antisense-dependent transcriptional activity. Absence of coverage for corresponding downstream regions. (h) Mapping of 30AS-capture RNA-seq hybrid reads (red coverage) to genomic region upstream of intergenic hotspot chr2: 242,107,500-242,240,229 reveals antisense-dependent chimeric transcription and interaction with oncogene ID3 (ref. 49). (i) Hybrid reads mapping to genomic region upstream of intergenic hotspot chr1: 175,765,639–175,927,608 (blue coverage) reveal ectopic expression of CCDC80 (not expressed in lymphocytes) consistent with a gain-of-function mechanism. NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications 10 ARTICLE NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications ARTICLE Sequencing was carried out on an Illumina MiSeq instrument with 2 150 bp reads (Reagent Kit v2). Primer sequences are available in Supplementary Data 5. Hybrid RNA-seq reads and chimeric transcripts were analysed as described above (standard RNA-seq). Hybrid reads upstream or downstream of intergenic hotspots were identiﬁed in genomic windows deﬁned as the region spanning from the hotspot extremity to the nearest protein-coding gene. HTS integration mapping and measure of clonal abundance. To identify HTLV-1 and BLV proviral integration sites we used a method similar to that outlined by Gillet9,29, but with a number of key changes to increase sensitivity and reduce costs by simplifying multiplexing24. In the case of asymptomatic samples, we used a modiﬁed method that included an extension step using a Hot start Taq Polymerase (Promega) with 25% of the dTTPs replaced with Biotin-11-dUTP (Thermo Scientiﬁc) followed by streptavidin-based selection (Dynabeads M-280 Streptavidin, Invitrogen/Life Technologies), allowing a reduced number of PCR cycles and removing the need for End Repair/dA-Tailing (15 cycles, annealing temperature of 66 C with a 30 s extension) prior to addition of Nextera XT indexes (Illumina). Primer sequences are available in Supplementary Data 5. Paired-end reads were aligned to a host–provirus hybrid genome using BWA. After quality trimming (average base quality Z30) only paired-end reads that fulﬁlled the following conditions (spanning LTR-host junctions) were retained: Read 1: BLV 5’LTR: 30 nts, BLV 3’LTR: 27nts, HTLV-1 50LTR: 29 nts, HTLV-1 30LTR: 45 nts of the read mapped to the relevant LTR extremity. Read 2: the read mapped to the host genome with r3 mismatches. Duplicates were removed based on reads that showed the same genomic insertion site and identical eight random nt tags. Read numbers were counted for each proviral integration site and reported using in-house R and Perl scripts. Clone abundance in tumours was determined as follows: if both 50 and 30LTR ﬂanking sites were identiﬁed, % ¼ average 30–50LTR. If only one ﬂanking site was detected: % ¼ % deﬁned by detected LTR ﬂanking site and provirus identiﬁed as either LTR-deleted if deletion is supported by evidence from RNA-seq hybrid read detection (non-canonical virus–host boundary) or full-length if RNA-seq-based data support the presence of LTR-host fusion reads. RNA sequencing. Total RNA was extracted using the Qiagen AllPrep DNA/RNA kit and strand-speciﬁc ribosomal RNA-depleted RNA-seq libraries were prepared using the Illumina TruSeq Total RNA stranded kit. ARTICLE The total animal tumour sample size (N ¼ 47) was deﬁned to match clinical sample numbers. No statistical test was used to determine adequate sample size. The study did not use blinding. Detailed HTLV-1 patients’ and animal samples’ information is available in Supplementary Data 1 and 2. YR2 and L267 are tumour B-cell lines derived from ovine B-cell tumours M395 (B-cell leukaemia) and T267 (B-cell lymphoma), respectively17,18. Cell cultures selected for RNA-seq were validated using FACS labelling as previously described and tested for the absence of mycoplasma contamination. their soft-clipped breakpoint was localized within a 5 nt window. Fusion reads sharing the same fusion point within a 5nt window were clustered. Fusions were classiﬁed as genic/intergenic and concordant/discordant using a combination of BEDTools 2.16.2 (ref. 62) and BEDOPS 2.30 (ref. 63). HTLV-1/BLV integration sites were determined based on identiﬁcation of chimeric transcripts that encom- passed LTR-host boundaries. Hybrid reads generated from 30AS-enriched RNA-seq data (asymptomatic samples) were additionally ﬁltered to exclude both tumour- speciﬁc chimeric reads and hybrid reads common to multiple asymptomatic samples to exclude artefacts from sample cross-contamination. RNA-seq read coverage, virus–host annotated junctions and hybrid reads were loaded onto IGV to visualize the viral and host-speciﬁc transcriptomes and assess the consequences of proviral integration/transcription on the host transcription patterns. 30AS-capture RNA-seq. Using total RNA as template, cDNA was produced with SuperScript III Reverse Transcriptase (Life technologies) following the manufacturer’s instruction and primed with an oligo-dT tailed with the Nextera reverse sequence attached (Integrated DNA Technologies). cDNA was treated with RNase H (New England Biolabs) and semi-nested PCR was carried out to enrich the BLV fusions. The ﬁrst PCR was performed using primers LTR1 (matches BLV AS exon 1) and NexRs (matches end of the Nextera-oligo dT) and Q5 High-Fidelity DNA Polymerase (New England Biolabs) with an annealing temperature of 66 C and a 4 min extension (25 cycles). In the second PCR the LTR1 primer was replaced by LTR2 and NexR was reused. The PCR product was sheared in a Bioruptor Pico (Diagenode) following the manufacturer’s instructions for fragments of B400 bp, treated with the NEBNext Ultra End Repair/dA-Tailing Module (New England Biolabs) and ligated to a Y adapter produced by annealing oligos corresponding to the Nextera forward and reverse sequences using T4 DNA Ligase (New England Biolabs). The resultant DNA was indexed with Nextera XT indexes (Illumina) and libraries were mixed in equal proportions. ARTICLE Libraries were analysed on an Agilent Bioanalyser DNA 1000 and quantiﬁed by qPCR using the KAPA kit (KAPA Biosystems). Sequencing was carried out on an Illumina HiSeq 2000 (2 100 bp paired-end reads) and generated B2 60 million raw paired-end reads per library. The RNA-seq raw reads were aligned using STAR (v2.3.1.u)53 and custom host–provirus genome references build using host reference genomes hg19 (human), UMD3.1 (bovine) or OAR3.1 (ovine), respectively, and proviral genome sequences BLV-YR2 (GenBank: KT122858) and HTLV-1-ATK-1 (GenBank: J02029), respectively. SAMtools and BAMtools were used to sort and index, and separate sense and antisense reads from the STAR output, respectively54,55. ENSEMBL v84 was used for host genome annotations and custom annotations for both HTLV-1 and BLV genomes. FeatureCounts56, R packages DESeq2 (ref. 57) and DEXSeq58 were used for read quantiﬁcation, normalization and differential gene and exon expression analysis, respectively. Integrative Genomic Viewer (IGV) was used for visualization of sequencing alignments on both the host and viral genomes59. RNA-seq reads stemming from any of the viral LTRs (HTLV-1 and BLV) systematically mapped to both the 50 and 30 LTR as both are identical in sequence. Reads were speciﬁcally assigned to one or the other LTR based on additional host-speciﬁc mapping data and fusion-read information (outlined below): upstream or downstream position of host-mapping mate pair alignment, 30AS RNA-host hybrid read identiﬁcation (antisense coverage, 30LTR alignment), host-LTR hybrid read identiﬁcation (sense coverage, 50LTR alignment), LTR-host read identiﬁcation (sense coverage, 30LTR alignment), host genomic environment (genic, intergenic) and evidence of 50LTR deletion from HTS-based integration mapping data. Identifying hotspots of proviral integration. Identical proviral integration sites (IS) across samples were removed to account for potential cross contamination between samples and the IS that showed the highest read count was retained. 50 and 30 LTR ﬂanking IS with same proviral orientation located within a 10 bp window were merged. IS distribution across the genome was ﬁrst examined by counting the number of IS in sliding consecutive bins of 100 kb (50 kb overlap) and bins showing IS numbers 499th percentile of the distribution of IS per bin (‘hotbins’) were visualized in individual chromosomes. Hotspots were then deﬁned as follows (contrary to hotbins that have a predetermined size of 100 kb, hotspots deﬁned by simulation show a range of sizes): for each IS we counted the number of IS located in a 50 kb window centred on that IS ¼ IS(i). ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 availability of sufﬁcient DNA/RNA of clinical samples available from the Necker Hospital at the time of initiation of the study (N ¼ 44). Samples consisted of peripheral blood mononuclear cells (PBMCs) from 21 acute ATLs, 3 chronic ATLs, 3 lymphoma-affected patients and 7 acute ATL patients that underwent therapy. Six samples were collected from lymphoma (lymph node and skin). Samples from asymptomatic carriers (AC) comprised three blood samples and one sample from an AC lymph node. Control samples consisted of uninfected CD4 þ T-cells. PBMCs were isolated from blood using Histopaque-1077 (Sigma). Primary leu- kaemic B-cells and lymphoid tumours from BLV-infected sheep (Suffolk and polled Dorset crossed with Arcott breeds of either sex; N ¼ 32) were collected at the acute stage of the disease (latency prior to tumour development range of 15–48 months). Sheep were housed at the Centre de Recherches Ve´te´rinaires et Agro- chimiques (Brussels, Belgium) and at the Vaccine and Infectious Disease Organi- zation (VIDO-Intervac, Saskatoon, Canada). Experimental procedures approved by the Comite´ d’Ethique Me´dicale de la Faculte´ de Me´decine, ULB were conducted in accordance with national and institutional guidelines for animal care and use. Asymptomatic sheep samples came from animals infected with the molecular clone pBLV3447 following experimental procedures approved by the University of Saskatchewan Animal Care Committee, following Canadian Council on Animal Care Guidelines (Protocol #19940212). PBMCs were isolated from EDTA-treated blood using standard Ficoll-Hypaque separation. Lymphoid tumours were minced through a nylon mesh cell strainer (Becton Dickinson) to obtain single-cell suspensions. B-cell percentages were measured by ﬂuorescence-activated cell sorting (FACS) using CD72-speciﬁc mAb DU2-104 (VMRD) and wereZ93% for all malignant samples used for further investigation. Samples from asymptomatic BLV-infected sheep were collected between 7 weeks to 34 month following experimental inoculation. Ovine control samples consisted of PBMCs of uninfected sheep assigned to the control group using physical randomization at the time of experimental inoculation with BLV (N ¼ 3). Bovine B-cell tumours (N ¼ 15) taken from our tumour collection stored at 80 C comprised samples from various geographical origins (Japan, France, United States, and Belgium). Tumour samples were collected from blood or B-cell lymphoid masses developing following natural BLV infection some of which have been characterized previously52. The latency period before tumour onset in these animals is not accurately documented. Bovine control samples consisted of PBMCs isolated from seronegative animals. Methods Samples. Samples from HTLV-1-infected individuals were collected after informed consent obtained in accordance with the Declaration of Helsinki and after institutional review board-approved protocol at the Necker Hospital, University of Paris, France in accordance with the ‘Comite´ d’e´thique Ile de France II’. We selected human ATL samples for sequencing on the basis of the In conclusion, although there is considerable evidence from previous work that the viral products—TAX and HBZ/AS—and 11 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications ARTICLE To assess the putative enrichment of abundant clones in hotspots of integration, we separated the total IS set identiﬁed in asymptomatic BLV-infected sheep (66,557 IS) in two classes (abundant IS: Ab þ and non-abundant IS: Ab-) according to the number of sequencing reads supporting each integration site. Increasing thresholds corresponding to sequencing read numbers per IS were used (six groups, from minimum 2 reads to minimum 100 reads) to assign an IS to the abundant or non-abundant class of clones. The number of Ab þ and Ab- clones located within or outside the 722 deﬁned hotspots respectively was reported for each test group and a one-tailed Fisher’s exact test was performed to assess the statistical enrichment of abundant clones in hotspots of integration. Validation of virus–host chimeric transcripts by RT–PCR. RNA was treated with TURBO DNA-free Kit and retrotranscribed using SuperScript First-Strand Synthesis System for RT–PCR (ThermoScientiﬁc) and random hexamers according to the manufacturer’s instructions. BLV-host chimeric transcripts were ampliﬁed from cDNA using combinations of a forward primer in BLV AS exon 1 and reverse primers binding downstream of the fusion breakpoint identiﬁed by RNA-seq (primers from Integrated DNA Technologies; Supplementary Data 5) and generated using Primer 3 (http://bioinfo.ut.ee/primer3/). Two microlitres cDNA was mixed with the PCR mix (4,125 ml H2O, 1,25 ml of each 2.5 nM reverse and forward primer, 0.2 ml 10 mM dNTP (Promega), 2 ml 5 Q5Reaction Buffer (New England BioLabs) and 0.1 ml Q5High-Fidelity DNA Polymerase (New England BioLabs)). PCR consisted of 35 cycles of 8 s at 98 C, 20 s at 67 C and 35 s at 72 C. Products were visualized on a 2% agarose gel and sequenced by conventional Sanger methods. Statistical assessment of host gene recurrence. The level of gene recurrence in the vicinity of provirus integration sites across tumours was assessed as follows: protein-coding genes in a 1 Mb genomic window upstream and downstream of each distinct proviral integration site (92 IS) were identiﬁed using Bedtools intersectBed tool62 with ENSEMBL v84 annotation. We calculated unweighted and weighted global recurrence scores by summing individual gene recurrence scores (unweighted score ¼ 1 regardless of the number of occurrences across the 92 IS window-based gene lists; weighted score ¼ number of occurrences of a particular gene across 92 IS window-based gene lists). ARTICLE To verify that genic hotspots were independent of gene size and gene expression level, the number of IS per gene in our data set (42,113 genic IS, corrected total number of unique genic integrations obtained after manual curation of hotspots deﬁned by simulation) was counted (IS.real(i) ¼ number of IS in gene i) and a simulation (N ¼ 100,000 iterations) was performed as follows: for each iteration IS were assigned to a subset of 10,679 genes matched for expression level with the 674 genic hotspot genes (considering bins of 100 genes centred on each hotspot gene in terms of expression level deﬁned by the average TPM across all ovine samples; ENSEMBL OAR3.1 v84: 25,197 genes), each IS having a probability to be assigned to a gene proportional to both gene size and gene expression level in ovine lymphocytes. The number of simulated IS per gene was counted at each iteration (i.e., IS.sim(i,j) ¼ number of simulated IS in each iteration j), and a P-value that reﬂects the independence of the number of IS and both gene size and gene expression level was computed for each gene and N ¼ 100,000 iterations comparing IS.real to the distribution of simulated IS.sim, i.e., for gene i : p(i) ¼ (# IS.sim(i,j)ZIS.real(i))/100,000. The proportion of genic hotspots not explainable by gene size and expression level alone (p11) was estimated on the set of nominal P-values obtained for the 674 genic hotspots64. RNA-seq based gene expression analysis. Gene and exon-speciﬁc read counts were processed according to a two-pass normalization step. First read counts were normalized to sequencing depth using the R package-implemented DESeq2. Normalized read counts of the fusion-affected host gene (or speciﬁc exons) in a particular tumour sample were divided by the mean normalized read count of this particular gene (or corresponding exons) in all remaining samples of the same host that do not have proviral integration in that locus. This resulted in an average expression value of 1 for the control tumour samples and a normalized expression value for the tumour containing the fusion-affected gene. Statistical signiﬁcance of expression levels was assessed using Mann–Whitney U-tests. Testing enrichment of abundant infected clones in hotspots. ARTICLE The provirus integration orientation relative to host gene transcription (concordant/discordant) was determined for hotspots that showed a signiﬁcant orientation bias. To measure the non-randomness of provirus orientation, we also determined the relative orientation and distance of each possible IS pair across all (i) genic and (ii) intergenic IS identiﬁed, grouped the results by distance in 100 bp increment bins and computed the orientation ratio (same/opposite) relative to each bin. Cancer driver gene enrichment analysis. Gene sets were tested for cancer-driver enrichment by calculating a cancer driver enrichment score (ES) using seven publicly available cancer-driver gene lists (CGL, Supplementary Table 2 and Supplementary Fig. 2). Each gene in a CGL is assigned a score that rates its cancer-driver potential. Observed scores were compared to simulated scores obtained from N ¼ 100,000 size-matched random or expression-matched gene sets, including information about paralogs (Random Para, Expr Para) or not (Random, Expr). Simulated expression-matched gene lists were generated as described for assessment of gene recurrence. Direct and paralog ES were obtained by summing each CGL-associated gene score (direct-score), or associated gene’s paralog score (paralog-score), respectively. Paralog scores were calculated by multiplying the direct score of the paralog gene with the percentage amino acid identity gene/paralog gene using the Paralog Mapping Table (ENSEMBL). We assessed enrichment by P-value by counting the frequency of simulated ES equal to or higher than the observed gene ES divided by the number of simulation iterations (N ¼ 100,000). In addition, we calculated a global enrichment score (‘meta-analysis’) that incorporated all seven CGL scores by summing the seven CGL-associated P-values and comparing this meta-P-value to N ¼ 100,000 simulated gene list meta-P-values. This resulted in ﬁve empirical ‘global P-values’. Score and P-value calculations, as well as simulations, were conducted using R 3.1.1. ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 that most closely matched the listed genes expression level based on average TPM computed using RSEM65 across all available human leukaemia samples. Additional scores were computed based on the recurrence of paralog genes. We assessed recurrence by P-value counting the frequency of simulated recurrence scores equal to or higher than the observed tumour/asymptomatic recurrence scores, divided by the number of simulated gene lists (N ¼ 100,000). Gene recurrence between genic hotspots of proviral integration in asymptomatic sheep samples and provirus-interacting host genes identiﬁed in either asymptomatic or tumour samples, as well as HTLV-1 genic integrations retrieved from the public RID database50, was assessed based on the same simulation method. The level of recurrence between asymptomatic individual samples was computed based on the same method testing unweighted and weighted scores (as deﬁned above) to N ¼ 100,000 size-matched random or expression level-matched simulated gene list scores. extreme IS ±5 kb determining the boundaries of the hotspot. Hotspots were deﬁned as genic if the median of the IS in the hotspot fell within a gene. Intergenic hotspots deﬁned by simulation were then manually curated to account for poor annotation of the ovine transcriptome (ENSEMBL OAR3.1 v84) and re-assigned to the genic class if they fell within transcribed regions (i.e., non annotated exon, 50 transcribed region of a mis-annotated gene). This was achieved by visual examination of each predicted hotspot (IS positions and orientation) in IGV combined with the corresponding RNA-seq alignments (RNA-seq data of ovine primary B cells). Predicted hotspots that overlapped multiple genes or novel unannotated transcripts were discarded for further analysis in this study. Of the genes uncovered in the genic hotspot class after curation, 618 showed evidence of 1 hotspot while 56 carried multiple hotspots (2–4). Multiple hotspots in the same gene were merged. This resulted in a curated list of 674 genic and 48 intergenic hotspots. Besides genic and intergenic types, we identiﬁed hotspots that combined genic and upstream intergenic integrations for the same gene. For each given hotspot a proviral orientation ratio was calculated by counting the number of proviruses integrated in the same orientation relative to the total number of IS within that hotspot, considering the predominant orientation. Hotspots signiﬁcantly biased towards non-random orientation were deﬁned as these associated with an FDRo0.1 (one-tailed binomial test). NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications ARTICLE We then randomly picked 66,557 IS in the ovine genome, performed the same counting procedure and retained the simulated IS window harbouring the highest number of IS ¼ max.simIS. We performed N ¼ 1,000,000 iterations of this procedure to generate a distribution reﬂecting random integration. We then assigned a P-value to each IS: P-value IS(i) ¼ # max simIS4 ¼ IS(i)/1,000,000. Hotspots were deﬁned by grouping consecutive IS windows that showed a P-value r0.05, the position of the two Detecting hybrid RNAs and insertion sites from RNA-seq data. Mispaired and soft-clipped reads supporting virus–host hybrid transcripts were identiﬁed using a custom two-pass alignment scheme as described in ViralFusionSeq60. RNA-seq paired-end reads were aligned to the host genome using BWA61 (default parameter except -k 19 and -L 1) and mispaired and soft-clipped reads (minimum 8 soft-clipped nts) were re-aligned to the proviral genome. Host and proviral alignments of each read were compared, and reads were ﬂagged as fusion reads if one read mapped to the host genome and the other to the proviral genome, or if 12 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 Testing statistical power of hotspot and cancer driver analysis. Assuming 20,000 genes of which y are cancer drivers (500–3,000) and z is their fraction present in cancer driver lists (25–100%), we ﬁrst randomly assigned x integration sites (IS, from 1,000 to 100,000) to G ¼ 20,000 genes and reported the maximum number of IS across all genes (max_hits_R) and the sum of IS across all cancer-driver genes (cancerDriverHits_R). We performed N ¼ 100,000 iterations resulting in two N ¼ 100,000-element vectors representing non-preferential integration (hotspot detection ¼ max_hits_R and cancer-driver enrichment ¼ cancerDriverHits_R). We then assumed the presence of two types of IS, w corre- sponding to expanded clones (from 0.01 to 100% of the x IS, integrated in cancer drivers) and 1 w to infected but non-expanded clones. We randomly assigned w IS to a fraction z of y cancer-driver genes and 1 w IS to G ¼ 20,000 genes and reported (i) the sum of IS across the y cancer-driver genes (cancerDriverHits_Exp) and (ii) the maximum number of IS across all G ¼ 20,000 genes (max_hits_Exp) for N ¼ 100,000. 16. Matsuoka, M. & Yasunaga, J. I. Human T-cell leukemia virus type 1: replication, proliferation and propagation by Tax and HTLV-1 bZIP factor. Curr. Opin. Virol. 3, 684–691 (2013). p 17. Van den Broeke, A. et al. Even transcriptionally competent proviruses are silent in bovine leukemia virus-induced sheep tumor cells. Proc. Natl Acad. Sci. USA 85, 9263–9267 (1988). 18. Merimi, M. et al. Complete suppression of viral gene expression is associated with the onset and progression of lymphoid malignancy: observations in Bovine Leukemia Virus-infected sheep. Retrovirology 4, 51 (2007). 19. Rose, N. J., Richardson, J. H., Desselberger, U. & Lever, A. M. L. Virus inactivation in a proportion of human T-cell leukaemia virus type I infected T-cell clones arises through naturally occurring mutations. J. Gen. Virol. 81, 97–104 (2000). 20. Takeda, S. et al. Genetic and epigenetic inactivation of tax gene in adult T-cell leukemia cells. Int. J. Cancer 109, 559–567 (2004). f max_hits_Exp and cancerDriverHits_Exp were calculated for all P-values of max_hits_Exp and cancerDriverHits_Exp were calculated for all parameter combinations: Pvaluehotspot ¼ PN¼100;000 i¼1 ðmax hits Expmax hits RÞ N¼100;000 and PvalueCancerDriverEnrichment ¼ PN¼100;000 i¼1 ðcancerDriverHits ExpcancerDriverHits RÞ N¼100;000 , respectively. References 1. Gessain, A. & Cassar, O. Epidemiological aspects and world distribution of HTLV-1 infection. Front. Microbiol. 3, 388 (2012). g gy 31. Lau, C. C. et al. Viral-human chimeric transcript predisposes risk to liver cancer development and progression. Cancer Cell 25, 335–349 (2014). 2. Ishitsuka, K. & Tamura, K. Human T-cell leukaemia virus type I and adult T-cell leukaemia-lymphoma. Lancet Oncol. 15, e517–e526 (2014). 32. Tamiya, S. et al. Two types of defective human T-lymphotropic virus type I provirus in adult T-cell leukemia. Blood 88, 3065–3073 (1996). 3. Burny, A. et al. in Viruses and Cancer (eds Minson, A. C., Neil, J. C. & McRae, M. A.) 213–234 (Cambridge University Press, 1994). & McRae, M. A.) 213–234 (Cambridge University Press, 1994 p 33. Fishel, R. et al. The human mutator gene homolog MSH2 and its 33. Fishel, R. et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 75, 1027–1038 (1993). 4. Gillet, N. et al. Mechanisms of leukemogenesis induced by bovine leukemia virus: prospects for novel anti-retroviral therapies in human. Retrovirology 4, 18 (2007). association with hereditary nonpolyposis colon cancer. Cell 75, 1027–1038 (1993). 34. Flores-Martin, J., Rena, V., Angeletti, S., Panzetta-Dutari, G. M. & Genti-Raimondi, S. The lipid transfer protein StarD7: structure, function, and regulation. Int. J. Mol. Sci. 14, 6170–6186 (2013). 5. Bartlett, P. C. et al. Options for the control of bovine leukemia virus in dairy cattle. J. Am. Vet. Med. Assoc. 244, 914–922 (2014). 6. Willems, L. et al. In vivo transfection of bovine leukemia provirus into sheep. Virology 189, 775–777 (1992). 35. Huang, D. et al. BRCC3 mutations in myeloid neoplasms. Haematologica 100, 1051–1057 (2015). 7. Van den Broeke, A. et al. Cytotoxic responses to BLV tax oncoprotein do not prevent leukemogenesis in sheep. Leuk. Res. 34, 1663–1669 (2010). 36. Yang, L., Rau, R. & Goodell, M. A. DNMT3A in haematological malignancies. Nat. Rev. Cancer 15, 152–165 (2015). 8. Bangham, C. R. M., Cook, L. B. & Melamed, A. HTLV-1 clonality in adult T-cell leukaemia and non-malignant HTLV-1 infection. Semin. Cancer Biol. 26, 89–98 (2014). 37. Hata, A., Lo, R. S., Wotton, D., Lagna, G. & Massague´, J. Mutations increasing autoinhibition inactivate tumour suppressors Smad2 and Smad4. Nature 388, 82–87 (1997). 9. Gillet, N. A. et al. The host genomic environment of the provirus determines the abundance of HTLV-1-infected T-cell clones. Blood 117, 3113–3122 (2011). 38. ARTICLE We performed T ¼ 10,000 iterations and computed the power of hotspot detection and cancer-driver enrichment as powerhotspot ¼ PT¼10;000 i¼1 ðPvaluehotspotPlimitÞ T¼10;000 and powercancerDriverEnrichment ¼ PT¼10;000 i¼1 ðPvalueCancerDriverEnrichmentPlimitÞ T¼10;000 for Plimit ¼ 0:01. P-values of max_hits_Exp and cancerDriverHits_Exp were calcula 21. Bangham, C. R. M. et al. The immune control of HTLV-1 infection: selection forces and dynamics. Front. Biosci. 14, 2889–2903 (2009). 22. Satou, Y. et al. HTLV-1 bZIP factor induces T-cell lymphoma and systemic inﬂammation in vivo. PLoS Pathogens 7, e1001274 (2011). 23. Mitobe, Y., Yasunaga, J., Furuta, R. & Matsuoka, M. HTLV-1 bZIP factor RNA and protein impart distinct functions on T-cell proliferation and survival. Cancer Res. 75, 4143–4152 (2015). 24. Durkin, K. et al. Characterization of novel Bovine Leukemia Virus (BLV) antisense transcripts by deep sequencing reveals constitutive expression in tumors and transcriptional interaction with viral microRNAs. Retrovirology 13, 33 (2016). Statistical analyses. Analyses of signiﬁcance for RNA-seq-based gene expression were performed using two-sided Mann–Whitney U-tests implemented in R 3.1.1, assuming equal variances. One-tailed Fisher’s exact-tests were used to assess recurrence between gene groups and the statistical enrichment of abundant clones in hotspots of integration. Values of Po0.05 were considered as statistically signiﬁcant. Continuous biological variables were assumed not to follow a normal distribution. 25. Rosewick, N. et al. Deep sequencing reveals abundant noncanonical retroviral microRNAs in B-cell leukemia/lymphoma. Proc. Natl Acad. Sci. USA 110, 2306–2311 (2013). 26. Schnurr, M. W. et al. Nonrandom chromosomal abnormalities in bovine lymphoma. Leuk. Res. 18, 91–99 (1994). 27. Dequiedt, F., Kettmann, R., Burny, A. & Willems, L. Mutations in the p53 tumor-suppressor gene are frequently associated with Bovine Leukemia Virus-induced leukemogenesis in cattle but not in sheep. Virology 209 ð1995Þ: Data availability. Sequence data that support the ﬁndings of this study have been deposited in the European Nucleotide Archive (ENA) hosted by the European Bioinformatics Institute (EMBL-EBI) and are accessible through study accession number PRJEB19394. All other relevant data are available within the article and its Supplementary Information ﬁles or from the corresponding author upon reasonable request. ð Þ 28. Melamed, A. et al. Genome-wide determinants of proviral targeting, clonal abundance and expression in natural HTLV-1 infection. PLoS Pathogens 9, e1003271 (2013). 29. Gillet, N. A. et al. Massive depletion of bovine leukemia virus proviral clones located in genomic transcriptionally active sites during primary infection. PLoS Pathogens 9, e1003687 (2013). g 30. Sokol, M., Wabl, M., Ruiz, I. R. ARTICLE & Pedersen, F. S. Novel principles of gamma-retroviral insertional transcription activation in murine leukemia virus-induced end-stage tumors. Retrovirology 11, 36 (2014). ARTICLE Observed scores were tested against N ¼ 100,000 simulated recurrence scores obtained from 92 random sets of adjacent genes of same size distribution. We assessed recurrence by P-value counting the frequency of simulated recurrence scores equal to or higher than the observed tumour provirus window-speciﬁc recurrence scores, divided by the number of simulated gene lists (N ¼ 100,000). Score calculation and simulations were conducted using R 3.1.1. ( g ) m Q g y y (New England BioLabs)). PCR consisted of 35 cycles of 8 s at 98 C, 20 s at 67 C and 35 s at 72 C. Products were visualized on a 2% agarose gel and sequenced by conventional Sanger methods. Proviral load quantiﬁcation. DNA was isolated using the Qiagen AllPrep DNA/ RNA/miRNA kit and proviral DNA was quantiﬁed by real-time PCR using primers targeting either the BLV or HTLV-1 30 region and RPS9 or Actin, respectively, for normalization (primers from Integrated DNA Technologies; Supplementary Data 5). Runs were performed in a 50 ml volume containing 1 mg of total DNA, primers and probe (200 nM concentration of each) in 1 PCR buffer (Platinum Quanti- tative PCR SuperMix-UDG) (HTLV-1) or 10 ml containing 50 ng DNA and 1 Universal PCR Master Mix, No AmpErase UNGa (ThermoScientiﬁc) (BLV). Thermocycling conditions were 10 min at 95 C, followed by 50 cycles at 95 C for 15 s and 60 C for 1 min. Standard curves were generated using serial dilutions of DNA from the YR2 cell line (BLV, two proviral copies) or the Tarl2 cell line (HTLV-1, single proviral copy). Proviral load in % PBMCs ¼ (Sample Average Quantity) 2/(Sample RPS9 or Actin quantity) 100. The YR2 chromosome that carries the BLV provirus integration appears to be duplicated. Gene recurrence between viral transcript-interacting host genes identiﬁed in asymptomatic sheep and tumour samples was assessed as follows: we calculated a gene recurrence score by counting the number of overlapping genes between the 82-gene tumour gene list and the 723-gene asymptomatic gene set and tested the statistical signiﬁcance of the overlap by simulation based on N ¼ 100,000 recurrence scores obtained from 82 and 723 random or expression level matched simulated gene lists, respectively. Simulated expression-matched gene lists were generated according to expression bins that each consisted of a group of 500 genes 13 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications Acknowledgements This work was supported by the Fonds de la Recherche Scientiﬁque (FRS), Te´le´vie, les Amis de l0Institut Bordet, the International Brachet Stiftung (IBS), the Fondation Lambeau-Marteaux, the Ligue contre le cancer, l’Institut National du Cancer (INCA), the Cance´ropole d’ıˆle de France and a Te´le´vie Grant to V.H. M.A. holds a Postdoctoral Researcher fellowship of the FRS, A.M. is recipient of a doctoral fellowship from the Institut National du Cancer (INCA), N.R. and K.D. are Scientiﬁc Research Worker of Te´le´vie. We thank Wouter Coppieters, Latifa Karim and the GIGA Genomics Platform for sequencing services and support, Franck Mortreux (ENS Lyon, France) for providing samples and cell lines used in this work, Charles Bangham (Imperial College, London, UK) for providing the Tarl2 cell line and Dominique Bron (Institut Jules Bordet, ULB) for comments on the manuscript. This work was supported by the Fonds de la Recherche Scientiﬁque (FRS), Te´le´vie, les Amis de l0Institut Bordet, the International Brachet Stiftung (IBS), the Fondation Lambeau-Marteaux, the Ligue contre le cancer, l’Institut National du Cancer (INCA), the Cance´ropole d’ıˆle de France and a Te´le´vie Grant to V.H. M.A. holds a Postdoctoral Researcher fellowship of the FRS, A.M. is recipient of a doctoral fellowship from the Institut National du Cancer (INCA), N.R. and K.D. are Scientiﬁc Research Worker of Te´le´vie. We thank Wouter Coppieters, Latifa Karim and the GIGA Genomics Platform for sequencing services and support, Franck Mortreux (ENS Lyon, France) for providing samples and cell lines used in this work, Charles Bangham (Imperial College, London, UK) for providing the Tarl2 cell line and Dominique Bron (Institut Jules Bordet, ULB) for comments on the manuscript. 48. Yuan, T. L. & Cantley, L. C. PI3K pathway alterations in cancer: variations on a theme. Oncogene 27, 5497–5510 (2008). g 49. Lasorella, A., Benezra, R. & Iavarone, A. The ID proteins: master regulators of cancer stem cells and tumour aggressiveness. Nat. Rev. Cancer 14, 77–91 (2014). 50. Shao, W. et al. Retrovirus Integration Database (RID): a public database for retroviral insertion sites into host genomes. Retrovirology 13, 47 (2016). 51. Cho, K.-J., Walseng, E., Ishido, S. & Roche, P. A. Ubiquitination by March-I prevents MHC class II recycling and promotes MHC class II turnover in antigen-presenting cells. Proc. Natl Acad. Sci. USA 112, 10449–10454 (2015). 52. Kettmann, R. et al. ARTICLE ARTICLE NATURE COMMUNICATIONS \| DOI: 10.1038/ncomms15264 44. Temiz, N. A. et al. RNA sequencing of Sleeping Beauty transposon-induced tumors detect transposon-RNA fusions allowing precision analyses of forward genetic cancer screens. Genome Res. 1, 119–129 (2016). 68. Liu, S., Nheu, T., Luwor, R., Nicholson, S. E. & Zhu, H.-J. SPSB1, a novel negative regulator of the transforming growth factor-b signaling pathway targeting the type II receptor. J. Biol. Chem. 290, 17894–17908 (2015). g 45. Sausen, M. et al. Integrated genomic analyses identify ARID1A and ARID1B alterations in the childhood cancer neuroblastoma. Nat. Genet. 45, 12–17 (2012). 46. Paolino, M. et al. The E3 ligase Cbl-b and TAM receptors regulate cancer metastasis via natural killer cells. Nature 507, 508–512 (2014). 45. Sausen, M. et al. Integrated genomic analyses identify ARID1A and ARID1B alterations in the childhood cancer neuroblastoma. Nat. Genet. 45, 12–17 (2012). 69. Xue, H., Wang, Z., Chen, J., Yang, Z. & Tang, J. Knockdown of reticulon 4C by lentivirus inhibits human colorectal cancer cell growth. Mol. Med. Rep. 12, 2063–2067 (2015). 46. Paolino, M. et al. The E3 ligase Cbl-b and TAM receptors regulate cancer metastasis via natural killer cells. Nature 507, 508–512 (2014). metastasis via natural killer cells. Nature 507, 508–512 (2014) 47. Song, M. S., Salmena, L. & Pandolﬁ, P. P. The functions and regulation of the PTEN tumour suppressor. Nat. Rev. Mol. Cell Biol. 13, 283 (2012). Acknowledgements Leukemogenesis by bovine leukemia virus: proviral DNA integration and lack of RNA expression of viral long terminal repeat and 30 proximate cellular sequences. Proc. Natl Acad. Sci. USA 79, 2465–2469 (1982). Author contributions K.D. performed HTS library preparations and sequencing validations, carried out all sample extractions and analysed the data. M.A. carried out HTS-based integration mapping and clonality analyses, V.H. determined proviral loads in animals and RT–PCR validations, N.R. processed the sequencing data and performed bioinformatics analyses, A.M. and O.H. collected clinical data and provided patient materials, V.A. determined patients’ proviral loads, P.G. and N.A. collected and provided animal samples, A.B. and C.C. contributed to data analysis and review of the manuscript, N.R., K.D., M.G. and A.V. performed data analyses and generated the text and the ﬁgures, A.V. and M.G. designed and supervised the study. All authors contributed to the ﬁnal manuscript. 53. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013). 54. Li, H. et al. The sequence alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009). 55. Barnett, D. W., Garrison, E. K., Quinlan, A. R., Stro¨mberg,, M. P. & Marth, G. T. Bamtools: a C þ þ API and toolkit for analyzing and managing BAM ﬁles. Bioinformatics 27, 1691–1692 (2011). f 56. Liao, Y., Smyth, G. K. & Shi, W. featureCounts: an efﬁcient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923–930 (2014). r The Author(s) 2017 Additional information 57. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). 58. Anders, S., Reyes, A. & Huber, W. Detecting differential usage of exons from 57. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). Supplementary Information accompanies this paper at http://www.nature.com/ naturecommunications Supplementary Information accompanies this paper at http://www.nature.com/ naturecommunications 58. Anders, S., Reyes, A. & Huber, W. Detecting differential usage of exons from RNA-seq data. Genome Res. 22, 2008–2017 (2012). Competing interests: The authors declare no competing ﬁnancial interests. Competing interests: The authors declare no competing ﬁnancial interests. 59. Thorvaldsdottir, H. et al. Integrative Genomics Viewer (IGV): high- performance genomics data visualization and exploration. Brief. Bioinform. 14, 178–192 (2013). Reprints and permission information is available online at http://npg.nature.com/ reprintsandpermissions/ 60. Li, J.-W. et al. ViralFusionSeq: accurately discover viral integration events and reconstruct fusion transcripts at single-base resolution. Bioinformatics 29, 649–651 (2013). How to cite this article: Rosewick, N. et al. Cis-perturbation of cancer drivers by the HTLV-1/BLV proviruses is an early determinant of leukemogenesis. Nat. Commun. 8, 15264 doi: 10.1038/ncomms15264 (2017). 61. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows- Wheeler transform. Bioinformatics 25, 1754–1760 (2009). Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. 62. Quinlan, A. R. & Hall, I. M. BEDTools: a ﬂexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010). 63. Neph, S. et al. BEDOPS: high-performance genomic feature operations. Bioinformatics 28, 1919–1920 (2012). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ References Narducci, M. G. et al. Regulation of TCL1 expression in B- and T-cell lymphomas and reactive lymphoid tissues. Cancer Res. 60, 2095–2100 ð2000Þ: 10. Cook, L. B. et al. The role of HTLV-1 clonality, proviral structure, and genomic integration site in adult T-cell leukemia/lymphoma. Blood 123, 3925–3931 (2014). 39. Dayyani, F. et al. Loss of TLE1 and TLE4 from the del(9q) commonly deleted region in AML cooperates with AML1-ETO to affect myeloid cell proliferation and survival. Blood 111, 4338–4347 (2008). 11. Kataoka, K. et al. Integrated molecular analysis of adult T cell leukemia/ lymphoma. Nat. Genet. 47, 1304–1315 (2015). 40. Cho, J.-Y. et al. Isoforms of the Ets transcription factor NERF/ELF-2 physically interact with AML1 and mediate opposing effects on AML1-mediated transcription of the B cell-speciﬁc blk gene. J. Biol. Chem. 279, 19512–19522 (2004). 12. Murakami, H. et al. Bovine leukemia virus integration site selection in cattle that develop leukemia. Virus Res. 156, 107–112 (2011). 13. Pozzatti, R., Vogel, J. & Jay, G. The human T-lymphotropic virus type I tax gene can cooperate with the ras oncogene to induce neoplastic transformation of cells. Mol. Cell. Biol. 10, 413–417 (1990). 41. Tang, W. et al. A genome-wide RNAi screen for Wnt/beta-catenin pathway components identiﬁes unexpected roles for TCF transcription factors in cancer. Proc. Natl Acad. Sci. USA 105, 9697–9702 (2008). 14. Willems, L. et al. Cooperation between bovine leukaemia virus transactivator protein and Ha-ras oncogene product in cellular transformation. EMBO J. 9, 1577–1581 (1990). 42. Koso, H. et al. Identiﬁcation of FoxR2 as an oncogene in medulloblastoma. Cancer Res. 74, 2351–2361 (2014). 15. Klener, P. et al. Insights into gene expression changes impacting B-cell transformation: cross-species microarray analysis of bovine leukemia virus tax-responsive genes in ovine B cells. J. Virol. 80, 1922–1938 (2006). 43. Li, X. et al. FOXR2 Interacts with MYC to promote its transcriptional activities and tumorigenesis. Cell Rep. 16, 487–497 (2016). 14 f 64. Storey, J. D. A direct approach to false discovery rates. J. R. Stat. Soc. Ser. B 64, 479–498 (2002). 65. Li, B. & Dewey, C. N. RSEM: accurate transcript quantiﬁcation from RNA-Seq data with or without a reference genome. BMC Bioinform. 12, 323 (2011). g f 66. Sancak, Y. et al. The Rag GTPases bind raptor and mediate amino acid i li t TORC1 S i 320 1496 1501 (2008) 66. Sancak, Y. et al. The Rag GTPases bind raptor and mediate amino acid signaling to mTORC1. Science 320, 1496–1501 (2008). 66. Sancak, Y. et al. The Rag GTPases bind raptor and media signaling to mTORC1. Science 320, 1496–1501 (2008). 67. Russell, S. E. H. & Hall, P. A. Do septins have a role in cancer? Br. J. Cancer 93, 499–503 (2005). r The Author(s) 2017 15 15 NATURE COMMUNICATIONS \| 8:15264 \| DOI: 10.1038/ncomms15264 \| www.nature.com/naturecommunications
https://openalex.org/W3215565758	https://www.nature.com/articles/s41467-021-27258-9.pdf	English	null	Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits	Nature communications	2,021	cc-by	24,584	ARTICLE ARTICLE Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits The effects are allocated into groups (1, …, Φ). Each group has a set of model parameters Θφ ¼ fβφ; πφ; σ2 Gφg, with βφ as a pφ × 1 vector of partial regression coefﬁcients, where βφj is the effect of a 1 SD where there is a single intercept term 1μ and a single error term, a vector (N × 1) of residuals ϵ, with ϵjσ2 ϵ N 0Iσ2 ϵ . An N by p matrix of single nucleotide polymorphism (SNP) genetic markers, centred and scaled to unit variance, which we denote as Xφ. The effects are allocated into groups (1, …, Φ). Each group has a set of model parameters Θφ ¼ fβφ; πφ; σ2 Gφg, with βφ as a pφ × 1 vector of partial regression coefﬁcients, where βφj is the effect of a 1 SD where there is a single intercept term 1μ and a single error term, a vector (N × 1) of residuals ϵ, with ϵjσ2 ϵ N 0Iσ2 ϵ . An N by p matrix of single nucleotide polymorphism (SNP) genetic markers, centred and scaled to unit variance, which we denote as Xφ. The effects are allocated into groups (1, …, Φ). Each group has a set of model parameters Θφ ¼ fβφ; πφ; σ2 Gφg, with βφ as a pφ × 1 vector of partial regression coefﬁcients, where βφj is the effect of a 1 SD change in the jth covariate within the φth group. The spike and slab prior, contains what is called a Dirac spike14,15 for βφ, which induces sparsity in the model through a Dirac-delta at zero, excluding variables from the model by setting their coefﬁcients to zero. A ﬁnite scale mixture of normal distributions centred at zero constitute the slab component. The slab shrinks the non-zero coefﬁcients towards zero according to the slab’s width, and by having a scale mixture of Gaussians, the distribution has heavier tails and can accommodate big and small effects16. Therefore, each βφj is distributed according to: g Furthermore, statistical inference usually follows a multi-step approach. Current mixed-linear association models such as those implemented in the software fastGWA7, BoltLMM8 and REGENIE9, use a two-step approach, ﬁrst estimating the variance contributed by the SNP markers without the use of MAF-LD- annotation information, and then estimating the marker effect sizes one-by-one as ﬁxed effects in a second step7,8,10. Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits Marion Patxot 1,11, Daniel Trejo Banos 1,11, Athanasios Kousathanas 1,11, Etienne J. Orliac2, Sven E. Ojavee 1, Gerhard Moser 3, Alexander Holloway1, Julia Sidorenko 4, Zoltan Kutalik 1,5,6, Reedik Mägi7, Peter M. Visscher 4, Lars Rönnegård 8,9 & Matthew R. Robinson 10✉ Marion Patxot 1,11, Daniel Trejo Banos 1,11, Athanasios Kousathanas 1,11, Etienne J. Orliac2, Sven E. Ojavee 1, Gerhard Moser 3, Alexander Holloway1, Julia Sidorenko 4, Zoltan Kutalik 1,5,6, Reedik Mägi7, Peter M. Visscher 4, Lars Rönnegård 8,9 & Matthew R. Robinson 10✉ We develop a Bayesian model (BayesRR-RC) that provides robust SNP-heritability estima- tion, an alternative to marker discovery, and accurate genomic prediction, taking 22 seconds per iteration to estimate 8.4 million SNP-effects and 78 SNP-heritability parameters in the UK Biobank. We ﬁnd that only ≤10% of the genetic variation captured for height, body mass index, cardiovascular disease, and type 2 diabetes is attributable to proximal regulatory regions within 10kb upstream of genes, while 12-25% is attributed to coding regions, 32–44% to introns, and 22-28% to distal 10-500kb upstream regions. Up to 24% of all cis and coding regions of each chromosome are associated with each trait, with over 3,100 independent exonic and intronic regions and over 5,400 independent regulatory regions having ≥95% probability of contributing ≥0.001% to the genetic variance of these four traits. Our open- source software (GMRM) provides a scalable alternative to current approaches for biobank data. 1 Department of Computational Biology, University of Lausanne, Lausanne, Switzerland. 2 Scientiﬁc Computing and Research Support Unit, University of Lausanne, Lausanne, Switzerland. 3 Australian Agricultural Company Limited, Brisbane, QLD, Australia. 4 Institute for Molecular Bioscience, University of Queensland, Brisbane, QLD, Australia. 5 University Center for Primary Care and Public Health, Lausanne, Switzerland. 6 Swiss Institute of Bioinformatics, Lausanne, Switzerland. 7 Estonian Genome Center, Institute of Genomics, University of Tartu, Tartu, Estonia. 8 School of Technology and Business Studies, Dalarna University, Falun, Sweden. 9 Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Uppsala, Sweden. 10 Institute of Science and Technology Austria, Klosterneuburg, Austria. 11These authors contributed equally: Marion Patxot, Daniel Trejo Banos, Athanasios Kousathanas. ✉email: matthew.robinson@ist.ac.at 1 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 A mixture priors, improving on the formulations of refs. 11–13. Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits CLφφ 2 6664 3 7775; with σ2 Gφ the phenotypic variance associated with the SNPs in group φ, which, like all the other parameters, is estimated directly from the data. Here, we use 78 MAF-LD-annotation SNP marker groups. SNPs are partitioned into seven location annotations preferentially to coding (exonic) regions ﬁrst, then to intronic regions, then to 1 kb upstream regions, then to 1–10 kb regions, then to 10–500 kb regions, then to 500–1 Mb regions. Remaining SNPs were grouped in a category labelled “others" and also included in the model so that variance is partitioned relative to these also. Thus, we assigned SNPs to their closest upstream region, for example if a SNP is 1 kb upstream of gene X, but also 10–500 kb upstream of gene Y and 5 kb downstream for gene Z, then it was assigned to be a 1 kb region SNP. This ensures that SNPs 10–500 kb and 500 kb–1 Mb upstream are distal from any known gene. We further partition upstream regions to experi- mentally validated promoters, transcription factor binding sites (tfbs) and enhancers (enh) using the HACER, snp2tfbs databases with σ2 Gφ the phenotypic variance associated with the SNPs in group φ, which, like all the other parameters, is estimated directly from the data. Here, we use 78 MAF-LD-annotation SNP marker groups. SNPs are partitioned into seven location annotations preferentially to coding (exonic) regions ﬁrst, then to intronic regions, then to 1 kb upstream regions, then to 1–10 kb regions, then to 10–500 kb regions, then to 500–1 Mb regions. Remaining SNPs were grouped in a category labelled “others" and also included in the model so that variance is partitioned relative to these also. Thus, we assigned SNPs to their closest upstream region, for example if a SNP is 1 kb upstream of gene X, but also 10–500 kb upstream of gene Y and 5 kb downstream for gene Z, then it was assigned to be a 1 kb region SNP. This ensures that SNPs 10–500 kb and 500 kb–1 Mb upstream are distal from any known gene. Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits We further partition upstream regions to experi- mentally validated promoters, transcription factor binding sites (tfbs) and enhancers (enh) using the HACER, snp2tfbs databases Here, we outline the fastest Bayesian penalised regression model to date, with a hybrid-parallel algorithm for analysing large-scale genomic biobank using a single command-line tool implemented in our grouped mixture regressions model (GMRM) software. We validate our approach in large-scale simulation study and provide an empirical example using four traits measured in both the UK Biobank and Estonian Biobank data. Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits Like these former methods, we consider a spike probability at zero (Dirac delta function), and a scale mixture of Gaussian dis- tributions as a slab probability density. Unlike these models, we have genetic markers grouped into MAF-LD-annotation speciﬁc sets, with independent hyper-parameters for the phenotypic variance attributable to each group, so that the mixture propor- tions, the variance explained by the SNP markers, and the mix- ture constants are all unique and independent across SNP marker groups. This enables estimation of the phenotypic variance attributable to the group-speciﬁc effects, and differences in the underlying distribution of the βφ effects among MAF-LD- annotation groups, with different degrees of sparsity. Assuming N individuals and p genetic markers, our model of an observed phenotype vector y is: A s whole-genomes are collected for hundreds of thousands of individuals, we require regression methods that are not only computationally efﬁcient, but which also provide improved inference. Rather than relying on subsets of the SNPs, methods should fully utilise the data, exploiting computational power to facilitate discovery of additional genomic regions, to improve understanding of the genomic architecture of common disease, and to provide more informative genomic prediction. p g p For example, when estimating the proportion of phenotypic variance attributable to different categories of genetic markers (the SNP-heritability, h2 SNP of a genomic region), recent studies1–4 highlight the importance of accounting for minor allele frequency (MAF) and LD structure of the genomic data. Generally, assess- ment of the relative contribution of different genomic regions is currently made assuming that markers within a category all contribute to the variance, with enrichment deﬁned as the esti- mated share of the variance explained divided by its expected share5,6. However ideally, the estimated distribution of marker effects for each category would be directly obtained, accounting for MAF and LD structure and allowing for some of the marker effects to be zero, as this would yield a better understanding of the polygenicity of genomic effects across different genomic anno- tation groups. y ¼ 1μ þ ∑ Φ φ¼1 Xφβφ þ ϵ; ð1Þ ð1Þ where there is a single intercept term 1μ and a single error term, a vector (N × 1) of residuals ϵ, with ϵjσ2 ϵ N 0Iσ2 ϵ . An N by p matrix of single nucleotide polymorphism (SNP) genetic markers, centred and scaled to unit variance, which we denote as Xφ. Probabilistic inference of the genetic architecture underlying functional enrichment of complex traits Following this initial mixed-model association step, statistical inference (variance components, ﬁne mapping and risk prediction) is then typically conducted on the summary statistics generated. The advantage of a multi-step approach is that large sample size can be easily obtained through meta-analyses, combining summary statistics from different studies and avoiding the need for individual-level data sharing. However, as large-scale biobank data is increasingly available, methods that provide joint estimates of the marker effects in a single step by estimating the effect sizes as random under ﬂexible prior formulations may become bene- ﬁcial as they: (i) can account for differences in the variance contributed across MAF, LD or annotation groups providing unbiased MAF-LD annotation-speciﬁc genetic effect size esti- mates and h2 SNP of different annotations, allowing for a con- trasting of the genetic architectures of complex traits; (ii) give the probability that each marker, genomic region, annotation, gene- coding region, or SNP is associated with a phenotype, alongside the proportion of phenotypic variation contributed by each, yielding test statistics that describe the gene architecture of complex traits and the uncertainty over the estimates; and (iii) provide improved genomic prediction, whilst providing a pos- terior predictive distribution for each individual. j change in the jth covariate within the φth group. The spike and slab prior, contains what is called a Dirac spike14,15 for βφ, which induces sparsity in the model through a Dirac-delta at zero, excluding variables from the model by setting their coefﬁcients to zero. A ﬁnite scale mixture of normal distributions centred at zero constitute the slab component. The slab shrinks the non-zero coefﬁcients towards zero according to the slab’s width, and by having a scale mixture of Gaussians, the distribution has heavier tails and can accommodate big and small effects16. Therefore, each βφj is distributed according to: φj βφj π0φδ0 þ π1φN 0; σ2 1φ þ π2φN 0; σ2 2φ þ ¼ þ πLφφN 0; σ2 Lφφ ; ð2Þ ð2Þ where for each SNP marker group fπ0φ; π1φ; ¼ ; πLφφg are the mixture proportions and fσ2 1φ; σ2 2φ; ¼ ; σ2 Lφφg are the mixture- speciﬁc variances proportional to p p σ2 1φ ... σ2 Lφφ 2 6664 3 7775 ¼ σ2 Gφ C1φ .. . NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications Results We determine the ability of our PPWV approach to correctly localise an association to LD blocks (deﬁned as groups of markers with LD R2 ≥0.1) that contain causal variants, and compare this to using LD to clump mixed- linear model association estimates obtained using the BoltLMM software (Fig. 2a). We ﬁnd that a PPWV approach identiﬁes associated LD blocks with higher probability as compared to clumped MLMA associations, for all genetic architectures, with the exception of simulated phenotypes with enrichment and low polygenicity, where the small numbers of relatively large effect size regions are better identiﬁed with a single-marker regression approach (Fig. 2a). Thus, BayesRR-RC provides an alternative to standard genome-wide association studies to localise SNP- phenotype associations at the regional level, especially for traits with high polygenicity. Simulation study. We ﬁrst compare the model performance of BayesRR-RC to existing approaches across 18 different genetic architectures. We randomly selected 40,000 unrelated UK Biobank individuals and used 596,741 imputed SNP markers from chro- mosomes 19 to 22. We randomly selected either 1000, 10,000 or 100,000 LD independent (LD R2 < 0.1) causal SNP markers. For each SNP marker set, we then simulated effect sizes from a normal distribution with zero mean and variance of 0.1, 0.3 or 0.6 divided by the number of causal variants and ∝N(0, [p(1−p)]−0.25), with p the allele frequency (see “Methods” section). This simulates stronger effect sizes for rare variants in line with recent empirical estimates and we simulated ten replicate phenotypes for each of the nine different genetic architectures. We then additionally repeat each simulation, sampling the SNP marker effects this time from 13 different distributions, one for each of 13 different genomic annotation groups with different proportions of h2 SNP to create nine further different genetic architectures. We compare our BayesRR-RC model to the following statistical models: (i) a restricted maximum likelihood (REML) model implemented in the software BoltREML17 with the same 78 MAF-LD-annotation groups enabling a direct comparison, (ii) a Haseman–Elston (HE) regression using the same 78 group model implemented in the software RHEmc18, (iii) summary statistic linkage disequilibrium score regression (LDSC)19, with LD scores calculated using the same data, and the same 78 non-overlapping annotations in a 78 component LDSC annotation model, and (iv) summary statistic SumHer6 (LDAK) with the same 78 non-overlapping annotations. Results Thus, BayesRR-RC provides an alternative to standard genome-wide association studies to localise SNP- phenotype associations at the regional level, especially for traits with high polygenicity. with the BoltREML model, with similar correlation of the estimated and simulated values within each simulation replicate (Fig. 1a). In comparison, RHEmc, which also uses individual-level data, yields estimates with lower correlation with the simulated value, but higher than both summary statistic approaches implemented in LDSC and Sumher (Fig. 1a). We calculate estimates of enrichment, deﬁned as the proportion of h2 SNP attributable to the annotation divided by the proportion of SNPs mapping to the annotation (for bayesRR-RC, because there is sparsity in the SNP effects, we deﬁne enrichment as the proportion of SNPs in the model that map to the annotation, see “Methods” section) and we compare these to the true simulated value. Compared to other approaches, we ﬁnd that BayesRR-RC gives a lower probability of false enrichment, calculated as the proportion of times within a simulation replicate that an annotation group was incorrectly assigned as having enrichment greater than 2 (Fig. 1b). Thus, BayesRR-RC provides accurate partitioning of genomic enrichment across the genome. (see “Code availability” section). All SNP markers assigned to 1 kb regions map to promoters; 1–10 kb SNPs, 10–500 kb SNPs, 500 kb–1 Mb SNPs are then split into enh, tfbs and others (unmapped SNPs) extending the model to 13 annotation groups (Supplementary Data 1). Within each of these annotations, we have three minor allele frequency groups (MAF ≤0.01, 0.01 < MAF ≤0.05, and MAF > 0.05), and then each MAF group is further split into two based on median LD score. This gives 78 non-overlapping groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects within, each group. For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to the group mul- tiplied by constants (0, 0.0001, 0.001, 0.01, 0.1). (see “Code availability” section). All SNP markers assigned to 1 kb regions map to promoters; 1–10 kb SNPs, 10–500 kb SNPs, 500 kb–1 Mb SNPs are then split into enh, tfbs and others (unmapped SNPs) extending the model to 13 annotation groups (Supplementary Data 1). Results A Bayesian model for large-scale genomic data. We derive a model that we call BayesRR-RC in Supplementary Note 1 and the “Methods” section, which is based on grouped effects with 2 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 with the BoltREML model, with similar correlation of the estimated and simulated values within each simulation replicate (Fig. 1a). In comparison, RHEmc, which also uses individual-level data, yields estimates with lower correlation with the simulated value, but higher than both summary statistic approaches implemented in LDSC and Sumher (Fig. 1a). We calculate estimates of enrichment, deﬁned as the proportion of h2 SNP attributable to the annotation divided by the proportion of SNPs mapping to the annotation (for bayesRR-RC, because there is sparsity in the SNP effects, we deﬁne enrichment as the proportion of SNPs in the model that map to the annotation, see “Methods” section) and we compare these to the true simulated value. Compared to other approaches, we ﬁnd that BayesRR-RC gives a lower probability of false enrichment, calculated as the proportion of times within a simulation replicate that an annotation group was incorrectly assigned as having enrichment greater than 2 (Fig. 1b). Thus, BayesRR-RC provides accurate partitioning of genomic enrichment across the genome. In Supplementary Note 3, we propose a posterior probability window variance (PPWV) approach20, which provides a probabilistic determination of association of a given LD block, genomic window, gene, or upstream region, relative to the amount of phenotypic variation attributable to that window. Our PPWV approach determines the posterior inclusion probability that each region and each gene contributes at least 0.001% to the h2 SNP, with theory outlined in Supplementary Note 3 suggesting well controlled FDR. We determine the ability of our PPWV approach to correctly localise an association to LD blocks (deﬁned as groups of markers with LD R2 ≥0.1) that contain causal variants, and compare this to using LD to clump mixed- linear model association estimates obtained using the BoltLMM software (Fig. 2a). We ﬁnd that a PPWV approach identiﬁes associated LD blocks with higher probability as compared to clumped MLMA associations, for all genetic architectures, with the exception of simulated phenotypes with enrichment and low polygenicity, where the small numbers of relatively large effect size regions are better identiﬁed with a single-marker regression approach (Fig. 2a). Results Within each of these annotations, we have three minor allele frequency groups (MAF ≤0.01, 0.01 < MAF ≤0.05, and MAF > 0.05), and then each MAF group is further split into two based on median LD score. This gives 78 non-overlapping groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects within, each group. For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to the group mul- tiplied by constants (0, 0.0001, 0.001, 0.01, 0.1). One of the major limitations preventing the application of Bayesian approaches to large-scale genomic data is the view that the computation of a posterior distribution is too expensive. In Supplementary Note 2, we derive a Bulk Synchronous hybrid- parallel (BSP) Gibbs sampling scheme for large-scale genomic data that allows both the data and the compute tasks to be split within and across compute nodes in a series of message-passing interface (MPI) tasks. We extend previous sparse residual updating schemes by deriving sampling steps to utilise whole genome sequence or SNP genetic data stored in mixed binary/ sparse-index representation (see Supplementary Note 2), redu- cing computational complexity of a single Gibbs step from OðnÞ to OðnzÞ, with nz the number of non-zero genotypes, as SNP- phenotype covariance estimation (dot product calculation) is conducted as a series of look-up tables. We provide publicly available open source software (GMRM) that requires as little as 22 s per MCMC sample to estimate 78 group-speciﬁc h2 SNP parameters, and the inclusion probabilities and effect sizes of 8,433,421 markers in 382,466 individuals on standard Intel Xeon CPU processors (see “Code availability” section, Supplementary Note 2). In Supplementary Note 3, we propose a posterior probability window variance (PPWV) approach20, which provides a probabilistic determination of association of a given LD block, genomic window, gene, or upstream region, relative to the amount of phenotypic variation attributable to that window. Our PPWV approach determines the posterior inclusion probability that each region and each gene contributes at least 0.001% to the h2 SNP, with theory outlined in Supplementary Note 3 suggesting well controlled FDR. NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 1 Simulation study for the performance of our BayesRR-RC model implemented in the GMRM software against existing approaches for variance component and genomic annotation enrichment estimation. a Correlation of the simulated and estimated SNP heritability across 13 genomic annotation groups within each of 20 replicates for ﬁve different statistical models: a mixture of regression model with multiple group-speciﬁc variance components described in this work (GMRM), Haseman–Elston regression with annotation-speciﬁc relationship matrices implemented in the RHEmc software (RHEmc), a multiple group-speciﬁc variance component REML model implemented in the software bolt (BOLT), and two annotation summary statistic models implemented in the software LDSC and LDAK. The column facets give the simulated heritability and rows give the number of causal variants. b Probability of falsely assigning one of the 13 genomic annotation groups as explaining 2 times greater proportion of variance given the proportion of SNPs mapping to the annotation. The column facets give the simulated heritability and rows give the number of causal variants. Boxplots give the median with 25th and 75th percentile and 95% credible intervals for n = 20 simulation replicates in both panels. z-scores of the BayesRR-RC estimates are generally stable across generative genetic models and that the MLMA estimates have higher estimation error, especially when the causal variant is rare, or in high-LD with many other SNPs (Supplementary Fig. 3). We also ﬁnd that our PPWV approach outperforms MLMA methods in their precision-recall curves across the range of genetic architectures (Supplementary Fig. 4). We conﬁrmed that popula- tion stratiﬁcation and relatedness are well-controlled for using a PPWV approach, as compared to an MLMA model with the leading PCs of the genomic data included (Supplementary Fig. 5). We compared the ability of our approach to identify candidate SNPs and to provide a probabilistic assessment of the most likely associated set of SNP markers. Finally, we show that our PPWV approach is analogous to the approach suggested in a recent paper (SuSiE22) of selecting credible sets of markers with high probability of association, ﬁnding that BayesRR-RC has higher power to localise associations to sets of SNP markers (Supple- mentary Fig. 6). The advantage of BayesRR-RC is also that assessment of associated regions is done genome-wide, with estimates obtained through simple summary of the posterior distribution instead of running numerous statistical models at different genomic regions. Results W ﬁd h B RR RC i h h i i i g p yg y We then also compare the prediction accuracy obtained in an independent sample when creating genomic predictors using (i) effect sizes estimated by BayesRR-RC, (ii) ﬁxed-effect SNP effect sizes estimated in the MLMA approach implemented in bolt, and (iii) effect size estimates obtained from four different genomic prediction models proposed in a recent paper21, implemented in the LDAK software, which are suggested to outperform all other current approaches. In comparison to the best LDAK predictor, we ﬁnd that BayesRR-RC obtains similar or improved prediction accuracy across all genetic architectures, with greater prediction accuracy gains observed under genetic architectures where the SNP effect distributions differed across genomic annotations (Fig. 2b). We ﬁnd that given sufﬁcient power, BayesRR-RC can obtain or even exceed the theoretical expectation of prediction accuracy under ridge regression assumptions (Fig. 2b, see “Methods” section). We then conduct a number of follow-up simulation studies. Recent work has highlighted differences in statistical model performance depending upon the relationship of SNP marker effect size, LD and MAF1,3,4. We explore the performance of our model in theory, with highly correlated genetic markers in Supplementary Note 4. We also conducted another large-scale, but well-powered, simulation study to explore the model performance of BayesRR-RC as compare to existing approaches We ﬁnd that BayesRR-RC estimates the phenotypic variation attributable to different genomic annotation groups comparable TURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 3 3 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.5 1.0 0.25 0.50 0.75 1.00 0.25 0.50 0.75 1.00 annotation variance accuracy a 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.0 0.2 0.4 false enrichment probability b BOLT GMRM RHEmc LDSC LDAK Fig. 1 Simulation study for the performance of our BayesRR-RC model implemented in the GMRM software against existing approaches for variance component and genomic annotation enrichment estimation. a Correlation of the simulated and estimated SNP heritability across 13 genomic annotation groups within each of 20 replicates for ﬁve different statistical models: a mixture of regression model with multiple group-speciﬁc variance components described in this work (GMRM), Haseman–Elston regression with annotation-speciﬁc relationship matrices implemented in the RHEmc software (RHEmc), a multiple group-speciﬁc variance component REML model implemented in the software bolt (BOLT), and two annotation summary statistic models implemented in the software LDSC and LDAK. The column facets give the simulated heritability and rows give the number of causal variants. b Probability of falsely assigning one of the 13 genomic annotation groups as explaining 2 times greater proportion of variance given the proportion of SNPs mapping to the annotation. The column facets give the simulated heritability and rows give the number of causal variants. Boxplots give the median with 25th and 75th percentile and 95% credible intervals for n = 20 simulation replicates in both panels. 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.5 1.0 0.25 0.50 0.75 1.00 0.25 0.50 0.75 1.00 annotation variance accuracy a 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.0 0.2 0.4 false enrichment probability b BOLT GMRM RHEmc LDSC LDAK 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.5 1.0 0.25 0.50 0.75 1.00 0.25 0.50 0.75 1.00 annotation variance accuracy a 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.0 0.2 0.4 false enrichment probability b BOLT GMRM RHEmc LDSC LDAK 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 0.0 0.2 0.4 false enrichment probability b 0.1 0.3 0.6 1000 10000 1e+05 0.0 0.5 1.0 0.25 0.50 0.75 1.00 0.25 0.50 0.75 1.00 annotation variance accuracy a BOLT GMRM b a 10000 10000 BOLT GMRM RHEmc LDSC LDAK Fig. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 2 Simulation study for the performance of our BayesRR-RC model implemented in the GMRM software against existing approaches for localisation of associations and genomic prediction. a Probability of detecting genomic regions containing simulated causal variants by a Bayesian regional ﬁne-mapping approach (GMRM: blue) versus standard mixed linear model association (MLMA) testing (BOLT: green). The column facets give the simulated heritability and rows give the number of causal variants and whether the effect sizes differed across genomic annotation groups (enrich) or were randomly assigned (random). b Correlation of a genomic predictor and a phenotype in an independent sample when the genomic predictor is created from GMRM effects sizes (blue), MLMA effect sizes using BOLT (green), and the optimal effect sizes obtained from individual-level and summary statistic models implemented in the Mega-PRS LDAK approach (purple). The column facets give the simulated heritability and the number of causal variants. The row facets give whether the effect sizes differed across genomic annotation groups (enrich) or were randomly assigned (random). The red lines give the expected prediction accuracy based on ridge regression theory. Error bars show the SD in both panels. UK Biobank data genotyped at 8,433,421 imputed SNP markers. These markers were selected as they overlap with the Estonian Genome Centre data (see “Methods” section) and have minor allele frequency >0.0002. We adjust each phenotype for age, sex, year of birth, genotype batch effects, UK Biobank assessment centre, and the leading 20 principal components of the SNP data. We conducted a series of convergence diagnostic analyses of the posterior distributions to ensure we obtained estimates from a converged set of four Gibbs chains, each run for 6000 iterations with a thin of ﬁve for each trait (Supplementary Figs. 7–10). Our estimates compare similarly to those obtained by RHEmc and SumHer, but differ to those obtained by LDSC (Table 1 and Supplementary Data 3, 4, and 5 for full results). In addition to providing variance component estimates, our model facilitates assessment of differences in the underlying effect size distribution across annotation groups. For each group, we modelled the SNP effects as coming from a series of ﬁve Gaussian mixtures, and we ﬁnd that at least 45% of the h2 SNP attributable to both introns and 500 kb upstream regions is underlain by many thousands of SNPs that on average each contribute 0.001% (estimates summed across MAF and LD groups in Fig. 3b and Supplementary Figs. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Taken together, these simulation results indicate that BayesRR-RC provides accurate estimates of the underlying effect size distribution for different genomic groups, yielding improved genomic prediction, across a wide range of different underlying generative genetic models. across a wide range of 20 different effect size, LD, and MAF relationships as described in Supplementary Table 1. For the estimation of h2 SNP and the proportion of h2 SNP attributable to different annotation groups, we ﬁnd that all statistical models other than BayesRR-RC are sensitive to the underlying generative genetic model, with no other approach providing consistent estimates across the 20 generative genetic models (Supplementary Fig. 1a). As in the previous simulation, BayesRR-RC estimates the variance attributable to different genomic regions on the correct scale, with higher correlation as compared with other approaches (Supplementary Fig. 1b), and this results in the estimated average effect size for each annotation group having high correlation with the simulated value (Supplementary Fig. 1c). Again, summary statistic approaches performed poorly for both variance compo- nent estimation (Supplementary Fig. 1b) and quantiﬁcation of enrichment as compared to individual-level methods, often even incorrectly selecting the group of highest average effect size (Supplementary Fig. 1c). We conﬁrmed our genomic prediction results, ﬁnding that BayesRR-RC outperforms all methods implemented in the LDAK software across all generative models, with BayesRR-RC very marginally outperforming a single variance component BayesR model in the enrichment simulations of each of the 20 generative genetic models (Supplementary Fig. 2). We further explored the ability of our PPWV approach to localise SNP-phenotype associations in the 20 generative models, by comparing the z-scores of the marker effect estimates from their true simulated value across the minor allele frequency spectrum (Supplementary Fig. 3) and the area under the precision-recall curve (AUPRC, Supplementary Fig. 4) for BayesRR-RC and a series of MLMA methods. We ﬁnd that the The genetic architecture of four complex traits in the UK Biobank. We apply BayesRR-RC to cardiovascular disease out- comes (CAD), type-2 diabetes (T2D), body mass index (BMI) and height measured for 382,466 unrelated individuals from the NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 4 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 0.1 0.3 0.6 enrich 1000 random 1000 enrich 10000 random 10000 enrich 1e+05 random 1e+05 0.00 0.25 0.50 0.75 1.00 0.0 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.00 0.05 0.10 0.15 0.20 0.00 0.05 0.10 0.15 0.00 0.01 0.02 0.03 0.04 0.05 probability of effect size localization a 0.1 1000 0.1 10000 0.1 1e+05 0.3 1000 0.3 10000 0.3 1e+05 0.6 1000 0.6 10000 0.6 1e+05 enrich random 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 prediction accuracy b BOLT GMRM LDAK Fig. 2 Simulation study for the performance of our BayesRR-RC model implemented in the GMRM software against existing approaches for localisation of associations and genomic prediction. a Probability of detecting genomic regions containing simulated causal variants by a Bayesian regional ﬁne-mapping approach (GMRM: blue) versus standard mixed linear model association (MLMA) testing (BOLT: green). The column facets give the simulated heritability and rows give the number of causal variants and whether the effect sizes differed across genomic annotation groups (enrich) or were randomly assigned (random). b Correlation of a genomic predictor and a phenotype in an independent sample when the genomic predictor is created from GMRM effects sizes (blue), MLMA effect sizes using BOLT (green), and the optimal effect sizes obtained from individual-level and summary statistic models implemented in the Mega-PRS LDAK approach (purple). The column facets give the simulated heritability and the number of causal variants. The row facets give whether the effect sizes differed across genomic annotation groups (enrich) or were randomly assigned (random). The red lines give the expected prediction accuracy based on ridge regression theory. Error bars show the SD in both panels. 0.1 1000 0.1 10000 0.1 1e+05 0.3 1000 0.3 10000 0.3 1e+05 0.6 1000 0.6 10000 0.6 1e+05 enrich random 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 prediction accuracy b BOLT GMRM LDAK 0.1 0.3 0.6 enrich 1000 random 1000 enrich 10000 random 10000 enrich 1e+05 random 1e+05 0.00 0.25 0.50 0.75 1.00 0.0 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.00 0.05 0.10 0.15 0.20 0.00 0.05 0.10 0.15 0.00 0.01 0.02 0.03 0.04 0.05 probability of effect size localization a b a performance of our BayesRR-RC model implemented in the GMRM software against existing approaches for Fig. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 a We partition SNP markers into seven location annotations (coding regions, intronic regions, and windows 1, 1–10, 10–500 kb and 500 kb–1 Mb upstream of genes, with other SNPs grouped in a category labelled “others"). Windows 1–10 kb, 10–500 kb and 500 kb–1 Mb upstream of genes are further split into SNPs mapped to enhancers (enh), transcription factor binding sites (tfbs) and others. Within each of the 13 annotations, we have three minor allele frequency groups (MAF ≤0.01 annotated as rare, 0.01 < MAF ≤0.05 annotated as low, and MAF > 0.05 annotated as common), and then each MAF group is further split into two based on median LD score. This gives 78 groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects within, each group. For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to the group multiplied by constants (mixture 0 = 0, mixture 1 = 0.0001, 2 = 0.001, 3 = 0.01, 4 = 0.1). b Posterior distribution of the proportion of the total phenotypic variance attributable to the SNP markers that is contributed by each of the four non-zero mixtures within each MAF-annotation group for HT, BMI, CAD and T2D. Within these, are boxplots of the posterior mean and 95% credible intervals. Values are summed over LD groups. c Bar plots with error bars giving the 95% credible intervals for the average effect size of markers in the model for each MAF-annotation group, split by mixture. Fig. 3 Genetic architecture of enrichment for height (HT), body mass index (BMI), cardiovascular disease (CAD) 382,466 unrelated European ancestry UK Biobank individuals genotyped at 8,430,446 SNP markers. a We partitio height (HT), body mass index (BMI), cardiovascular disease (CAD) and type-2 diabetes (T2D) for also did not ﬁnd evidence that the allele substitution effect size differed across frequency groups for transcription factor binding sites, distal SNPs 1 MB upstream of genes, or those not mapping to an annotation group (Supplementary Fig. 12b). a speciﬁc category, with perhaps the exception of high MAF variants (Fig. 3c). Generally, all phenotypes simply appear to be predominantly underlain by very many common variants, with SNPs within distal regulatory regions, coding and intronic regions contributing more to the variance. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 11 and 12). In contrast, the variance is spread more evenly across the mixtures for the other groups, implying that 10–500 kb upstream regions and introns are more polygenic than other groups. This is especially so for BMI where 35% of the h2 SNP is attributable to many thousands of intronic variants (Fig. 3 and Supplementary Fig. 12). Therefore, we ﬁnd that the polygenicity of the genetic effects varies across different genomic regions, with remarkably consistent patterns across traits in the partitioning of h2 SNP across the genome. ( pp y g ) We ﬁnd that 32–44% of the h2 SNP is attributable to intronic regions, 12–25% is attributable to exonic regions, 22–28% is attributable to markers 10–500 kb upstream of genes, with proximal (within 10 kb) promotors, enhancers and transcription factor binding sites cumulatively contributing <10% (Fig. 3b and Supplementary Fig. 11, with estimates summed across MAF and LD groups Table 1, and full results in Supplementary Data 2). The large contribution of exonic and intronic annotations to variation is in-line with the fact that these annotations account for ~40% of the total genome length. All four traits show the same pattern of group-speciﬁc variation, with the exception of height, where the proportion of h2 SNP attributable to exons is almost twice as large as the other phenotypes (Fig. 3b; Table 1 and Supplementary Fig. 11 and Supplementary Data 2). For all annotation groups in exons, introns, and within 500 kb of genes across all traits, ≥60% of the h2 SNP attributable to these groups is contributed by many thousands of common variants, each of small effect (Fig. 3b and Supplementary Figs. 11 and 12). Across traits, posterior mean effect sizes scale to their differences in h2 SNP, and we ﬁnd that exonic and intronic region effect sizes were higher than the rest of the genome, across all mixture groups, followed by 10–500 kb upstream regions (Fig. 3c). We ﬁnd little evidence that SNPs located in proximal promotors, enhancers, and transcription factor binding sites within 10 kb of genes showed average effect sizes that were higher than SNPs located 1 MB away from genes, or those that were not mapped to 5 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Within each of the 13 annotations, we have three minor allele frequency groups (MAF ≤0.01 annotated as rare 0.01 < MAF ≤0.05 annotated as low, and MAF > 0.05 annotated as common), and then each MAF group is further split into two based on median L Fig. 3 Genetic architecture of enrichment for height (HT), body mass index (BMI), cardiovascular disease (CAD) and type-2 diabetes (T2D) for 382,466 unrelated European ancestry UK Biobank individuals genotyped at 8,430,446 SNP markers. a We partition SNP markers into seven location annotations (coding regions, intronic regions, and windows 1, 1–10, 10–500 kb and 500 kb–1 Mb upstream of genes, with other SNPs grouped in a category labelled “others"). Windows 1–10 kb, 10–500 kb and 500 kb–1 Mb upstream of genes are further split into SNPs mapped to enhancers (enh), transcription factor binding sites (tfbs) and others. Within each of the 13 annotations, we have three minor allele frequency groups (MAF ≤0.01 annotated as rare, 0.01 < MAF ≤0.05 annotated as low, and MAF > 0.05 annotated as common), and then each MAF group is further split into two based on median LD score. This gives 78 groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects within, each group. For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to the group multiplied by constants (mixture 0 = 0, mixture 1 = 0.0001, 2 = 0.001, 3 = 0.01, 4 = 0.1). b Posterior distribution of the proportion of the total phenotypic variance attributable to the SNP markers that is contributed by each of the four non-zero mixtures within each MAF-annotation group for HT, BMI, CAD and T2D. Within these, are boxplots of the posterior mean and 95% credible intervals. Values are summed over LD groups. c Bar plots with error bars giving the 95% credible intervals for the average effect size of markers in the model for each MAF-annotation group, split by mixture. Fig. 3 Genetic architecture of enrichment for height (HT), body mass index (BMI), cardiovascular disease (CAD) and type-2 diabetes (T2D) for 382,466 unrelated European ancestry UK Biobank individuals genotyped at 8,430,446 SNP markers. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Fig. 3 Genetic architecture of enrichment for height (HT), body mass index (BMI), cardiovascular disease (CAD) and type-2 diabetes (T2D) for 382,466 unrelated European ancestry UK Biobank individuals genotyped at 8,430,446 SNP markers. a We partition SNP markers into seven locatio annotations (coding regions, intronic regions, and windows 1, 1–10, 10–500 kb and 500 kb–1 Mb upstream of genes, with other SNPs grouped in a catego labelled “others"). Windows 1–10 kb, 10–500 kb and 500 kb–1 Mb upstream of genes are further split into SNPs mapped to enhancers (enh), transcriptio factor binding sites (tfbs) and others. Within each of the 13 annotations, we have three minor allele frequency groups (MAF ≤0.01 annotated as rare 0.01 < MAF ≤0.05 annotated as low, and MAF > 0.05 annotated as common), and then each MAF group is further split into two based on median L score. This gives 78 groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects withi each group. For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to th group multiplied by constants (mixture 0 = 0, mixture 1 = 0.0001, 2 = 0.001, 3 = 0.01, 4 = 0.1). b Posterior distribution of the proportion of the tot phenotypic variance attributable to the SNP markers that is contributed by each of the four non-zero mixtures within each MAF-annotation group for H BMI, CAD and T2D. Within these, are boxplots of the posterior mean and 95% credible intervals. Values are summed over LD groups. c Bar plots with err bars giving the 95% credible intervals for the average effect size of markers in the model for each MAF-annotation group, split by mixture. Fig. 3 Genetic architecture of enrichment for height (HT), body mass index (BMI), cardiovascular disease (CAD) and type-2 diabetes (T2D) for 382,466 unrelated European ancestry UK Biobank individuals genotyped at 8,430,446 SNP markers. a We partition SNP markers into seven locati annotations (coding regions, intronic regions, and windows 1, 1–10, 10–500 kb and 500 kb–1 Mb upstream of genes, with other SNPs grouped in a catego labelled “others"). Windows 1–10 kb, 10–500 kb and 500 kb–1 Mb upstream of genes are further split into SNPs mapped to enhancers (enh), transcripti factor binding sites (tfbs) and others. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 We also re-scaled the marker effects by the standard deviation of each marker, to give effect sizes on the allele substitution effect size scale, and again we ﬁnd that rare variants have higher average allele substitution effects than common variants for exonic, intronic, promotors and enhancers (Supplementary Fig. 12b). An exception to these patterns were BMI-associated intronic and 10–500 kb group SNPs, where we ﬁnd no evidence that the allele substitution effect size differs across frequency groups (Supplementary Fig. 12b). We a speciﬁc category, with perhaps the exception of high MAF variants (Fig. 3c). Generally, all phenotypes simply appear to be predominantly underlain by very many common variants, with SNPs within distal regulatory regions, coding and intronic regions contributing more to the variance. We also re-scaled the marker effects by the standard deviation of each marker, to give effect sizes on the allele substitution effect size scale, and again we ﬁnd that rare variants have higher average allele substitution effects than common variants for exonic, intronic, promotors and enhancers (Supplementary Fig. 12b). An exception to these patterns were BMI-associated intronic and 10–500 kb group SNPs, where we ﬁnd no evidence that the allele substitution effect size differs across frequency groups (Supplementary Fig. 12b). We Discovery of associated genomic regions. We then partitioned the variance attributed to SNP markers across 50kb regions of the genome, then across SNPs annotated to genes, and then to LD blocks of the DNA using our PPWV approach. We ﬁnd 1660 50 kb regions for height with ≥95% posterior probability of explaining 0.001% of the h2 SNP, 520 regions for BMI, 70 regions for CAD and 87 regions for T2D (Fig. 4a and Table 2). We then map NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 6 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Table 1 Proportion of genetic variance attributable to different genomic regions for height (HT), body mass index (BMI), type-2 diabetes (T2D) and cardiovascular disease (CAD). NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Table 1 Proportion of genetic variance attributable to different genomic regions for height (HT), body mass index (BMI), type-2 diabetes (T2D) and cardiovascular disease (CAD). etic variance attributable to different genomic regions for height (HT), body mass index (BMI), type-2 ovascular disease (CAD). calculate the number of exons, introns, promotors and cis regions with ≥95% posterior probability of explaining 0.001% of the h2 SNP, as a proportion of the total number within each chromosome, we ﬁnd that up to 24% of the genes on each chromosome are associated with each of the four traits (Fig. 4b). Generally, we ﬁnd that only 1% or less of the available exons and promotor regions of genes per chromosome show an association with each of the phenotypes, but up to 14% of the available intronic regions and up to 10% of the cis-regions surrounding genes contribute to the phenotypic variance with ≥95% probability (Fig. 4b). The var- iance contributed by each exonic, intronic, promotor, or cis region is typically only a small fraction of a percent, with largest effect sizes being the exonic region of GDF5 contributing 0.26% SNPs to their closest gene (+/−50 kb from SNP position) and we use our annotations to label them (see “Methods” section). We ﬁnd 243 independent coding regions for height with ≥95% pos- terior probability of explaining at least 0.001% of the h2 SNP, 29 independent coding regions for BMI, 5 for CAD and 13 for T2D. We ﬁnd many more associations in the cis region of genes with 1254 independent cis-regions for height with ≥95% posterior probability of explaining 0.001% of the h2 SNP, 1765 independent cis-regions for BMI, 1166 for CAD and 1221 for T2D. We additionally ﬁnd 9 independent promoter regions and 1072 independent introns for height with ≥95% posterior probability of explaining at least 0.001% of the h2 SNP, 1162 independent intronic gene regions for BMI, 307 for CAD and 347 for T2D. When we SNPs to their closest gene (+/−50 kb from SNP position) and we use our annotations to label them (see “Methods” section). We ﬁnd 243 independent coding regions for height with ≥95% pos- terior probability of explaining at least 0.001% of the h2 SNP, 29 independent coding regions for BMI, 5 for CAD and 13 for T2D. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Group Trait BayesRR-RC RHE-mca sLDSCa SumHera Posterior mean (95% CI) h2 obs (se) % h2 obs (se) % h2 obs (se) % Variance attributable to SNP markers genome-wide HT 57.66 (56.09, 59.14) 63.28 (3.57) 64.16 (2.86) 98.58 (0.69) BMI 28.74 (27.62, 30.0) 26.76 (1.06) 31.03 (0.9) 44.98 (0.53) CAD 5.94 (5.30, 6.67) 4.49 (>100) 4.73 (0.28) 7.33 (0.43) T2D 8.45 (7.83, 9.18) 6.90 (0.47) 6.53 (0.3) 11.65 (0.44) Proportion of genetic variance attributable to exonic regions of genes HT 24.75 (23.39, 26.071) 27.09 3.00 16.74 BMI 12.98 (10.98, 14.84) 12.62 4.37 7.60 CAD 13.23 (8.40, 18.84) 18.68 1.69 15.34 T2D 14.49 (10.74, 18.54) 14.60 2.46 10.12 Proportion of genetic variance attributable to intronic regions of genes HT 41.54 (39.91, 43.39) 41.60 46.07 43.03 BMI 44.17 (41.36, 47.25) 47.87 44.61 48.19 CAD 32.05 (24.98, 39.51) 41.15 47.22 41.94 T2D 37.28 (32.22, 42.57) 48.66 38.52 48.02 Proportion of genetic variance attributable to snps 1 kb upstream of genes HT 2.81 (2.24, 3.42) 1.76 1.46 1.74 BMI 1.62 (0.75, 2.69) 0.36 1.90 1.15 CAD 4.20 (1.71, 7.55) 2.49 <0.00 1.26 T2D 3.58 (1.77, 5.86) 3.40 <0.00 1.57 Proportion of genetic variance attributable to snps 10 kb upstream of genes HT 6.60 (5.84, 7.40) 6.73 4.29 12.87 BMI 5.28 (3.92, 6.87) 3.19 6.58 4.10 CAD 13.06 (8.70, 18.16) 5.70 6.02 8.91 T2D 9.08 (5.90, 13.28) 4.02 20.44 7.56 Proportion of genetic variance attributable to snps 500 kb upstream of genes HT 22.13 (21.00, 23.40) 21.53 37.23 24.14 BMI 28.58 (26.41, 31.01) 28.81 35.86 31.17 CAD 28.02 (21.24, 35.04) 30.23 38.90 29.58 T2D 27.42 (22.68, 32.36) 24.33 32.49 27.47 Proportion of genetic variance attributable to exonic regions that is explained by common variants HT 72.09 (69.77, 74.14) 62.62 75.35 51.22 BMI 69.41 (62.60, 76.42) 59.67 16.43 54.31 CAD 64.97 (43.08, 83.16) 61.72 >100 49.17 T2D 68.57 (56.00, 79.82) 66.33 >100 64.11 Proportion of genetic variance attributable to intronic regions that is explained by common variants HT 81.19 (79.30, 83.02) 79.96 70.88 66.12 BMI 85.05 (78.28, 91.49) 86.10 70.62 69.68 CAD 84.68 (65.64, 95.91) 96.55 61.11 78.17 T2D 87.62 (75.65, 94.85) 87.63 67.93 71.39 Proportion of genetic variance attributable to snps 500 kb upstream of genes that is explained by common variants HT 81.59 (78.91, 83.96) 80.66 71.86 77.28 BMI 86.78 (80.56, 91.60) 89.95 67.38 74.81 CAD 66.49 (49.11, 81.79) 88.51 60.52 79.91 T2D 72.35 (58.71, 83.75) 94.91 69.48 75.12 aRHEmc18, LDSC19 and SumHer6 provide the total SNP heritability observed (%) and single heritability estimates per genetic component (see Supplementary Data 2–5) that we summarised to obtain the proportion of genetic variance attributed to exonic regions, intronic regions and windows 1, 1–10 and 10–500 kb upstream of genes. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 CAD T2D HT BMI 1 3 5 7 9 11 13 15 17 19 21 1 3 5 7 9 11 13 15 17 19 21 0 50 100 0 50 100 chromosome number of regions PPWV (0.8,0.9] (0.9,0.95] (0.95,1] a CAD T2D HT BMI 1 3 5 7 9 11 13 15 17 19 21 1 3 5 7 9 11 13 15 17 19 21 0.00 0.05 0.10 0.00 0.05 0.10 chromosome proportion of regions PPWV > 0.95 exons introns 1kb prom cis b HT BMI CAD T2D 0.00 0.05 0.10 0.15 0.20 0.0 0.2 0.4 0.6 0.00 0.05 0.10 0.15 0.0 0.1 0.2 correlation exons,introns exons,prom exons,cis introns,prom introns,cis prom,cis d ADAMTS17 GDF5 DIS3L2 CABLES1 BDNF−AS FTO CELSR2 PHACTR1 LPA CDKN2B−AS1 PAM HNF1A TCF7L2 T2D CAD BMI HT 0.00000 0.00005 0.00010 0.00015 0.00020 0.00000 0.00005 0.00010 0.00015 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00100 0.00200 0.00300 0.00400 0.00500 0.00000 0.00020 0.00040 0.00060 0.00000 0.00100 0.00200 by exons mean genetic variance explained by introns PPWV > 0.95 exons introns c ADAMTS17 GDF5 IGF1R BMP2 IGF2BP2 BDNF−AS SEC16B uc021ukz.1 TMEM18 ALKAL2 CELSR2 LDLR PLG LINC00310 PAM HNF1A CDKN2B−AS1 T2D CAD BMI HT 0.00000 0.00005 0.00010 0.00015 0.00020 0.00000 0.00005 0.00010 0.00015 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00025 0.00050 0.00075 0.00000 0.00005 0.00010 0.00015 0.00000 0.00030 0.00060 0.00090 by exons mean genetic variance explained by cis regions PPWV > 0.95 exons cis Fig. 4 Contribution of genes and 50kb regions to height (HT), body-mass-index (BMI), cardiovascular disease (CAD) and type-2-diabetes (T2D). a We grouped SNPs in 50 kb-regions genome-wide and estimated the sum of the squared regression coefﬁcient estimates for each 50 kb-region. We then select the number of 50 kb regions that explain at least 0.001% of the variance attributed to all SNP markers in 80, 90 and 95% of the iterations. This gives a measure called the posterior probability that the window variance (PPWV)20 exceeds 1/10,000 of the phenotypic variation attributed to SNP markers. b We mapped SNPs to the closest gene +/−50 kb from the SNP position and labelled them as located in a coding region, an intron, 1 kb upstream of a gene using our functional annotations (Fig. 3a). Remaining snps are labelled as located in a cis-region (up to +/−50 kb from a gene). NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 With the exception of height, there was no evidence for a relationship among the following groups: (i) SNPs in the exons of each gene and SNPs +/−50 kb outside of the exon and promotor regions; (ii) SNPs in the exons of each gene and SNPs in proximal promotors; and (iii) intronic SNPs and SNPs in promotor regions (Fig. 4d). This implies that trait associated SNPs in proximal and distal regulatory regions are largely independent of the effects of SNPs in their closest exon, as they do not align in terms of the variance they explain (Fig. 4d). For height, small weakly positive correlations across all gene regions in their contribution to variance, implies a degree of alignment across genes in regulatory variants and the closest exon (Fig. 4d). These results suggest a regulatory link between introns and distal cis regions outside of the promotor, or that introns may be correlated with structural variation. They also imply that the variance contributed by regulatory regions and those in the closest coding regions are not strongly coupled for these common complex traits. (95% CI 0.21, 0.32) to the phenotypic variance of height, the intronic region of FTO contributing 0.48% (95% CI 0.29, 1.12) to BMI, both the exonic-region and intronic-region of LPA con- tributing a combined 0.08% (95% CI 0.04, 0.13) to the risk of CAD, and the intronic region of TCF7L2 contributing 0.28% (95% CI 0.23, 0.35) to the risk of T2D (Fig. 4c, full results in Supplementary Data 6–9). Taken together, these results support an inﬁnitesimal contribution of many thousands of genes to common complex trait variation and give joint estimates of the proportions of variance contributed by each gene and their probability of association. (95% CI 0.21, 0.32) to the phenotypic variance of height, the intronic region of FTO contributing 0.48% (95% CI 0.29, 1.12) to BMI, both the exonic-region and intronic-region of LPA con- tributing a combined 0.08% (95% CI 0.04, 0.13) to the risk of CAD, and the intronic region of TCF7L2 contributing 0.28% (95% CI 0.23, 0.35) to the risk of T2D (Fig. 4c, full results in Supplementary Data 6–9). Taken together, these results support an inﬁnitesimal contribution of many thousands of genes to common complex trait variation and give joint estimates of the proportions of variance contributed by each gene and their probability of association. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 4 Contribution of genes and 50kb regions to height (HT), body-mass-index (BMI), cardiovascular disease (CAD) and type-2-diabetes (T2D). a We grouped SNPs in 50 kb-regions genome-wide and estimated the sum of the squared regression coefﬁcient estimates for each 50 kb-region. We then select the number of 50 kb regions that explain at least 0.001% of the variance attributed to all SNP markers in 80, 90 and 95% of the iterations. This gives a measure called the posterior probability that the window variance (PPWV)20 exceeds 1/10,000 of the phenotypic variation attributed to SNP markers. b We mapped SNPs to the closest gene +/−50 kb from the SNP position and labelled them as located in a coding region, an intron, 1 kb upstream of a gene using our functional annotations (Fig. 3a). Remaining snps are labelled as located in a cis-region (up to +/−50 kb from a gene). We then select the number of regions where PPWV is higher than 95% and explains at least 0.001 % of the phenotypic variance attributed to all SNP markers. We then calculate the number of signiﬁcant coding regions, introns, 1 kb regions and cis regions as a proportion of the total number of genes for each chromosome. Genic associations that explain at least 0.001% of the phenotypic variance attributed to all SNP markers are again spread across chromosomes according to the chromosome length. c Shows the mean of the phenotypic variance attributed to intron and cis regions (y-axis) and coding regions (x-axis) that explain at least 0.001% of the phenotypic variance attributable to SNP markers in ≥95% of the iterations (PPWV > 0.95). These results provide joint estimates of the proportions of variance contributed by different gene bodies and automatic ﬁne-mapping of gene bodies and their cis-regulatory regions. For example, introns and cis-regulatory regions of FTO respectively contribute 0.48% (95% CI 0.29, 1.12) and 0.01% (95% CI 0, 0.01) to the phenotypic variance of BMI. d We calculated the phenotypic variance contributed by exonic, intronic, promoter region and SNPs +/−50 kb outside of the exon and promotor regions (cis) for each gene. Bar plots show the correlation among the variance explained by the groups across genes. Error bars show the SD. s to height (HT), body-mass-index (BMI), cardiovascular disease (CAD) and type-2-diabetes (T2D). (Fig. 4d). NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 CAD T2D HT BMI 1 3 5 7 9 11 13 15 17 19 21 1 3 5 7 9 11 13 15 17 19 21 0 50 100 0 50 100 chromosome number of regions PPWV (0.8,0.9] (0.9,0.95] (0.95,1] a CAD T2D HT BMI 1 3 5 7 9 11 13 15 17 19 21 1 3 5 7 9 11 13 15 17 19 21 0.00 0.05 0.10 0.00 0.05 0.10 chromosome proportion of regions PPWV > 0.95 exons introns 1kb prom cis b HT BMI CAD T2D 0.00 0.05 0.10 0.15 0.20 0.0 0.2 0.4 0.6 0.00 0.05 0.10 0.15 0.0 0.1 0.2 correlation exons,introns exons,prom exons,cis introns,prom introns,cis prom,cis d ADAMTS17 GDF5 DIS3L2 CABLES1 BDNF−AS FTO CELSR2 PHACTR1 LPA CDKN2B−AS1 PAM HNF1A TCF7L2 T2D CAD BMI HT 0.00000 0.00005 0.00010 0.00015 0.00020 0.00000 0.00005 0.00010 0.00015 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00100 0.00200 0.00300 0.00400 0.00500 0.00000 0.00020 0.00040 0.00060 0.00000 0.00100 0.00200 by exons mean genetic variance explained by introns PPWV > 0.95 exons introns c ADAMTS17 GDF5 DIS3L2 CABLES1 BDNF−AS FTO CELSR2 PHACTR1 LPA CDKN2B−AS1 PAM HNF1A TCF7L2 T2D CAD BMI HT 0.00000 0.00005 0.00010 0.00015 0.00020 0.00000 0.00005 0.00010 0.00015 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00100 0.00200 0.00300 0.00400 0.00500 0.00000 0.00020 0.00040 0.00060 0.00000 0.00100 0.00200 by exons mean genetic variance explained by introns PPWV > 0.95 exons introns c IGF BMP2 IGF2BP2 SE TMEM18 A LDLR PLG LI CDKN2B− 0.00000 0.00005 0 0.00000 0.00005 0.00000 0.00010 0.00000 0.001 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00025 0.00050 0.00075 0.00000 0.00005 0.00010 0.00015 0.00000 0.00030 0.00060 0.00090 mean genetic variance explained by cis regions PPWV > 0 ADAMTS17 GDF5 IGF1R BMP2 IGF2BP2 BDNF−AS SEC16B uc021ukz.1 TMEM18 ALKAL2 CELSR2 LDLR PLG LINC00310 PAM HNF1A CDKN2B−AS1 T2D CAD BMI HT 0.00000 0.00005 0.00010 0.00015 0.00020 0.00000 0.00005 0.00010 0.00015 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00025 0.00050 0.00075 0.00000 0.00005 0.00010 0.00015 0.00000 0.00030 0.00060 0.00090 by exons mean genetic variance explained by cis regions PPWV > 0.95 exons cis ADAMTS17 GDF5 DIS3L2 CABLES1 BDNF−AS FTO CELSR2 PHACTR1 LPA CDKN2B−AS1 CAD BMI HT 00 05 10 15 0.00000 0.00010 0.00020 0.00030 0.00000 0.00100 0.00200 0.00000 0.00050 0.00100 0.00150 0.00200 0.00000 0.00100 0.00200 0.00300 0.00400 0.00500 0.00000 0.00020 0.00040 0.00060 mean genetic variance explained by introns PPWV > 0.95 exons introns c HT BMI CAD T2D 0.00 0.05 0.10 0.15 0.20 0.0 0.2 0.4 0.6 0.00 0.05 0.10 0.15 0.0 0.1 0.2 correlation exons,introns exons,prom exons,cis introns,prom introns,cis prom,cis d exons,introns exons,prom exons,cis introns,prom introns,cis prom,cis 0 0 by exons 0 0 by exons Fig. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 We ﬁnd many more associations in the cis region of genes with 1254 independent cis-regions for height with ≥95% posterior probability of explaining 0.001% of the h2 SNP, 1765 independent cis-regions for BMI, 1166 for CAD and 1221 for T2D. We additionally ﬁnd 9 independent promoter regions and 1072 independent introns for height with ≥95% posterior probability of explaining at least 0.001% of the h2 SNP, 1162 independent intronic gene regions for BMI, 307 for CAD and 347 for T2D. When we 7 7 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications ARTICLE NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 We then select the number of regions where PPWV is higher than 95% and explains at least 0.001 % of the phenotypic variance attributed to all SNP markers. We then calculate the number of signiﬁcant coding regions, introns, 1 kb regions and cis regions as a proportion of the total number of genes for each chromosome. Genic associations that explain at least 0.001% of the phenotypic variance attributed to all SNP markers are again spread across chromosomes according to the chromosome length. c Shows the mean of the phenotypic variance attributed to intron and cis regions (y-axis) and coding regions (x-axis) that explain at least 0.001% of the phenotypic variance attributable to SNP markers in ≥95% of the iterations (PPWV > 0.95). These results provide joint estimates of the proportions of variance contributed by different gene bodies and automatic ﬁne-mapping of gene bodies and their cis-regulatory regions. For example, introns and cis-regulatory regions of FTO respectively contribute 0.48% (95% CI 0.29, 1.12) and 0.01% (95% CI 0, 0.01) to the phenotypic variance of BMI. d We calculated the phenotypic variance contributed by exonic, intronic, promoter region and SNPs +/−50 kb outside of the exon and promotor regions (cis) for each gene. Bar plots show the correlation among the variance explained by the groups across genes. Error bars show the SD. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 For each gene, we also calculated the phenotypic variance contributed by exonic, intronic, promotor region, and cis SNPs and then calculated the correlation among the variances explained by the groups across genes. Across traits, we ﬁnd small positive correlations of the variance attributable to exonic and intronic regions of 0.17 (0.09, 0.24 95% CI) for height, 0.02 (0.001, 0.05 95% CI) for BMI, 0.103 (−0.007, 0.71 95% CI) for CAD, and 0.064 (0.01, 0.19 95% CI) for T2D. Similarly, we ﬁnd small positive correlations between introns and cis regions Finally, our approach provides automatic ﬁne-mapping of SNP loci, and of these region-level and gene-level associations, 360 8 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 individuals to their overall genetic value, as expected under a highly polygenic model. Table 2 Summary of ﬁndings for height (HT), body mass index (BMI), type-2 diabetes (T2D) and cardiovascular disease (CAD). Findings Method HT BMI CAD T2D Associated SNPs COJO-plink2 1673 517 34 85 COJO- BoltLMM 2131 565 34 84 COJO- Regenie 2134 555 34 82 50 kb regions (PPWV ≥95%) BayesRR-RC 1660 520 70 87 Genic regions (PPWV ≥95%) BayesRR-RC 2578 2956 1478 1581 Exons 243 29 5 13 Introns 1072 1162 307 347 cisa 1254 1765 1166 1221 SNPs (PIP ≥95%) BayesRR-RC 360 20 2 9 Exons 216 16 1 4 Introns 73 2 1 5 10–500 kb 48 1 0 0 LD clumps with r2 = 0.1 (PPWV ≥95%) BayesRR-RC 1220 206 16 19 aSNPs located up to +/−50 kb from the closest gene. Table 2 Summary of ﬁndings for height (HT), body mass index (BMI), type-2 diabetes (T2D) and cardiovascular disease (CAD). g y p yg Additionally, for height and BMI we also determined the proportion of the posterior predictive distribution for each individual that was within +/−1 SD of their true phenotypic value. On average 67.5% of an individuals posterior predictive distribution is within +/−1 SD of their true phenotype for BMI and 75% for height, with similar prediction accuracy across individuals (Supplementary Fig. 14c). For T2D and CAD, we extended the PCF metric, typically deﬁned as the proportion of cases with larger estimated risk than the top pth percentile of the distribution of genetic risk in the general population. For each individual, we calculated the proportion of their posterior predictive distribution that falls above the top 25% of the distribution of genetic risk in the general population. Discussion There is no single statistical model appropriate for all settings and thus there will always be a situation where a model poorly ﬁts the data. We have provided theoretical and empirical evidence that a grouped Dirac spike-and-slab model (which we term BayesRR- RC), has a prior that is ﬂexible enough to show robust model performance across the data analysed here, improving inference in many settings over commonly applied approaches. We develop a range of computational and statistical approaches which allow this, or any similar Gibbs sampling algorithm, to scale to whole genome sequence data on many hundreds of thousands of indi- viduals. This has enabled us to compare and contrast the inferred underlying genetic distribution for four complex phenotypes under this prior, providing novel insight into the genetic archi- tecture of these traits. We observe that all phenotypes simply appear to be predominantly underlain by very many common variants, with SNPs within distal regulatory regions, coding and intronic regions each contributing more to the phenotypic var- iance and having higher allele substitution effects. Out-of-sample prediction into another European healthcare system. We generated a full posterior predictive distribution for each trait in each of 32,500 individuals from the Estonian Gen- ome Centre data, which allows the transmission of uncertainty in the marker effect estimates from the UK Biobank to the genomic predictors created in Estonia. First, despite this study having almost half the sample size, we show improved genomic predic- tion as compared to recently proposed summary statistic approaches23, when taking the mean of the predictor across iterations and correlating this with the phenotype with correla- tion of 0.62 for height, 0.34 for BMI, 0.16 for T2D, and 0.07 for CAD (Supplementary Fig. 14a). The area under the receiver operator curve (AUC) for T2D was 0.67 and 0.57 for CAD. In comparison, using the 64 BLD-LDAK annotations recommended by a recent study21, the highest prediction accuracy obtained from MegaPRS was 0.55 for height, 0.32 for BMI, 0.10 for T2D, and 0.05 for CAD. We then estimated the distribution of the partial correlations between the trait and genomic predictors created from our different annotation groups and ﬁnd that exonic, intronic, and 10–500 kb upstream regions contribute proportionally more to the prediction accuracy than other genomic groups, replicating our results from the UK Biobank (Supplementary Fig. 14). NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 The distribution of these probabilities is shown for conﬁrmed cases and those without diagnosis in the Estonian Biobank (Supple- mentary Fig. 14d). We ﬁnd 25 individuals for T2D and 15 individuals for CAD where ≥90% of their posterior predictive distribution is within the high risk group of which 40 and 18% are currently deﬁned as cases for T2D and CAD, respectively based on recent medical records. This is compared to 1% and 2% case rate for those with ≤10% probability of being in the high risk group for T2D and CAD respectively, giving an odds ratio of 20 and 18 between the ≥90% and ≤10% groups. However, our results clearly show that the individual-level sensitivity and speciﬁcity of genomic prediction for these common complex diseases is very poor, as 75% of T2D cases and 92% of CAD cases have ≤50% of their distribution within the high-risk category. These results highlight how variation contained within a posterior predictive distribution that is typically ignored in human genomic prediction can be used. We show that genomic prediction for personalised medicine with patient-speciﬁc predictions or stratiﬁcation of patients is currently extremely limited. SNPs for height, 20 for BMI, 2 for CAD and 9 for T2D could be mapped to a single SNP with greater than 95% inclusion probability across all four chains (Supplementary Data 10 and Supplementary Fig. 13). Of these ﬁne-mapped SNPs, only 53.45% are top loci with a p-value < 5 × 10−8 from the fastGWAS UK Biobank summary statistic data for standing height, BMI, angina/ heart attack and type-2 diabetes (fastGWA, see “Code avail- ability”). This highlights that selecting on the top SNP markers identiﬁed by standard association studies would give a different set of variants than those obtained from selecting high PIP SNPs. NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications Methods BayesRR-RC model. We extend the BayesR model to a BayesRR-RC model as follows y ¼ 1μ þ ∑ Φ φ¼1 Xφβφ þ ϵ; ð3Þ ð3Þ where there is a single intercept term 1μ and a single error term ϵ but now SNPs are allocated into groups (φ1, …, φΦ), each of which having it’s own set of model parameters Θφ ¼ fβφ; πβφ; σ2 Gφg. As such, each βφj is distributed according to: βφj π0φδ0 þ π1φN 0; σ2 1φ þ π2φN 0; σ2 2φ þ ¼ þ πLφN 0; σ2 Lφ ; ð4Þ ð4Þ where for each SNP marker group fπ0φ; π1φ; ¼ ; πLφg are the mixture proportions and fσ2 1φ; σ2 2φ; ¼ ; σ2 Lφg are the mixture-speciﬁc variances prop ortional to σ2 1φ ... σ2 Lφ 2 6664 3 7775 ¼ σ2 βφ C1φ ... CLφ 2 6664 3 7775 l d Additionally, in this work we do not extend past a limited number of functional annotations and thus we do not provide a model capable of further partitioning the variation into speciﬁc regulatory functions (eQTL, mQTL, pQTL etc.) or directly modelling the relationships among components. LDSC functional methods take the approach that SNPs can be assigned to different categories (e.g., both coding and conserved), with the categories competing against each other to explain the signal, with the downside that enrichment is relative and that the total variance is not partitioned. Here, the total variance is partitioned but this is based on preferential allocation of SNPs to coding regions, then introns, and then to their nearest upstream gene position. These SNPs are most likely to be allocated accurately, with 1 and 1–10 kb groups being more ambiguous in high gene density regions and likely mislabelled. However, if this was the case then variance would still be partitioned to these mislabelled groups and it would just be evenly split across them, with experimentally validated promotor, enhancer and tfbs regions assisting to some degree in alleviating this. Rather, here we see a clear pattern of increasing variance contributed, increasing average effect size, and an increasing pattern of higher rare allele substitution effects by individual markers as distance from the nearest gene increases. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 phenotypic variance remains a substantial challenge and our results suggest a predominant role for introns and for distal, and thus likely more global enhancers, rather than locally dominant proximal expression QTL. The recent “omnigenic” model24, suggests that trait- associated variants in regulatory regions inﬂuence a local gene which is not directly causal to the disease, and also co-regulate other disease causal genes (or “core” gene). Our ﬁndings of little correlation of exonic and proximal regulatory variance and a large number of trait- associated intronic and cis regions do not rule this out, but suggest a more complex inﬁnitesimal picture with differences occurring among traits, potentially due to their evolutionary history. upstream region variance is allocated to, our inference that genic regions are uncorrelated in their contribution to variance with the promotor and upstream regions still holds as does our prob- abilistic inference on the associations of each gene and their contribution to the phenotypic variation. Our results provide evidence for an inﬁnitesimal contribution of many thousands of common genomic regions to common complex trait variation and for a predominant role of intronic, exonic, and distal regulatory regions. This highlights the immense challenge of understanding the molecular underpinning of each association and the difﬁculties in improving the estimation of many tens of thousands of small-effect associations that are required to improve genomic prediction. This work represents a step toward maximising the probabilistic inference that can be obtained from large-scale Biobank studies. p y y y There are important caveats and limitations to consider. Here, we present an approach for analysing large-scale biobank data, which is becoming increasingly available, However, a substantial number of GWAS have already been conducted, with associated published genome-wide summary association statistic estimates. Many methods have been developed to take advantage of these estimates, with downstream analysis models making use of var- ious summary statistics resources in efﬁcient and ﬂexible ways. We show here that two leading summary statistic approaches perform poorly as compared to individual-level models for esti- mation of enrichment and genomic prediction. Despite this, the sample sizes obtained in consortia study meta-analyses will exceed those from single biobanks, especially for disease, and thus the genomic prediction accuracy of consortia study meta-analysis summary statistic prediction models may exceed those from individual-level analyses. Combining the posterior distribution obtained from BayesRR-RC across different individual-level bio- bank studies would alleviate this issue. ARTICLE ARTICLE Methods p g We can sketch the difference in the models by looking at the respective conditional posteriors, again, assuming a single component for simpliﬁcation purposes We have a BayesRC or BayesRS estimator by assuming different groups g can sketch the difference in the models by looking at the respect We can sketch the difference in the models by looking at the respective conditional posteriors, again, assuming a single component for simpliﬁcation purposes. We have a BayesRC or BayesRS estimator by assuming different groups of effects as described in Supplementary Note 4 Eq. 35, which yields: f α; γjπβφ; σ2 β; σ2 ϵ; y / exp 1 2σ2ϵ jjy Xγ≠0αγ≠0jj2 2 1 2σ2 β jjαjj2 2 log 1πβφ πβφ jjγφjj0 ; ð5Þ where πβφ are the group-speciﬁc mixture proportions and ∣∣γφ∣∣0 is the cardinality of the group. The corresponding MAP estimate would amount to adding extra penalisation on sparsity through the πφ terms, while keeping the same level of shrinkage as the baseline BayesR. g y In our model the conditional posterior is: f α; γjπβφ; σ2 βφ; σ2 ϵ; y / exp 1 2σ2 ϵ jjy Xγ≠0αγ≠0jj2 2 1 2σ2 βφ jjαjj2 2 log 1 πβφ πβφ ! jjγφjj0 ( ) f α; γjπβφ; σ2 βφ; σ2 ϵ; y / exp 1 2σ2 ϵ jjy Xγ≠0αγ≠0jj2 2 1 2σ2 βφ jjαjj2 2 log 1 πβφ πβφ ! jjγφjj0 ( ) ð6Þ ð6Þ now each marker has a group-speciﬁc shrinkage σ2 βφ, which translates to a speciﬁc λφ per group in the MAP estimate. This amounts to markers being shrunk according to the scale of the effects of their group, instead of the scale of all other markers. So instead of solving a single model selection and regularisation problem we are solving Φ model selection and regularisation problems, with shared information only through the residuals If we subset by MAF and LD bins the now each marker has a group-speciﬁc shrinkage σ2 βφ, which translates to a speciﬁc λφ per group in the MAP estimate. This amounts to markers being shrunk according to the scale of the effects of their group, instead of the scale of all other markers. Methods 10–500 kb distal regions may contribute more variance as marker density and marker coverage is higher in these regions, with missing variation within 10 kb upstream as causal variants are poorly correlated with SNPs. The posterior distributions for the variance explained by 1 kb, 1–10 kb regions, and 10–500 kb regions are negatively correlated (Supplementary Fig. 8, meaning that these groups are competing with each other, as if variance goes to one then it is being taken away from the other because they are in LD), and thus there is the risk that the model cannot separate these effectively. However, this is true of any enrichment analysis conducted to date and we can only make inference in the data that we have currently available. Resolving this requires the application of this model to whole genome sequence data where the total variance can be partitioned across upstream regions without marker coverage concerns. Irrespective of exactly which Thus the mixture proportions, variance explained by the SNP markers, and mixture constants are all unique and independent across SNP marker groups. This extends previous models (known as BayesRC25 and BayesRS26), which have used dditi l i t f diff t SNP b t k t i l l b l i Thus the mixture proportions, variance explained by the SNP markers, and mixture constants are all unique and independent across SNP marker groups. This extends previous models (known as BayesRC25 and BayesRS26), which have used additional mixtures for different SNP groups, but kept a single global variance component. Importantly, a single variance component with more mixtures serves only to change the amount of mass allocated at different sizes of the distribution, but does not alter the sizes of the effects themselves as there is still a single distribution. In contrast, the formulation presented here of having an independent variance parameter σ2 βφ per group of markers, and independent mixture variance components, enables estimation of the amount of phenotypic variance attributable to the group-speciﬁc effects and enables differences in the distribution of effects among groups. In this work, we use 78 SNP marker groups, each with ﬁve mixture components (including 0). Discussion We ﬁnd evidence for zero/low correlations of genomic predictors created from different annotation groups, which supports our results from the UK Biobank (Supplementary Fig. 14e). This suggests that individuals have a different portfolio of risk variants, with different genomic regions contributing for different There has been debate on how to best estimate SNP heritability1,3,4 and here we validate that one approach could be to split SNP mar- kers by LD to improve genetic effect size estimates. Our results suggest that the proportion of genomic variation attributable to mutations in regulatory regions and mutations in the closest genic regions are largely independent. Additionally our model tests asso- ciation within groups in a probabilistic way and we ﬁnd 290 inde- pendent coding, 2888 independent intronic, and 5406 independent cis regions with ≥95% probability of contributing at least 0.001% of the SNP heritability. Understand how these coding, intronic and proximal and distal regulatory regions combine to contribute to 9 9 Simulation study Genetic architecture. We ﬁrst compare the model performance of BayesRR-RC to existing approaches across 18 different genetic architectures. We randomly selected 40,000 unrelated UK Biobank individuals and used 596,741 imputed SNP markers from chromosomes 19 to 22. We randomly selected either 1000, 10,000, or 100,000 LD independent (LD R2 < 0.1) causal SNP markers. For each SNP marker set there were two settings. We then ran the following prediction models, using a testing set of 10,000 UK Biobank unrelated individuals, that were also unrelated to the training data, and focusing on the models proposed in a recent paper21. These methods contain two approximations to our BayesRR-RC model and the authors claim to outperform all other existing methods, including individual-level models. The models are: g In the ﬁrst setting, we simulated effect sizes from a normal distribution with zero mean and variance of 0.1, 0.3, or 0.6 divided by the number of causal variants ∝N(0, [p(1−p)]−0.25), with p the allele frequency. We sampled individual-level environmental (residual) variance from a normal distribution with zero mean and variance equal to 1 minus either 0.1, 0.3, or 0.6 to give phenotypes with zero mean and unit variance. This gave h2 SNP = 0.1, 0.3, or 0.6 and simulates stronger effect sizes for rare variants in line with recent empirical estimates. We simulated ten replicate phenotypes for each of the nine different genetic architectures. In the second setting, we repeat each simulation, sampling the SNP marker effects from 13 different normal distributions, one for each of 13 different genomic annotation groups described in the main text. The 13 groups were allocated different proportions of the h2 SNP as follows: for exonic variants Pðh2 SNPÞ = 0.167, intronic variants Pðh2 SNPÞ = 0.334, 1 kb promotor variants Pðh2 SNPÞ = 0.0835, 1–10 kb enhancer variants Pðh2 SNPÞ = 0.04175, 1–10 kb transcription factor binding sites Pðh2 SNPÞ = 0.04175, 1–10 kb other variants Pðh2 SNPÞ = 0, 10–500 kb enhancers Pðh2 SNPÞ = 0.0835, 10–500 kb transcription factor binding sites Pðh2 SNPÞ = 0.0835, 10–500 kb other variants Pðh2 SNPÞ = 0, 500 kb–1 Mb enhancers Pðh2 SNPÞ = 0.0835, 500 kb–1 Mb transcription factor binding sites Pðh2 SNPÞ = 0.0835, 500 kb–1 Mb other variants Pðh2 SNPÞ = 0, and other non-annotated SNPs Pðh2 SNPÞ = 0. Simulation study For each of the 13 groups marker effects were simulated as ∝N(0, [p(1−p)]−0.25) to give h2 SNP = 0.1, 0.3, or 0.6, with stronger effect sizes for rare variants. Four of these 13 groups had zero variance indicating that no associations were created for these groups. h h ﬁ l l d b ● An individual-level bayesR model using genomic annotation SNP variance estimates from the SumHer models as implemented in the software MegaPRS21. This provides estimates of the SNP marker effects for creating a genetic risk predictor. g p ● An individual-level boltREML model using genomic annotation SNP variance estimates from the SumHer models as implemented in the software MegaPRS21. This provides estimates of the SNP marker effects for creating a genetic risk predictor. g g p ● A summary statistic bayesR model using genomic annotation SNP variance estimates from the SumHer models as implemented in the software MegaPRS21. This provides estimates of the SNP marker effects for creating a genetic risk predictor. g p ● A summary statistic boltREML model using genomic annotation SNP variance estimates from the SumHer models as implemented in the software MegaPRS21. This provides estimates of the SNP marker effects for creating a genetic risk predictor. First, we compared the correlation of the simulated and estimated proportion of phenotypic variance attributable to the 13 genomic annotation groups across all models in Fig. 1. We determined the ability of the approaches to correctly identify enriched regions of the DNA by estimating the probability within each simulation replicate that a SNP marker group would have an estimated enrichment of ≥2 (i.e., being described as having average effect sizes that are twice as large as expected) when the simulated value was ≤1.1. As BayesRR-RC induces sparsity in the SNP effect estimates, with some markers always remaining in the variance = 0 spike, we propose a different enrichment deﬁnition where the proportion of h2 SNP is divided by the proportion of markers that are in the model for the SNP group, rather than the proportion of markers mapping to the SNP group. Thus, in the ﬁrst setting we simulate variance explained by annotation groups that is on average proportional to the number of SNPs within each annotation (due to the random allocation of SNPs and effect sizes). In the second setting, the variance and average effect size differ across annotation groups. Simulation study We refer to these as two different enrichment settings: “random”, or “enriched”. g For these 180 phenotypes, we ran the following individual-level models: For these 180 phenotypes, we ran the following individual-level models: ● A restricted maximum likelihood model implemented in the software GCTA with a single relationship matrix providing an estimate of the variance attributable to SNPs genome-wide. In Supplementary Note 3, we propose a posterior probability window variance (PPWV) approach20, which provides a probabilistic determination of association of a given LD block, genomic window, gene, or upstream region, relative to the amount of phenotypic variation attributable to that window. Our PPWV approach determines the posterior inclusion probability that each region and each gene contributes at least 0.001% to the h2 SNP, with theory and small-scale simulations outlined in Supplementary Note 3 suggesting well controlled FDR. We partitioned the 596,741 imputed SNP markers in LD blocks, deﬁned as groups of markers with LD R2 ≥0.1. Within each simulation replicate, we estimated the probability that LD blocks containing a causal variant were identiﬁed by PPWV. We compared this to MLMA estimates obtained using the BoltLMM software, by estimating the probability that LD blocks containing a causal variant were identiﬁed as having a SNP with p-value ≤5 × 10−8, the standard genome-wide signiﬁcance threshold. We present these results in Fig. 2a. ● A restricted maximum likelihood model implemented in the software BoltREML17. Here, we used a 78 MAF-LD-annotation group model using the non-overlapping genomic annotation groups described below in the UK Biobank analysis providing an estimate of the variance attributable to SNPs genome-wide and an estimate of the variance attributable to SNP markers of each annotation group. g p ● A Haseman-Elston regression using the same 78 group model implemented in the software RHEmc18, providing an estimate of the variance attributable to SNPs genome-wide and an estimate of the variance attributable to SNP markers of each annotation group. g p ● Mixed linear association model (MLMA), which is a two-stage approach where the variance attributable to the SNP markers genome-wide is estimated and this estimate is then used in a second generalised least squares step to test for SNP-phenotype associations one marker at a time. There are two forms of this model. In the ﬁrst, the SNP is ﬁtted twice as it is included in both the ﬁxed and random terms (MLMAi). ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 calculating the taggings as we found this gave the best estimates we could obtain from the simulated data across all scenarios. We set the relationship of effect size and minor allele frequency to be −0.25 as suggested by the authors and which matches the simulation setting. This model is intended to approximate an individual-level REML analysis with 78 annotations, but using a different scaling of the relationship matrix, and provides an estimate of the variance attributable to SNPs genome-wide and an estimate of the variance attributable to SNP markers of each annotation group. it is showed that the former case requires weaker conditions in order to recover the true vector β consistently than the latter. Although the sampling scheme was different, we have shown that a similar model with only two groups: genetic markers and epigenetic markers, is successful in identifying BMI and smoking epigenetic signatures13. The baseline model derivations for this model are outlined in Supplementary Note 1, a BSP Gibbs sampling scheme and an assessment of its performance is outlined in Supplementary Note 2, and an assessment of the model performance with correlated covariates is outlined in Supplementary Note 4. g p ● Our BayesRR-RC model implemented in GMRM with 78 SNP-marker groups and run for 5000 iterations with a burn-in period of 2000 iterations. ● Our BayesRR-RC model implemented in GMRM with only a single SNP- marker group, which is equivalent to BayesR, run for 5000 iterations with a burn-in period of 2000 iterations. g p ● Our BayesRR-RC model implemented in GMRM with 78 SNP-marker groups and run for 5000 iterations with a burn-in period of 2000 iterations. ● Our BayesRR-RC model implemented in GMRM with only a single SNP- marker group, which is equivalent to BayesR, run for 5000 iterations with a burn-in period of 2000 iterations. Simulation study In the second, the SNP to be tested as ﬁxed is removed from the random term alongside those on the same chromosome (MLMA). We used the software BoltLMM8, Regenie9, and GCTA to ﬁt these models. These approaches provided estimates of the SNP regression coefﬁcients (marker effect sizes). 28 We then compare the prediction accuracy obtained in a testing set of 10,000 UK Biobank unrelated individuals, that were also unrelated to the training data. We predicted phenotype using SNP marker effect sizes obtained from BayesRR-RC, MLMA effect sizes from BoltLMM, and the four MegaPRS methods outlined above implemented in the LDAK software. While we would suggest that ﬁxed-effect MLMA estimates are improper for prediction we include this comparison as polygenic risk scores have often been created from ﬁxed-effect SNP estimates. We calculate the correlation between the simulated phenotype in the testing set and the genomic predictor within each simulation replicate and we compare the mean correlation across the 18 different genomic annotations in Fig. 2. Additionally, to provide a benchmark, we compare to the theoretical expectation under ridge regression approximations29, with the number of markers set to the number of causal variants. estimates of the SNP regression coefﬁcients (marker effect sizes). ● Single marker marginal least squares regression using plink228, whilst ﬁtting 20 principal components of the marker data as covariates. 19 ● Linkage disequilibrium score regression (LDSC19), with LD scores calculated using the same data, and the same 78 non-overlapping annotations in a 78 component LDSC annotation model. We included SNPs with MAF > 1% following the software instructions. This model is intended to approximate an individual-level REML analysis with 78 annotations and provides an estimate of the variance attributable to SNPs genome-wide and an estimate of the variance attributable to SNP markers of each annotation group. Relationship between effect size, minor allele frequency and LD. We then conducted another large-scale, but this time well-powered simulation study, where we ascertained the causal variant SNP markers in different ways and varied the rela- tionship between effect size, minor allele frequency and LD. We used the same randomly selected 40,000 unrelated individuals and all 596,741 imputed (version 3) genetic markers from chromosomes 19 through 22 from the UK Biobank. We g p ● We used the software SumHer6. We calculated marker taggings under the same 78 component annotation model. Methods So instead of solving a single model selection and regularisation problem now each marker has a group-speciﬁc shrinkage σ2 βφ, which translates to a speciﬁc now each marker has a group-speciﬁc shrinkage σ2 βφ, which translates to a speciﬁc λφ per group in the MAP estimate. This amounts to markers being shrunk according to the scale of the effects of their group, instead of the scale of all other markers. So instead of solving a single model selection and regularisation problem we are solving Φ model selection and regularisation problems, with shared information only through the residuals. If we subset by MAF and LD bins, the resulting groups of columns will have a correlation pattern similar to an exponential decay (LD decays with distance). If we take the whole genotype matrix, the pattern would be closer to a block diagonal matrix of correlations, in refs. 16,27 g p p g βφ p λφ per group in the MAP estimate. This amounts to markers being shrunk according to the scale of the effects of their group, instead of the scale of all other markers. So instead of solving a single model selection and regularisation problem we are solving Φ model selection and regularisation problems, with shared information only through the residuals. If we subset by MAF and LD bins, the resulting groups of columns will have a correlation pattern similar to an exponential decay (LD decays with distance). If we take the whole genotype matrix, the pattern would be closer to a block diagonal matrix of correlations, in refs. 16,27 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 10 ARTICLE Selecting only the highest frequency marker per LD block creates a setting where for each set of markers in LD with each other, there is only one causal genetic variant, and where the distribution of associated markers differs to that of the SNP markers as a whole. We then extend our model comparisons in a number of ways. While direct comparisons of frequentist and Bayesian approaches are difﬁcult and often ill advised, we wished to show that BayesRR-RC provides accurate effect size estimation in the presence of LD. We provide three simple comparable metrics to assess model performance of BayesRR-RC against frequentist mixed linear association models (MLMA) applied as two-stage approaches, where either the SNP is ﬁtted twice as it is included in both the ﬁxed and random terms (MLMAi implemented in GCTA), or the SNP to be tested as ﬁxed is removed from the random term alongside those on the same chromosome (MLMA implemented in BoltLMM and Regenie). ● Having created four different ways of selecting associated markers (5000 or 10,000 and high-MAF or random) we then created ﬁve different ways of assigning effect sizes to them: ● Having created four different ways of selecting associated markers (5000 or 10,000 and high-MAF or random) we then created ﬁve different ways of assigning effect sizes to them: First, we calculated z-scores of the marker effect estimates from their true simulated value. As MLMA approaches estimate marker effects one-at-a-time, we calculated the z-score of the estimate from the true simulated value for the causal variants in each simulation replicate, across generative genetic models. For the Bayesian methods, at any one iteration, LD among the markers is controlled for (see Supplementary Note 4). However across iterations as the chain mixes, markers in LD will enter and leave the model, with their posterior inclusion probabilities reﬂecting their association with the trait. Thus, we summed the squared regression coefﬁcient estimates of SNPs in the model at each iteration for each LD block (markers in LD R2 ≥0.1 within 1 MB) of each simulation replicate, took the posterior mean across iterations, and then calculated the z-score of the estimate from the simulated value. This metric provides an assessment of the ability of BayesRR-RC to accurately estimate the contribution of a genomic region to the phenotypic variance as compared to MLMA approaches. We present these results in Supplementary Fig. ARTICLE 2, where we ﬁnd that the z-scores of the estimated BayesRR- RC effects are generally stable across generative genetic models and comparable to those obtained from BayesR but with slightly elevated variance in many cases as the model is less sparse (Supplementary Fig. 2a). We ﬁnd that SNP effect size estimates from MLMA models have higher estimation error, especially when the causal variant is rare, or in high-LD with many other SNPs (Supplementary Fig. 2a). MLMAi models show lower estimation error than MLMA approaches, likely as they control for both distant and local LD (Supplementary Fig. 2a). We explore this further in Supplementary Note 4. – We simulated effect sizes from a normal distribution with zero mean and variance 0.6 divided by the number of markers (5000 or 10,000) with no relationship to the LD or MAF of the markers. Thus, effects had variance ∝N(0, w0[p(1−p)]0) with w the LD score of the marker and p the allele frequency. y – We simulated effect sizes from a normal distribution with zero mean and variance 0.6 divided by the number of markers (5000 or 10,000) ∝N(0, w −0.25[p(1−p)]−0.25). This simulates stronger effect sizes for rare variants and those in low LD. – We simulated effect sizes from a normal distribution with zero mean and variance 0.6 divided by the number of markers (5000 or 10,000) ∝N(0, w0.25[p(1−p)]−0.25). This simulates stronger effect sizes for rare variants and those in high LD. – We simulated effect sizes from a normal distribution with zero mean and variance 0.6 divided by the number of markers (5000 or 10,000) ∝N(0, w −0.25[p(1−p)]0.75). This simulates equivalent effect sizes for common and rare variants, and greater effects for markers in low LD. – We simulated effect sizes from a normal distribution with zero mean and variance 0.6 divided by the number of markers (5000 or 10,000) ∝N(0, w0.25[p(1−p)]0.75). This simulates equivalent effect sizes for common and rare variants, and greater effects for markers in high LD. Second, to further test our PPWV approach we calculated precision-recall curves, where associations are deﬁned as LD blocks with PPWV of ≥95% at 0.001% proportion of variance explained. True positives were the number of identiﬁed 5000 or 10,000 LD blocks that contained a causal variant. False positives were the number of identiﬁed LD blocks that did not contain a causal variant. ARTICLE In the second setting, for each of the ﬁve different effect size sampling schemes we simulated effects from 13 different distributions, one for each of 13 different genomic annotation groups with different proportions of total SNP heritability (h2 SNP). For each of the ﬁve different effect size sampling schemes the relationship to LD and MAF remained the same, but the total variance attributed to the SNP markers was partitioned across annotation groups as follows for exonic variants (h2 SNP = 0.1), intronic variants (h2 SNP = 0.2), 1 kb promotor variants (h2 SNP = 0.05), 1–10 kb enhancer variants (0.025), 1–10 kb transcription factor binding sites (h2 SNP = 0.025), 1–10 kb other variants (h2 SNP = 0), 10–500 kb enhancers (h2 SNP = 0.05), 10–500 kb transcription factor binding sites (h2 SNP = 0.05), 10–500 kb other variants (h2 SNP = 0), 500 kb–1 Mb enhancers (h2 SNP = 0.05), 500 kb–1 Mb transcription factor binding sites (h2 SNP = 0.05), 500 kb-1 Mb other variants (h2 SNP = 0), and other non-annotated SNPs (h2 SNP = 0). Four of these distributions had zero variance indicating that no associations were created for these groups. In the ﬁrst setting, this simulates variance explained by annotation groups that is on average proportional to the number of SNPs within each annotation. In the second scheme, the variance and average effect size differs across annotation groups. We refer to these as two different enrichment settings: “random”, or “enriched”. ● This created four different sets of associated markers (5000 or 10,000 and high- MAF or random), each with ﬁve different marker effect size sampling schemes, which we refer to in the main text as the 20 different generative genetic models (Table 1), each of which has two enrichment settings. This gave 40 different sampling schemes for the genetic effects and we simulated ten replicates for each setting, giving a total set of 400 simulated phenotypes. ● This created four different sets of associated markers (5000 or 10,000 and high- MAF or random), each with ﬁve different marker effect size sampling schemes, which we refer to in the main text as the 20 different generative genetic models (Table 1), each of which has two enrichment settings. This gave 40 different sampling schemes for the genetic effects and we simulated ten replicates for each setting, giving a total set of 400 simulated phenotypes. ARTICLE Precision was deﬁned as the ratio of true positives to the sum of true positives and false positives. Recall was deﬁned as the ratio of true positives to the sum of true positives plus false negatives. The FDR was deﬁned as the proportion of LD blocks with PPWV of ≥95% at 0.001% proportion of variance explained that did not contain a causal variant. For the MLMA methods, following standard practice, we clumped the marker effect estimates using Plink, as local LD is not controlled for, selecting LD independent markers (LD R2 ≤0.01 with other markers) across the genome. True associations were deﬁned as selected SNPs that were in LD with a simulated causal variant (LD R2 ≥0.01). False associations were deﬁned as selected SNPs that were not in LD (LD R2 ≤0.01) with a simulated causal variant. Precision and recall were calculated across thresholds of the chi-squared statistics of the selected markers, and the area under the curve was calculated using the trapezoid rule for calculating the integrals, assuming the curve is linear between the points. FDR is then calculated as the proportion of markers with p-value ≤5 × 10−8 that were not in LD with a causal variant (LD R2 ≥0.01). This provides a way to directly compare model performance for the discovery of associated genomic regions across Bayesian and frequentist approaches and tests our hypothesis that a PPWV approach controls FDR well in comparison with Bonferroni p-value correction (Supplementary Fig. 2b, c). For both MLMA and Bayesian approaches our deﬁnition of FDR is not strictly the FDR. Markers in LD R2 ≤0.01 with the clumped selected markers may still show a weak correlation with the simulated causal variants, and likewise blocks of SNPs in LD R2 ≤0.1 may still be in weak LD with the causal variants. Our approach instead captures the ability of MLMA and Bayesian approaches to localise an effect within R2 ≥0.01 and R2 ≥0.1 respectively. We present these results in Supplementary Fig. 2. ● For each of the four different sets of markers, each with ﬁve different effect size sampling schemes, we then created two additional settings. In the ﬁrst setting markers were sampled from the various normal distribution, as described above, for the ﬁve different effect size sampling schemes. ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 simulated a wide-range of different possible underlying genetic effect size dis- tributions as follows: variance attributable to the SNP markers, for testing association of genetic markers with phenotypes genome-wide, and for genomic prediction. simulated a wide-range of different possible underlying genetic effect size dis- tributions as follows: y g g This simulation provides a range of different scenarios for which we can explore the model performance of BayesRR-RC and compare it to existing approaches. In Supplementary Fig. 1, we compare the h2 SNP estimation, estimation of the annotation genetic variance along with the RMSE of the estimates, and the estimated average effect size. ● We chose either 5000 or 10,000 imputed SNP markers for which to assign a genetic effect size, providing two different levels of polygenicity. ● We chose either 5000 or 10,000 imputed SNP markers for which to assign a genetic effect size, providing two different levels of polygenicity. g p g p yg y ● We selected these 5000 or 10,000 markers in two different ways. Either, we selected SNPs at random, or we selected the marker of highest minor allele frequency per LD block of the genome, with an LD block deﬁned as a group of SNP markers with absolute LD of at least 0.05. Randomly allocating markers creates a set of associated variants with the same distribution of LD and MAF as the SNP data, which is composed of predominantly low frequency variants. Selecting only the highest frequency marker per LD block creates a setting where for each set of markers in LD with each other, there is only one causal genetic variant, and where the distribution of associated markers differs to that of the SNP markers as a whole. ● We selected these 5000 or 10,000 markers in two different ways. Either, we selected SNPs at random, or we selected the marker of highest minor allele frequency per LD block of the genome, with an LD block deﬁned as a group of SNP markers with absolute LD of at least 0.05. Randomly allocating markers creates a set of associated variants with the same distribution of LD and MAF as the SNP data, which is composed of predominantly low frequency variants. Simulation study We ignored the LD weights when 11 TURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 We then investigated the importance of controlling for multicollinearity for the control of population genetic and data structure effects. In principle, a MLMA approach will control for bias with correlated markers (either local or long-range LD) ﬁtted as random when testing for the effects of a focal SNP. For two markers, X1 and X2 in LD correlation ρX1;X2, with β2 = 0 we can express the MLMA ﬁxed effect solution as a partial regression coefﬁcient of the phenotype regressed onto the focal SNP after adjusting for X2 estimated as uX2 ¼ XT 2 y XT 2 X2þλI. Following our derivation above for a shrinkage esti- mator of a partial regression coefﬁcient the effect size of X1 is estimated as ρ 1X y ^βy;X1jX2 ¼ N XT 1 X1 ´ ρy;X1 ρX1;X2 1 NX2y 1ρX1;X2 and in this two-SNP example the bias is ^βy;X1jX2 ¼ N XT 1 X1 ´ ρy;X1 ρX1;X2 NX2y 1ρX1;X2 and in this two-SNP example the bias is accounted for in the term ρX1;X2 1 NX2y 1ρX1;X2 when the ﬁxed effect is estimated. Multi- T accounted for in the term ρX1;X2 1 NX2y 1ρX1;X2 when the ﬁxed effect is estimated. Multi- T accounted for in the term ρX1;X2 1 NX2y 1ρX1;X2 when the ﬁxed effect is estimated. Multi- T accounted for in the term ρX1;X2 NX2y 1ρX1;X2 collinearity acts to increase the σG term of λ, reducing the denominator XT 2 X2 þ λI in the estimation of uX2, and increasing the variance of the estimates of common markers in high LD, those with the highest average FST. collinearity acts to increase the σG term of λ, reducing the denominator XT 2 X2 þ λI in the estimation of uX2, and increasing the variance of the estimates of common markers in high LD, those with the highest average FST. g g We also focused on the ability of our approach to ﬁne-map associated regions to identify candidate SNPs and to provide a probabilistic assessment of the most likely associated set of SNP markers. To do this we used our large-scale simulation data and focused on seven focal regions within a blocks of chromosome 22. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 We allocated effect sizes to the following SNPs: rs131529 with MAF 0.32 which had LD R2 ≥0.15 with 348 other SNPs, rs2096537 with MAF 0.14 which had LD R2 ≥0.15 with 295 other SNPs, rs131538 with MAF 0.05 which had LD R2 ≥0.15 with 82 other SNPs, rs141962840 with MAF 0.007 which had LD R2 ≥0.15 with 11 other SNPs, rs117873986 with MAF 0.02 which had LD R2 ≥0.15 with 12 other SNPs, rs9606483 with MAF 0.005 which had LD R2 ≥0.15 with 1 other SNP, and rs78881648 with MAF 0.009 which had LD R2 ≥0.15 with 1 other SNP. To these seven SNPs, we assigned the same effect sizes in four different scenarios, either 0.05, 0.025, 0.0125, or 0.01 on the SD scale. On the remainder of chromosomes 19, 20, 21 and 22, we randomly selected 1000 SNPs as causal variants to give a polygenic background, sampling their effects from a normal distribution with zero mean and variance 0.5/1000. We repeated each of the four scenarios 20 times. We selected these regions to compare the performance of BayesRR-RC to the ﬁne- mapping approach SuSiE as outlined in a recent paper22. For BayesRR-RC, we calculate the PPWV of the LD blocks containing the seven focal SNPs, and then prune these blocks based on the LD among the markers in the block (described as “purity" in the SuSiE paper22) to identify a credible set with LD R2 ≥0.9. We then count the proportion of times across the simulations that each causal variant was contained with one of the credible sets. For SuSiE, we ran the model from the individual-level data of the whole block of chromosome 22 using the suggested settings and setting K = 10. We then calculate the proportion of times that the identiﬁed credible sets contained one of the seven causal variants. We present these results in Supplementary Fig. 6c. 2 markers in high LD, those with the highest average FST. g g g ST We conducted a simulation study using real genomic data from chromosome 22 where 10,000 individuals were selected from two UK Biobank assessment centres (Glasgow and Croydon). First, causal variants were allocated to 5000 high-LD SNPs with effect sizes simulated from a normal distribution with variance proportional to the FST among the two populations at each SNP. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Second, we selected the same high-LD SNPs as the causal variants, but simulated effect sizes to have correlation 0.5 with the allele frequency differences of the SNPs among the two populations, and thus not only is the effect size proportional to the FST, but there is also directional differentiation (trait increasing loci tend to be those with higher allele frequency in Croydon, trait decreasing alleles have lower frequency in Croydon). For each of these two scenarios, we simulated 50 replicate phenotypes where the phenotypic variance attributable to the causal SNPs is 0.5, there is a phenotypic difference in which Croydon individuals have a phenotype that is 0.5 SD higher than Glasgow individuals (contributing variance 0.05), and residual variance was simulated from a normal with variance 0.45, to give a phenotype with mean of zero and variance of 1. The data were then analysed using a mixed-linear model association (MLMAi implemented in GCTA) and a grouped Bayesian dirac spike and slab models (BayesR implemented in GMRM). In the analysis, we either adjusted the phenotype by the ﬁrst 20 PCs of the genetic data used in the simulation study, or we did not adjust the phenotype for the PCs, to examine the effects of this common methods of population stratiﬁcation control. In a two-population scenario the leading eigenvector encapsulates the allele frequency differentiation between the populations and thus the expectation is that this should adjust for these differences when estimating the marker associations. The results are presented in Fig. S5a, where we ﬁnd that an MLMA approach overestimates the variance attributable to the SNPs under all scenarios, both with and without adjustment for PCs. BayesR returns accurate estimates when the variance of the marker effects is proportional to FST and underestimates the variance when there is a directional associations, with this underestimation being less severe with PC adjustment. UK Biobank data. We restricted our discovery analysis of the UK Biobank to a sample of European-ancestry individuals. To infer ancestry, we used both self- reported ethnic background (UK Biobank data code 21000-0) selecting coding 1 and genetic ethnicity (UK Biobank data code 22006-0) selecting coding 1. We also took the 488,377 genotyped participants and projected them onto the ﬁrst two genotypic principal components (PC) calculated from 2504 individuals of the 1000 Genomes project with known ancestries. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 with four other variants with LD = 0.95, with the remaining 5000 variants having zero effect size and LD = 0. For each scenario, we simulate effect sizes as an equally spaced sequence from an effect size of −0.04 SD, to 0.04 SD giving genetic variance of 0.55, and we simulate residual variance from a normal distribution with zero mean and variance 0.45, to give a phenotype with zero mean and unit variance. For the ﬁrst scenario, we calculate the posterior inclusion probability of each causal SNP. For the second scenario, we calculate the PPWV for each 5-SNP group. Across the 500 replicates of each scenario, we take the mean PPWV and mean PIP for each of the 1000 different effect sizes and compare these in Fig. S6a. Addi- tionally, we grouped SNPs in 50kb regions and selected the number of regions that explain at least 0.1, 0.01 and 0.001% of the variance attributed to all SNP markers in 0.8–100% of the iterations using the simulated data described above for the multiple group enrichment scenario for chromosome 22 in the UK Biobank. We then calculated the false discovery rate (FDR), deﬁned as the proportion of 50 kb regions identiﬁed that do not contain a causal variant, at PPWV thresholds ranging from 0.8 to 100%. We compare these in Supplementary Fig. 6b where as we lower the PPWV variance threshold, the number of false discoveries in the model increases but remains at ≤5% when the PPWV is ≥95%. This further demonstrates that our proposed PPWV approach is an appropriate metric of summarising the posterior distribution to identify associated genomic regions, irrespective of the genomic region used. within the 10,000 UK Biobank individual selected for out-of-sample prediction. We also used the MegaPRS methods implemented in the software LDAK running the four different models described above. We compared the correlation of predicted and simulated genetic value across approaches for each of the 400 simulated phenotypes (Supplementary Fig. 2d). within the 10,000 UK Biobank individual selected for out-of-sample prediction. We also used the MegaPRS methods implemented in the software LDAK running the four different models described above. We compared the correlation of predicted and simulated genetic value across approaches for each of the 400 simulated phenotypes (Supplementary Fig. 2d). The inﬂuence of population structure and relatedness. ARTICLE Third, we wished to determine the out-of-sample phenotypic prediction performance of BayesRR-RC. We used the same randomly selected 10,000 individuals from the UK Biobank that were unrelated to those used in the simulation. Using the same SNP markers and the simulated marker effects we calculated a simulated genetic value for each individual across the replicates. Then, using the effects generated by BayesR and BayesRR-RC, we calculated the predicted genetic value and determined the correlation with the simulated genetic value. We took the marker effect estimates from the MLMA approaches and conducted LD clumping with p-value thresholding using Plink to ﬁnd the set of markers that gave the highest correlation of the genetic predictor and the simulated genetic value g g g p yp ● For each generative model the total genetic variance was 0.6 and we sampled individual-level environmental (residual) variance from a normal distribution with zero mean and variance 0.4 to give phenotypes with zero mean and unit variance. ● For each generative model the total genetic variance was 0.6 and we sampled individual-level environmental (residual) variance from a normal distribution with zero mean and variance 0.4 to give phenotypes with zero mean and unit variance. This range covers generative genetic models discussed in the literature and provides models that both ﬁt and violate the assumptions of the range of variance component statistical models. This includes both individual-level and summary statistic approaches, that are currently applied in the literature for estimation of the 12 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications ARTICLE ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Here, we selected 2000 LD blocks with a single causal SNP per block at random, where an LD block is deﬁned as a group of SNP markers with absolute LD of at least 0.01. We assigned effect sizes to these 2000 selected SNPs, drawing them from a normal distribution with zero mean and variance 0.5/ 2000. We then multiplied effect sizes by the simulated marker values scaled to zero mean and unit variance to create the genetic values with variance 0.5. In addition to the genetic component, we added a common environment component to simulate effects coming from shared familial environment. We simulated four scenarios p To facilitate contrasting the genetic basis of different phenotypes, we then removed closely related individuals as identiﬁed in the UK Biobank data release. While the BayesRR model can accommodate relatedness similar to mixed linear models, we wished to simply compare phenotypes at markers that enter the model due to LD with underlying causal variants. Relatedness leads to the addition of markers within the model to capture the phenotypic covariance of closely related individuals, and this will vary across traits in accordance with the genetic and environmental covariance for each phenotype. For these unrelated individuals, we used the imputed autosomal genotype data of the UK Biobank provided as part of the data release. We used the genotype probabilities to hard-call the genotypes for variants with an imputation quality score above 0.3. The hard-call-threshold was 0.1, setting the genotypes with probability ≤0.9 as missing. From the good quality markers (with missingness less than 5% and p-value for Hardy–Weinberg test larger than 10-6, as determined in the set of unrelated Europeans) were selected those with minor allele frequency (MAF) > 0.0002 and rs identiﬁer, in the set of European-ancestry participants, providing a data set 9,144,511 SNPs, short indels and large structural variants. From these, we took the overlap with the Estonian Genome centre data to give a ﬁnal set of 8,430,446 markers. From the UK Biobank European data set, samples were excluded if in the UKB quality control procedures they (i) were identiﬁed as extreme heterozygosity or missing genotype outliers; (ii) g where each family was assigned the same common environment effect drawn from a normal distribution with variance 0 (no common environment), 0.1, 0.2, and 0.3. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Using the obtained PC loadings, we then assigned each participant to the closest population in the 1000 Genomes data: European, African, East-Asian, South-Asian or Admixed, selecting individuals with PC1 projection < absolute value 4 and PC 2 projection < absolute value 3. This gave a sample size of 456,426 individuals. g j Finally, we also assess the inﬂuence of relatedness on the estimates obtained from a BayesR model using real genomic data from chromosome 21 and 22 (226,662 SNP markers) and 10,000 families randomly selected from the UK Biobank (26,034 individuals). Here, we selected 2000 LD blocks with a single causal SNP per block at random, where an LD block is deﬁned as a group of SNP markers with absolute LD of at least 0.01. We assigned effect sizes to these 2000 selected SNPs, drawing them from a normal distribution with zero mean and variance 0.5/ 2000. We then multiplied effect sizes by the simulated marker values scaled to zero mean and unit variance to create the genetic values with variance 0.5. In addition to the genetic component, we added a common environment component to simulate effects coming from shared familial environment. We simulated four scenarios where each family was assigned the same common environment effect drawn from a normal distribution with variance 0 (no common environment), 0.1, 0.2, and 0.3. Finally, we added an environmental component simulated from a normal distribution with mean zero and variance 1 minus the genetic variance and minus the common environment variance. We analysed 20 replicates of each of the four scenarios with BayesRR-RC with six MAF-LD groups (terciles of MAF, each split into two groups based on median LD score within each MAF tercile). In Supplementary Fig. 5, we summarise 800 samples of the posterior distribution from 5000 iterations with a thin of ﬁve and removing the ﬁrst 1000 iterations as burn-in. We ﬁnd that the variance attributable to the SNPs, the regression coefﬁcients and the posterior probability of window variance (PPWV) remain unchanged with relatedness and with increasing family effects. g j Finally, we also assess the inﬂuence of relatedness on the estimates obtained from a BayesR model using real genomic data from chromosome 21 and 22 (226,662 SNP markers) and 10,000 families randomly selected from the UK Biobank (26,034 individuals). NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunicati ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 had a genetically inferred gender that did not match the self-reported gender; (iii) were identiﬁed to have putative sex chromosome aneuploidy; (iv) were excluded from kinship inference. Information on individuals who had withdrawn their consent for their data to be used was also removed. These ﬁlters resulted in a data set with 382,466 individuals. region (up to +/−50 kb from a gene, Supplementary Data 6–9). Finally, we mapped SNPs with greater than 50% posterior inclusion probability (PIP) across all four chains labelling them using our seven location annotations (Supplementary Fig. 13). We report SNPs with PIP > 95% and their corresponding p-value from reported GWAS summary statistics (fastGWA, see “Code availability”) with “body mass index” entry for BMI, “standing height” for HT, “angina/heart attack” for CAD and “diabetes” for T2D (Supplementary Data 10). We then selected the recorded measures of BMI (UK Biobank variable identiﬁer 21001-0.0) and height (variable identiﬁer 50-0.0) collected during initial assessment visit (year 2006-2010). BMI and height phenotypes six standard deviations (SD) away from the mean were not included in the analyses. For Type 2 Diabetes (T2D) in UKB, we selected cases very broadly as individuals who have main or secondary diagnosis (UKB ﬁelds 41202-0.0–41202-0.379 and 41204-0.0–41204-0.434) of “non- insulin-dependent diabetes mellitus” (ICD 10 code E11) or self-reported non-cancer illness (UKB ﬁeld 20002-0.0–20002-2.28) “type 2 diabetes” (code 1223). From respondents self-reporting just “diabetes” (code 1220), we selected as cases those who did not self-report “type 1 diabetes” (code 1222) and had no Type 1 Diabetes (ICD code E10) diagnosis. Individuals with self-reported “diabetes” and “type 1 diabetes”/E10 were also left out from controls. We also deﬁned coronary artery disease (CAD) cases broadly as participants with one of the following primary or secondary diagnoses or cause of death: ICD 10 codes I20 to I28; self-reported angina (code 1074) or self-reported heart attack/myocardial infarction (code 1075). Participants with self-reported “heart/cardiac problem” (code 1066) were not included as cases but also excluded from controls. This gave a sample size for each trait of 25,773 T2D cases and 359,730 T2D controls, 39,766 CAD cases and 344,054 CAD controls, 382,402 measures of height and 381,899 measures of BMI. pp y We then compared our BayesRR-RC estimates for height, BMI, T2D and CAD to RHEmc18 which also relies on individual level data. ARTICLE For the association testing, we split the 8,430,446 SNP markers in blocks of 400 consecutive SNP markers and set the Firth correction p-value threshold to 0.01. We then applied an approximate and joint association analysis called GCTA-COJO31 to the summary statistics obtained with Bolt-LMM, Regenie and plink2. We ran GCTA-COJO using a subset of 20,000 individuals randomly selected from the 382,466 individuals as reference with a window size of 10,000 kb and a r2 cutoff value of 0.5 for the LD among the SNPs in the data. Finally, we set a p-value threshold to 5e−8 to report signiﬁcant SNPs associated with height, BMI, CAD an T2D in Table 2. p yp p We partition SNP markers into seven location annotations using the knownGene table from the UCSC browser data (see “Code availability” section). We preferentially assigned SNPs to coding (exonic) regions ﬁrst, then in the remaining SNPs, we preferentially assigned them to intronic regions, then to 1 kb upstream regions, then to 1–10 kb regions, then to 10–500 kb regions, then to 500–1 Mb regions. Remaining SNPs were grouped in a category labelled “others” and also included in the model so that variance is partitioned relative to these also. Thus, we assigned SNPs to their closest upstream region, for example if a SNP is 1 kb upstream of gene X, but also 10–500 kb upstream of gene Y and 5 kb downstream for gene Z, then it was assigned to be a 1 kb region SNP. This means that SNPs 10–500 kb and 500 kb–1 Mb upstream are distal from any known nearby genes. We further partition upstream regions to experimentally validated promoters, transcription factor binding sites (tfbs) and enhancers (enh) using the HACER, snp2tfbs databases (see “Code availability” section). All SNP markers assigned to 1 kb regions map to promoters; 1–10 kb SNPs, 10–500 kb SNPs, 500 kb–1 Mb SNPs are split into enh, tfbs and others (un-mapped SNPs) extending the model to 13 annotation groups. Within each of these annotations, we have three minor allele frequency groups (MAF < 0.01, 0.01 > MAF > 0.05, and MAF > 0.05), and then each MAF group is further split into two based on median LD score. This gives 78 non-overlapping groups for which our BayesRR-RC model jointly estimates the phenotypic variation attributable to, and the SNP marker effects within, each group. ARTICLE Data from this project were held under UK Biobank project ID 35520. All phenotypes were adjusted for age of attending assessment centre (UKB code 21003-0.0, factor with levels for each age), year of birth (UKB ﬁeld 34-0.0, factor with levels for each year), UK Biobank recruitment centre (UKB ﬁeld 54-0.0, factor with levels for each centre), Genotype batch (UKB ﬁeld 22000, factor with levels for each batch) and ﬁnal 20 leading principal components of 1.2 million LD clumped markers from the 8,430,446 markers included in the analysis, calculated using ﬂashPCA (see “Code availability” section). The residuals were then converted to z- scores with 0 mean and variance of 1. Similarly as for relatedness, population stratiﬁcation is also accounted for within the BayesRR model through the addition of a background of marker effects entering the model, however we also wished to account for this in the standard manner by adjusting for the leading 20 PCs of the SNP data to get as close as possible to the inclusion of markers in the model that reﬂect LD with the causal variants. We note that as with any association model, while we take steps to adjust for known spatial (UKB centre), batch, and ancestry effects, and that the effects of each SNP is estimated jointly (and thus conditionally on the effects of all the other SNPs) environmentally induced covariance between SNP markers and a phenotype is still possible. y In addition to plink228 summary statistics, we also applied Bolt-LMM8 and Regenie9 to height, BMI, T2D and CAD. In the ﬁrst step, we pruned SNPs using plink30 with a pairwise r2 threshold of 0.5 and a window size of 1000 kb, resulting in a subset of 1,362,013 SNPs markers. We restricted the random effects in the mixed model for bolt-LMM and the ridge regression predictors for Regenie to this subset of pruned SNPs. In the second step, all 8,430,446 SNPs from the full genotype data were then tested for association in both methods. Following recommendations, we used the provided hg19 genetic map ﬁle and 1000 Genomes LD scores reference for Bolt-LMM and performed the default mixed linear model association test. For Regenie, the 1,362,013 SNP markers are split in blocks of 1000 consecutive SNP markers and ridge regression predictors are computed for a range of ﬁve shrinkage parameters within each block. ARTICLE We ran RHEmc with ten independent random vectors and 100 jackknife blocks on the 382,466 individuals and 8,430,446 SNP markers assigned to our 78 non-overlapping groups. SNP heritability estimates, enrichment and standard errors per genetic component are reported in Supplementary Data 3. We intended to include SNP heritability estimates from Bolt-REML17 in the method comparison but the run time and memory usage exceeded 7 days and 900 GB which is the limiting run-time and memory for our HPC system. Among the summary statistic methods, we ran sLDSC19 and SumHer6. To do so, we created summary statistics containing marginal associations for each of the 8,430,446 markers using plink228 for height, BMI, T2D and CAD. For sLDSC, we computed univariate LD scores and annotation-speciﬁc LD scores for the 78 non-overlapping groups using a window size of 10,000 kb and a subset of 20,000 individuals randomly selected from the full data set. We then partitioned heritability with our annotations and no restriction on MAF. SNP heritability estimates, proportions of heritability, enrichment and standard errors per genetic component are reported in Supplementary Data 4. For SumHer, we computed LDAK weightings and created tagging ﬁles separately by chromosomes using the full data set (M = 8,430,446 and N = 382,466) as reference and a window size of 1000 kb. Because SNPs included in groups others and rare 1Mb tfbs are not present in all chromosomes, tagging ﬁles are constructed using 70 non-overlapping annotations only. The remaining SNPs are modelled together in an extra partition. Finally, we merged the tagging ﬁles and regressed the summary statistics onto this ﬁle assuming the LDAK model. SNP heritability estimates, proportions of heritability, enrichment and standard errors per genetic component are reported in Supplementary Data Table 5. The proportion of genetic variance estimated genome-wide with RHE-mc, sLDSC, and SumHer are shown in Table 1. We also report the proportion of genetic variance attributed to SNPs located in exons, introns, 1, 1–10 and 10–500 kb regions and the proportion of common SNPs located in exons, introns and 10–500 kb regions computed from the single heritability estimates observed (Table 1). g UK Biobank has approval from the North West Multi-centre Research Ethics Committee (MREC) to obtain and disseminate data and samples from the participants (http://www.ukbiobank.ac.uk/ethics/), and these ethical regulations cover the work in this study. Written informed consent was obtained from all participants. ARTICLE For each of the 78 groups, SNPs were modelled using ﬁve mixture groups with variance equal to the phenotypic variance attributable to the group multiplied by constants (mixture 0 = 0, mixture 1 = 0.0001, 2 = 0.001, 3 = 0.01, 4 = 0.1). We conducted a series of convergence diagnostic analyses of the posterior distributions to ensure we obtained estimates from a converged set of four Gibbs chains, each run for 6000 iterations with a thin of ﬁve and burn-in of 500 for each trait (Supplementary Figs. 7–10). Estonian Genome Centre data. For the Estonian Genome Centre Data, 32,594 individuals were genotyped on Illumina Global Screening (GSA) arrays and we imputed the data set to an Estonian reference, created from the whole genome sequence data of 2244 participants32. From 11,130,313 markers with imputation quality score >0.3, we selected SNPs that overlapped with the UK Biobank, resulting in a set of 8,433,421 markers. g We selected height and BMI measures from the Estonian Genome Centre data, in 32,594 individuals genotyped on GSA array and converted them to sex-speciﬁc z-scores after applying the same outlier removal procedure as in UKB and adjusting for the age at agreement. Prevalent cases of CAD and T2D in the Estonian Biobank cohort were ﬁrst identiﬁed on the basis of the baseline data collected at recruitment, where the information on prevalent diseases was either retrieved from medical records or self-reported by the participant. The cohort was subsequently linked to the Estonian Health Insurance database that provided additional information on prevalent cases (diagnoses conﬁrmed before the date of recruitment) as well as on incident cases during the follow-up. We calculate PPWV for LD blocks of the genome, by ﬁrst calculating the minor allele frequency of each SNP (p) and using 1 −p in a Plink clumping procedure to select LD independent (correlation2 ≤0.1) blocks of SNPs. We then repeat the estimation of the PPWV of 50 kb regions across the genome, then map SNPs to the coding region of genes, and to the closest gene +/−50 kb from the SNP position. These are labelled as located in a coding region, an intron, 1 kb upstream of a gene using our functional annotations. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Finally, we added an environmental component simulated from a normal distribution with mean zero and variance 1 minus the genetic variance and minus the common environment variance. We analysed 20 replicates of each of the four scenarios with BayesRR-RC with six MAF-LD groups (terciles of MAF, each split into two groups based on median LD score within each MAF tercile). In markers (with missingness less than 5% and p-value for Hardy–Weinberg test larger than 10-6, as determined in the set of unrelated Europeans) were selected those with minor allele frequency (MAF) > 0.0002 and rs identiﬁer, in the set of European-ancestry participants, providing a data set 9,144,511 SNPs, short indels and large structural variants. From these, we took the overlap with the Estonian Genome centre data to give a ﬁnal set of 8,430,446 markers. From the UK Biobank European data set, samples were excluded if in the UKB quality control procedures they (i) were identiﬁed as extreme heterozygosity or missing genotype outliers; (ii) Localisation and ﬁne-mapping of SNP-phenotype associations. We further validate the use of PPWV in an another simulation study with 500 replicate data sets of 10,000 SNP markers for 5000 individuals for each of two scenarios. In the ﬁrst scenario, 1000 SNPs are randomly selected to be causal variants and all 10,000 SNP markers are LD independent. In the second, the 1000 causal variants are each in LD TURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 13 References As a comparison, we also estimated a boltLMM prediction model using MegaPRS21 as recommended by the authors and as shown to have the best prediction performance out of the MegaPRS approaches in our simulation study. We clumped SNPs with r2 threshold of 0.5 resulting in 1,508,624 SNP markers to be included in the analysis and randomly selected 20,000 individuals to compute the LDAK weights. We then computed the tagging ﬁle using the same data set as reference and the 64 BLD-LDAK annotations. Here, weights are models as an extra annotation and we save the heritability matrix. We then regress the plink228 summary statistics for height, BMI, CAD and T2D onto the tagging ﬁle, saving the per-predictor heritabilities. We then created four reference panels with the same 1,508,624 SNP markers but randomly selecting different 5000 related individuals from the UK Biobank and we used these to: (i) calculate predictor-predictor correlations with a window size of 3000 kb to estimate the LD structure; (ii) compute pseudo summaries from the plink2 summary statistics including ambiguous alleles, which creates pseudo training and test summary statistics to be used in the construction of the prediction model; (iii) estimate effect sizes specifying a Bolt-LMM model for height, BMI, CAD and T2D, using the predictor–predictor correlations, the per-predictor heritabilities, the plink2 summary statistics and training pseudo summary statistics, whilst including ambiguous allele and specifying a 1000 kb window; (iv) test prior distributions to determine the most accurate model and obtain the best effect sizes. These steps resulted in 1,397,514 predictors for height, 1,471,586 for BMI, 1,397,514 for CAD and 1,389,364 for T2D and we ensured that at no point was the Estonian genome centre data used, nor was any overlapping individuals in the UK Biobank subsets used to train the models and the data used to generate the summary statistics. Finally, we then calculated genomic predictors for each individual in the Estonian Biobank using the best effect sizes. We report the squared correlations between the genomic predictor and phenotypes. As a comparison, we also estimated a boltLMM prediction model using MegaPRS21 as recommended by the authors and as shown to have the best prediction performance out of the MegaPRS approaches in our simulation study. We clumped SNPs with r2 threshold of 0.5 resulting in 1,508,624 SNP markers to be included in the analysis and randomly selected 20,000 individuals to compute the LDAK weights. Received: 10 February 2021; Accepted: 5 November 2021; scale. Thus when we create a genomic predictor, for say coding SNPs, by multiplying SNPs mapped to coding regions genotyped in Estonia to the effect sizes obtained in the UK Biobank for each iteration, we obtain a genetic predictor for each iteration, providing a posterior predictive distribution that is also on the SD scale. For each trait, we created 2000 genomic predictors for each individual in the Estonian Biobank, at each of the 13 annotation groups, by selecting effect size estimates obtained every tenth iteration from the last 3000 iterations of each of the four Gibbs chains and combining them together in a single posterior. We calculated prediction accuracy as the proportion of phenotypic variation explained by the genomic predictor, and area under the receiver operator curve (AUC) for T2D and CAD using each individual’s mean genetic predictor. For each of the 13 annotation groups, we calculated the partial correlation of the genetic predictor of each of the 2000 iterations and the phenotype. We then used this to estimate the independent proportional contribution of each group to the total prediction accuracy, providing a metric of replication for our UK Biobank enrichment results. F h i ht d BMI d t i d th b bilit th t h E t i References These steps resulted in 1,397,514 predictors for height, 1,471,586 for BMI, 1,397,514 for CAD and 1,389,364 for T2D and we ensured that at no point was the Estonian genome centre data used, nor was any overlapping individuals in the UK Biobank subsets used to train the models and the data used to generate the summary statistics. 18. Pazokitoroudi, A. et al. Efﬁcient variance components analysis across millions of genomes. Nat Commun 11, 4020 (2020). 19. Finucane, H. K. et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat. Genet. 47, 1228–1235 (2015). 20. Fernando, R., Toosi, A., Wolc, A., Garrick, D. & Dekkers, J. Application of whole-genome prediction methods for genome-wide association studies: a bayesian approach. J. Agric. Biol. Environ. Stat. 22, 172–193 (2017). Finally, we then calculated genomic predictors for each individual in the Estonian Biobank using the best effect sizes. We report the squared correlations between the genomic predictor and phenotypes. 21. Zhang, Q., Privé, F., Vilhjálmsson, B. et al. Improved genetic prediction of complex traits from individual-level data or summary statistics. Nat Commun 12, 4192 (2021). Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. 22. Wang, G., Sarkar, A., Carbonetto, P. & Stephens, M. A simple new approach to variable selection in regression, with application to genetic ﬁne mapping. J. R. Stat. Soc. 82, 1273–1300 (2020). Data availability 23. Lloyd-Jones, L. R. et al. Improved polygenic prediction by bayesian multiple regression on summary statistics. Nat. Commun. 10, 5086 (2019). This project uses UK Biobank data under project 35520. The Estonian Genome Centre data are protected and are not available due to data privacy laws. The Estonian Genome Centre data can be made available under restricted access upon request from the cohort author R.M. with appropriate research agreements. Summaries of all posterior distributions generated in this study are provided in Supplementary Data tables. Full posterior distributions of the SNP marker effects sizes and estimated variance components for each trait are deposited on Dryad with https://doi.org/10.5061/ dryad.sqv9s4n51. This project uses UK Biobank data under project 35520. The Estonian Genome Centre data are protected and are not available due to data privacy laws. The Estonian Genome Centre data can be made available under restricted access upon request from the cohort author R.M. with appropriate research agreements. Summaries of all posterior g y 24. Boyle, E. A., Li, Y. I. & Pritchard, J. K. An expanded view of complex traits: From polygenic to omnigenic. Cell 169, 1177–1186 (2017). From polygenic to omnigenic. Cell 169, 1177–1186 (2017) 25. MacLeod, I. M. et al. Exploiting biological priors and sequence variants enhances qtl discovery and genomic prediction of complex traits. BMC Genomics 17, 144 (2016). 26. Brøndum, R. F. et al. Genome position speciﬁc priors for genomic prediction. BMC Genom. 13, 543 (2012). 27. Zhao, P. & Yu, B. On model selection consistency of lasso. J. Mach. Learn. Res. 7, 2541–2563 (2006). References We then computed the tagging ﬁle using the same data set as reference and the 64 BLD-LDAK annotations. Here, weights are models as an extra annotation and we save the heritability matrix. We then regress the plink228 summary statistics for height, BMI, CAD and T2D onto the tagging ﬁle, saving the per-predictor heritabilities. We then created four reference panels with the same 1,508,624 SNP markers but randomly selecting different 5000 related individuals from the UK Biobank and we used these to: (i) calculate predictor-predictor correlations with a window size of 3000 kb to estimate the LD structure; (ii) compute pseudo summaries from the plink2 summary statistics including ambiguous alleles, which creates pseudo training and test summary statistics to be used in the construction of the prediction model; (iii) estimate effect sizes specifying a Bolt-LMM model for height, BMI, CAD and T2D, using the predictor–predictor correlations, the per-predictor heritabilities, the Quantitative and Binary Traits (Nature Publishing Group, 20 10. Zhou, W. et al. Efﬁciently controlling for case-control imbalance and sample relatedness in large-scale genetic association studies. Nat. Genet. 50, 1335–1341 (2018). 11. Erbe, M. et al. Improving accuracy of genomic predictions within and between dairy cattle breeds with imputed high-density single nucleotide polymorphism panels. J. Dairy Sci. 95, 4114–4129 (2012). 12. Moser, G. et al. Simultaneous discovery, estimation and prediction analysis of complex traits using a bayesian mixture model. PLoS Genet. 11, 1–22 (2015). 13. Banos, D. T. et al. Bayesian reassessment of the epigenetic architecture of complex traits. Nat. Commun. 11, 1–14 (2020). p 14. George, E. I. & McCulloch, R. E. Variable selection via gibbs sampling. J. Am. Stat. Assoc. 88, 881–889 (1993). 15. Malsiner-Walli, G. & Wagner, H. Comparing spike and slab priors for bayesian variable selection. Austrian J. Stat. 40, 241–264 (2016). 16. Castillo, I. et al. Bayesian linear regression with sparse priors. Ann. Stat. 43, 1986–2018 (2015). 17. Loh, P.-R. et al. Contrasting genetic architectures of schizophrenia and other complex diseases using fast variance-components analysis. Nat. Genet. 47, 1385 (2015). plink2 summary statistics and training pseudo summary statistics, whilst including ambiguous allele and specifying a 1000 kb window; (iv) test prior distributions to determine the most accurate model and obtain the best effect sizes. References 1. Evans, L. M. et al. Comparison of methods that use whole genome data to estimate the heritability and genetic architecture of complex traits. Nat. Genet. 50, 737–745 (2018). 2. Speed, D. et al. Reevaluation of snp heritability in complex human traits. Nat. Genet. 49, 986 (2017). 3. Speed, D., Holmes, J. & Balding, D. J. Evaluating and improving heritability models using summary statistics. Nat. Genet. 52, 458–462 (2020). g y 4. Hou, K. et al. Accurate estimation of snp-heritability from biobank-scale dat irrespective of genetic architecture. Nat. Genet. 51, 1244–1251 (2019). For height and BMI, we determined the probability that each Estonian individual’s predictor accurately reﬂected their phenotypic value. To do this, we calculated the proportion of posterior samples with abs ð^g yÞ of less than 1 for each individual, which gives a measure of the degree to which each posterior predictive distribution overlaps with the phenotype within +/−1 SD. 5. Gazal, S., Marquez-Luna, C., Finucane, H. K. & Price, A. L. Reconciling s-ldsc and ldak functional enrichment estimates. Nat. Genet. 51, 1202–1204 (2019). 6. Speed, D. & Balding, D. J. SumHer better estimates the SNP heritability of complex traits from summary statistics. Nat. Genet. 51, 277–284 (2019). p p p yp For T2D and CAD, we extended the PCF metric, typically deﬁned as the proportion of cases with larger estimated risk than the top pth percentile of the distribution of genetic risk in the general population. We calculated the proportion of posterior samples for each individual with values in the top 25% of the distribution of genomic predictors for each trait. Thus for each individual, we calculate the probability that the posterior predictive distribution is in the top 25% of the distribution of genetic risk in the general population. 7. Jiang, L. et al. A resource-efﬁcient tool for mixed model association analysis of large-scale data. Nat. Genet. 51, 1749–1755 (2019). 8. Loh, P.-R. et al. Efﬁcient bayesian mixed-model analysis increases association power in large cohorts. Nat. Genet. 47, 284 (2015). 9. Mbatchou, J. et al. Computationally Efﬁcient Whole Genome Regression for Quantitative and Binary Traits (Nature Publishing Group, 2020). 9. Mbatchou, J. et al. Computationally Efﬁcient Whole Genome Regression Quantitative and Binary Traits (Nature Publishing Group, 2020). o t e d st but o o ge et c s t e ge e a popu at o . ARTICLE Remaining SNPs are labelled as located in a cis- All Estonian biobank participants have signed a broad informed consent form and the study was carried out under ethical approval 1.1 12/2856 from the Estonian Committee on Bioethics and Human Research (Estonian Ministry of Social Affairs). As the UK Biobank marker effects are estimated from traits that were standardised to a z-score prior to analysis, all effect sizes obtained are on the SD 14 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-021-27258-9 Received: 10 February 2021; Accepted: 5 November 2021; Received: 10 February 2021; Accepted: 5 November 2021; Additional information 31. Yang, J. et al. Conditional and joint multiple-snp analysis of gwas summary statistics identiﬁes additional variants inﬂuencing complex traits. Nat. Genet. 44, 369–375 (2012). Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41467-021-27258-9. 32. Tasa, T. et al. Genetic variation in the estonian population: pharmacogenomics study of adverse drug effects using electronic health records. Eur. J. Hum. Genet. 27, 442–454 (2019). Correspondence and requests for materials should be addressed to Matthew R. Robinson. Peer review information Nature Communications thanks Luke O’Connor and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Reprints and permission information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. (Stockholm, Sweden). We also would like to acknowledge Simone Rubinacci, Oliver Delanau, Alexander Terenin, Eleonora Porcu, and Mike Goddard for their useful comments and suggestions. (Stockholm, Sweden). We also would like to acknowledge Simone Rubinacci, Oliver Delanau, Alexander Terenin, Eleonora Porcu, and Mike Goddard for their useful comments and suggestions. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Acknowledgements This project was funded by an SNSF Eccellenza Grant to MRR (PCEGP3-181181), and by core funding from the Institute of Science and Technology Austria. We would like to thank the participants of the cohort studies, and the Ecole Polytechnique Federal Lausanne (EPFL) SCITAS for their excellent compute resources, their generosity with their time and the kindness of their support. P.M.V. acknowledges funding from the Australian National Health and Medical Research Council (1113400) and the Australian Research Council (FL180100072). L.R. acknowledges funding from the Kjell & Märta Beijer Foundation Author contributions M.R.R. conceived and designed the study. M.P., D.T.B. and A.K. contributed to the study design. M.P. and M.R.R. conducted the experiments and analyses with input from D.T.B., A.K., S.E.O., A.H., J.S., P.M.V., R.M. and L.R. M.R.R., D.T.B., S.E.O. and L.R. derived the equations and the algorithm. EJO and DTB developed the software, with contributions from M.R.R., M.P., S.E.O., A.K. and G.M. M.R.R., M.P. and DTB wrote the paper. RM and ZK provided study oversight and contributed data to the analysis. All authors approved the ﬁnal manuscript prior to submission. Code availability Our BayesRR-RC model is implemented within the software GMRM, with full open source code available at: https://github.com/medical-genomics-group/gmrm. UCSC Table Browser https://genome.ucsc.edu/cgi-bin/hgTables. ﬂashPCA https://github.com/ gabraham/ﬂashpca. Plink1.90 https://www.cog-genomics.org/plink2/. GCTA https:// cnsgenomics.com/content/software. HACER database http://bioinfo.vanderbilt.edu/AE/ HACER/. snp2tfbs database https://ccg.epﬂ.ch//snp2tfbs/. fastGWA database http:// fastgwa.info/ukbimp/phenotypes/. Computing environment https://www.epﬂ.ch/ research/facilities/scitas/hardware/helvetios/. 28. Chang, C. C. et al. Second-generation plink: rising to the challenge of larger and richer datasets. Gigascience 4, s13742–015 (2015). 29. Daetwyler, H. D., Villanueva, B. & Woolliams, J. A. Accuracy of predicting the genetic risk of disease using a genome-wide approach. PLoS ONE 3, 1–8 (2008). 30. Purcell, S. et al. Plink: a tool set for whole-genome association and population- based linkage analyses. Am. J. Hum. Genet. 81, 559–575 (2007). 15 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications Competing interests © The Author(s) 2021 The authors declare no competing interests. NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunications 16 NATURE COMMUNICATIONS \| (2021) 12:6972 \| https://doi.org/10.1038/s41467-021-27258-9 \| www.nature.com/naturecommunicatio
https://openalex.org/W2193388533	https://www.fupress.com/redir.ashx?RetUrl=1737_21393.pdf	Italian	null	Gli erbari aretini da Andrea Cesalpino ai giorni nostri	Cataloghi e collezioni	2,008	cc-by	57,530	Cataloghi e collezioni 4 4 Gli erbari aretini da Andrea Cesalpino ai giorni nostri a cura di Chiara Nepi Enrico Gusmeroli Gli erbari aretini da Andrea Cesalpino ai giorni nostri / a cura di Chiara Nepi e Enrico Gusmeroli. – Firenze : Firenze University Press, 2008. (Cataloghi e collezioni ; 4) http://digital.casalini.it/9788884538031 ISBN 978-88-8453-765-2 (print) ISBN 978-88-8453-803-1 (online) Gli erbari aretini da Andrea Cesalpino ai giorni nostri / a cura di Chiara Nepi e Enrico Gusmeroli. – Firenze : Firenze University Press, 2008. (Cataloghi e collezioni ; 4) http://digital.casalini.it/9788884538031 ISBN 978-88-8453-765-2 (print) ISBN 978-88-8453-803-1 (online) Il volume è stato finanziato dalla Provincia di Arezzo. L’erbario di Andrea Cesalpino 3 L’erbario di Andrea Cesalpino Guido Moggi 3 I primi erbari della storia 4 La vita e l’opera scientifica di Andrea Cesalpino nel campo della botanica 8 L’erbario di Cesalpino: vicende storiche 12 L’erbario di Cesalpino: descrizione Referenze fotografiche Tutte le immagini del volume sono di Saulo Bambi, ad eccezione di quelle diversamente indicate nelle rispettive didascalie. Per eventuali citazioni bibliografiche, si raccomanda la seguente dizione: Chiara Nepi e Enrico Gusmeroli (a cura di), 2008 – Erbari Aretini da Andrea Cesalpino ai giorni nostri. 208 pagg., Firenze University Press Per eventuali citazioni bibliografiche, si raccomanda la seguente dizione: Chiara Nepi e Enrico Gusmeroli (a cura di), 2008 – Erbari Aretini da Andrea Cesalpino ai giorni nostri. 208 pagg., Firenze University Press Progetto grafico di Alberto Pizarro Fernández © Copyright 2008 Firenze University Press. Università degli Studi di Firenze Firenze University Press Borgo Albizi, 28, 50122 Firenze, Italy http://www.fupress.com/ Università degli Studi di Firenze Firenze University Press Borgo Albizi, 28, 50122 Firenze, Italy http://www.fupress.com/ Sommario IX Presentazione Angelo Maria Cardone, Assessore della Provincia di Arezzo XI Premessa Donato Chiatante, Presidente della Società Botanica Italiana XIII Introduzione Chiara Nepi e Enrico Gusmeroli L’erbario di Andrea Cesalpino 3 L’erbario di Andrea Cesalpino Guido Moggi 3 I primi erbari della storia 4 La vita e l’opera scientifica di Andrea Cesalpino nel campo della botanica 8 L’erbario di Cesalpino: vicende storiche 12 L’erbario di Cesalpino: descrizione L’erbario Coltellini 23 L’erbario Coltellini della Sezione Botanica del Museo di Storia Naturale Chiara Nepi L’Hortus Siccus Pisanus 31 L’Hortus Siccus Pisanus di Castiglion Fiorentino Leonardo Magionami 32 Descrizione dell’Hortus Siccus Pisanus 38 Vicende dell’acquisizione 41 Elenco delle specie presenti nell’Hortus Siccus Pisanus Paolo Emilio Tomei e Francesca Malfanti L’erbario di Mattia Moneti 53 Appunti sulla Società Botanica di Cortona e su Mattia Moneti Bruno Gialluca 63 L’erbario dipinto di Mattia Moneti: note botaniche Maria Adele Signorini con la collaborazione di Laura Vivona Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University P IX Presentazione Angelo Maria Cardone, Assessore della Provincia di Arezzo XI Premessa Donato Chiatante, Presidente della Società Botanica Italiana XIII Introduzione Chiara Nepi e Enrico Gusmeroli L’Hortus Siccus Pisanus 31 L’Hortus Siccus Pisanus di Castiglion Fiorentino Leonardo Magionami 32 Descrizione dell’Hortus Siccus Pisanus 38 Vicende dell’acquisizione 41 Elenco delle specie presenti nell’Hortus Siccus Pisanus Paolo Emilio Tomei e Francesca Malfanti L’erbario di Mattia Moneti Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press VI Gli erbari aretini da Andrea Cesalpino ai giorni nostri L’erbario della Biblioteca Rilliana di Poppi 71 L’erbario della Biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi L’erbario Venturini del Santuario della Verna 79 Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna Secondino Gatta 83 L’erbario Venturini: note botaniche Lorenzo Lastrucci e Guido Moggi L’erbario egiziano di Jacob Corinaldi 91 L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio Paolo Emilio Tomei e Lucia Amadei 95 L’erbario 96 Appendice Gli erbari contemporanei 101 Erbari di oggi in provincia di Arezzo Michele Padula e Vincenzo Gonnelli 101 Herbarium M. Padulae 103 Erbario del Museo Forestale «Carlo Siemoni» di Badia Prataglia (Poppi, Arezzo) 103 Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) 105 Erbario Gonnelli 106 Erbario Istituto Professionale di Stato per l’Agricoltura e l’Ambiente «A.M. Camaiti» (Pieve di S. Stefano) Appendice. Illustratio in hortum siccum Andreae Caesalpini 111 Introduzione all’edizione di Teodoro Caruel (1858) Traduzione a cura di Leonardo Magionami 114 Edizione Anastatica 187 Note sugli autori 189 Indice dei nomi Gli erbari aretini da Andrea Cesalpino ai giorni nostri Gli erbari contemporanei 101 Erbari di oggi in provincia di Arezzo Michele Padula e Vincenzo Gonnelli 101 Herbarium M. Padulae 103 Erbario del Museo Forestale «Carlo Siemoni» di Badia Prataglia (Poppi, Arezzo) 103 Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) 105 Erbario Gonnelli 106 Erbario Istituto Professionale di Stato per l’Agricoltura e l’Ambiente «A.M. Camaiti» (Pieve di S. Stefano) 111 Introduzione all’edizione di Teodoro Caruel (1858) Traduzione a cura di Leonardo Magionami 114 Edizione Anastatica 187 Note sugli autori 189 Indice dei nomi Presentazione Angelo Maria Cardone Assessore della Provincia di Arezzo alla Difesa del suolo, Acque e Demanio Idrico, Conservazione della Natura e Protezione Civile Presentazione Angelo Maria Cardone Assessore della Provincia di Arezzo alla Difesa del suolo, Acque e Demanio Idrico, Conservazione della Natura e Protezione Civile Angelo Maria Cardone Assessore della Provincia di Arezzo alla Difesa del suolo, Acque e Demanio Idrico, Conservazione della Natura e Protezione Civile Q Q uesto volume completa il lavoro di co- noscenza e valorizzazione degli erbari realizzati da illustri aretini o conservati nel territorio della provincia di Arezzo, avviato con il convegno Da Andrea Cesalpino ai nostri giorni, evoluzione delle conoscenze botaniche in provincia di Arezzo che inaugurò, nella pri- mavera del 2005, una mostra presso il Museo d’Arte Medievale e Moderna di Arezzo, sui più importanti erbari aretini che vengono oggi descritti all’interno di questo libro. le Università di Firenze e Siena, come pure i professionisti e i cultori della materia aretini. le Università di Firenze e Siena, come pure i professionisti e i cultori della materia aretini. Tutto questo lavoro non sarebbe stato possibile senza la lungimiranza e la grande apertura culturale del Dott. Amedeo Bigazzi, ora in pensione, all’epoca Dirigente dell’Area Difesa del Suolo, Risorse Idriche e Natura- li della Provincia di Arezzo, che ha saputo coniugare la passione per le testimonianze storiche con i moderni criteri di analisi e conservazione del territorio. Tutto questo lavoro non sarebbe stato possibile senza la lungimiranza e la grande apertura culturale del Dott. Amedeo Bigazzi, ora in pensione, all’epoca Dirigente dell’Area Difesa del Suolo, Risorse Idriche e Natura- li della Provincia di Arezzo, che ha saputo coniugare la passione per le testimonianze storiche con i moderni criteri di analisi e conservazione del territorio. L’idea di raccontare l’opera dell’illustre scienziato aretino Andrea Cesalpino, consi- derato tra l’altro uno dei padri della botanica sistematica, a più di quattro secoli dalla mor- te, ci ha permesso di verificare, se ce ne fosse stato ancora bisogno, di quanto sia radicato nel nostro territorio l’interesse per lo studio della natura e della botanica in particolare. Alle numerose raccolte di «Erbe Secche», che negli ultimi quattro secoli sono state messe insieme da vari studiosi, si aggiunge il lavoro dei botanici di oggi che con passione non smettono di studiare il nostro territorio. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Ringraziamenti PROVINCIA DI AREZZO La realizzazione di questo volume ha visto la disponibilità, oltre che degli autori, di molte persone e Istituzioni che hanno col- laborato attivamente con suggerimenti ed informazioni, oltre a concedere l’autorizza- zione a riprodurre le immagini degli erbari qui presentati. In particolare si ringrazia- no: Vincenzo Ceccarelli (Presidente della Provincia di Arezzo), Amedeo Bigazzi (già Dirigente della Provincia di Arezzo), la Bi- blioteca del Comune e dell’Accademia Etru- sca di Cortona, la Biblioteca Rilli-Vettori di Poppi, la Biblioteca Comunale di Castiglion Fiorentino, la Biblioteca dell’Istituto e Mu- seo di Storia della Scienza di Firenze, la Biblioteca di Scienze-Sezione Botanica del- l’Università di Firenze, la Sezione Botanica del Museo di Storia Naturale dell’Università di Firenze, il Dipartimento di Biologia del- l’Università di Pisa, Padre Paolo Fantacci- ni (Ministro Provinciale per la Toscana dei Frati Minori), Fra’ Massimo Grassi (Guar- diano del Santuario Francescano della Verna), Paolo Salvi, Gianni Bedini, Piero Cuccuini, Enrico Venturi, Paolo Giulierini, Anna Bernardini, Piero Fusi, Leandro Ra- dicchi, Mauro Frosini, Saulo Bambi. la pubblicazione di questo volume è stata promossa dalla Provincia di Arezzo e dal Museo di Storia Naturale dell’Università di Firenze. Un grato ricordo va a Padre Fiorenzo Lo- catelli, recentemente scomparso, per molti anni Ministro Provinciale per la Toscana dei Frati Minori nonché Padre Guardiano del Santuario Francescano della Verna. la pubblicazione di questo volume è stata promossa dalla Provincia di Arezzo e dal Museo di Storia Naturale dell’Università di Firenze. Presentazione Angelo Maria Cardone Assessore della Provincia di Arezzo alla Difesa del suolo, Acque e Demanio Idrico, Conservazione della Natura e Protezione Civile Premessa Donato Chiatante Presidente della Società Botanica Italiana L a storia dell’umanità contiene molteplici esempi di come la cultura di un popolo sia influenzata profondamente, e spesso in modo determinante, dalle caratteristiche spe- cifiche del territorio: anche in riferimento ai suoi aspetti di naturalità. È pertanto perfet- tamente comprensibile come le genti toscane siano sempre state affascinate ed influenzate dalla dolcezza e dalla bellezza naturalistica dei propri luoghi. In questo particolare an- golo della penisola italiana, le scienze bota- niche hanno avuto grande rilievo e continuo sviluppo, come è testimoniato dalla presenza in questi luoghi di grandi scienziati e dal- la nascita di iniziative culturali di pregio e fortemente innovative quali: la realizzazione degli orti botanici (vedi tra i primi al mon- do quello di Pisa) o la costituzione di società scientifiche (anche queste prime al mondo), quali la Società Botanica Fiorentina e la So- cietà Botanica Cortonese. Questo filo storico si dipana per arrivare fino ai giorni nostri, che vedono ancora la botanica primeggiare tra gli aspetti culturali della regione e dar vita a stu- di ed iniziative di grande rilievo e spessore culturale come la mostra degli erbari areti- ni organizzata dalla Provincia di Arezzo. La presenza a Firenze dell’Erbario Centrale Ita- liano voluto da Parlatore, costituisce sicura- mente il motore principale e più recente che alimenta questo movimento culturale. noscere molti degli aspetti scientifici che hanno motivato la realizzazione di uno spe- cifico erbario. I commenti presentati dagli esperti su tutti gli erbari storici inclusi in questa rassegna, se presi in considerazione in modo sequenziale, potranno anche rap- presentare il percorso storico che metterà il lettore in grado di comprendere come si è evoluta la scienza botanica nei secoli che vanno dal 1500 fino al 1900 in questo ter- ritorio. Ci sono, ad esempio, circostanziate argomentazioni a riguardo dell’erbario di Andrea Cesalpino che mettono in evidenza le stupefacenti intuizioni che questo grande botanico aretino ha utilizzato nel realizza- re la sua prestigiosa collezione. Intuizioni, quelle fatte da Andrea Cesalpino, che pre- cedono di quasi due secoli quelle simili che saranno portate a fondamento dell’opera di riordino tassonomico fatta da Linneo. Non meno interessanti e stimolanti sono anche le altre considerazioni ed i commenti fatti a seguito dello studio degli altri erbari storici presentati in questo volume. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Presentazione Angelo Maria Cardone Assessore della Provincia di Arezzo alla Difesa del suolo, Acque e Demanio Idrico, Conservazione della Natura e Protezione Civile Mi preme sottolineare che l’attività del- l’Assessorato nell’ambito della riscoperta de- gli antichi erbari non si è limitata solo alla mostra ed alla pubblicazione degli atti del Convegno, ma è proseguita anche con il fi- nanziamento del lavoro di restituzione conser- vativa dell’Erbario di Andrea Cesalpino alla forma antecedente gli interventi di fascicola- tura realizzati per volere di Filippo Parlatore nell’800, e del restauro conservativo dell’Er- bario anonimo della Rilliana di Poppi. p L’iniziativa ha rappresentato, quindi, l’oc- casione per poter fare il punto della situazio- ne sui tanti lavori eseguiti in ambito botanico nella provincia di Arezzo e in particolare per quelli contenuti nel progetto di conoscenza e tutela delle risorse naturali dal titolo Carta della Natura della Provincia di Arezzo. Que- sto strumento di analisi del territorio è stato utilizzato per la tutela del patrimonio natura- listico all’interno di un importante strumento di pianificazione quale il Piano territoriale di coordinamento provinciale (PTCP). Il progetto, coordinato dai tecnici della Provincia di Arez- zo, ha rappresentato tra l’altro un momento di importante collaborazione tra l’Amministra- zione che rappresento e i tanti ricercatori del- Sappiamo che ancora molto c’è da fare per far crescere la cultura del rispetto e della tu- tela delle risorse naturali, e vogliamo sperare che anche questa pubblicazione in qualche modo sia uno spunto di riflessione sull’im- portanza di conoscere la flora del nostro ter- ritorio e di operare di conseguenza per la sua conservazione. Gli erbari aretini da Andrea Cesalpino ai giorni nostri XII ovunque, non solo in Toscana, un crescente interesse per gli erbari. Questo interesse è testimoniato dal sempre più alto numero di visitatori che richiedono di visionare gli er- bari nei musei e negli orti botanici univer- sitari. Si susseguono, inoltre, in diverse città italiane numerose mostre, iniziative di ca- talogazione degli erbari quest’ultime spesso sponsorizzate da enti locali o semplici privati cittadini. Certamente non siamo ancora ar- rivati ad avere il numero di visitatori degli erbari che si augurerebbe di ottenere ogni responsabile di un erbario; tuttavia, sareb- be sciocco non cogliere l’importanza anche dei piccoli incrementi. Che esista un grande interesse scientifico per gli erbari da parte dei botanici ‘di professione’ o dei botanici ‘per passione’ è cosa perfettamente com- prensibile e quasi scontata, se si parte dalla constatazione che l’erbario era, e continua ad essere attualmente, un valido, efficiente ed indispensabile strumento di lavoro. Ma sa- rebbe sbagliato pensare che i visitatori che si fermano ad esaminare un erbario siano solo ed esclusivamente dei botanici. Viene allora spontaneo chiedersi cosa attrae un visitato- re non botanico che si trova ad esaminare un erbario. Personalmente ritengo possibile l’esistenza di due motivazioni ben precise e distinte: la prima riguarda certamente l’in- trinseca bellezza ed il fascino estetico di un ‘bene culturale’ come quello rappresentato da un erbario ben fatto e ben conservato; la seconda riguarda la curiosità di conoscere meglio questo strumento di lavoro che or- mai i mezzi di comunicazione hanno portato alla ribalta del pubblico e fuori dalle mura, fino ad ora invalicabili, della «cittadella del- la conoscenza universitaria» Entrambe le motivazioni inducono a chiedersi se non sia utile sfruttare questa opportunità per lan- ciare ai visitatori degli erbari, un messaggio culturale forte. Potrebbe, ad esempio, essere utile mettere in risalto prioritariamente la funzione di catalogazione della biodiversità ricoperta dagli erbari. Nella gente comune comincia a farsi largo la consapevolezza di quanto profondamente la società moderna abbia devastato in pochi decenni il proprio territorio mettendo a rischio la sopravvivenza di molte specie animali e vegetali. La perdita di biodiversità è un concetto che viene col- to nella pienezza del suo significato da tutti. Ecco allora che mostrare i fogli di un erbario potrebbe essere proposto al visitatore come un modo per vedere documentato ‘l’esisten- te’. Premessa L La realizzazione di questa rassegna e poi la decisione di pubblicare questo volume, mi offrono questo piccolo spazio di presentazio- ne per fare qualche brevissima riflessione sul valore attuale degli erbari nella società moderna. Devo partire in questa riflessio- ne dalla premessa che durante i miei anni di impegno alla presidenza della Società Botanica Italiana, ho avuto modo di notare Nelle pagine che seguono questa mia breve premessa, il lettore avrà modo di co- Gli erbari aretini da Andrea Cesalpino ai giorni nostri Introduzione Chiara Nepi e Enrico Gusmeroli N egli ultimi dieci anni si è andato con- solidando sempre più il rapporto di collaborazione tra le Università degli studi di Firenze e Siena con la Provincia di Arez- zo: le prime due con il significativo aumento delle loro attività di ricerca sulla flora e la vegetazione nonché sulla fauna del territorio aretino e l’ultima con la intensificazione de- gli sforzi per la realizzazione di Aree Protette e per la tutela della biodiversità, utilizzando proprio i risultati dell’attività dei ricercatori universitari. presentazione dell’Assessore, ebbe l’idea di raccogliere in una mostra e in un convegno i frutti di questa sorta di ‘censimento’ di tutte le collezioni botaniche che riguardassero il territorio della Provincia, o perché costituite da piante ivi raccolte o perché allestite da botanici locali ovvero fossero conservate in istituzioni aretine. Inoltre, come recitava lo stesso titolo della manifestazione Da Andrea Cesalpino ai nostri giorni, evoluzione delle conoscenze botaniche in provincia di Arezzo, venivano illustrati non solo gli antichi erbari, ma anche quelli moderni, allestiti da botani- ci contemporanei, a significare una sorta di continuum nella ricerche floristiche dal XVI secolo fino ad oggi. N Nell’anno 2005 si è voluto dare visibilità a questa collaborazione e ai risultati delle ri- cerche, con l’organizzazione di un convegno sull’evoluzione delle conoscenze botaniche in provincia di Arezzo, a partire da quel- li che vengono considerati i veri documenti archivistici della flora di un dato territorio: gli erbari. L’argomento risultava ancora più stimolante perché era stato proprio un areti- no, il celebre Andrea Cesalpino, a realizzare l’erbario a carattere sistematico più antico al mondo, nel 1563. E questo erbario si conser- va per l’appunto presso l’Università di Firen- ze, nel suo Museo di Storia Naturale. La mostra, organizzata presso il Museo d’Arte Medievale e Moderna, grazie anche alla collaborazione con la Soprintendenza per i Beni Architettonici e per il Paesaggio, per il Patrimonio Storico, Artistico ed Etnoantropo- logico di Arezzo, vide quindi la riscoperta di un patrimonio archivistico e scientifico sco- nosciuto ai più e, difatti, un notevole successo di pubblico premiò gli organizzatori. Tra l’al- tro, fu anche in seguito a questa manifesta- zione che la Provincia di Arezzo provvide a finanziare una grande campagna fotografica di tutti gli erbari esposti, nonché il restauro conservativo dell’erbario di Poppi e la sle- gatura dell’erbario più antico, il Cesalpino, per assicurarne la migliore conservazione. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Gli erbari aretini da Andrea Cesalpino ai giorni nostri La successione storica di erbari realizzati in tempi diversi diventerebbe nella sua com- plessità come una ‘emeroteca’ della vegeta- zione presente in un determinato territorio. In questo modo il visitatore coglierebbe con immediatezza l’importanza della conservazio- ne degli erbari come mezzo che ci permette di poter confrontare e documentare qualita- tivamente e quantitativamente la storia della vegetazione. L’informatizzazione degli erbari per mezzo delle nuove tecnologie, oltre che velocizzare lo studio degli erbari, potrebbe offrire nuove modalità di presentazione degli stessi ai visitatori.i In definitiva, iniziative come quella rea- lizzata per la presentazione degli erbari storici aretini sono veramente utili per ac- crescere l’interesse della società per gli er- bari: questa è l’unica speranza che rimane ai botanici per reperire i fondi necessari per il loro recupero, la loro valorizzazione e la loro conservazione. Gli erbari aretini da Andrea Cesalpino ai giorni nostri XIV teriormente a rafforzare i rapporti tra l’Ente locale e l’Università di Firenze, in partico- lare con la Sezione Botanica del Museo di Storia Naturale nella quale si conserva anche un altro prezioso erbario aretino, il Coltelli- ni, risalente al XVIII secolo. che attiravano la loro attenzione, le hanno studiate e hanno dato loro dei nomi, a costi- tuire dei veri e propri archivi di dati che in diversi casi possono essere confrontati con quelli attuali, per scoprire eventuali diffe- renze tra passato e presente nella compo- sizione floristica di certe zone. Non solo, il volume vuole rendere testimonianza anche della cura con cui queste collezioni sono state realizzate, sia che si tratti di raccol- te vere e proprie con piante essiccate, da quelle più antiche fino alle moderne, sia di tavole dipinte: quasi che le conoscenze scientifiche non potessero e non possano ancora oggi prescindere dalla bellezza della disposizione o raffigurazione del campione vegetale, come risulta evidente dalle imma- gini a corredo dei singoli capitoli. Questo volume vuole quindi ‘chiudere’ idealmente il percorso che, iniziato più di tre anni fa con la ricerca e l’ostensione degli er- bari, si completa con la descrizione accurata dei loro contenuti e, quando conosciuta, della loro storia. Il tutto arricchito ed impreziosito dalle immagini dei fogli che contengono le piante, talvolta raffigurate, talvolta essiccate. Il volume è il frutto di collaborazioni tra autori di diversa formazione culturale (bota- nici, paleografi, storici, conservatori di beni librari) che hanno studiato gli erbari da an- golature diverse. Ne deriva un volume non omogeneo nella trattazione degli argomenti, che mette in evidenza via via aspetti e consi- derazioni diverse sugli erbari, talvolta eviden- ziando la loro collocazione storica, talvolta il loro valore scientifico, talvolta la semplicità ‘narrativa’ della flora del territorio aretino. Il volume, infine, è reso ancora più pre- zioso dalla ristampa anastatica dell’opera di Teodoro Caruel (1830–1898) dedicata pro- prio all’erbario Cesalpino, la Illustratio in hortum siccum Andreae Caesalpini, pubbli- cata nel 1858. In questo libro Caruel, che succederà alla direzione dell’Erbario di Fi- renze dopo Filippo Parlatore, non solo elenca il contenuto dell’intero erbario, con la fedele trascrizione dei nomi delle piante dati da Ce- salpino in greco, latino e italiano, ma ne cita anche il riferimento, quando presente, al- l’opera del medico aretino del 1583, dal titolo De Plantis libri XVI. Introduzione Proprio quest’ultimo intervento contribuì ul- Dal conosciuto erbario di Andrea Cesal- pino alla scoperta che nel territorio aretino sono presenti molti erbari e, tra questi, alcu- ni con un notevole interesse storico oltre che scientifico, il passo è stato breve. L’allora diri- gente della Provincia di Arezzo, il dott. Ame- deo Bigazzi, come già ricordato anche nella Gli erbari aretini da Andrea Cesalpino ai giorni nostri Gli erbari aretini da Andrea Cesalpino ai giorni nostri Inoltre Caruel descrive il campione e il suo stato di conservazione e, elemento importantissimo, fa l’aggiorna- mento nomenclaturale della specie, talvolta con considerazioni sulla sua identificazione. Si tratta, in definitiva, dello strumento fon- damentale per gli studiosi per facilitare la ‘lettura’ dell’erbario cinquecentesco ed oggi viene riproposto in questa edizione anche con la traduzione della presentazione in lati- no che l’autore pose all’inizio, cosa utilissima per quanti vorranno avvicinarsi alla cono- scenza di questa pietra miliare della storia della botanica sistematica. Il carattere disomogeneo viene poi ac- centuato anche dalle differenze insite nelle collezioni stesse: si passa dall’erbario Cesal- pino, realizzato dallo scienziato aretino ma con piante raccolte per lo più lungo la costa toscana e nel pisano in particolare, all’erbario Moneti, costituito da tavole raffiguranti piante proprie del cortonese o all’erbario Corinaldi, che seppure contenente piante raccolte addi- rittura in Egitto, è tuttavia conservato a Mon- tevarchi, fino ad arrivare agli erbari moderni con le piante dell’Appennino aretino. Si tratta, in definitiva, dello strumento fon- damentale per gli studiosi per facilitare la ‘lettura’ dell’erbario cinquecentesco ed oggi viene riproposto in questa edizione anche con la traduzione della presentazione in lati- no che l’autore pose all’inizio, cosa utilissima per quanti vorranno avvicinarsi alla cono- scenza di questa pietra miliare della storia della botanica sistematica. Al di là comunque delle differenze an- che nell’importanza storica e scientifica dei singoli erbari, questo volume ha voluto ren- der conto del lavoro, sempre appassionato e molto spesso motivato solo dall’interesse per- sonale, di quanti – medici, farmacisti, inse- gnanti, preti o, finalmente, botanici – hanno pazientemente raccolto o illustrato le piante I primi erbari della storia di questi codici sono stati realizzati fra il IV- V secolo e il XVI secolo per lo più nei mo- nasteri, nei conventi o comunque nei luoghi dove esistevano delle «spezierie», cioè dove le piante venivano coltivate e utilizzate come medicamenti (i cosidetti «semplici»). Le piante che venivano usate a scopo medicina- le (o talora anche quelle ad uso alimentare) venivano perciò riprodotte in appositi mano- scritti che spesso sono giunti fino a noi, come i codici derivati dall’opera di Dioscoride, dei secoli V-IX o i manoscritti figurati dei secoli X-XV, come l’Herbarium di Apuleius Pla- tonicus (XI sec.), l’Herbolaire francese (XV sec.) (Fig. 1) o l’Ortus sanitatis del 14912. L’illustrazione dell’erbario di Andrea Cesal- pino non può prescindere dalla storia di que- sto tipo di collezioni scientifiche, gli erbari o «Orti secchi» (Horti sicci) come venivano definiti nel XVI secolo, in contrapposizione con gli orti botanici o «Orti vivi» (Horti vivi), cioè con le collezioni di piante viventi. L’erbario è in effetti una collezione di piante secche, cioè di campioni vegetali es- siccati e pressati in modo che possano esse- re conservati su di un foglio di cartoncino, generalmente incollati o fermati con spilli. Negli erbari moderni ogni esemplare è ac- compagnato da un’etichetta sulla quale è in- dicato il nome della pianta, il luogo dove è stata raccolta, la data di raccolta e il nome di chi ha prelevato il campione. I campioni vegetali così essiccati si possono conservare per molti anni e anche per secoli, purché sia- no mantenuti al riparo da attacchi di insetti o di muffe1. Le immagini di questi codici sono tuttavia spesso molto sommarie e talora anche poco aderenti alla realtà, in quanto non sempre sono state realizzate dipingendo piante vive ma ricopiando figure preesistenti. Per tali motivi l’erbario di piante essiccate ha mol- to maggior valore documentario in quanto conserva le piante nella loro realtà, anche se in parte deformate dall’essiccazione e dalla compressione. L’idea di seccare le piante per conservar- le indefinitamente è relativamente recente, poiché si fa risalire al XVI secolo; sembra infatti che prima di quel periodo non esistes- se l’usanza di seccare le piante per poterle avere a disposizione per studio o per consul- tazione. Fino alla fine del XV secolo erano invece molto diffusi i cosiddetti Herbaria che erano dei codici dipinti, dove le piante veni- vano raffigurate, generalmente a colori. 1 G. Moggi, Storie di collezioni di piante: gli erbari fiorentini, «Atti Soc. Leonardo da Vinci», Ser. 5, 3, 1984, pp. 49-66; Id., L’erbario. Origine, evoluzione storica, significato, in F. Montacchini (ed.), Erbari e iconografia botanica. Storia delle collezioni dell’Orto Botanico dell’Università di Torino, U. Allemandi & C., Torino 1896, pp. 24-28. Fig. 1 Due pagine del codice dipinto Herbolaire o Grant Herbier, manoscritto illustrato del XV secolo. 2 F.J. Anderson, An illustrated history of the herbals, Columbia Univ. Press, New York, 1977; A. Arber, Herbals. Their origin and evolution, Cambridge University Press, Cambridge 1990 (3rd ed.); M. Collins, Medieval Herbals. The Illustrative Traditions, The British Library, London 2000. L’erbario di Andrea Cesalpino Guido Moggi L’erbario di Andrea Cesalpino Fig. 1 Fig. 1 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press U. Allemandi & C., Torino 1896, pp. 24-28. I primi erbari della storia Molti Uno dei più antichi erbari sembra sia sta- to quello di Gherardo Cibo (1512-1600), tut- tora conservato presso la Biblioteca Angelica di Roma e la cui realizzazione viene fatta ri- salire al 1532. Alla metà del ’500 si fanno risalire anche gli erbari dell’abate lucchese Michele Merini, che si trova a Firenze al Mu- seo di Storia Naturale, quello di Ulisse Al- Guido Moggi approfondita delle piante si poteva ottenere soltanto se queste potevano essere conserva- te indefinitamente e potevano essere esami- nate anche da altri studiosi, ciò che avrebbe permesso lo scambio di opinioni e quindi un maggior approfondimento delle idee. Ciò si può realizzare soltanto conservando le pian- te sotto forma di campioni essiccati, i cui esemplari possono essere distribuiti ai vari corrispondenti nelle diverse parti del mon- do. Ciò è proprio quello che farà Ghini con Aldrovandi, Mattioli, Petrollini, Calzolari ed altri. Nasce in questo modo l’erbario, inteso appunto come collezione di piante essicca- te e pressate, incollate su fogli di carta o di cartoncino5. La paternità di questa idea, come si è detto, viene attribuita a Luca Ghini, il quale tuttavia non tratterrà per sè i campioni es- siccati ma li distribuirà ai suoi studenti, agli amici e ai collaboratori, stimolandoli a fare altrettanto. Non è un caso che gli erbari più antichi che oggi si conoscano, come quelli citati di Merini, di Cesalpino, di Aldrovandi, ecc., siano tutti opera di allievi o corrispon- denti di Ghini. Anche il Mattioli sfruttò ade- guatamente questo sistema di conservazione delle piante per poterle descrivere nei suoi Discorsi e nei Commentarii; è documentato che egli si faceva mandare da Ghini campio- ni d’erbario per poter descrivere le diverse specie nelle sue opere6. Fig. 2 Ritratto di Luca Ghini (1490-1556). Museo Botanico del Dipartimento di Biologia, Università di Pisa. Fig. 2 Ritratto di Luca Ghini (1490-1556). Museo Botanico del Dipartimento di Biologia, Università di Pisa. drovandi (1522-1605), conservato presso la Biblioteca Universitaria di Bologna, quello di Cesalpino, del 1563, ecc., e, all’estero, quel- lo dell’inglese William Turner (1510-1568), del tedesco Caspar Ratzenberger (di cui è menzionato un erbario iniziato nel 1556), del francese Jehan Girault (1558), ecc.3 Fig. 2 Ritratto di Luca Ghini (1490-1556). Museo Botanico del Dipartimento di Biologia, Università di Pisa. Ma fra tutti questi erbari cinquecenteschi il più importante resta senza dubbio quello di Cesalpino per i motivi che vedremo più avanti. I primi erbari della storia Gli erbari di piante essiccate quindi prendono presto il sopravvento sugli «erbari dipinti» e pian piano andranno a sostituir- li come elementi di documentazione di una realtà scientifica molto più attendibili di questi ultimi. 3 Arber, op. cit.; G. Moggi, Andrea Cesalpino (1525-1603) e il suo erbario, in E. Gusmeroli, A. Bigazzi (a cura di), Da Andrea Cesalpino ai nostri giorni. Erbari aretini in mostra, Catalogo della mostra, Arezzo, 4 marzo-27 maggio 2005, 2005, pp. 3-5; G. Moggi, Andrea Cesalpino “fondatore” della bo- tanica sistematica, in E. Gusmeroli, L. Lastrucci (a cura di), Atti del Convegno “Evoluzione delle cono- scenze botaniche e problematiche della conservazione in provincia di Arezzo da Andrea Cesalpino ad oggi”, Arezzo, 4 marzo 2005, Uni- versità degli Studi di Firenze-Arti Grafiche Cianferoni, Stia (Arezzo) 2006, pp. 8-18. 6 Cristofolini, op. cit., p. 214; S. Ferri, Il “Dioscoride”, i “Discorsi”, i “Commentarii”: gli amici e i nemici, in S. Ferri (a cura di), Pietro Andrea Mattioli (Siena 1501- Trento 1578). La vita, le opere. Con l’identificazione delle piante, Quattroemme, Perugia 1997, pp. 15-48; F. Garbari, Luca Ghini a Pisa, cardine della cultura botanica del XVI secolo, «Museol. Sci.», 8, 1992 (1991), pp. 223-236. 7 U. Viviani, Tre medici aretini (A. Cesalpino, F. Redi e F. Folli), R. Accad. Petrarca, Arezzo 1936. 4 G. Cristofolini, Luca Ghini a Bologna: la nascita della scienza moderna, «Museol. Sci.», 8, 1992 (1991), pp. 207-221. 5 Moggi, L’erbario, cit. 3 Arber, op. cit.; G. Moggi, Andrea Cesalpino (1525-1603) e il suo erbario, in E. Gusmeroli, A. Bigazzi (a cura di), Da Andrea Cesalpino ai nostri giorni. Erbari aretini in mostra, Catalogo della mostra, Arezzo, 4 marzo-27 maggio 2005, 2005, pp. 3-5; G. Moggi, Andrea Cesalpino “fondatore” della bo- tanica sistematica, in E. Gusmeroli, L. Lastrucci (a cura di), Atti del Convegno “Evoluzione delle cono- scenze botaniche e problematiche della conservazione in provincia di Arezzo da Andrea Cesalpino ad oggi”, Arezzo, 4 marzo 2005, Uni- versità degli Studi di Firenze-Arti Grafiche Cianferoni, Stia (Arezzo) 2006, pp. 8-18. Arezzo da Andrea Cesalpino ad oggi”, Arezzo, 4 marzo 2005, Uni- versità degli Studi di Firenze-Arti Grafiche Cianferoni, Stia (Arezzo) 2006, pp. 8-18. pp 4 G. Cristofolini, Luca Ghini a Fig. 2 Ritratto di Luca Ghini (1490-1556). Museo Botanico del Dipartimento di Biologia, Università di Pisa. La vita e l’opera scientifica di Andrea Cesalpino nel campo della botanica Non è ben chiaro come sia nata l’idea del- l’erbario come strumento di conservazione di campioni di piante né si conosce chi sia stato veramente il primo che abbia adottato questo metodo; tuttavia tutti gli studiosi sono concordi nel ritenere Luca Ghini l’ideatore di questo sistema di conservazione delle piante, anche se nessun erbario a lui attribuibile è pervenuto fino a noi4 (Fig. 2). Andrea Cesalpino (Fig. 3) nacque nel 1525 ad Arezzo o forse, secondo alcune opinioni, nella campagna intorno alla città7. Poco si conosce sulla sua vita nel periodo aretino; vi è stata tuttavia a lungo una controversia in merito all’anno di nascita, che in mol- te delle opere scritte su Cesalpino nel XIX secolo risulta essere il 1519. Tuttavia, studi Ghini infatti, nel corso del suo insegna- mento della botanica medica all’Università di Pisa, si rese conto che una conoscenza 5 l’ErBArIo dI ANdrEA CESAlPINo 1583. Nominato «Professore ordinario di me- dicina pratica», continuò nell’insegnamento fi no al 1591 quando venne chiamato a Roma da papa Clemente VIII per assumere la cari- ca di insegnante di medicina alla «Sapienza» e di archiatra pontifi cio, e qui restò fi no alla sua morte, avvenuta nel 160311. più recenti8, in base all’esame di documenti originali, stabiliscono il 1525 come data di nascita, che è stata confermata anche dagli studiosi che si sono occupati di Cesalpino nel XX secolo9. Nel 1545, quindi all’età di 20 anni, si sa- rebbe iscritto all’Università di Pisa, dove si laureò in medicina intorno al 155110. A Pisa Cesalpino seguì le lezioni di botanica medica di Ghini, che era appunto «Lettore de’ Sem- plici», terminologia che veniva usata a quel tempo. Quando Ghini si trasferì a Bologna nel 1555, Cesalpino lo sostituì nell’insegnamento della «materia medica» e nella conduzione dell’orto botanico pisano, di cui terrà la guida formale fi no al 1558, come «Prefetto» dell’or- to. Con lo spostamento dell’orto botanico dalla sede originaria (presso l’arsenale) alla zona di S. Marta venne affi data di nuovo a Cesalpino la «prefettura» dell’Orto che terrà dal 1563 al Durante il periodo pisano Cesalpino deve avere svolto una intensa attività di studioso, di ricercatore e di insegnante. Giovanni Tar- gioni Tozzetti, nei suoi manoscritti conservati alla Biblioteca Nazionale di Firenze12, affer- ma che Cesalpino effettuò frequenti viaggi in tutta la Toscana e altrove per incrementare le sue conoscenze sulla fl ora locale. Fig. 3 ritratto di Andrea Cesalpino (1525-1603). Museo Botanico del dipartimento di Biologia, Università di Pisa. 8 G. lais, Documenti inediti di Andrea Cesalpino, «Atti Accad. Pontif. Nuovi lincei», 35, 1882, pp. 95-102. 9 Viviani, Tre medici aretini, cit.; r. Pazzagli, Andrea Cesalpino e la scoperta della circolazione del sangue nel terzo centenario della morte di Guglielmo Harvey, «Boll. Mem. Soc. Tosco-Umbra Chir». 19(5), 1958, pp. 454-464; r. Pazzagli, Andrea Cesalpino, «Il Cesalpino, Boll. Soc. Med.-Chir. Aretina», 1, 1960, pp. 9-15; r. Pazzagli, Il Cesalpino oggi, «Atti Mem. Acc. Petrarca lett. Arti Sc Arezzo», n.s. 40, 1970-72, 1974, pp. 206-212. 10 G. Moggi, Andrea Cesalpino botanico, «Atti Mem. Acc. Petrarca lett. Arti Sci. Arezzo», n.s. 42: 235-249, 1981 (1976-78); Pazzagli, Andrea Cesalpino e la scoperta della circolazione, cit.; Pazzagli, Andrea Cesalpino, cit. 8 G. lais, Documenti inediti di Andrea Cesalpino, «Atti Accad. Pontif. Nuovi lincei», 35, 1882, pp. 95-102. 12 G. Targioni Tozzetti, Selva di notizie spettanti all’origine de’ progressi e miglioramenti delle Scienze fi siche in Toscana, per uso del dottore Ottaviano suo fi glio, Ms. 189, 17 volumi («le Selve»), Biblioteca Nazionale Centrale di Firenze, Fondo Targioni Tozzetti (BNCF-Fondo TT). 11 U. Viviani, L’iconografi a, la vita e le opere di Andrea Cesalpino, «Il Cesalpino», Arezzo, 13 (n. 5, 6, 10, 11, 15, 16), pp. 1-95, tavv. 1-13, 1917; U. Viviani, Vita ed opere di Andrea Cesalpino, Viviani, Arezzo 1927; Moggi, Andrea Cesalpino botanico, cit.; Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit. Fig. 4 Il frontespizio dell’opera di Cesalpino De Plantis Libri XVI, stampata a Firenze nel 1583 (Biblioteca di Scienze-Sezione Botanica dell’Università di Firenze foto di Egildo Luccioli). Per comprendere il significato dell’opera botanica di Cesalpino e l’importanza del suo erbario è necessario risalire agli inizi del XVI secolo e ricapitolare quelle che erano le conoscenze scientifiche dell’epoca. È oppor- tuno ricordare infatti che fino ai primi del ’500 tutta la scienza naturalistica risentiva ancora delle opere dell’antichità greca e ro- mana. Aristotele, Teofrasto, Dioscoride, Pli- nio, Galeno erano ancora i grandi ispiratori degli scienziati a cavallo fra ’400 e ’500 sia dal punto di vista teorico-filosofico che da quello pratico17. Apuane, alla Verna, a Vallombrosa, nei din- torni di Arezzo, in Val Tiberina, oltre che in Maremma, all’isola d’Elba e nella campagna Romana13. I risultati di questi viaggi furono sicura- mente numerosi campioni di piante da lui raccolti per essere in parte messi in colti- vazione nell’Orto botanico pisano e in parte essiccati; questi ultimi con molta probabilità andarono a costituire il nucleo principale de- gli erbari da lui realizzati. Anche Cesalpino si dimostra fondamen- talmente un seguace della filosofia aristote- lica, dimostrando come le nuove concezioni ed i nuovi modi di affrontare i problemi della scienza abbiano inizialmente convissuto con gli antichi concetti dei naturalisti-filosofi gre- ci e romani. Ad esempio nell’analogia aristo- telica fra animali e piante, attraverso la quale il filosofo greco vede in ogni organismo un centro propulsore delle attività vitali (il cuo- re), anche Cesalpino riconosce nelle piante un organo analogo, il cor medullae o sempli- cemente cor, posto alla base della pianta fra il fusto e la radice, al quale si deve lo svilup- po di tutti gli organi aerei (fusto, foglie, fiori, ecc.) e sotterranei (radice, bulbi, ecc.) della pianta. Egli non conosce ancora l’esistenza e la funzione dell’embrione, e pertanto questo «centro propulsore» della pianta viene iden- tificato nel punto dal quale sembrano sorge- re tutte le parti che costituiscono l’organismo vegetale. In queste somiglianze fra animali Pichi Sermolli ci ricorda due viaggi di Cesalpino, uno alla Verna e l’altro sulle Alpi Apuane14, dei quali però non si conoscono né le date né altri particolari. Infatti, come si dirà più avanti, né nel suo libro De Plantis Libri XVI né sui fogli del suo erbario ancora esistente sono citate località di raccolta; per- tanto possiamo ricostruire le destinazioni dei suoi viaggi solo attraverso il resoconto che ci ha trasmesso Giovanni Targioni Tozzetti ne «Le Selve». 13 Viviani, L’iconografia, la vita e le opere di Andrea Cesalpino, cit.; Viviani, Vita ed opere di Andrea Cesalpino, cit.; Viviani, Tre medici aretini, cit.; R.E.G. Pichi Sermolli, Contributo alla storia della Botanica in Toscana. I precursori dell’esplorazione floristica delle Alpi Apuane, «Museol. sci.». 15(2), Suppl., 1999, pp. i-v, 1-289. 14 R.E.G. Pichi Sermolli, Da Cesalpino a Fra’ Ginepro. Cenni sulla storia dell’esplorazione floristica della Verna, in E. Ferrarini, R.E.G. Pichi Sermolli (a cura di), La Verna, Cantico delle Creature. I fiori del Monte di Francesco visti da Fra’ Ginepro, pp. 37-52, La Verna, 1998; Pichi Sermolli, Contributo alla storia della Botanica in Toscana, cit. 15 Pazzagli, Andrea Cesalpino, cit.; Pazzagli, Il Cesalpino oggi, cit. 16 Pazzagli, Andrea Cesalpino e la scoperta della circolazione, cit. 17 A.G. Morton, History of Botanical Science, Academic Press, London 1981; Moggi, Andrea Cesalpino botanico, cit.; Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit.; F. Garbari, I “prefetti” del Giardino, dalle origini, in F. Garbari, L. Tongiorgi Tomasi, A. Tosi, Giardino dei Semplici: l’Orto Botanico di Pisa dal XVI al XIX secolo, Cassa di Risp. di Pisa, Pacini, Ospedaletto (Pisa) 1991, pp. 27-114. La vita e l’opera scientifica di Andrea Cesalpino nel campo della botanica Anche se non abbiamo una documentazione diretta dei suoi viaggi, dalle notizie riferite da Targioni Tozzetti emerge che egli erborizzò special- mente nei dintorni di Pisa, in Lucchesia, nel Pistoiese, nell’Appennino Ligure, sulle Alpi Guido Moggi Fig. 4 Il frontespizio dell’opera di Cesalpino De Plantis Libri XVI, stampata a Firenze nel 1583 (Biblioteca di Scienze-Sezione Botanica dell’Università di Firenze foto di Egildo Luccioli). Pubblicò tre libri sui minerali, dedicandosi in particolare ai marmi e alle gemme, tentò di spiegare l’origine dei fossili, intuì la pre- senza dell’ossigeno nell’aria, e così via15. Non è questa la sede per ricordare i meriti di Ce- salpino come medico (è noto che a lui si deve l’intuizione che permise la scoperta della circolazione del sangue)16 e come naturalista edotto in altre discipline. Va invece ricordato come profondo studioso e importante capo- saldo nel campo della botanica, in quanto è ormai riconosciuto da tutti come colui che per primo ha introdotto i concetti di base della sistematica delle piante ed ha tentato una classificazione degli organismi vegetali conosciuti a quell’epoca. L’erbario di Andrea Cesalpino (plantarum genera) in base al portamento ed ai caratteri riproduttivi; con i caratteri dei frutti e dei semi quindi distingue all’interno dei gruppi categorie inferiori fino alla ultima species. Va notato per inciso che i termini di genere e di specie non avevano ancora assun- to il significato che noi oggi diamo loro, tanto che per Cesalpino genera sono i gruppi più elevati, oggi riconducibili a ordini e classi. e piante, ancora confermate da Cesalpino, si riconosce l’ispirazione aristotelica della sua scienza; ed è su questi punti che si soffer- marono in seguito coloro che criticarono le idee di Cesalpino, come ad esempio lo sto- rico della botanica Sachs nel 1890, il quale, pur ammirando per alcuni aspetti l’opera di Cesalpino, lo incolpa di essere stato troppo influenzato dalle idee di Aristotele. Tuttavia Sachs e coloro che ne seguirono le idee sot- tovalutarono gli aspetti sistematici della bo- tanica di Cesalpino, che sono quelli che ne qualificano l’opera innovativa. Nel suo sistema, Cesalpino separa le Crit- togame dalle Fanerogame definendo le prime per l’assenza di frutti e di semi; e la suddi- visione delle Crittogame è condotta con cri- teri così accurati che resterà inalterata per quasi tre secoli. Sarà infatti necessario il microscopio per approfondire adeguatamen- te le conoscenze, in modo da modificare in maniera sostanziale il sistema classificatorio usato fino allora. Le idee sistematiche di Cesalpino sono tutte compendiate nell’unico libro che egli ha scritto su questo argomento, intitolato De Plantis Libri XVI (Fig. 4). Questo libro si sud- divide come dice il titolo in 16 capitoli: nel primo egli espone tutte le sue idee sulla biolo- gia e la sistematica vegetale; negli altri 15 de- scrive più di 1300 specie di piante, suddivise in «gruppi» e categorie secondo criteri origi- nali, da lui esposti capitolo per capitolo18. Nelle piante a seme (le attuali Fanero- game), Cesalpino si accorge dell’esistenza di un preciso rapporto fra il frutto e le parti fio- rali e introduce nella classificazione un crite- rio ancora oggi di grande valore sistematico, ossia ciò che modernamente viene definito come ovario supero e ovario infero. Va pre- cisato che Cesalpino non ha ancora chiaro il significato di ovario, di frutto e di seme nella concezione attuale e quindi ciò che lui chia- ma «seme» è in realtà talora il frutto o anche una sua parte, o addirittura l’embrione20. 18 A. Cesalpino, De Plantis Libri XVI, apud G. Marescottum, Florentiae 1583; T. Caruel, Andrea Cesalpino e il libro De Plantis. «N. Giorn. Bot. Ital.», 4(1), 1872, pp. 23-48; C.E.B. Bremekamp, A re-examination of Cesalpino’s classification, «Acta Bot. Neerl.», 1, 1953, pp. 580-593; Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit. 19 Moggi, Andrea Cesalpino botanico, cit.; Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit. 20 Bremekamp, A re-examination of Cesalpino’s classification, cit.; Morton, History of Botanical Science, cit.; Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit. 21 Bremekamp, A re-examination of Cesalpino’s classification, cit. Anche se Cesalpino è conosciuto preva- lentemente per la sua attività di medico, va ricordato che, come avveniva spesso in quei tempi, egli si occupò di molte discipline na- turalistiche, dalla mineralogia alla paleon- tologia, dalla chimica alla botanica, ecc. L’erbario di Andrea Cesalpino L’erbario di Andrea Cesalpino Nel primo capitolo Cesalpino non si limi- ta a trattare della classificazione, ma affronta altri aspetti dello studio della pianta, come la nutrizione, l’assorbimento dell’acqua, la circolazione, ecc. Inoltre introduce concetti nuovi rifiutando i sistemi basati sull’uso delle piante («gruppi farmacologici»), sull’aspetto generale o sui caratteri utilitaristici (sapore, odore, ecc.). Egli afferma che la classificazio- ne si deve basare sui caratteri, sulla struttura degli organi, ecc. e non sulle proprietà della pianta o sull’uso che ne vien fatto. I caratteri fondamentali devono essere quelli morfologici perché sono più stabili; e, fra questi, quelli da usare per definire i gruppi superiori devo- no essere i caratteri relativi alla riproduzione (fiori, frutti, semi), mentre quelli desunti dalle foglie, dal fusto, dalle radici possono servi- re per definire le specie. Per ogni carattere Cesalpino inoltre stabilisce che occorre dare maggiore importanza al numero delle parti, alla loro posizione, alla forma, ecc. a cui si possono aggiungere caratteri qualitativi, come la consistenza, il colore, ecc. Tutti i caratteri sono combinati dalla natura in vario modo per realizzare le varie specie di piante19. p Sulla base dei caratteri dei ‘semi’ Cesal- pino perciò distingue dei gruppi di largo va- lore tassonomico in molti dei quali possiamo riconoscere quelle che oggi vengono definite come famiglie, anche se il concetto di fami- glia ancora non è presente nella botanica cinquecentesca poichè farà la sua comparsa solo alla fine del ’600. Ad esempio i ‘gruppi’ 6 e 17 definiscono l’attuale famiglia delle Le- guminosae, il ‘gruppo’ 19 le Umbelliferae, il ‘gruppo’ 28 le Labiatae, il ‘gruppo’ 27 le Bo- raginaceae, e così anche per le Compositae, le Liliaceae, le Cruciferae, ecc.21. Un altro aspetto del libro De Plantis meri- ta di essere sottolineato. In un periodo in cui cominciano ad essere pubblicate opere bota- niche abbondantemente illustrate, il libro di Cesalpino non possiede figure di nessun tipo. Intorno alla metà del ’500 numerosi testi di botanica provvisti di illustrazioni vengono Facendo seguito alle premesse sopra illu- strate, Cesalpino suddivide i gruppi superiori Guido Moggi pubblicati in Italia (come il Mattioli a partire dal 1554) e all’estero (come i ben noti testi di Brunfels, del 1530 e di Fuchs, del 1542, in Germania), poichè l’immagine, redatta ormai con accuratezza scientifica e non più somma- ria come erano quelle dei secoli precedenti, era ritenuta essenziale per meglio illustrare le specie descritte. L’erbario di Andrea Cesalpino Ciò potrebbe far pensare che Cesalpino non ritenesse necessario cor- redare la sua opera con illustrazioni espli- cative. Invece un attento esame del libro De Plantis dimostra come Cesalpino avrebbe voluto corredate la sua opera con figure e che addirittura ne aveva fatte preparare alcune che oggi purtroppo sono andate perdute. via sempre più ampie in base alla conver- genza di determinati caratteri23. Valutando quindi queste convergenze ed i caratteri che hanno determinato le categorie cesalpiniane, si può ricostruire uno schema di classifica- zione che ricalca né più e né meno le attuali ‘chiavi analitiche’ dicotomiche, come quelle che oggi si trovano in qualsiasi testo di bota- nica sistematica. Ecco quindi in sintesi i tre elementi essen- ziali che costituiscono il fondamento dell’im- portanza scientifica dell’opera di Cesalpino: il riconoscimento dei caratteri differenziali come elementi di base della biodiversità; il raggruppamento degli organismi vegetali in gruppi sistematici omogenei; il confronto re- ciproco fra i vari raggruppamenti e quindi la loro gerarchizzazione in categorie sempre più ampie e generali. Questi sono gli elementi es- senziali che stanno alla base della botanica sistematica. Ed è appunto il riconoscimento dell’opera di Cesalpino in questo campo che ci permette di definirlo come il ‘fondatore’ di questa disciplina. Se si vuole riassumere in breve il signi- ficato dell’opera scientifica di Andrea Ce- salpino nel campo della botanica, possiamo rilevare tre aspetti fondamentali22. Anzitutto egli per la prima volta mette l’accento sull’importanza dei caratteri distin- tivi, specialmente quelli morfologici, di cui segnala in particolare il significato nel nu- mero, nella rispettiva posizione e nell’aspetto (numerus, situs, figura). Rileva il valore fon- damentale dei caratteri riproduttivi, ai quali dà un significato primario come elementi di- stintivi. Ma in particolare riconosce come le caratteristiche di un sistema naturale si deb- bano basare sulle affinità e le differenze dei caratteri. Egli infatti afferma come scientia omnis in similium collectione et dissimilium distinctione consistat: è questo in sintesi il fondamento della biodiversità naturale. Di questi elementi resterà traccia nella sua unica opera botanica di un certo valore, cioè il libro De Plantis Libri XVI del 1583; ma già nell’erbario (datato 1563) possiamo trovare le basi pratiche della sua classificazione. 22 Moggi, Andrea Cesalpino “fondatore” della botanica sistematica, cit., p. 13. 23 Bremekamp, A re-examination of Cesalpino’s classification, cit. 24 Cesalpino, op. cit., p. (9); P.A. Micheli, Descriptio et Illustratio Horti Sicci quem Ccl: Andreas Caesalpinus Alphonso Tornabonio Episcopo Biturgensi paravit nunc vero in Bibliotheca D: D: Pandulphinorum asservatur, Ms. 9, Biblioteca di Botanica dell’Università, Manoscritti Micheli (BUB-Mich.). 25 Targioni Tozzetti, Selva di notizie, cit., vol. V, p. 41; Pichi Sermolli, Contributo alla storia della Botanica in Toscana, cit., p. 27. L’erbario di Cesalpino: vicende storiche L’erbario di Andrea Cesalpino, tuttora con- servato presso il Museo di Storia Naturale dell’Università di Firenze, è l’unico di questo grande botanico che è pervenuto fino a noi ed è appunto l’espressione più evidente delle idee e dei concetti che esporrà venti anni più tardi, nel 1583, nel libro De Plantis Libri XVI. Un secondo aspetto per il quale Cesalpino deve essere considerato innovatore risiede nel fatto che, come si è visto, per la prima volta egli raggruppa le piante in categorie (basate appunto sui criteri di affinità) che non hanno nulla a che vedere con i gruppi di piante che erano riconosciuti da Teofrasto o da Plinio. In questo modo egli viene a realizzare una clas- sificazione delle piante interamente nuova e fondata su criteri strettamente scientifici. È sicuro che Cesalpino avesse realizzato almeno due erbari: di uno, dedicato al gran- duca Cosimo I, ci parla lui stesso nella pre- fazione del libro citato24, ma fin dalla prima metà del ’700 non se ne hanno più notizie. Infatti anche Giovanni Targioni Tozzetti nei suoi manoscritti afferma «non si è potuto in questi ultimi tempi ritrovare»25. Il terzo elemento nell’opera di Cesalpino che merita di essere sottolineato riguarda i criteri di ordinamento con cui egli mette a confronto i diversi ‘gruppi’ da lui identificati. Cesalpino infatti identifica nei 15 capito- li del suo libro 34 «gruppi» che a loro volta possono essere raggruppati in categorie via Il terzo elemento nell’opera di Cesalpino che merita di essere sottolineato riguarda i criteri di ordinamento con cui egli mette a confronto i diversi ‘gruppi’ da lui identificati.i Cesalpino infatti identifica nei 15 capito- li del suo libro 34 «gruppi» che a loro volta possono essere raggruppati in categorie via L’erbario del Museo di Storia Naturale, dedicato dall’autore al vescovo di Borgo San L’erbario di Andrea Cesalpino Fig. 5 La parte iniziale della lettera, indirizzata al vescovo Tornabuoni, che apre l’erbario di Cesalpino. ta («E desiderando V. S. Rma che io gli faces- si una ragunata de semplici ataccati sopra e fogli per riconoscerli…»). Dopo la morte di Alfonso Tornabuoni non si hanno più notizie dell’erbario per più di un secolo. L’anno 1714, mentre il Signr: Guglielmo Sherard di- morava in Smirne ebbi l’onore di alcune sue lettere toccanti varie Piante dell’Opera De Plantis d’Andrea Cesalpino stampata in Firenze, e presso Giorgio Ma- rescotto l’anno 1583; e con le quali sue erudite lette- re non cessava di dare a Noi impeto di ricercare i due Orti secchi fatti dal medesimo Cesalpino, rammen- tati da esso nella Prefazione dedicatoria di detta Sua Opera. L’anno 1717, nel mese di maggio, furono mag- giori gl’impulsi, mentre nel ritorno che faceva detto Sigr: G. Sherard dalle Smirne in Inghilterra, passò per Firenze, e di gran lunga accalorò la cosa, e di tal Fig. 5 La parte iniziale della lettera, indirizzata al vescovo Tornabuoni, che apre l’erbario di Cesalpino. L’erbario di Cesalpino: vicende storiche Giovanni Targioni Tozzetti nel suo Prodromo della Co- rografia della Toscana28 afferma che Stefano Rosselli, speziale della Corte Medicea alla fine del XVI secolo, avrebbe avuto occasione di esaminarlo e ne avrebbe redatto un cata- logo delle piante, con annotazioni e commen- ti, con l’intendimento di pubblicarlo29. Non vi sono tuttavia conferme su questa ipotesi e nessuna altra notizia esiste su ciò che può es- sere avvenuto fino al XVIII secolo. Bisogna giungere quindi ai primi del ’700 quando Pier’Antonio Micheli, spronato anche dal suo amico William Sherard che si era interessa- to all’opera De Plantis Libri XVI, si impegnò attivamente per la ricerca dell’erbario. Così infatti egli dice nei suoi manoscritti30: Sepolcro Alfonso Tornabuoni e datato 1563, è da considerarsi una delle più antiche col- lezioni di piante essiccate esistenti al mon- do e, in assoluto, la più antica nella quale le piante sono ordinate con criteri sistematici. Sappiamo che altri erbari furono realizzati prima di Cesalpino, ad esempio dallo stesso Luca Ghini, dai suoi allievi Petrollini, Me- rini, Cibo, Aldrovandi, probabilmente negli anni fra il 1530 e il 1560, ma nessuno con particolari intendimenti scientifici. La storia dell’erbario di Cesalpino ci viene riferita da Parlatore nel 1856 e 187426 e an- cora dallo stesso autore nelle sue Memorie27; a Parlatore infatti, come si vedrà più avanti, va il merito di averlo recuperato per il Museo di Storia Naturale nel 1844. L’erbario porta all’inizio una lunga let- tera autografa (in italiano), indirizzata da Cesalpino al vescovo di Borgo San Sepolcro Alfonso Tornabuoni (Fig. 5); in questa let- tera egli riassume le ragioni della prepara- zione di questa collezione di piante secche e dimostra che l’erbario era già preparato a quell’epoca, il che fa pensare che l’abbia realizzato durante gli anni di insegnamento a Pisa, cioè fra il 1555 e il 1563. 26 F. Parlatore, Elogio di Filippo Barker Webb, Le Monnier, Firenze 1856; Ph. Parlatore, Les collections botaniques du Musée Royal de Physique et d’Histoire Naturelle de Florence au printemps de MDCCCLXXIV, Imp. Succ. Le Monnier, Florence 1874. 29 G. Targioni Tozzetti, Notizie della vita e delle opere di Pier’Antonio Micheli botanico fiorentino, di Giovanni Targioni Tozzetti pubblicate a cura di Adolfo Targioni Tozzetti, Le Monnier, Firenze 1858, p. 115. 30 Micheli, Descriptio et Illustratio, cit., c.238r-239r.; Pichi Sermolli, Contributo alla storia della Botanica 28 G. Targioni Tozzetti, Prodromo della corografia e della topografia fisica della Toscana, Stamperia Imperiale, Firenze 1754, pp. 94, 106. 26 F. Parlatore, Elogio di Filippo Barker Webb, Le Monnier, Firenze 1856; Ph. Parlatore, Les collections botaniques du Musée Royal de Physique et d’Histoire Naturelle de Florence au printemps de MDCCCLXXIV, Imp. Succ. Le Monnier, Florence 1874. 27 F. Parlatore, Mie memorie, (a cura di A. Visconti), Sellerio, Palermo 1992, p. 111. 28 G. Targioni Tozzetti, Prodromo della corografia e della topografia fisica della Toscana, Stamperia Imperiale, Firenze 1754, pp. 94, 106. 29 G. Targioni Tozzetti, Notizie della vita e delle opere di Pier’Antonio Micheli botanico fiorentino, di Giovanni Targioni Tozzetti pubblicate a cura di Adolfo Targioni Tozzetti, Le Monnier, Firenze 1858, p. 115. 30 Micheli, Descriptio et Illustratio, cit., c.238r-239r.; Pichi Sermolli, Contributo alla storia della Botanica in Toscana, cit., p. 106. 30 Micheli, Descriptio et Illustratio, cit., c.238r-239r.; Pichi Sermolli, Contributo alla storia della Botanica in Toscana, cit., p. 106. 27 F. Parlatore, Mie memorie, (a cura di A. Visconti), Sellerio, Palermo 1992 p 111 L’erbario di Cesalpino: vicende storiche Dopo la morte di Micheli, an- che Giovanni Targioni Tozzetti e il suo figlio Ottaviano ebbero occasione di osservare e studiare l’erbario (il primo intorno al 1737-38 ed il secondo nel 1796), che nel frattempo era passato nelle mani della famiglia fiorentina Nencini, eredi dei Pandolfini. È indicativa a questo proposito una lettera del botanico bas- sanese Giambattista Brocchi (scritta nel marzo 1818 all’amico Giuseppe Moretti, professore di botanica e agraria a Pavia32), nella quale, in occasione di una sua visita a Firenze all’amico Ottaviano Targioni Tozzetti, egli afferma: Presso la Biblioteca Palatina l’osservò an- che Ottaviano Targioni Tozzetti, il quale nel 1822 ebbe occasione di esaminarlo e stu- diarlo, provvedendo ad aggiungere note ed appunti a quelli già formulati da Micheli nel suo manoscritto34. Per il destino dell’erbario fu determinante la venuta a Firenze nel 1842 del grande bota- nico siciliano Filippo Parlatore il quale già nel 1843 sollecitò il granduca Leopoldo II a trasfe- rire l’erbario al Museo di Storia Naturale, nel quale Parlatore aveva creato appunto nel 1842 l’Erbario Centrale Italiano, per svilupparne la parte botanica35. Parlatore infatti riteneva che il Museo di Storia Naturale fosse per l’erbario una collocazione più logica piuttosto che la Bi- blioteca Palatina, trattandosi di una collezione scientifica e non di un’opera libraria. Egli così ci riferisce nelle sue Memorie36: «Chiesi ed ottenni dal granduca Leopoldo di conservare nel gabinetto botanico l’erbario che Andrea Cesalpino fece e donò a Monsignor Tornabuo- ni e che, passato di mano in mano, era venuto finalmente nella Biblioteca Palatina, dove era tenuto più come cosa curiosa che scientifica e andava sensibilmente a deperire». Il trasferi- mento dell’erbario al Museo di Storia Naturale fu effettuato nel gennaio 184437. Poiché l’er- bario era in non buone condizioni, Parlatore lo fece disinfettare accuratamente e ne fece interfogliare i fogli; inoltre, poiché in un tomo unico come era rimasto fino allora era molto voluminoso e difficilmente consultabile (con pregiudizio per la conservazione dei campio- ni), lo fece suddividere in 3 volumi (Fig. 6) e rilegare in elegante marocchino rosso, come è stato conservato fino a poco tempo fa. L’erbario di Cesalpino: vicende storiche Questo er- bario era stato preparato in un unico volume (come è rimasto fino al 1844) ed era stato appositamente commissionato a Cesalpino dal vescovo Tornabuoni, come risulta da una frase dello stesso Cesalpino nella lettera cita- L’anno 1714, mentre il Signr: Guglielmo Sherard di- morava in Smirne ebbi l’onore di alcune sue lettere toccanti varie Piante dell’Opera De Plantis d’Andrea Cesalpino stampata in Firenze, e presso Giorgio Ma- rescotto l’anno 1583; e con le quali sue erudite lette- re non cessava di dare a Noi impeto di ricercare i due Orti secchi fatti dal medesimo Cesalpino, rammen- tati da esso nella Prefazione dedicatoria di detta Sua Opera. L’anno 1717, nel mese di maggio, furono mag- giori gl’impulsi, mentre nel ritorno che faceva detto Sigr: G. Sherard dalle Smirne in Inghilterra, passò per Firenze, e di gran lunga accalorò la cosa, e di tal 10 Guido Moggi La biblioteca privata di S.A.I. il Serenissimo Ferdi- nando III, Granduca di Toscana, tra i doviziosi ac- quisti, che in ogni maniera di scienze e di lettere va continuamente facendo, annovera il prezioso erbario, che Andrea Cesalpino raccolse, e donò a Monsigno- re Alfonso de’Tornabuoni. Questo erbario, mancati i Tornabuoni, passò nella casa Pandolfini, indi fu ereditato da quella de’Nencini, da cui la biblioteca Granducale ne fece l’acquisto. sorte, che subito dopo la sua partenza da questa città, mi diedi a tal ricerca; e parlandone, e riparlandone con vari studiosi, finalmente mi fu data notizia, che nella celebre Libreria dell’Ill.mo e Claris. mo Sig: Se- natore Pandolfo Pandolfini si ritrova un certo libro di Piante Secche che veniva giudicato di Andrea Cesal- pino. Che però partitomi a visitarlo, non solo rimasi assicurato dall’Ill.mo Sigr: Senatore che il libro era del d. Autore perchè la sua casa lo aveva ricevuto dall’eredità Tornaboni, ma che il medesimo libro, ce ne assicurava per eservi una lunga ed erudita lettera del medesimo Cesalpino… E più avanti afferma: «Passando io non ha guari per Firenze, e valendomi della clemen- za, colla quale Sua Altezza Imperiale mi ha permesso di visitare la sua biblioteca, volli esaminare l’erbario…». Appare chiaro quindi come l’erbario sia passato per eredità dai Tornabuoni a Pandolfo Pandolfini, nella cui biblioteca lo vide Micheli nel 1717 e potè quindi esaminarlo e studiar- lo con cura, come riferisce egli stesso nel suo manoscritto31. 33 A. Bertoloni, Memoria del Prof. Antonio Bertoloni sopra l’erbario ed una lettera del Cesalpino, «Opuscoli scientifici», 3, Bologna 1819, pp. 271-275. 34 Micheli, Descriptio et Illustratio, cit. 31 Micheli, Descriptio et Illustratio, cit. 32 G.B. Brocchi, Lettera inedita di Andrea Cesalpino, e notizie intorno al suo erbario che si conserva in Firenze in casa Nencini, con ragguaglio di alcune Opere inedite del Micheli e del Targioni, e di un Codice miniato di storia naturale che è nella Galleria di Firenze, «Bibliot. Ital.», 10, (Firenze) 1818, pp. 203-215. 35 C. Nepi, La “slegatura” dell’erbario di A. Cesalpino (1525- 1603), «Museol. sci.», n.s. 1, 2007, pp. 50-54. p , , p 38 Nepi, La “slegatura” dell’erbario di A. Cesalpino, cit., p. 53. 36 Parlatore, Mie memorie, cit., p. 111. p 37 Nepi, La “slegatura” dell’erbario di A. Cesalpino, cit., p. 52. 31 Micheli, Descriptio et Illustratio, cit. 35 C. Nepi, La “slegatura” 35 C. Nepi, La “slegatura” l’erbario di A Cesalpino (1525 L’erbario di Cesalpino: vicende storiche Nel 2003, prendendo lo spunto da nuove ricerche Venni da quest’opera [si riferisce al manoscritto di Giovanni Targioni Tozzetti «Dei progressi delle Scienze fisiche in Toscana durante il regno del gran Duca Cosimo I», allora nelle mani di Ottaviano e oggi conservato alla Biblioteca Nazionale Centrale di Firenze] in lume che l’erbario del Cesalpino era a Firenze in casa Pandolfini al tempo del Targioni seniore. Non indugiai a farne inchiesta in compagnia del prof. Ottaviano, che lo aveva esso stesso veduto in sua gioventù; ma siccome la casa Pandolfini è ora spenta, fui così fortunato di rinvenire questo prezioso codice presso gli eredi Nencini. È questo un erbario che aveva il Cesalpino allestito per ordine di Monsig. Alfonso de’Tornabuoni, a cui fu regalato. Il granduca Ferdinando III di Asburgo- Lorena, che era appassionato di scienza, poco dopo il suo arrivo a Firenze (1815) si interessò a questo importante erbario e fra il 1818 e il 1819 lo fece acquistare per la Bi- blioteca granducale in Palazzo Pitti, dove lo vide nel 1819 Antonio Bertoloni durante una sua visita a Firenze, come ci riferisce in una nota di quell’anno33: 11 l’ErBArIo dI ANdrEA CESAlPINo Fig. 7 Fig. 8 Fig. 7 Fig. 7 Fig. 6 Fig. 6 Fig. 7 Fig. 8 Fig. 8 sull’erbario in occasione del 4° centenario della morte di Cesalpino, Chiara Nepi, a quel tempo responsabile delle collezioni botaniche del Museo di Storia Naturale, prese in esa- me la possibilità di una nuova sistemazione dell’erbario, anche in vista di una eventuale riproduzione fotografi ca digitale di tutti i cam- pioni. In effetti anche nella forma rilegata vo- luta da Parlatore, cioè nei tre volumi di cui si è parlato, l’erbario era diffi cilmente consultabi- le: l’esame dei vari fogli obbligava a voltare le pagine come in un libro, con grave pregiudizio per la conservazione dei campioni sopra incol- lati. Chi lavora negli erbari sa bene che i fogli devono essere conservati sempre separati e la consultazione deve avvenire per traslazione orizzontale dei fogli stessi per non danneggia- re i campioni vegetali (Fig. 7). Fig. 6 l’erbario Cesalpino, rilegato in tre volumi, come è rimasto dal 1844 al 2006 (foto di Egildo luccioli). Fig. 7 I volumi dell’erbario nella versione rilegata come sono rimasti fi no al 2006. Si noti la diffi coltà dell’apertura delle pagine per una loro appropriata consultazione (foto di Egildo luccioli). Fig. L’erbario di Cesalpino: vicende storiche 8 Un momento del processo di ‘slegatura’ dell’erbario (foto di Egildo luccioli). Fig. 9 l’erbario nell’allestimento attuale, collocato in tre scatole (foto di Egildo luccioli). Fig. 9 Dopo accurate indagini sull’opportunità di eseguire un’operazione del genere, confortata dal parere di esperti del Laboratorio di Re- stauro della Biblioteca Nazionale Centrale di Firenze e del Settore Musei e Biblioteche del- la Regione Toscana, sotto la guida di Chiara Nepi è avvenuta la ‘slegatura’ dell’erbario: i tre volumi sono stati smontati (Fig. 8), i fogli d’er- bario sono stati interfogliati con carta adatta e tutti i fogli sciolti sono stati poi disposti, nell’ordine originale, in tre scatole di cartone costruite all’uopo, che richiamano quindi la suddivisione in tre parti operata da Parlatore38 (Fig. 9). In questa collocazione è sistemato Fig. 6 l’erbario Cesalpino, rilegato in tre volumi, come è rimasto dal 1844 al 2006 (foto di Egildo luccioli). Fig. 7 I volumi dell’erbario nella versione rilegata come sono rimasti fi no al 2006. Si noti la diffi coltà dell’apertura delle pagine per una loro appropriata consultazione (foto di Egildo luccioli). Fig. 8 Un momento del processo di ‘slegatura’ dell’erbario (foto di Egildo luccioli). Fig. 9 l’erbario nell’allestimento attuale, collocato in tre scatole (foto di Egildo luccioli). 12 Guido Moggi Fig. 10 oggi l’erbario in un armadio della sezione Bo- tanica del Museo di Storia Naturale dell’Uni- versità di Firenze, ed è con questa disposizio- ne che è oggi consultabile, senza il pericolo di danneggiamenti per la piegatura dei fogli. Tutta l’operazione, insieme con la riproduzio- ne fotografica digitale di tutti i fogli, è stata realizzata tramite l’appoggio finanziario della Provincia di Arezzo, che nel 2005 ha anche organizzato una mostra su questo tema39. Fig. 11 La prima pagina della lettera di Cesalpino al vescovo Tornabuoni che si trova collocata all’inizio dell’erbario. Fig. 10 Alcuni fogli dell’erbario nell’allestimento attuale (foto di Egildo Luccioli). Fig. 11 La prima pagina della lettera di Cesalpino al vescovo Tornabuoni che si tro L’erbario di Cesalpino: descrizione L’erbario, oggi sistemato come si è detto a fo- gli sciolti, è collocato in tre scatole contenenti rispettivamente i fogli 1-90 (scat. 1), 91-180 (scat. 2) e 181-266 (scat. 3); ogni foglio misu- ra cm 30 x 45 (Fig. 10). L’erbario si apre con dieci carte non nu- merate: le prime due contengono la lettera con cui Cesalpino dedica l’erbario al vescovo Alfonso Tornabuoni (Figg. 11 e 12); seguono quindi 8 carte, scritte sia sul recto che sul verso, comprendenti gli indici delle specie citate (il primo per i nomi in greco (Fig. 13), il secondo per quelli in latino e in volgare), secondo i nomi ed i numeri che sono riportati nel testo dell’erbario. Dopo queste carte iniziali segue il vero e proprio erbario, comprendente 768 campioni di piante incollati su 266 carte. È interessante un esame della lettera di apertura dell’erbario. Questa, come si è det- to, è dedicata al vescovo Alfonso Tornabuoni, personaggio di nobile famiglia fiorentina, no- minato nel 1546 dal papa Paolo III vescovo di Borgo San Sepolcro (l’attuale Sansepolcro in Val Tiberina). Uomo erudito, amico di molti scienziati del suo tempo ed anche di Cesal- pino, fu esperto di botanica ed appassionato sperimentatore. Fu il primo ad introdurre in Toscana il tabacco, che qui fu appunto chia- mato «Erba Tornabuona». In questa lettera Cesalpino spiega quale deve essere lo scopo di un erbario come stru- mento di confronto e di identificazione delle piante e chiarisce perchè ha ritenuto neces- sario raggruppare i campioni secondo deter- minati criteri. A quanto ci riferisce Brocchi, che ha consultato l’erbario di Cesalpino nel 13 L’erbario di Andrea Cesalpino Fig. 13 Parte della pagina contenente i nomi greci delle piante col numero di richiamo alla carta relativa. 181840, di questa lettera esisteva anche una copia in latino di cui però si sono perse le tracce nel XVIII secolo. Ne esiste però una trascrizione effettuata da Micheli quando potè consultare l’erbario (cfr. più avanti), tra- scrizione che è stata riportata nei manoscritti di Micheli41 di cui ci parla appunto Brocchi. che «l’ordine adunque di Dioscoride serve solamente per quelle piante, de quali si san- no le virtu: ma per havere una general’ co- gnitione di tutte, non è suffitiente, percioche è impossibile haver’havuto esperientia di tut- te quelle, che ci si rappresentano di nanzi». Fig. 13 Parte della pagina contenente i nomi greci delle piante col numero di richiamo alla carta relativa. Fig. 12 La parte finale della lettera con la firma autografa di Cesalpino e la data 14 settembre 1563. 39 E. Gusmeroli, A. Bigazzi (a cura di), Da Andrea Cesalpino ai nostri giorni. Erbari aretini in mostra, Catalogo della mostra, Arezzo, 4 marzo-27 maggio 2005, 2005; E. Gusmeroli, L. Lastrucci (a cura di), Atti del Convegno “Evoluzione delle conoscenze botaniche e problematiche della conservazione in provincia di Arezzo da Andrea Cesalpino ad oggi”, Arezzo, 4 marzo 2005, Provincia di Arezzo- Università degli Studi di Firenze, Arti Grafiche Cianferoni, Stia (Arezzo) 2006. 40 Brocchi, Lettera inedita di Andrea Cesalpino cit 41 Micheli, Descriptio et Illustratio, cit. 41 Micheli, Descriptio et Illustratio, cit 40 Brocchi, Lettera inedita di Andrea Cesalpino, cit. Fig. 15 Un altro passo della medesima lettera in cui Cesalpino dichiara che si ripromette in seguito di esporre i criteri sistematici da lui seguiti («[…] Ma perché quivi si ricercheria una lunga dichiaratione, per sapere quali & quanti siano questi generi, & come siano multiplicate le spetie di ciaschuna sorte, non sendo hora commodo, mi riservo in altro tempo a farlo …»), cosa che farà appunto nel libro De Plantis Libri XVI del 1583. Seguono alcuni accenni sulla classificazione da lui adottata. L’erbario di Cesalpino: descrizione q pp Successivamente, nell’illustrare i partico- lari dell’erbario, esprime chiaramente i criteri con cui l’ha realizzato: «… essendomi messo innanzi tutti e semplici, quali infino a qui mi sono venuti alle mani, gli ho distribuiti per questa prima volta grossamente, facendone le schiatte separate l’una dall’altra secondo il mio primo proponimento…» (Fig. 14). Più avanti espone sommariamente quali sono secondo lui i caratteri da prendere in consi- derazione per differenziare le piante: «Gl’arti- fitii mirabili & varii instrumenti appariscono in quella parte che serve alla generatione; … con tanta varietà, che pare non si trovi fine Nella lettera inoltre Cesalpino illustra al vescovo le basi di una moderna conoscenza delle piante, in contrasto con quanto era noto ai tempi di Teofrasto e Dioscoride, e introdu- ce i primi concetti di sistematica e di clas- sificazione. Rilevando come nell’antichità si tenesse conto prevalentemente del valore terapeutico delle piante, piuttosto che delle loro caratteristiche intrinseche, cita come esempio Dioscoride, il quale «ridusse insie- me quelle che hanno simiglanza nelle virtu, & le separò da quelle che in ciò sono disso- miglanti». E più avanti afferma chiaramente 14 Guido Moggi Guido Moggi Fig. 14 Un passo della lettera di Cesalpino al vescovo Tornabuoni in cui l’autore afferma di avere disposto i campioni secondo un ordine prestabilito da lui previsto: «[…] Però essendomi messo innanzi tutti e semplici, quali infino a qui mi sono venuti alle mani, gli ho distribuiti per questa prima volta grossamente, facendone le schiatte separate l’una dall’altra secondo il mio primo proponimento: & desiderando V.S. Rma che io gli facesi una ragunata de semplici ataccati sopra e fogli per riconoscerli, quelli dei quali ho possuto haverne il saggio, ho atacchati in questo libro secondo quell’ordine» . d’intorno alla moltitudine delle spetie. Perchè alcune mostrano fuori il lor seme quasi ignu- do… altre lo tengono racchiuso in varie sorti d’involti & di vasi, chi piu semplici, chi piu composti, chi soli, …». E ancora: «Adunque da e modi varii del produrre e semi, o quel- lo che ha proportione con e semi genitali, & dalla simiglanza di quelli ho rintracciato e generi & le spetie delle Piante…» Ecco perché questo erbario assume gran- dissima importanza nella storia della bota- nica: esso è un primo esempio di raccolta ‘sistematica’, cioè realizzata secondo precisi criteri di classificazione. Numerosi studiosi cercarono di esaminare questo erbario e di indagare sulle piante ivi conservate. 42 Targioni Tozzetti, Prodromo della corografia e della topografia fisica della Toscana, cit., p. 94. 43 Targioni Tozzetti, Notizie della vita e delle opere di Pier’Antonio Micheli, cit.; Micheli, Descriptio et Illustratio, cit. 44 S. Ragazzini, I manoscritti di Pier Antonio Micheli conservati nella Biblioteca Botanica dell’Università di Firenze, Giunta Regionale Toscana, Ed. Bibliografica, Firenze 1993. 44 S. Ragazzini, I manoscritti di Pier Antonio Micheli conservati nella Biblioteca Botanica dell’Università di Firenze, Giunta Regionale Toscana, Ed. Bibliografica, Firenze 1993. 42 Targioni Tozzetti, Prodromo della corografia e della topografia fisica della Toscana, cit., p. 94. 43 Targioni Tozzetti, Notizie della vita e delle opere di Pier’Antonio Micheli, cit.; Micheli, Descriptio et Illustratio, cit. L’erbario di Cesalpino: descrizione Come abbiamo già visto, secondo Giovanni Targioni Tozzetti alla fine del XVI secolo Stefano Rosselli, speziale dei Medi- ci, avrebbe redatto un catalogo delle piante dell’erbario42, ma è un’ipotesi che non trova ulteriori conferme. Infine conclude la lettera illustrando come le piante sono ordinate nell’erbario: «Ho ra- gunato gl’Alberi & Arbusti tutti insieme,… Dipoi seguono quelle piante, che producono il seme nudo senza alcuno involto. Appres- so vengono quelle, che l’hanno racchiuso nei vasi,… In ultimo ho messe quelle che non fanno seme qual’ si conosca». Il primo che, dopo averlo ritrovato intor- no al 1717, ne fece uno studio critico appro- fondito fu Pier’Antonio Micheli, ma la sua ricerca rimase inedita ed è tuttora presen- te nei manoscritti micheliani conservati a Firenze43. Questo manoscritto è di grande importanza per comprendere la storia ed il significato dell’erbario (Fig. 16). Ragazzini ha condotto un’accurata indagine su di esso illustrandone le varie parti (numerate da I a XIII) e le varie calligrafie autografe di tutti coloro che aggiunsero notizie con annota- zioni e commenti44. Il manoscritto consta di 276 carte e la parte più voluminosa è il catalogo delle piante dell’erbario (Parte V: Catalogus Plantarum Horti Sicci Andreae Cesalpini), che comprende le carte da c.23r L’ordinamento delle piante nell’erbario e la disposizione nelle singole pagine rispec- chiano già le idee che saranno poi sviluppate da Cesalpino nel libro De Plantis Libri XVI del 1583. E che nel 1563 Cesalpino avesse idea di completare e perfezionare i suoi con- cetti sistematici appare chiaro anche da un passo della lettera in cui dice «Ma perche quivi si ricercheria una lunga dichiaratione, per sapere quali & quanti siano questi ge- neri, & come siano multiplicate le spetie di ciaschuna sorte, non sendo hora commodo, mi riservo in altro tempo a farlo» (Fig. 15). 15 l’ErBArIo dI ANdrEA CESAlPINo ig. 17 Fig. 16 Fig. 16 Ottaviano Targioni Tozzetti) ed aggiunse un elevato numero di annotazioni e commenti. a c.215r (Fig. 17). Interessanti sono comun- que anche tutte le altre parti dalle quali si ricavano notizie sulla storia dell’erbario (Parti I, II, III, VI, VIII) e commenti su- gli elenchi di piante (Parti IV, VI, IX-XIII). Dall’analisi effettuata da Ragazzini si evin- ce che, oltre a Giovanni Targioni Tozzetti, esaminarono l’erbario ed aggiunsero note di proprio pugno anche Ottaviano Targioni Tozzetti, Antonio Targioni Tozzetti ed infi ne Teodoro Caruel. Fig. 16 Frontespizio del manoscritto Ms.9 di Pier’Antonio Micheli (1679-1737), da lui interamente dedicato alla illustrazione dell’erbario di Cesalpino. (Biblioteca di Scienze- Sezione Botanica dell’Università di Firenze, foto di Egildo luccioli). Fig. 17 Manoscritto Ms.9 di P.A. Micheli: la pagina iniziale del catalogo dell’erbario Cesalpino (c.23r). (Biblioteca di Scienze- Sezione Botanica dell’Università di Firenze, foto di Egildo luccioli). 45 ragazzini, I manoscritti di Pier Antonio Micheli, cit., p. 19. 46 T. Caruel, Illustratio in hortum siccum Andreae Caesalpini, le Monnier, Florentiae 1858. Fig. 17 Manoscritto Ms.9 di P.A. Micheli: la pagina iniziale del catalogo dell’erbario Cesalpino (c.23r). (Biblioteca di Scienze- Sezione Botanica dell’Università di Firenze, foto di Egildo luccioli). L’erbario di Cesalpino: descrizione Dopo Giovanni, anche Ottaviano Targioni Tozzetti ebbe occasione di contultare l’erba- rio, dapprima nel 1796 come risulta da una sua postilla autografa visibile sul frontespi- zio del citato Ms.9. Quindi lo vide ancora nel 1818, quando era ancora nelle mani della famiglia Nencini, in occasione della visita di Brocchi a Firenze, ed infi ne ne fece un esame accurato nel 1822 (dopo che era stato acquisito dal granduca Ferdinando III per la Biblioteca Palatina), aggiungendo commen- ti e note, forse nella speranza di pubblicare uno studio analitico45 (Fig. 18). Che Otta- viano Targioni Tozzetti avesse intenzione di pubblicare i suoi risultati sullo studio dell’er- bario di Cesalpino ci viene confermato dal fi glio Antonio, il quale nel Ms.9 di Micheli riferisce che Ottaviano aveva già redatto la dedica del suo lavoro al granduca, dedica che è riportata alla c.4r del Ms.9. Questo suo desiderio rimase però incompiuto. Micheli dopo il ritrovamento dell’erbario (e quindi fra il 1717 e il 1737, anno della sua morte) eseguì uno studio accuratissimo cer- cando di identifi care tutti i campioni ed at- tribuendo loro la nomenclatura di C. Bauhin (Pinax Theatri Botanici, 1623), di P. Bocco- ne (Museo di Piante Rare, 1697), di J. Ray (Historia plantarum, 1686-1704) ed altri bo- tanici secenteschi, ma in particolare quella di J. P. de Tournefort (Institutiones Rei Her- bariae, 1700). In ciò fu probabilmente aiuta- to dal suo allievo Giovanni Targioni Tozzetti che lo esaminò negli anni 1737-38 (come ap- pare dal frontespizio del Ms.9, realizzato da Per un’indagine approfondita e completa bi- sognerà attendere al 185846, quando Teodoro 16 GUIdo MoGGI Caruel pubblicherà la sua opera Illustratio in Hortum Siccum Andreae Caesalpini, che vie- ne ora riproposta in questo volume in ristam- pa anastatica, insieme con la traduzione della sua prefazione latina De horto sicco Andreae Caesalpini (p. vii-xii) e con la trascrizione del- la lettera al vescovo Alfonso Tornabuoni. Fig. 18 Caruel, utilizzando il manoscritto miche- liano (Fig. Fig. 21 Nella c.11 dell’erbario sono rappresentati alcuni arbusti: il lillatro, Phyllirea media (n. 33), il ligustro, Ligustrum vulgare (n. 34), il lentisco, Pistacia lentiscus (n. 35) e l’alaterno, Rhamnus alaternus (n. 36). Da notare che probabilmente i nomi per i n. 33 e 36 sono stati invertiti. L’erbario di Cesalpino: descrizione Già ai primi del ‘700, dopo il “ritrovamento” da parte del Micheli, le condizioni dell’erbario erano precarie, tanto che Giovanni Targioni Tozzetti nel suo manoscritto Le Selve affer- ma: «Esso Orto secco del Cesalpino, fatto per Monsigr Tornabuoni, è un grosso volu- me in foglio di carta Reale, coperto di Car- tapecora, ed è di carte 266, alle quali sono incollate 769 Scheletri di Piante, molti dei quali sono stati rosi dalle Tarme partico- larmente nei fiori e negli embrioni di frut- ti, ma non però tanto, che non si possano bastantemente distinguere, a riserva di po- chissimi, dei quali non vi resta quasi vesti- gio»47. Anche Brocchi, che vide l’erbario nel 1818, affermava: «… un grosso volume in foglio di carte 266, le quali comprendono 768 piante attaccate con colla, alcune delle quali sono malconce, ma tutte nulladimeno abbastanza riconoscibili». Secondo Caruel, ancora nel 1858 le condizioni dei campioni erano talvolta precarie («… Tunc plantae, jam aliquid detrimenti a tarmetibus passae …»), ma erano comunque identificabili con sicurezza («Et quamvis iis, de quibus supra memini detrimentis affectum sit herbarium, plantae tali sunt conditione ut plerumque certissime agnoscere possis»). Fig. 18 Manoscritto Ms.9 di P.A. Micheli (c.175r): pagina del catalogo dell’erbario Cesalpino relativa alla c.224 dell’erbario (contenente un Allium e quattro Orchidee) con le annotazioni di Ottaviano Targioni Tozzetti, il quale ha aggiunto, nella col. sinistra, l’identificazione secondo la nomenclatura linneana. (Biblioteca di Scienze-Sezione Botanica dell’Università di Firenze, foto di Egildo Luccioli). Fig. 19 Altre pagine del catalogo dell’erbario Cesalpino nel Ms.9. (c.67v e c.68r) dedicate ad alcune Composite; sono presenti anche due foglietti volanti con annotazioni manoscritte di T. Caruel. (Biblioteca di Scienze-Sezione Botanica dell’Università di Firenze, foto di Egildo Luccioli). Fig. 20 La c.209 dell’erbario, che una volta conteneva tre campioni di Euforbie, dei quali uno solo (n. 579, Euphorbia pithyusa) è rimasto, sia pure in pessime condizioni. Gli altri due, pur essendo andati distrutti, hanno lasciato impronte molto evidenti per cui Caruel, anche con l’aiuto dei nomi, ha potuto identificarli come E. characias (il n. 578) e E. cyparissias (il n. 580). Oggi alcuni campioni sono completa- mente distrutti, ma la maggior parte sono conservati quasi per intero (sia che si tratti di campioni costituiti da piante complete oppure da porzioni, come rametti, foglie o infiorescenze). Per questo motivo sono an- cora perfettamente identificabili, anche se spesso appaiono anneriti dal tempo. Fig. 20 La c.209 dell’erbario, che una volta conteneva tre campioni di Euforbie, dei quali uno solo (n. 579, Euphorbia pithyusa) è rimasto, sia pure in pessime condizioni. Gli altri due, pur essendo andati distrutti, hanno lasciato impronte molto evidenti per cui Caruel, anche con l’aiuto dei nomi, ha potuto identificarli come E. characias (il n. 578) e E. cyparissias (il n. 580). 47 Targioni Tozzetti, Selva di notizie, cit., Vol. VI, p. 45; Pichi Sermolli, Contributo alla storia della Botanica in Toscana, cit., p. 22. L’erbario di Cesalpino: descrizione 19) con le note di Giovanni e di Ot- taviano Targioni Tozzetti, pubblicò il catalogo completo dei campioni conservati nell’erbario, riportando per ogni esemplare il numero d’or- dine dato da Cesalpino e la pagina dell’erba- rio su cui il campione è posto, la trascrizione dei nomi apposti da Cesalpino (greco, latino e/o volgare), il ‘libro’ e il capitolo dell’opera De Plantis Libri XVI dove la pianta è men- zionata, l’indicazione delle parti della pianta con cui la specie è rappresentata nell’erbario e spesso anche lo stato del campione; inoltre viene sempre riportato il nome scientifi co se- condo la nomenclatura linneana. Ma veniamo ora alla descrizione del- l’erbario. Questo si presenta come un erbario mo- derno, cioè organizzato più o meno con gli Fig. 18 Fig. 19 Fig. 19 Fig. 19 Fig. 19 17 17 L’erbario di Andrea Cesalpino Fig. 20 Fig. 21 stessi criteri che si usano ancora oggi. Le condizioni di conservazione dei campioni sono in generale discrete; purtroppo in al- cuni casi gli esemplari sono molto danneg- giati o addirittura scomparsi (come ad es. in molte Ombrellifere e nelle Euforbiaceae – da c.209 a c.214) (Fig. 20), ma ciò è com- prensibile viste le lunghe vicende ed i tra- sferimenti subiti dall’erbario e le numerose persone che vi hanno messo le mani. Già ai primi del ‘700, dopo il “ritrovamento” da parte del Micheli, le condizioni dell’erbario erano precarie, tanto che Giovanni Targioni Tozzetti nel suo manoscritto Le Selve affer- ma: «Esso Orto secco del Cesalpino, fatto per Monsigr Tornabuoni, è un grosso volu- me in foglio di carta Reale, coperto di Car- tapecora, ed è di carte 266, alle quali sono incollate 769 Scheletri di Piante, molti dei quali sono stati rosi dalle Tarme partico- larmente nei fiori e negli embrioni di frut- ti, ma non però tanto, che non si possano bastantemente distinguere, a riserva di po- chissimi, dei quali non vi resta quasi vesti- gio»47. Anche Brocchi, che vide l’erbario nel stessi criteri che si usano ancora oggi. Le condizioni di conservazione dei campioni sono in generale discrete; purtroppo in al- cuni casi gli esemplari sono molto danneg- giati o addirittura scomparsi (come ad es. in molte Ombrellifere e nelle Euforbiaceae – da c.209 a c.214) (Fig. 20), ma ciò è com- prensibile viste le lunghe vicende ed i tra- sferimenti subiti dall’erbario e le numerose persone che vi hanno messo le mani. Fig. 19 Altre pagine del catalogo dell’erbario Cesalpino nel Ms.9. (c.67v e c.68r) dedicate ad alcune Composite; sono presenti anche due foglietti volanti con annotazioni manoscritte di T. Caruel. (Biblioteca di Scienze-Sezione Botanica dell’Università di Firenze, foto di Egildo Luccioli). L’erbario di Cesalpino: descrizione (Fig. 23). E così Graminacee, Ciperacee e Giuncacee a causa della loro affinità sono raggruppate nei fogli da c.103 a c.111, mentre le Labiate si tro- vano da c.113 a c.130 (Fig. 24), con qualche intromissione di specie oggi attribuite ad altre famiglie, come il Lythrum salicaria (Lythra- ceae) alla c.118 (n. 302) o il Myriophyllum ver- ticillatum (Haloragaceae) alla c.119 (n. 306). Ancora le Solanacee sono alle c.143-148 (Fig. 25), mentre quasi tutte le Leguminose si tro- vano riunite fra la c.158 e la c.168 (Fig. 26), le Scrofulariacee fra c.169 e c.178, le Crucifere da c. 193 a c. 202, le Ranuncolacee da c. 248 a c.256, le felci (sensu lato) nelle c.263, 264, 266, e così via (Fig. 27). Che Cesalpino avesse identificato l’unità d i i i f i li è f In questo erbario possiamo trovare rias- sunte le teorie di Cesalpino sulla classifi- cazione delle piante che saranno poi da lui esposte nel suo libro. I ‘gruppi sistematici’, da lui delineati nel libro De Plantis, sono già chiaramente identificabili nelle pagine dell’erbario; ciò significa che le idee che poi Cesalpino pubblicherà nel 1583 nel suo libro erano già nella sua mente al momento della preparazione dell’erbario, come del resto lui accenna nella lettera al vescovo Tornabuoni. Interessante è anche l’esame dei nomi usati da Cesalpino per identificare le piante (Figg. 28 e 29). Come si è detto, egli usa spes- so nomi greci, ma anche nomi latini e italiani. È chiaro che la nomenclatura greca deriva di- rettamente da Teofrasto o anche da Dioscori- de, della cui opera (Περì ὕληϚ ἰατρικῆϚ o De Materia Medica) Cesalpino aveva certamente dimestichezza, avendola illustrata all’Univer- sità di Pisa nei corsi di «materia medica». Del resto il riferimento a questi autori è esplicito anche in alcuni campioni dell’erbario, come ad esempio al n. 4, Φίλυρα, al n. 17, Κέδρος, o al n. 91, Γεράνιον, dove vengono citati espres- samente «Theoph.» o «Diosc.». L’esame dettagliato dell’erbario ci confer- ma il preciso ordine seguito da Cesalpino nel sistemare le piante. Infatti ad esempio da c.1 a c.20 troviamo prima alberi e poi arbusti (fag- gio, tiglio, frassino, ontano, ginepro, abete, tas- so, alaterno, mirto, viburno, lauroceraso, ecc.) (Fig. 21), da c.21 a c.44 le attuali Ombrellife- re, da c.50 a c.54 le Borraginacee (Fig. 22), da c.69 a c.102 molte Composite, ecc. (Fig. 23). Fig. 22 La c.50 comprende tre campioni, oggi tutti attribuibili alla famiglia delle Borraginacee (n. 119, l’eliotropio, Heliotropium europaeum; n. 120, il non-ti-scordar-di-me, Myosotis scorpioides; n. 121, la lingua di cane, Cynoglossum creticum). Fig. 23 Un foglio contenente tre Composite (c.94: n. 237, Carlina corymbosa; n. 238, Scolymus hispanicus; n. 239, Centaurea solstitialis, o calcatreppola). Fig. 24 I campioni qui riuniti nella c.126 sono tutti rappresentanti della famiglia delle Labiate o Lamiacee (a sin. n. 326, Calamintha grandiflora; al centro n. 328, Nepeta cataria o erba gatta; a destra n. 327, Calamintha nepeta o nepitella;). Fig. 28 Esempio di nomi di piante usati da Cesalpino: il pungitopo (Ruscus aculeatus L., c.136, n. 370) è indicato con il nome greco, la trascrizione latina del nome greco, il nome latino e due nomi italiani. Fig. 29 Un altro esempio di nomi: l’equiseto comune (Equisetum arvense L., c.68, n. 170) porta anche qui cinque nomi. Fig. 29 Un altro esempio di nomi: l’equiseto comune (Equisetum arvense L., c.68, n. 170) porta anche qui cinque nomi. Fig. 22 La c.50 comprende tre campioni, oggi tutti attribuibili alla famiglia delle Borraginacee (n. 119, l’eliotropio, Heliotropium europaeum; n. 120, il non-ti-scordar-di-me, Myosotis scorpioides; n. 121, la lingua di cane, Cynoglossum creticum). Fig. 23 Un foglio contenente tre Composite (c.94: n. 237, Carlina corymbosa; n. 238, Scolymus hispanicus; n. 239, Centaurea solstitialis, o calcatreppola). Fig. 24 I campioni qui riuniti nella c.126 sono tutti rappresentanti della famiglia delle Labiate o Lamiacee (a sin. n. 326, Calamintha grandiflora; al centro n. 328, Nepeta cataria o erba gatta; a destra n. 327, Calamintha nepeta o nepitella;). Fig. 25 Due Solanacee: a sinistra la dulcamara, Solanum dulcamara (n. 390) e a destra Withania somnifera (n. 389) (c.144). Fig. 26 Alcune Leguminose suffruticose od arbustive della c.158, tutte spinose: a sinistra la ginestra spinosa, Genista germanica (n. 430), al centro la ginestrella, Genista tinctoria (n. 432) e a destra il ginestrone, Ulex europaeus (n. 431). Fig. 27 Tre caratteristici convolvoli (c.190): il comune convolvolo, Convolvulus cantabrica (n. 525), il convolvolo delle rocce costiere, C. cneorum (n. 526), e il convolvolo delle spiagge, Calystegia soldanella (n. 527). L’erbario di Cesalpino: descrizione Come si è già detto, l’erbario consta di 266 carte, numerate da Cesalpino stesso sul recto in alto a destra, su ciascuna delle quali sono incollati da uno a tre campioni 18 Guido Moggi per foglio (qualche volta anche 4 o 5); in to- tale l’erbario comprende 768 esemplari che corrispondono a circa 760 specie di piante, un numero veramente notevole se si pensa che le piante note a quell’epoca erano circa 1300 (almeno tante ne cita Cesalpino nel li- bro De Plantis e più o meno altrettante sono menzionate da Mattioli nelle prime edizioni dei suoi Discorsi e dei Commentarii). Tutti i campioni sono ordinati ovviamente secondo il sistema di Cesalpino, anche se l’erbario è stato preparato ben venti anni prima della pubblicazione del libro De Plantis. nell’erbario, anche da come poi tratterà le famiglie nel libro De Plantis Libri XVI. In- fatti un confronto fra l’erbario ed il libro ci mostra ad esempio come tutte le Ombrellifere presenti nell’erbario siano citate nel Liber VII dell’opera De Plantis, mentre le Boraginacee sono trattate unitariamente nel Liber XI. Così troviamo anche le Leguminose nel Liber VI, quasi tutte le Scrofulariacee nel Liber VIII e la maggior parte delle Ranuncolacee nel Liber XIV. Il confronto fra l’erbario (1563) e il libro (1583) conferma quindi l’omogeneità dei prin- cipi concettuali espressi da Cesalpino, che re- stano i medesimi a distanza di 20 anni. campioni sono ordinati ovviamente secondo il sistema di Cesalpino, anche se l’erbario è stato preparato ben venti anni prima della pubblicazione del libro De Plantis. In questo erbario possiamo trovare rias- sunte le teorie di Cesalpino sulla classifi- cazione delle piante che saranno poi da lui esposte nel suo libro. I ‘gruppi sistematici’, da lui delineati nel libro De Plantis, sono già chiaramente identificabili nelle pagine dell’erbario; ciò significa che le idee che poi Cesalpino pubblicherà nel 1583 nel suo libro erano già nella sua mente al momento della preparazione dell’erbario, come del resto lui accenna nella lettera al vescovo Tornabuoni. L’esame dettagliato dell’erbario ci confer- ma il preciso ordine seguito da Cesalpino nel sistemare le piante. Infatti ad esempio da c.1 a c.20 troviamo prima alberi e poi arbusti (fag- gio, tiglio, frassino, ontano, ginepro, abete, tas- so, alaterno, mirto, viburno, lauroceraso, ecc.) (Fig. 21), da c.21 a c.44 le attuali Ombrellife- re, da c.50 a c.54 le Borraginacee (Fig. 22), da c.69 a c.102 molte Composite, ecc. L’erbario di Cesalpino: descrizione E così Graminacee, Ciperacee e Giuncacee a causa della loro affinità sono raggruppate nei fogli da c.103 a c.111, mentre le Labiate si tro- vano da c.113 a c.130 (Fig. 24), con qualche intromissione di specie oggi attribuite ad altre famiglie, come il Lythrum salicaria (Lythra- ceae) alla c.118 (n. 302) o il Myriophyllum ver- ticillatum (Haloragaceae) alla c.119 (n. 306). Ancora le Solanacee sono alle c.143-148 (Fig. 25), mentre quasi tutte le Leguminose si tro- vano riunite fra la c.158 e la c.168 (Fig. 26), le Scrofulariacee fra c.169 e c.178, le Crucifere da c. 193 a c. 202, le Ranuncolacee da c. 248 a c.256, le felci (sensu lato) nelle c.263, 264, 266, e così via (Fig. 27). I nomi latini derivano molto probabilmen- te da Plinio (Naturalis historia): anche qui li ritroviamo su alcuni esemplari dell’erbario, come al n. 41, Laburnum, al n. 157, Lupus sa- lictarius, o al n. 312, Lamium. Nel complesso, i campioni che portano sul foglio la citazione di Teofrasto sono 29, quelli di Dioscoride 11, mentre le citazioni riferite a Plinio sono 55. Non rare sono le citazioni di nomi in vol- gare (italiano), spesso associate a nomi greci o latini, come ad esempio al n. 224, «Σόxχος, Sonchus levis, Cicerbita» (per Mycelis mura- lis [L.] Dumort., la lattuga dei boschi) o al n. 536, «Σίνηπι, Sinapis, Senapa» (per Brassi- ca nigra [L.] Koch, la senape). Questo sistema per denominare le piante, in uso ai primi del ’500, sarà poi modificato nel corso dei secoli per giungere quindi alla metà del XVIII secolo alla cosidetta «no- menclatura binomia» (cioè con due termini latini), introdotta da Linneo, e che è quella in uso ancora oggi. Che Cesalpino avesse identificato l’unità dei caratteri per ogni famiglia è conferma- to, oltre che dalla disposizione dei campioni 19 L’erbario di Andrea Cesalpino Fig 23 Fig. 23 Fig. 24 Fig. 25 Fig. 24 Fig. 25 20 Guido Moggi Fig 26 Fig. 28 Fig. 29 Fig. 29 Fig. 28 L’erbario Coltellini 1 Archivio di Stato di Firenze, Fondo Imperiale e Real Corte, Inventario del Reale Gabinetto di Fisica e Storia Naturale, 1793, voll. VII (filza 5265), VIII (filza 5266) e IX (filza 5267); ibidem, Inventario dell’I. e R. Museo di Fisica e Storia Naturale, 1820, filze 5315, 5316 e 5317; Sezione Botanica Museo di Storia Naturale, Catalogo della collezione dei prodotti vegetali, 1904, voll. 1 e 2. L’erbario Coltellini della Sezione Botanica del Museo di Storia Naturale Chiara Nepi U n piccolo ed elegante erbario, allestito nella seconda metà del XVIII secolo da certo Agostino Coltellini di Cortona è cu- stodito insieme alle altre collezioni storiche possedute dalla Sezione Botanica del Museo di Storia Naturale dell’Università di Firenze. Negli inventari del Museo, a partire da quelli più antichi1 non si è trovata traccia di questa collezione che, donata dall’autore al grandu- ca Pietro Leopoldo di Lorena, passò tra le collezioni dell’Imperiale e Regio Museo di Fisica e Storia Naturale di Firenze, fondato dal granduca stesso nel 1775. Il passaggio è testimoniato dalla presenza dell’antico tim- bro del Museo nelle prime pagine dell’erba- rio. Esso poi vi è rimasto insieme alle altre più note e prestigiose collezioni storiche di Andrea Cesalpino, Michele Merini e Pier Antonio Micheli. U co di raso verde e adornato da tralci fogliari finemente ritagliati. Tutti i fogli hanno anche una sorta di doppia cornice colorata di verde e rosso chiaro, gli stessi colori dell’ovale con il nome. Ogni foglio, infine, possiede un pic- colo nastro, sempre di seta verde, che facilita la consultazione dell’erbario. Un’altra particolarità che caratterizza questa collezione è la ‘verniciatura’ con una sostanza trasparente dei campioni, che per questo motivo hanno un aspetto particolar- mente prezioso e brillante. Forse proprio a causa di questa protezione, al contrario di ciò che si verifica normalmente nei campioni di piante essiccate, quelli dell’erbario Coltellini hanno conservato praticamente integri i loro colori originali, suggerendo una grande cura da parte del preparatore ed anche il possesso da parte sua di certe conoscenze, proprio per la migliore conservazione di strutture, tessuti e pigmenti delicati come quelli delle piante. L’erbario è costituito da 28 fogli di piccole dimensioni: la base misura cm 27 e l’altezza cm 39. La coperta è cartonata e rivestita di carta marmorizzata, con gli angoli e la costo- la rinforzati da inserti di pelle, decorati da dorature (Fig. 1). Tutti i fogli sono rilegati a libro, tranne uno che è sciolto. Su quest’ul- timo e su altri dodici fogli sono attaccati, mediante striscioline di carta incollate, tre- dici campioni vegetali, tutti accuratamente preparati e ben disposti. L’erbario Coltellini Fig. 1 Fig. 1 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Fig. 1 L’esterno dell’erbario Coltellini, con l’elegante copertina. L’erbario Coltellini della Sezione Botanica del Museo di Storia Naturale Ogni campione è identificato con il nome della specie scritto in un elegante ovale a foggia di cartellino, tenuto attaccato al foglio da un piccolo fioc- I nomi usati per indicare i campioni sono tutti nomi linneani ed anche questo costi- tuisce un particolare interessante, in quan- to l’erbario è da datarsi senza alcun dubbio, come si vedrà più avanti, ante-1790, cioè al massimo 37 anni dopo che Carlo Linneo, fondatore della moderna Sistematica, aveva stabilito la nomenclatura binomia nel suo Species Plantarum del 1753, nomenclatura che a poco a poco, a partire da questa data, aveva conquistato la preminenza nella de- nominazione delle piante, sostituendo altri 24 Chiara Nepi Fig. 2 Fig. 3 Fig. 3 Fig. 2 Fig. 3 Fig. 3 Fig. 2 borea presente nell’erbario e chiamato «falso pepe», originario del Perù, venne introdotto in Europa e, precisamente in Spagna, intor- no al 1570. Arrivò in Italia nei primi anni del secolo successivo3, ma venne coltivato in Toscana soltanto molti anni dopo, quando se ne trovano le prime tracce nel catalogo del- l’Orto Botanico di S. Maria Nuova a Firenze nel 17804. E ancora, l’Asclepias curassavica L. (Fig. 3), il cui fiore viene chiamato «blood flower» dagli anglosassoni per il colore, an- ch’essa originaria dell’America meridionale come il «falso pepe», era arrivata in Italia a metà del XVII secolo, mentre il Polygonum orientale L. proveniente, come dice il nome, da oriente, era stato introdotto già nel XVI secolo, analogamente al Dracocephalum moldavica L. (Fig. 4). sistemi fino ad allora adottati, come ad esem- pio quello di J.P. de Tournefort, sicuramente uno dei più diffusi in Europa. Infine, ultima caratteristica interessante di questo piccolo erbario sono proprio le piante prescelte. Le specie sono dodici, nonostan- te, come si è detto, i campioni siano uno di più, ma la specie Zinnia pauciflora è ripetuta inspiegabilmente due volte. La maggioranza delle piante appartiene a specie coltivate nei giardini, non di rado da non molto tempo in- trodotte in Italia e in Toscana in particolare. Si possono citare come esempi il bel Gera- nium inquinans L. (oggi Pelargonium inqui- nans [L.] Aiton), detto volgarmente «geranio chermisino» per il colore dei petali, il cui campione è l’unico posto su un foglio sciolto. La specie, proveniente dal Sud Africa e in- trodotta in Inghilterra nel 1714, viene citata per la prima volta come coltivata in Toscana nel Giardino Botanico di Pisa nel 17232. Fig. 2 Il campione di Schinus molle L. o «falso pepe». Fig. 3 Il campione di Asclepias curassavica L. 2 Ant. Targioni Tozzetti, Cenni storici sulla introduzione di varie piante nell’agricoltura ed orticultura toscana, Tip. Galileiana, Firenze 1853. 5 Vedi in questo stesso volume il contributo di B. Gialluca sulla Società di Cortona. 3 F. Maniero, Fitocronologia d’Italia, Leo S. Olschki, Firenze 2000. 4 Ant. Targioni Tozzetti, ibidem, 1853. 3 F. Maniero, Fitocronologia d’Italia, Leo S. Olschki, Firenze 2000. 2 Ant. Targioni Tozzetti, Cenni storici sulla introduzione di varie piante nell’agricoltura ed orticultura toscana, Tip. Galileiana, Firenze 1853. 3 F. Maniero, Fitocronologia d’Italia, Leo S. Olschki, Firenze 2000. 4 Ant. Targioni Tozzetti, ibidem, 1853. 5 Vedi in questo stesso volume il contributo di B. Gialluca sulla Società di Cortona. Fig. 4 Il campione di Dracocephalum moldavica L. Fig. 5 Il campione di Matricaria parthenium L. 6 Si veda, a tal proposito, ancora il saggio di Bruno Gialluca in questo volume. L’erbario Coltellini della Sezione Botanica del Museo di Storia Naturale Lo Schinus molle L. (Fig. 2), unica specie ar- È probabile che il costitutore dell’erbario le abbia scelte, magari raccogliendole nel giardino della Società Botanica Cortonese5 oppure semplicemente nel suo giardino pri- 25 L’erbario Coltellini della Sezione Botanica del Museo di Storia Naturale Fig. 5 Fig. 4 Fig. 5 Fig. 5 Fig. 4 granduca, con l’aggiunta della provenienza, la città di Cortona. Ora, proprio a Cortona la botanica aveva visto fin dalla metà del secolo XVIII un grande fervore nei suoi confronti, culminato con la fondazione, nel 1754, del- la Società Botanica Cortonese, seconda so- cietà di questo tipo in Europa dopo quella Fiorentina, fondata nel 1716 da Pier Antonio Micheli ed altri studiosi6. Di questa Società Cortonese furono soci molti uomini di scien- za del tempo, come Antonio Cocchi, Giovan- ni Lami e Saverio Manetti, solo per citarne alcuni provenienti dall’ambiente culturale fiorentino. Tra i cittadini di Cortona, poi, ne fecero parte anche alcuni appartenenti alla famiglia Coltellini, come Augusto, Tommaso e suo fratello Lodovico, il più noto di tutti, perché a lungo segretario della Società. Ma non sembra esserci traccia di questo Agosti- no, che pure si definisce, sempre nella lette- ra dedicatoria a Pietro Leopoldo, «Studiosus Chemiae Pharmaciae et Obstetriciae», quin- vato, proprio per il loro ‘sapore esotico’; al- cune, come la Zinnia pauciflora o gli stessi Geranium inquinans e Schinus molle, per il loro carattere di novità nei giardini toscani. Accanto a queste, troviamo anche specie più comuni, ma sempre coltivate, come Medica- go sativa L. (la nota «erba medica») ed altre Compositae, al pari della già rammentata Zinnia, come Matricaria parthenium L. (Fig. 5), Achillea ageratum L. e Tanacetum vulga- re L., caratterizzate dal possedere un forte odore aromatico. Ma chi era il costitutore di questo erba- rio? Gli unici indizi che abbiamo sono nel frontespizio con dedica a Pietro Leopoldo e su quell’unico foglio staccato dagli altri e su cui è posto il campione di Geranium inqui- nans (Fig. 6). Su quest’ultimo, infatti, esiste una sorta di firma: Augustinus Cultellinius Excogitator et Artifex, cioè a dire Agostino Coltellini Ideatore e Realizzatore e queste stesse parole sono ripetute nella dedica al Fig. 7 28 ChIArA NEPI per il gusto raffi nato con cui sono rilegati a libro i fogli dei campioni, sia per la cura con cui sono stati preparati questi ultimi, a te- stimonianza di una precisa conoscenza delle migliori tecniche per conservarne i pigmenti durante il processo di essiccazione, cono- scenza derivata senza dubbio dalla sua pro- fessione di chimico e farmacista (Fig. 8). ig. 8 L’erbario non riporta alcuna data, ma noi possiamo risalire indirettamente ad una sua datazione considerando il fatto che l’autore usa la nomenclatura binomia di Linneo e, soprattutto, come si è già detto, dedica que- sta piccola collezione a Pietro Leopoldo di Asburgo Lorena, che fu granduca di Toscana dal 1765 al 1790. Questi dati ci suggerisco- no che l’erbario possa essere stato allestito molto probabilmente verso la fi ne del gran- ducato di Pietro Leopoldo, quando ormai la nomenclatura linneana aveva cominciato a diffondersi ed a ‘scalzare’ i precedenti siste- mi di denominazione delle piante. Non solo, ma questa ipotesi è suffragata anche dal campione di Schinus molle, la cui coltiva- zione in Toscana è accertata uffi cialmente dal 1780, come si è già visto. Una piccolissima collezione, quindi, ma i cui campioni suggeriscono diverse con- siderazioni: innanzi tutto la preparazione accurata dal punto di vista metodologico, sia nell’allestimento che nella preoccupa- zione per la conservazione dei reperti, poi l’aggiornamento scientifi co per l’uso della ‘nuova’ nomenclatura, senza dubbio facili- tato dalla frequentazione di un ambiente, come quello cortonese di metà ’700, molto vivace dal punto di vista culturale. Infi ne, una grande attenzione per il risultato este- tico dell’allestimento, che avvicina questa sorta di divertissement del chimico-farmaci- sta Coltellini ai vasi della manifattura Gino- ri contenenti le famose piante in cera della stessa Sezione Botanica in cui l’erbario è conservato: come ispirati dal medesimo gusto settecentesco, gli eleganti cartelli- ni ovali, sia sulla carta dei fogli d’erbario che sulla porcellana dei vasi, riportano scientifi camente e in bella scrittura i nomi di quelle piante che sempre più numerose giungevano ad abbellire i giardini di tutta Europa dalle regioni più lontane. Fig. 8 di sicuramente un erudito anch’egli e, soprat- tutto, un conoscitore delle piante, aggiornato sui nuovi sistemi di nomenclatura (Fig. 7). 7 M.A. Morelli Timpanaro, A Livorno nel Settecento. Medici, mercanti, abati, stampatori: Giovanni Gentili (1700-1784) ed il suo ambiente, Belforte ed., livorno 1997. devo a Bruno Gialluca, che qui ringrazio, il suggerimento di questo lavoro per il reperimento di notizie su Agostino Coltellini. Fig. 6 Il campione di Geranium inquinans l. sul foglio non rilegato dell’erbario. Fig. 7 dedica dell’autore dell’erbario al granduca Pietro leopoldo. da notarsi in alto al centro il timbro dell’I. e r. Museo. Fig. 8 Uno dei due fogli con campioni di Zinnia paucifl ora l. Fig. 6 Il campione di Geranium inquinans l. sul foglio non rilegato dell’erbario. Fig. 6 Il campione di Geranium inquinans l. sul foglio non rilegato dell’erbario. Fig. 7 dedica dell’autore dell’erbario al granduca Pietro leopoldo. da notarsi in alto al centro il timbro dell’I. e r. Museo. Fig. 8 Uno dei due fogli con campioni di Zinnia paucifl ora l. In effetti, proprio Tommaso Coltellini, notaio e cancelliere della Curia vescovile di Cortona, considerato addirittura tra i primi ideatori di quella Società Botanica Corto- nese, dal matrimonio con Caterina Fabbri- ni aveva avuto un fi glio, di nome Agostino. Questi venne ricordato nel numero 50 del 15 dicembre 1804 della Gazzetta Toscana come «Professore di Farmacia e Chimica in Corto- na», in occasione del ricevimento di un’alta onorifi cenza per un suo lavoro «relativo alla conservazione della salute»7. Sicuramente, l’erbario che Agostino ha lasciato è davvero un piccolo gioiello, sia L’Hortus Siccus Pisanus di Castiglion Fiorentino Leonardo Magionami* Oltra questo avrebbe ancora bisogno per aiumento di questa facultà, essiccare ogni sorte di piante pellegri- ne, ponendole nelle carte straccie che non sono atte a scrivergli sopra, et si pongono in modo ch’una non tocchi l’altra interponendossi fra ciascuna pianta tre carte, acciò che l’humidità d’una non corrompa l’altra prossima. Appreso avvertendo che ogni tre giorni si mutano perche quella carta sorbisce, tutto l’humido et bagnandosi per l’humore facilmente saria causa di putredine, et questa mutation si fa cinque o sei volte, a talche uno spacio de quindeci giorni si secca. Et facendo a questo modo restono talmente essiccate, e verdi col suo colore naturale, che di poi agevolmente si possono agglutinare ne’ libri, com’ho fatto io nelle mie. Et si potria da quelle agglutinate come da veri exemplari originali, per la difficoltà di molte che non così agevolmente portare si possono et portate che siano vivere in queste nostre non è così facile. Ghini3, docente prima presso l’Ateneo di Bo- logna e dal 1543 a Pisa dove costituì a fini di studio il famoso Giardino Botanico, cui seguì due anni dopo la fondazione del «Giardino dei Semplici» di Firenze. Oltre all’Hortus vi- vus la scuola di Ghini, grazie ai suoi allievi, come l’Aldrovandi appunto e anche Andrea Cesalpino, diffuse e approfondì la conoscen- za botanica mediante lo studio dei vegetali, introducendo l’utilizzo di raccolte di piante, essiccate e descritte in speciali libri. Gli Horti sicci, infatti, consentivano di disporre di esem- plari con il fiore anche in quei periodi dell’an- no in cui le piante vive ne erano sprovviste. La preparazione di essi inoltre era per lo più direttamente operata dal botanico e proprio questa pratica dava la possibilità allo studioso di conoscere al meglio i singoli esemplari. Fig. 1 Costola del volume dell’Hortus Siccus Pisanus. C on queste parole Ulisse Aldovrandi spie- gava in una nota di un suo manoscritto1, oggi conservato a Bologna, il modo di essic- care le piante, ossia il sistema di essiccazione e preparazione di esemplari vegetali al fine dell’allestimento di un Hortus siccus a scopo scientifico e didattico. L’Hortus Siccus Pisanus L’Hortus Siccus Pisanus Fig. 1 5 Per l’illustrazione dei vegetali nelle edizioni a stampa dei secoli XV e XVI e la descrizione del programma PLANT (Plantarum Aetatis Novae Tabulae) si veda il lavoro di Alain Touwaide dal titolo L’illustrazione botanica negli erbari a stampa del XV e XVI secolo. Il programma di ricerca PLANT e il suo contributo all’analisi delle rappresentazioni di Piante in Erbe e speziali. I laboratori della salute, a cura di Margherita Breccia Fratadocchi e Simonetta Buttò, Sansepolcro 2007. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press * Colgo l’occasione per ringraziare qui due persone senza le quali il lavoro di ricostruzione storica che si presenta in queste pagine non sarebbe stata possibile: Piero Fusi, Direttore dell’Istituzione Culturale ed Educativa Castiglionese di Ca- stiglion Fiorentino per la disponi- bilità e la calorosa amicizia che mi ha sempre dimostrato e Antonella Moriani per i costanti consigli che non mi ha mai lesinato. il testo della sua ricerca: M.È. Boutroue, Pourquoi faire un herbier de plantes sèches à la renaissance?, * Colgo l’occasione per ringraziare qui due persone senza le quali il lavoro di ricostruzione storica che si presenta in queste pagine non sarebbe stata possibile: Piero Fusi, Direttore dell’Istituzione Culturale ed Educativa Castiglionese di Ca- stiglion Fiorentino per la disponi- bilità e la calorosa amicizia che mi ha sempre dimostrato e Antonella Moriani per i costanti consigli che non mi ha mai lesinato. 1 Biblioteca Universitaria di Bologna, ms.91 c. 356r. 2 Sugli sviluppi degli studi botanici in epoca medievale e rinascimentale attraverso la conoscenza approfondita delle piante e l’allestimento di codici erbari vedi K.M. Reeds, Botany in Medieval and Renaissance Universities, Garland Publishing, Inc. New York-London 1991. 3 Per ulteriori notizie su questo argomento, vedi anche il saggio di G. Moggi in questo stesso volume. 4 Per quanto riguarda l’allestimento di erbari confezionati mediante la tecnica di impressione naturale si veda l’interessante lavoro di Marie-Èlisabeth Boutroue dell’IRHT-CNRS France che qui mi permetto di ringraziare per avermi fornito in anteprima il testo della sua ricerca: M.È. Boutroue, Pourquoi faire un herbier de plantes sèches à la renaissance?, in Le thèâtre des curiosités, Paris 2008, pp. 91-107. 5 Per l’illustrazione dei vegetali nelle edizioni a stampa dei secoli XV e XVI e la descrizione del programma PLANT (Plantarum Aetatis Novae Tabulae) si veda il lavoro di Alain Touwaide dal titolo L’illustrazione botanica negli erbari a stampa del XV e XVI secolo. Il programma di ricerca PLANT e il suo contributo all’analisi delle rappresentazioni di Piante in Erbe e speziali. I laboratori della salute, a cura di Margherita Breccia Fratadocchi e Simonetta Buttò, Sansepolcro 2007. 1 Biblioteca Universitaria di in Le thèâtre des curiosités, Paris 2008, pp. 91-107. 3 Per ulteriori notizie su questo argomento, vedi anche il saggio di G. Moggi in questo stesso volume. 5 Per l’illustrazione dei vegetali nelle edizioni a stampa dei secoli Fig. 1 Costola del volume dell’Hortus Siccus Pisanus. Fig. 1 Costola del volume dell’Hortus Siccus Pisanus. 2 Sugli sviluppi degli studi botanici in epoca medievale e rinascimentale attraverso la conoscenza approfondita delle piante e l’allestimento di codici erbari vedi K.M. Reeds, Botany in Medieval and Renaissance Universities, Garland Publishing, Inc. New York-London 1991. 6 Hortus siccus pisanus. L’erbario settecentesco della biblioteca di Castiglion Fiorentino, presentazione, edizione del testo latino, traduzione in italiano di Leonardo Magionami, Montepulciano 2005. Fig. 2 Prefazione all’erbario. gli studi naturalistici sono ancor più spinti dal desiderio di una conoscenza organizzata e ap- profondita della medicina e della botanica. dio delle piante, è sicuramente il volume oggi conservato presso l’Istituzione Culturale ed Educativa Castiglionese di Castiglion Fioren- tino segnato n. 525 e conosciuto come Hor- tus Siccus Pisanus6. Tale manoscritto, infatti, conserva nelle sue carte sia la descrizione di esemplari vegetali, sia vere e proprie piante essiccate. Si tratta di un volume cartaceo di dimensioni medie (324 mm di altezza e 227 mm di base) ricoperto con una legatura in L’Hortus Siccus Pisanus di Castiglion Fiorentino Questo genere di libro si configurava come uno strumento partico- larmente efficace per la conoscenza botanica, dal momento che dava la possibilità di con- sultazione tutto l’anno di esemplari autentici, anche se nella loro forma essiccata, al con- trario dei giardini botanici, i cosiddetti Horti vivi, che all’epoca non erano in grado di forni- re una produzione perenne delle piante2. C Nel caso degli Horti sicci, il fatto di con- servare la pianta vera diventa anche un ele- mento di distinzione rispetto agli Herbaria, erbari dipinti o «codici erbari», libri in cui le piante sono solamente raffigurate con un disegno o tutt’al più impresse naturalmente come nel caso presente nel codice atlantico di Leonardo da Vinci4. Con la diffusione del- la stampa inoltre si presentò la possibilità di raffigurare i vegetali grazie ad incisioni, spes- so colorate ad acquarello5; tuttavia la pratica di allestimento di raccolte di piante vere es- siccate, perdurò per tutta l’età moderna. Il primo botanico ad allestire un Hortus vi- vus fu proprio il maestro di Aldrovandi, Luca Questa pratica trovò un momento partico- larmente fertile nel Settecento, secolo in cui 32 Leonardo Magionami Fig. 2 Prefazione all’erbario. L’Hortus Siccus Pisanus di Castiglion Fiorentino Tra i vari riferimenti e indicazioni bibliografiche, abbondantemente presenti nel testo, compaiono le Institutiones Rei Herbariae7 di Tournefort, l’Hortus Regius Parisiensis8 di Jonquet, l’Hortus Amsteloda- mensis9 di Commelin, e l’Hortus Lugduni- Batavorum10 di Boeravio. per avvalorare le proprietà terapeutiche dei singoli farmaci preparati con le piante, si ci- tano Plinio e Dioscoride ma anche Tournefort e Hoffmann che fanno da cornice ad altre in- dicazioni di carattere storico11 e si legano ad aneddoti sull’utilizzo del vegetale: come esso sia stato adoperato nella storia o addirittura come esso venga chiamato dal volgo o dagli alchimisti12 (Fig. 4). Nella seconda parte invece, che ricomin- cia con una numerazione delle carte che va da 1 a 160, si trovano le piante nella loro for- ma essiccata ossia 300 esemplari che sono corrispondenti alle schede esplicative poste nelle pagine precedenti. Le piante ad oggi sono ben conservate e ben mantengono il colore, in alcuni casi si possono vedere ancora gli esemplari fioriti. È qui che si conserva il vero e proprio Hortus siccus in cui le piante sono disposte general- mente in gruppi di tre o cinque, ma a volte anche di sei o sette per pagina, fermate me- diante l’inserimento di parti di esse, quali lo stelo o la foglia, in una sorta di asola creata appositamente sulla pagina mediante due ta- gli paralleli della carta. Tale sistema risulta un ottimo modo di fermare le piante senza comprometterne la struttura e la conserva- zione. L’inserimento dei vegetali in questa maniera è del tutto originale e piuttosto in- solita rispetto ad altri Horti sicci in cui le piante sono o ancorate con fili di cucitura o, come nella maggior parte dei casi, incollate o meglio agglutinate, per dirla con le parole che proprio Aldrovandi utilizza per spiega- re come si confezionano gli Horti sicci. Sotto ogni esemplare è stato scritto il nome della pianta, il numero progressivo di riferimento della scheda posta sulla prima parte del ma- noscritto e di seguito il numero della pagina in cui si trova la scheda stessa. La seconda parte è anticipata dall’indice che occupa cin- que carte non numerate. 11 Per avvalorare le preparazioni che si possono comporre con alcune piante si cita come fonte addirittura Cesare nel De Bello Gallico: si veda a questo proposito la scheda n. 88 a pag. 57 del ms. 12 Si veda a questo proposito la scheda n. 79 a pag. 52 del ms dove viene riportata la pianta detta ferrum equinum conosciuta volgarmente come Sferracavallo e dagli alchimisti come Lunaria minor: ad essa si attribuiscono proprietà prodigiose quali appunto quella di essere utilissima nel togliere i ferri dagli zoccoli dei cavalli che la calpesteranno e secondo gli alchimisti quello di coagulare l’argento vivo. 10 H. Boerhaave, Historia plantarum, quae in horto academico Lugduni-Batavorum crescunt cum earum characteribus, & medicinalibus virtutibus, Roma 1727. 7 J. De Tournefort, Institutiones rei herbariae, Parigi 1700. 8 D. Jonquet, Hortus regius Parisiensis, Parigi 1665. 9 J. Commelin, Horti medici Amstelodamensis rariorum plantarum descriptio et icones, Amsterdam 1697. L’Hortus Siccus Pisanus di Castiglion Fiorentino 33 piena pelle coeva al manoscritto (Fig. 1). Sul dorso della legatura compaiono, oltre al tito- lo con cui il manoscritto è conosciuto Hortus Siccus Pisanus, due talloncini, quello che ri- porta l’attuale segnatura 525, e un altro più antico dove è riportata la segnatura O-VI-12. Il libro nell’insieme sembra essere concepito come un’unica unità codicologica anche se è strutturato in due parti ben definite e distinte. L’introduzione, anticipata dal titolo In Botha- nicas institutiones praefatio (Fig. 2), è una vera e propria premessa metodologica nella quale si spiegano le varie parti costitutive delle piante e viene definita la nomenclatura botanica delle singole partizioni dei vegetali. Dopo questa spiegazione si sviluppa la prima sezione che va da p. 3 a p. 177. In questa parte compare la descrizione di 300 piante diverse, numerate progressivamente in cifre arabe ma non disposte in ordine alfabetico. Ogni singola descrizione, come in una sor- ta di scheda, si articola partendo dal nome scientifico attribuito alla pianta, seguito dal- la nomenclatura attribuita da botanici come Mattioli, L’Obel, Caspar e Johan Bauhin e Cesalpino (Fig. 3). Tra i vari riferimenti e indicazioni bibliografiche, abbondantemente presenti nel testo, compaiono le Institutiones Rei Herbariae7 di Tournefort, l’Hortus Regius Parisiensis8 di Jonquet, l’Hortus Amsteloda- mensis9 di Commelin, e l’Hortus Lugduni- Batavorum10 di Boeravio. piena pelle coeva al manoscritto (Fig. 1). Sul dorso della legatura compaiono, oltre al tito- lo con cui il manoscritto è conosciuto Hortus Siccus Pisanus, due talloncini, quello che ri- porta l’attuale segnatura 525, e un altro più antico dove è riportata la segnatura O-VI-12. Il libro nell’insieme sembra essere concepito come un’unica unità codicologica anche se è strutturato in due parti ben definite e distinte. L’introduzione, anticipata dal titolo In Botha- nicas institutiones praefatio (Fig. 2), è una vera e propria premessa metodologica nella quale si spiegano le varie parti costitutive delle piante e viene definita la nomenclatura botanica delle singole partizioni dei vegetali. Dopo questa spiegazione si sviluppa la prima sezione che va da p. 3 a p. 177. In questa parte compare la descrizione di 300 piante diverse, numerate progressivamente in cifre arabe ma non disposte in ordine alfabetico. Ogni singola descrizione, come in una sor- ta di scheda, si articola partendo dal nome scientifico attribuito alla pianta, seguito dal- la nomenclatura attribuita da botanici come Mattioli, L’Obel, Caspar e Johan Bauhin e Cesalpino (Fig. 3). Descrizione dell’Hortus Siccus Pisanus In area aretina una testimonianza tangibile di questa produzione di raccolte per lo stu- L’Hortus Siccus Pisanus di Castiglion Fiorentino 9 J. Commelin, Horti medici Amstelodamensis rariorum plantarum descriptio et icones, Amsterdam 1697. Fig. 3 La carta 6 dell’erbario in cui è rammentato Mattioli. Fig. 4 Indice delle piante e della loro collocazione nel volume. 13 Fondo Ghizzi in Fondi speciali della biblioteca di Castiglion Fiorentino, a cura di Piero Fusi (quaderno didattico n. 2), Castiglion Fiorentino 2003, pp.10-12. 14 A. Moriani, Il Fondo “Ghizzi” della Biblioteca Comunale di Castiglion Fiorentino in Vivo solo e nego il saluto. Giuseppe Ghizzi e la sua raccolta documentaria, (quaderno di biblioteca 12), Arezzo 1996, pp. 55-64. 15 A. Sorbelli, Castiglione Fiorentino, Biblioteca Comunale (Fondo Ghizzi), in Inventari dei manoscritti delle biblioteche d’Italia, a cura di G. Mazzatinti, vol. XXVI, 1918. 16 A. Sorbelli, Castiglione Fiorentino, Biblioteca Comunale (Fondo Ghizzi), in Inventari dei manoscritti delle biblioteche d’Italia, a cura di G. Mazzatinti, vol. XXIV, 1930. L’Hortus Siccus Pisanus di Castiglion Fiorentino L’indice è composto dai nomi delle piante disposti in ordine al- fabetico seguiti da tre numeri: quello posto nella prima colonna corrisponde alla scheda descrittiva di ogni singola erba o pianta, il numero successivo fa riferimento alla pagina in cui si trova la scheda del vegetale e infine il terzo numero indica la pagina in cui si tro- va la pianta essiccata. In alcuni casi il nome latino del vegetale è seguito da quello volgare, lemma generalmen- te sottolineato come forma di richiamo che tende ad evidenziare e distinguere tale paro- la rispetto al testo stesso. Nel caso le piante non siano considerate autoctone ne viene ri- portata la provenienza; segue la descrizione della pianta partendo dallo stelo o dal fusto per poi proseguire con le foglie, l’apparato radicale e per finire si descrivono la forma e le proprietà del frutto o del fiore. Queste spe- cifiche indicazioni sono nella maggior parte dei casi seguite dalla spiegazione dell’utilizzo e della funzione della singola pianta nei trat- tamenti terapeutici. Si descrivono non solo la preparazione di composti, unguenti, decotti o farmaci in genere, ma anche le modalità di assunzione di tali medicamenti e gli effetti benefici che porterà il suo utilizzo. Spesso, Fig. 3 Fig. 4 38 Leonardo Magionami Vicende dell’acquisizione Il libro, oggi conservato nel fondo Ghizzi13, non entrò a far parte della collezione della biblioteca pubblica per volere di Giuseppe Ghizzi primo bibliotecario della biblioteca pubblica, istituita nel 1867. Infatti lo stu- dioso, scomparso nel 1893, aveva lasciato al Comune solo i suoi libri a stampa, esclu- dendo la restante parte della sua biblioteca costituita da manoscritti e carte d’archivio, perché riteneva che i suoi concittadini non avessero compreso fino in fondo l’importanza della sua raccolta14. La parte della collezione esclusa dal lascito testamentario fu donata al Comune dalla vedova di Ghizzi, la signora Eleonora Gnagnoni, nell’anno 1895. I pezzi donati furono 465 e ancora l’Hortus siccus non era tra questi, come dimostra anche il catalogo redatto dal Sorbelli nel 1918 in cui il libro non è presente15. Successivamente, nel 1929, don Angelo Nunziati, bibliotecario del Comune di Castiglion Fiorentino, fece acqui- stare all’amministrazione comunale un’altra ottantina di pezzi trovati in casa di Giusep- pe Gnagnoni, nipote di Ghizzi. In questa occasione i manoscritti furono riuniti nella collezione che porta il nome del primo rac- coglitore e vennero segnati, secondo la con- suetudine che aveva contraddistinto la prima donazione, con una sigla alfanumerica che ne determinava la collocazione descrivendo anche lo scaffale, il palchetto e il numero di catena; nello specifico l’Hortus Siccus Pisa- nus fu segnato O VI 12. Questi manoscritti furono descritti nell’aggiornamento al cata- logo, redatto, sempre da Sorbelli, nell’anno 193016. In esso, al numero 525, compare la descrizione dell’ Hortus Siccus in cui si riporta, oltre al titolo e alla spiegazione del contenuto, la nota che compare sulla prima carta: del P. Liborio Tommasini come fosse una sorta di nota di possesso. In effetti l’an- notazione è presente nel codice anche se la lettura esatta è: «del D. Liborio Tommasini» intendendo la lettera D. come il titolo di Dot- tore (Fig. 5). La nota ci riporta ad altri mano- Oltre alla particolare struttura del libro sono però interessanti le vicende che hanno porta- to il manoscritto nella collezione dell’Istitu- zione Culturale ed Educativa Castiglionese. Il libro, oggi conservato nel fondo Ghizzi13, non entrò a far parte della collezione della biblioteca pubblica per volere di Giuseppe Ghizzi primo bibliotecario della biblioteca pubblica, istituita nel 1867. Vicende dell’acquisizione scritti della collezione Ghizzi che in qualche maniera si legano alla famiglia Tommasini, di cui l’Hortus Siccus faceva sicuramente parte. Al numero 527 ne compare addirittu- ra un altro appartenuto anche questo proprio a Liborio Tommasini, che ne risulta, oltre che possessore, anche autore. Il manoscritto in questione è intitolato Institutiones Medi- cae et de Historia medica, e del resto questo tipo di testo è perfettamente in sintonia con gli interessi professionali dell’autore. Di par- ticolare importanza a questo punto è l’osser- vazione della scrittura di questo manoscritto, riconducibile alla stessa mano che ha vergato anche il ms. 525 ossia l’Hortus Siccus. Una mano fortemente calligrafica, che utilizza il medesimo sistema interpuntivo e abbreviati- vo e dispone il testo all’interno della pagi- na mantenendo gli stessi spazi e gli stessi margini. L’unica differenza è che la scrittura dell’Hortus, pur mantenendo lo stesso modu- lo, nella parte centrale riduce l’interlineo e mantiene il modulo della scrittura rendendo le linee grafiche più compatte e meno leggi- bili. Anche la maniera di numerare le carte è la medesima sia nel 525 che nel 527. L’ipo- tesi che i due manoscritti siano della stessa mano e dello stesso autore è avvalorata anche dal testo, in quanto nel trattato medico ricor- rono continuamente riferimenti all’utilizzo delle piante per scopi terapeutici e si parla proprio di alcune piante conservate nell’Hor- tus Siccus. Di rimando nella raccolta di ve- getali, all’interno della schede delle singole piante, nei punti in cui si spiega l’utilizzo a scopi terapeutici, spesso si fa riferimento a personaggi di rilievo nel campo della botani- ca e della medicina che sono più volte citati e presi come esempio nel trattato di medici- na. L’insieme di tutte queste osservazioni e riflessioni avvalora ancor di più l’ipotesi che entrambi i manoscritti siano appartenuti al dottor Liborio Tommasini e alcune espres- sioni presenti nell’Hortus Siccus, soprattut- to nella premessa metodologica, come per esempio diximus anno superiore, et hoc anno repetimus oppure explicavimus anno superio- re potrebbero essere allusioni ad una sorta di corso universitario. Sembra proprio che si faccia riferimento ad un approfondimen- to della disciplina botanica che ha previsto, Oltre alla particolare struttura del libro sono però interessanti le vicende che hanno porta- to il manoscritto nella collezione dell’Istitu- zione Culturale ed Educativa Castiglionese. Vicende dell’acquisizione Infatti lo stu- dioso, scomparso nel 1893, aveva lasciato al Comune solo i suoi libri a stampa, esclu- dendo la restante parte della sua biblioteca costituita da manoscritti e carte d’archivio, perché riteneva che i suoi concittadini non avessero compreso fino in fondo l’importanza della sua raccolta14. La parte della collezione esclusa dal lascito testamentario fu donata al Comune dalla vedova di Ghizzi, la signora Eleonora Gnagnoni, nell’anno 1895. I pezzi donati furono 465 e ancora l’Hortus siccus non era tra questi, come dimostra anche il catalogo redatto dal Sorbelli nel 1918 in cui il libro non è presente15. Successivamente, nel 1929, don Angelo Nunziati, bibliotecario del Comune di Castiglion Fiorentino, fece acqui- stare all’amministrazione comunale un’altra ottantina di pezzi trovati in casa di Giusep- pe Gnagnoni, nipote di Ghizzi. In questa occasione i manoscritti furono riuniti nella collezione che porta il nome del primo rac- coglitore e vennero segnati, secondo la con- suetudine che aveva contraddistinto la prima donazione, con una sigla alfanumerica che ne determinava la collocazione descrivendo anche lo scaffale, il palchetto e il numero di catena; nello specifico l’Hortus Siccus Pisa- nus fu segnato O VI 12. Questi manoscritti furono descritti nell’aggiornamento al cata- logo, redatto, sempre da Sorbelli, nell’anno 193016. In esso, al numero 525, compare la descrizione dell’ Hortus Siccus in cui si riporta, oltre al titolo e alla spiegazione del contenuto, la nota che compare sulla prima carta: del P. Liborio Tommasini come fosse una sorta di nota di possesso. In effetti l’an- notazione è presente nel codice anche se la lettura esatta è: «del D. Liborio Tommasini» intendendo la lettera D. come il titolo di Dot- L’Hortus Siccus Pisanus di Castiglion Fiorentino 39 L’Hortus Siccus Pisanus di Castiglion Fiorentino Fig. 5 Particolare della prima carta dell’erbario in cui compare il nome del proprietario. nistrazione economica familiare19, sia quei libri di studio riferibili alla pratica medica che aveva contraddistinto gli uomini della famiglia Tommasini. E proprio il bisnipote di Giovanna Teresa, Giuseppe Ghizzi, tre generazioni dopo, partendo anche da questo nucleo di libri, si fece zelante collezionista di carte e insieme di tutte le altre testimo- nianze storiche e scientifiche che si riferi- vano al territorio di Castiglion Fiorentino e alla zona circostante, a tal punto che la sua opera e il suo impegno in questo senso lo portò ad autodefinirsi «il più gran raccogli- tore del Paese»20. 17 La genealogia della famiglia Tommasini, originaria di Poppi, è ricostruita nel ms. 473. L’albero genealogico qui raffigurato dalla mano dello stesso Giuseppe Ghizzi è arricchito di commenti sui singoli componenti della famiglia. Si dice che il capostipite Tommaso era originario di Poppi e viveva nel 1641, inoltre sulla linea parentale di Liborio si aggiunge che il fratello, il dott. Tommaso Ranieri, fu autore di una piccola dissertazione sull’Aurora Boreale. 18 La vita della suora è descritta nel ms. 9 conservato all’interno dello stesso fondo. 19 Si tratta dei mss. 222, 473, 475 conservati all’interno dello stesso fondo. 20 Tali parole sono presenti nel ms. 476 contenente la sua autobiografia completata nel 1892. Fig. 5 Particolare della prima carta dell’erbario in cui compare il nome del proprietario. 20 Tali parole sono presenti nel ms. 476 contenente la sua autobiografia completata nel 1892. 19 Si tratta dei mss. 222, 473, 475 conservati all’interno dello stesso fondo. 1 S. Pignatti, Flora d’Italia, voll. 1-3, Edagricole. Bologna 1982. 2 F. Conti, G. Abbate, A. Alessandrini, C. Blasi, An annotated checklist of the Italian Vascular Flora, Palombi e Partner S.r.l., Roma 2005. Elenco delle specie presenti nell’Hortus Siccus Pisanus Paolo Emilio Tomei e Francesca Malfanti L L ’interesse di questo erbario, racchiuso nel suo titolo Hortus Siccus Pisanus, è relativo al fatto che, come pare, si tratta di una raccolta di specie per lo più spontanee del territorio pisano; nel testo questo è speci- ficato dall’autore che quando tratta di piante non autoctone ne specifica la provenienza. L’erbario pisano quindi, pur essendo stato preparato con finalità di carattere medico, permette anche una lettura delle informazio- ni in chiave fitogeografica. – Achillea millefolium L. s.l. = Millefolium purpureum maius Gaspari Bavini (87/187) – Achillea millefolium L. s.l. = Millefolium purpureum maius Gaspari Bavini (87/187) – Achillea millefolium L. s.l. = Millefo- lium nobile Thrasii, seu Achillęa Mattioli (87/188) – Achillea millefolium L. s.l. = Millefo- lium nobile Thrasii, seu Achillęa Mattioli (87/188) – Adianthum capillus-veneris L. = Adiantum Mattioli (149/285) – Adianthum capillus-veneris L. = Adiantum Mattioli (149/285) – Aesculus hyppocastanum L. = Hippocasta- num vulgare Institutionum Rei Herbarię, seu Castanea equina Mattioli (39/100) – Aesculus hyppocastanum L. = Hippocasta- num vulgare Institutionum Rei Herbarię, seu Castanea equina Mattioli (39/100) – Agrimonia eupatoria L. s.l. = Agrimonia Officinarum Institutionum Rei Herbarię seu Eupatorium Mattioli (117/235) Di seguito viene presentato una lista delle specie identificate, che deve intendersi del tutto preliminare, in primo luogo perché il riconoscimento è stato eseguito su immagini fotografiche (fino ad ora non è stato possibile lavorare direttamente sugli exsiccata), inoltre perché diversi campioni mancano di parti essenziali per una valida diagnosi. Di seguito viene presentato una lista delle specie identificate, che deve intendersi del tutto preliminare, in primo luogo perché il riconoscimento è stato eseguito su immagini fotografiche (fino ad ora non è stato possibile lavorare direttamente sugli exsiccata), inoltre perché diversi campioni mancano di parti essenziali per una valida diagnosi. – Agrimonia procera Wallr. = Agrimonia odorata Lemonari (117/236) – Agrimonia procera Wallr. = Agrimonia odorata Lemonari (117/236) – Ajuga chamaepitys (L.) Schreb. s.l. = Ca- mepitis prima Mattioli (29/78) – Ajuga chamaepitys (L.) Schreb. s.l. = Ca- mepitis prima Mattioli (29/78) – Ajuga reptans L. cfr. Ajuga pyramidalis L. = Bubula Dodonei, seu Consolida media Mattioli (3/1) Le specie sono state determinate facendo riferimento a Pignatti1, conformandosi alla nomenclatura di Conti et al.2 Esse vengono elencate in ordine alfabetico, seguite dal po- linomio prelinneano indicato nel manoscrit- to, dalla tavola su cui l’exsiccatum è fissato e il suo numero di identificazione. – Alcea rosea L. Vicende dell’acquisizione È per questo motivo che tra tutte queste testimonianze sono presenti anche quei manoscritti che in qualche ma- niera si legano alle vicende della famiglia Tommasini. Questo particolare aspetto po- trebbe essere un ulteriore elemento a dimo- strazione dell’importanza del manoscritto, non soltanto come fonte diretta per gli studi di storia della botanica o per la storia della produzione libraria e della cultura scritta in genere, ma anche come elemento aggiuntivo per ricostruire vicende di storia sociale, es- sendo proprio questo manoscritto un oggetto che si ricollega agli episodi personali e agli interessi di uomini studiosi ed eruditi che, sin da un lontano passato, attraverso i loro studi hanno reso grande l’intero territorio to- scano e nello specifico aretino. l’anno precedente, una sorta di introduzione e successivamente una forma di approfondi- mento pratico attraverso l’allestimento di un repertorio di vegetali sia autoctoni che, per così dire, esotici. Questo non ci meraviglie- rebbe affatto, vista la sostanziale necessità per una persona che all’epoca volesse affron- tare gli studi medici di approfondire a scopi terapeutici la conoscenza delle piante e la fruizione di esse in questo senso. l’anno precedente, una sorta di introduzione e successivamente una forma di approfondi- mento pratico attraverso l’allestimento di un repertorio di vegetali sia autoctoni che, per così dire, esotici. Questo non ci meraviglie- rebbe affatto, vista la sostanziale necessità per una persona che all’epoca volesse affron- tare gli studi medici di approfondire a scopi terapeutici la conoscenza delle piante e la fruizione di esse in questo senso. q Ma per ricostruire le vicende di queste carte e ripercorrere gli avvenimenti che hanno portato il manoscritto a far parte della raccolta dell’Istituzione Castiglione- se si dovranno prendere in esame anche altri elementi. Il dottor Liborio Tommasini (1719-1767), figlio di Raffaello Gaetano, fu medico condotto a Castiglion Fiorentino17 ed ebbe un fratello, Tommaso Ranieri, morto in giovane età nel 1742, e tre sorelle Anna Caterina18, Isabella Vittoria e Giovanna Te- resa. Come risulta dal ms. 475, dove sono riportate notizie sulla famiglia, due sorelle si fecero suore e di conseguenza, alla morte di Liborio, i suoi beni passarono alla terza sorella, Giovanna Teresa, che nel 1738 ave- va sposato il castiglionese Leonardo Ghizzi. Probabilmente con l’eredità di Liborio per- vennero alla sorella anche tutte le carte di famiglia, sia le filze archivistiche, il cui con- tenuto è riferibile prevalentemente all’ammi- Fig. 6 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Fig. 6 Campione di Achillea millefolium L. s.l. (n. 186). Elenco delle specie presenti nell’Hortus Siccus Pisanus Paolo Emilio Tomei e Francesca Malfanti = Malva vulgaris flore maio- re, folio sinuato Ioannis Bavini (43/105) – Alliaria petiolata (Bieb.) Cavara et Grande = Hęsperis allium redolens Morisoni, seu Alliaria Mattioli (7/14)i – Althaea officinalis L. = Althęa Mattioli, seu Bismalva Officinarum (47/114) I polinomi delle specie non identifica- te o dei campioni mancanti non sono stati riportati. – Anagallis arvensis L. s.l. = Anagallis fęniceo flore Gaspari Bavini seu Anagallis mas Mattioli (155/295) A – Anagallis foemina Miller = Anagallis ce- ruleo flore Gaspari Bavini seu Anagallis fęmina Mattioli (155/296) – Achillea millefolium L. s.l. = Millefolium minus Mattioli (87/186) (Fig. 6) – Achillea millefolium L. s.l. = Millefolium minus Mattioli (87/186) (Fig. 6) Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 42 Paolo Emilio Tomei e Francesca Malfanti – Anagyris foetida L. = Anagiris fętida Gaspari Bavini (113/230) – Anagyris foetida L. = Anagiris fętida Gaspari Bavini (113/230) – Anchusa azurea Mill. = Buglossum vulgare Mattioli (93/192) – Anemone apennina L. = Ranunculus vernus rotundifo- lius minor Institutionum Rei Herbarię, seu Cheidonia rotundifolia minor Gaspari Bavini (111/224) – Angelica arcangelica L. = Angelica montana perennis Aquilegie foliis Institutionum Rei Herbarię (19/41) – Angelica sylvestris L. s.l. = Angelica silvestris minor, sive erratica Gaspari Bavini, seu Podagraria Lobelli (17/40) – Angelica sylvestris L. s.l. = Imperatoria pratensis maior Institutionum Rei Herbarię seu Angelica sylvestris Mat- tioli (73/164) – Angelica sylvestris L. s.l. = Imperatoria pratensis maior Institutionum Rei Herbarię seu Angelica sylvestris Mat- tioli (73/164) – Anthriscus cerefolium (L.) Hoffm. = Cherophillum sati- vum Gaspari Bavini, seu cerefolium Mattioli (51/123) – Anthriscus sylvestris (L.) Hoffm. = Cherophillum sylvestre perennę cicutę folio Institutionum Rei Herbarię (51/124) – Anthyllis barba-jovis L. = Barba Iovis pulchre lucens Ioannis Bavini (13/26) – Apium graveolens L. = Apium palustre Mattioli (119/241) – Apium nodiflorum (L.) Lag., il campione e la descrizione si riferiscono a questo nome, ma la nomenclatura di Mattioli corrisponde a Berula erecta (Hudson) Coville = Sium verum Mattioli (121/243)l Fig. 7 – Aquilegia sp. = Aquilegia flore semplici Gaspari Bavini, seu Aquilegia Mattioli (35/96)l Fig. 8 Fig. 8 – Aquilegia sp. = Aquilegia flore albo semplici Gaspari Bavini (35/97) – Aquilegia sp. (cultivar) = Aquilegia hortensis multiplex flore magno albo Gaspari Bavini (35/99) – Aquilegia sp. (cultivar) = Aquilegia hortensis multiplex flore magno albo Gaspari Bavini (35/99) – Aquilegia sp. (cultivar) = Aquilegia hortensis multiplex flore magno violaceo Gaspari Bavini (35/98) – Aquilegia sp. (cultivar) = Aquilegia hortensis multiplex flore magno violaceo Gaspari Bavini (35/98) l g p ( – Arbutus unedo L. = Arbutus Mattioli (3/3) – Arisarum vulgare Targ.-Tozz. = Arisarum Mattioli (109/226) – Arisarum vulgare Targ.-Tozz. = Arisarum Mattioli (109/226) – Aristolochia clematitis L. Fig. 7 Campione di Artemisia abrotanum L. (n. 175). Fig. 8 Campione di Calamintha nepeta (L.) Savi s.l. (n. 52). Fig. 9 Campione di Cochlearia officinalis L. (n. 251). Fig. 10 Campione di Cornus mas L. (n. 293). A = Aristholochia longa vera Ga- spari Bavini seu Aristholochia longa Mattioli (55/136) – Aristolochia clematitis L. = Aristholochia longa vera Ga- spari Bavini seu Aristholochia longa Mattioli (55/136) – Aristolochia pistolochia L. Cfr. = Aristolochia Pistolochia altera Ihoannis Bavini, seu Bistolochia crętica Gaspari Bavini (55/138) – Artemisia abrotanum L. Cfr. = Abrotanum mas Mattioli (81/175) (Fig. 7) – Artemisia absinthium L. Cfr. = Absynthium Mattioli (77/173) – Artemisia pontica L. Cfr. = Absynthium ponticum Mat- tioli (81/174) – Artemisia sp. = Arthęmisia italica eluptior Delphini fo- lio viridi, caule atrorubente, inodora Accademię Regię Scientiarum (25/61) – Artemisia sp. = Arthęmisia italica eluptior Delphini fo- lio viridi, caule atrorubente, inodora Accademię Regię Scientiarum (25/61) Fig. 8 Elenco delle specie presenti nell’Hortus Siccus Pisanus 43 = Iacobea marina, sive cineraria Ioannis Bavini (67/153) Fig. 7 Campione di Artemisia abrotanum L. (n. 175). Fig. 8 Campione di Calamintha nepeta (L.) Savi s.l. (n. 52). Fig. 9 Campione di Cochlearia officinalis L. (n. 251). Fig. 10 Campione di Cornus mas L. (n. 293). Fig. 10 Fig. 9 = Iacobea marina, sive cineraria Ioannis Bavini (67/153) – Artemisia vulgaris L. = Arthęmisia Mattio- li (25/60) – Arum italicum Miller = Harum vulgare non maculatum Gaspari Bavini (105/218) B B – Ballota nigra L. s.l. = Ballotthi Mattio- li seu Marrubium nigrum, sive Ballotthi Iohannis Bavini (49/121) – Arum maculatum L. = Harum maculatum maculis candidis vel nigris Gaspari Bavi- ni, seu Harum Mattioli (105/217) – Bellis perennis L. = Bellis minor Mattioli (97/201) – Asphodeline lutea (L.) Rchb. = Asphodelus spiralis, luteus, italicus, flore magno Horti Regi Parisiensis (139/266) – Berberis vulgaris L. s.l. = Berberis dume- torum Gaspari Bavini, seu Crespinus Mat- tioli (23/59) – Asphodelus sp. = Asphodelus albus, ramo- sus, mas Gaspari Bavini (137/264) – Bistorta officinalis Delarbre = Bistorta Mattioli (137/263) – Asphodelus sp. = Asphodelus altissimus albus Arundinis Indicę Variegatę foliis Catalogi Horti Parisiensis (139/265) – Borago officinalis L. = Borrago floribus ceruleis Iohanni Bavini seu Buglossum sive Borrago Mattioli (115/237)il – Asplenium adianthum-nigrum L. = Fili- cula quę Adianthum nigrum Officinarum pinnulis obtusioribus Institutionum Rei Herbarię (149/284) – Borago officinalis L. = Borrago floribus al- bis Iohanni Bavini (119/238) – Bryonia dioica Jacq. = Brionia aspera, sive Alba baccis rubris Gaspari Bavini seu Brionia Alba Dodonei, seu pitis alba Mat- tioli (62/141) – Asplenium ruta-muraria L. s.l. = Rutha muraria Gaspari Bavini seu Paronichia Mattioli (149/283) – Asplenium trichomanes L. s.l. = Tricoma- nes Mattioli (149/281) – Bryonia sp., il campione è così identifica- bile, ma la descrizione è quella di Ecbal- lium elaterium (L.) A. Rich. = Cucunus sylvestris Asininus dictus Gaspari Bavini, seu Cucunus sylvestris Mattioli, seu Ela- therium Officinarum (49/122) – Buglossoides arvensis (L.) Johnston = Lyti- spermum minus Mattioli (47/113) – Bryonia sp., il campione è così identifica- bile, ma la descrizione è quella di Ecbal- lium elaterium (L.) A. Rich. = Cucunus sylvestris Asininus dictus Gaspari Bavini, seu Cucunus sylvestris Mattioli, seu Ela- therium Officinarum (49/122) – Atriplex halimus L. = Atriplex latifolia sive Alimus fruticosus Morioni (157/299) – Atriplex halimus L. = Atriplex marittima angustissimo folio Morisoni seu Alimus vulgaris Mattioli (159/300) Buglossoides arvensis (L.) Johnston = Lyti- spermum minus Mattioli (47/113) – Atriplex portulacioides L., ma il campione è identificabile con Portulaca oleracea L. Paolo Emilio Tomei e Francesca Malfanti 44 – Bupleurum fruticosum L. = Bupleurum ar- borescens salicifolium Institutionum Rei Herbarię, seu Seseli etiopicum alterum Mattioli (33/92) – Cistus creticus L. subsp. eriocephalus (Viv.) Greuter & Burdet = Cystus mas Mattioli (123/245) – Cistus monspeliensis L. = Cystus Ledon foliis oleę, sed angustioribus Gaspari Ba- vini, seu Ledon Mattioli (123/247) – Bupleurum rotundifolium L. B = Bupleurum perfoliatum rotundifolium annuum Institu- tionum Rei Herbarię, seu Berfoliata Mat- tioli (39/91) – Cistus salvifolius L. = Cystus fęmina Mat- tioli (123/246) – Cnicus benedictus L. = Nicus sylvestris hirsutior sive Cardus Benedictus Gaspa- ri Bavini seu Cardus Sanctus Mattioli (31/257) C – Calamintha nepeta (L.) Savi s.l. = Cala- minta pulegi odore, sive Nepeta Bavini, sive Calaminta montana Mattioli (23/52) (Fig. 8) – Calamintha nepeta (L.) Savi s.l. = Cala- minta pulegi odore, sive Nepeta Bavini, sive Calaminta montana Mattioli (23/52) (Fig. 8) – Cochlearia officinalis L. = Coclearia folio cubitali Institutionum Rei Herbarię, seu Raphanus vulgaris et rusticanus Mattioli (129/251) (Fig. 9) – Calendula arvensis L. = Caltha arvensis Gaspari Bavini (57/133)i – Calendula arvensis L. = Caltha arvensis Gaspari Bavini (57/133) – Cochlearia officinalis L. = Coclearia folio subrotondo Gaspari Bavini (132/252) – Calendula officinalis L. = Caltha vul- garis Gaspari Bavini seu caltha Mattioli (57/132) – Conium maculatum L. = Cicuta Mattioli (19/43) – Caltha palustris L. = Populago flore maiore Institutionum Rei Herbarię, seu Tussilago altera Mattioli (17/37) – Convallaria majalis L. = Lilium conval- lium Mattioli (13/27) – Cornus mas L. = Cornus Mattioli (157/293) (Fig. 10) – Capsella bursa-pastoris (L.) Medicus = Bursa Pastoris Mattioli (97/202) – Cotinus coggygria Scop. = Cotinus Mattio- li (155/288) – Cotinus coggygria Scop. = Cotinus Mattio- li (155/288) – Celtis australis L. = Celtis fructu nigrican- te Institutionum Rei Herbarię, seu Lotus arbor Mattioli (31/77) – Cruciata laevipes Opiz = Cruciata irsuta Gaspari Bavini (11/22) – Centaurium erythraea Rafn s.l. = Centau- rinum minus Mattioli (89/184) – Centaurium erythraea Rafn s.l. = Centau- rinum minus Mattioli (89/184) – Cymbalaria muralis Gaertn., Mey. et Sch. s.l. = Linaria hederaceo folio glabro, seu Cimbalaria vulgaris Gaspari Bavini seu Cimbalaria Mattioli (53/135)i – Centranthus ruber (L.) DC. = Valeriana ru- bra Gaspari Bavini (55/130) – Centranthus ruber (L.) DC. = Valeriana ru- bra Gaspari Bavini (55/130) – Ceratonia siliqua L. = Siliqua ędulis Ga- spari Bavini seu Siliqua Mattioli (37/291)l – Ceratonia siliqua L. = Siliqua ędulis Ga- spari Bavini seu Siliqua Mattioli (37/291)l – Cynoglossum officinale L. Cfr. = Cinoglos- sum vulgare Mattioli (147/277) – Cercis siliquastrum L. = Siliquastrum flore albo Institutionum Rei Herbarię (39/93) – Cercis siliquastrum L. = Siliquastrum flore albo Institutionum Rei Herbarię (39/93) D – Cercis siliquastrum L. = Siliquastrum Ca- storis durantis Istitutionum Rei Herbarię (39/94) – Cercis siliquastrum L. – Emerus majus Mill. s.l. = Emerus Cisalpi- ni (11/25) B = Siliquastrum Ca- storis durantis Istitutionum Rei Herbarię (39/94) – Dipsacus fullonum L. = Dipsacus sylve- stris aut Virga pastoris maior Gaspari Ba- vini seu Labium Veneris alterum Mattioli (89/189) – Cerinthe major L. s.l. = Cerinte, seu Cino- glossum Montanum maius Gaspari Bavini (3/2) – Dipsacus laciniatus L. = Dipsacus syl- vestris laciniato folio Gaspari Bavini (93/191) – Ceterach officinarum DC. s.l. = Asplenium Mattioli (149/287) – Ceterach officinarum DC. s.l. = Asplenium Mattioli (149/287) – Dipsacus sativus (L.) Honckeny = Dipsa- cus sativus Gaspari Bavini, seu Labium Veneris Mattioli (91/189) – Chelidonium majus L. = Chęlidonium maius vulgare Gaspari Bavini, seu Cheli- donium maius Mattioli (103/211) – Chelidonium majus L. = Chęlidonium maius vulgare Gaspari Bavini, seu Cheli- donium maius Mattioli (103/211) – Dracunculus vulgaris Schott = Dracun- culus minor Mattioli (155/294) – Chelidonium majus L. = Chelidonium maius folia quernis Gaspari Bavini (101/212) E – Chenopodium bonus-enricus L. = Atriplex hortensis alba Iohannis Bavini (159/298) – Emerus majus Mill. s.l. = Emerus Cisalpi- ni (11/25) – Emerus majus Mill. s.l. = Emerus Cisalpi- ni (11/25) 45 Elenco delle specie presenti nell’Hortus Siccus Pisanus – Euphorbia peplus L. = Titimalus rotundis foliis non crenatis hortus Luĝduni Batavo- rum seu Peplus Mattioli (5/10) Fig. 11 Campione di Plantago media L. s.l. (n. 232). Fig. 12 Campione di Polypodium vulgare L. (n. 282). Fig. 12 – Erodium cfr. cicutarium (L.) L’Hér. = Ge- ranium cicutę folio acu longissima Gaspari Bavini (63/143) Eryngium campestre L = Eringium mon – Euphorbia peplus L. = Titimalus rotundis foliis non crenatis hortus Luĝduni Batavo- rum seu Peplus Mattioli (5/10) Fig. 11 Campione di Plantago media L. s.l. (n. 232). Fig. 12 Campione di Polypodium vulgare L. (n. 282). Fig. 11 Fig. 12 Fig. 12 Fig. 11 Campione di Plantago media L. s.l. (n. 232). Fig. 12 Campione di Polypodium vulgare L. (n. 282). – Euphorbia peplus L. = Titimalus rotundis foliis non crenatis hortus Luĝduni Batavo- rum seu Peplus Mattioli (5/10) – Erodium cfr. cicutarium (L.) L’Hér. = Ge- ranium cicutę folio acu longissima Gaspari Bavini (63/143) – Erodium cfr. cicutarium (L.) L’Hér. = Ge- ranium cicutę folio acu longissima Gaspari Bavini (63/143) – Eryngium campestre L. = Eringium mon- tanum, sive campestre Mattioli (31/83) F – Ferula communis L. = Ferula Galbanifera Lobelli (29/71) – Ferula communis L. = Ferula Galbanifera Lobelli (29/71) – Eryngium planum L. Fig. 11 Campione di Plantago media L. s.l. (n. 232). Fig. 12 Campione di Polypodium vulgare L. (n. 282). – Galium aparine L. = Aparine Mattioli, seu Aparine vulgaris Gaspari Bavini (29/44) – Galium aparine L. = Aparine Mattioli, seu Aparine vulgaris Gaspari Bavini (29/44) – Galium sp. = Gallium montanum Latifo- lium ramosum Institutionum Rei Herbarię (21/45) B = Eringium planum Mattioli (31/85) – Eryngium sp. = Eringium montanum ametistinum Gaspari Bavini (31/84) – Ferulago campestris (Besser) Grec. = Fe- rula Mattioli (29/72) – Ferulago campestris (Besser) Grec. = Fe- rula Mattioli (29/72) – Erysimum cheiri (L.) Crantz = Leucoium hęsperidisfolio Institutionum Rei Herbarię (23/56) – Filipendula vulgaris Moench = Filipendu- la Mattioli (141/273) – Filipendula vulgaris Moench = Filipendu- la Mattioli (141/273) – Fumaria capreolata L. =Fumaria maior scandens foliorum, pediculis, flore maiore pallidiore Morioni (99/206) – Euonymus europaeus L. = Evonimus Mat- tioli (147/280) – Fumaria cfr. officinalis L. s.l. = Fumaria Mattioli (97/205) – Eupatorium cannabinum L. s.l. = Eupato- rium vulgare Mattioli (25/64) – Euphorbia ceratocarpa Ten. Cfr. = Titima- lus orientalis salicis folio, caule purpureo, flore magno, Tournefort Corollario (5/7) – Fumaria sp. = Fumaria foliis tenuissimis, Floris albi; circa monpelium nascuntur Gaspari Bavini (101/207) – Euphorbia characias L. = Titimalus Cara- cias Mattioli (5/6) G – Euphorbia helioscopia L. = Titimalus elio- scopius Mattioli (5/9) – Galium aparine L. = Aparine Mattioli, seu Aparine vulgaris Gaspari Bavini (29/44) – Galium sp. = Gallium montanum Latifo- lium ramosum Institutionum Rei Herbarię (21/45) – Euphorbia lathyris L. = Titimalus latifo- lius Cathaputia dictus Horti Lubudni Ba- tavorum, seu Latiris Mattioli (7/11) 46 Paolo Emilio Tomei e Francesca Malfanti – Galium verum L. s.l. = Gallium Mattioli, seu Gallium luteum Gaspari Bavini (21/46) – Galium verum L. s.l. = Gallium Mattioli, seu Gallium luteum Gaspari Bavini (21/46) p ( ) – Geranium sanguineum L. = Geranium nodosum Gaspari Bavini (61/142) – Geranium sanguineum L. = Geranium nodosum Gaspari Bavini (61/142) – Geranium tuberosum L. = Geranium tuberosum maius Ga- spari Bavini, seu Geranium primum Mattioli (63/146) – Geum urbanum L. = Cariophillata Mattioli (137/262) – Glechoma hederacea L. = Calaminta humilior, folio ro- tundiore Institutionum Rei Herbarię seu Hędera terre- stri Mattioli (23/51) – Hedera helix L. s.l. = Hędera arborea Gaspari Bavini, seu Hędera arborea Mattioli (107/216) – Helleborus foetidus L. = Helleborus niger fętidus Gaspa- ri Bavini (97/203) – Helleborus viridis L. s.l. = Helleborus niger alter Mattio- li (97/204) – Hibiscus syriacus L. = Alcea vulgaris maior flore ex ru- bro roseo Gaspari Bavini, seu Alcea Mattioli (85/185)i – Hieracium pilosella L. = Dens Leonis, qui Pilosella Offi- cinarum Istitutionum Rei Herbarię, seu Pilosella Mattio- li (13/29) – Hippocrepis biflora Spreng. = Ferrum equinum Mattioli (29/79) Fig. 13 – Hyoschyamus niger L. B = Iosciamus Mattioli (133/258) – Hyoseris radiata L. = Dens Leonis minor foliis radiatis Gaspari Bavini (13/30) Fig. 14 – Hypericum androsaemum L. Cfr. = Androsenum maxi- mum frutescens Gaspari Bavini (3/4) – Hypericum perforatum L. = Hipericum Mattioli (32/89) – Hypericum perforatum L. = Hipericum Mattioli (32/89) – Ilex aquifolium L. = Aquifolium Mattioli seu Agrifolium Dodonei (135/259) – Isatis tinctoria L. = Hisatis domestica, sive Glastum Mattioli (33/88) – Jasminum fruticans L. = Iasminum luteum vulgo dicitur Bacciferum Gaspari Bavini (37/292) – Lamium maculatum L. = Lamium folio oblungo, flore rubro Institutionum Rei Herbarię, seu Galeopsis Mattioli (103/213) – Lamium purpureum L. = Lamium purpureum fętidum, folio subrotundo, sive Galeopsis Dioscoridis Gaspari Ba- vini (107/215) – Lavandula stoechas L. = Stechas purpurea Gaspari Ba- vini, seu Stechas Mattioli (41/101) Fig. 14 Elenco delle specie presenti nell’Hortus Siccus Pisanus 47 – Leonurus cardiaca L. = Cardiaca Mattioli (145/278) – Leonurus cardiaca L. = Cardiaca Mattioli (145/278) – Leonurus cardiaca L. = Cardiaca Mattioli (145/278) ta, et graveolens, Horti Regi Pariesiensis (17/39) – Minuartia sp. =Alsine Mattioli (111/227) – Molopospermum peloponnesiacum (L.) Koch = Cicutaria latifolia fętida Gaspari Bavini, seu Seseli peloponense Mattioli (33/90) ta, et graveolens, Horti Regi Pariesiensis (17/39) – Linaria vulgaris Miller = Linaria vulgaris Lutea flore maiore Gaspari Bavini seu Osi- ris Mattioli (53/134) – Minuartia sp. =Alsine Mattioli (111/227) – Molopospermum peloponnesiacum (L.) Koch = Cicutaria latifolia fętida Gaspari Bavini, seu Seseli peloponense Mattioli (33/90) ( ) – Lonicera caprifolium L. = Caprifolium Ita- licum Dodonei, seu Periclimium Mattioli (11/23) N – Lonicera cfr. xylosteum L. = Caprifolium Germanicum Dodonei (11/24) – Nasturtium officinale R. Br. = Sisimbrium acquaticum Mattioli (117/239) – Nasturtium officinale R. Br. = Sisimbrium acquaticum Mattioli (117/239) – Lunaria annua L. = Lunaria maior siliqua rutundiore Iohannis Bavini (41/102) – Nepeta cataria L. = Cattaria Maior vulga- ris Institutionum Rei Herbarię seu herba gattaria Mattioli (13/28) – Nepeta cataria L. = Cattaria Maior vulga- ris Institutionum Rei Herbarię seu herba gattaria Mattioli (13/28) – Lunaria rediviva L. = Lunaria Lenchoii folio siliqua oblunga maiori Institutionum Rei Herbarię (41/103) O – Lysimachia nummularia L. = Lysimachia humi fusa, folio rotundiore, flore luteo In- stitutionum Rei Herbarię, seu Nummula- ria Mattioli (33/81) – Oxalis acetosella L. = Ocxys flore albo Institutionum Rei Herbarię, seu Trifolium acetosum Mattioli (41/104) – Lysimachia vulgaris L. = Lisimachia Mat- tioli (31/80) P – Papaver cfr. dubium L. Fig. 13 Campione di Salvia officinalis L. (n. 180). Fig. 14 Campione di Thapsia garganica L. (n. 274). B = Plantago media Mattioli (113/232) (Fig. 11) – Ranunculus sp. = Ranunculus nemorosus, flore ceruleo, foliis maioribus Appennini montis Menzeli (109/223) – Ranunculus sp. = Ranunculus nemorosus, flore ceruleo, foliis maioribus Appennini montis Menzeli (109/223) – Plantago sp. = Plantago angustifolia albi- da, ispanica Institutionum Rei Herbarię (113/234) ( ) – Rhamnus alaternus L., ma il campione corrisponde ad altra specie non identifica- ta = Alaternus prima Clusii (67/155)i – Polygonatum odoratum (Miller) Druce = Poligonatum Mattioli, seu Pligonatum vulgo sigillum salomonis Iohannis Bavini (65/150) – Polygonatum odoratum (Miller) Druce = Poligonatum Mattioli, seu Pligonatum vulgo sigillum salomonis Iohannis Bavini (65/150) – Rosmarinus officinalis L. = Rosmarinus coronarius Mattioli (157/297) – Rubia tinctorum L. = Rubia sativa Mattioli (19/42) – Polypodium vulgare L. = Polipodium Mat- tioli (151/282) (Fig. 12) – Rumex acetosa L. = Acetosa pratensis Ga- spari Bavini seu Oxalis Mattioli (149/289) – Potentilla anserina L. = Luteola herba Sa- licis folio Gaspari Bavini seu Dostrutium Mattioli (143/275) – Ruscus aculeatus L. = Ruscus Mattioli (7/15) – Potentilla anserina L., ma il campione corrisponde ad altra specie non identifica- ta = Pentaphilloides argenteum alatum seu Potentilla Institutionum Rei Herbarię seu Potentilla Mattioli (151/279) – Ruscus hypoglossum L. = Ruscus angusti- folius fructu folio innascente Institutionum Rei Herbarię, seu Bonifacia, sive Bislingua Ioannis Bavini, seu Hippoglossum Mattioli (9/16) – Potentilla sp. = Quinque folium monta- num, erectum, hirsutum, foliis profonde, et eleganter incisis, flore luteo, pętalis corda- tis Cathalogi Horti Parisiensis (95/196) – Ruscus hypoglossum L. = Ruscus angu- stifolio, fructo folio in nascente Cathalogi Horti Pisani (9/17) – Ruscus hypoglossum L. = Ruscus angu- stifolio, fructo folio in nascente Cathalogi Horti Pisani (9/17) – Ruscus racemosus L. = Ruscus angustifo- lius, fructu summis ramulis in nascente, Instututionum rei Herbarię, seu Laurus Alexandrina racemosa, fructu e summitate caulium prodeunte Horti Regii Parisiensis (11/19) – Potentilla sp. = Quinque folium alpinum erectum latifolium, hirsutum, flore luteo parvo, petalis angustioribus, cordiformi- bus Gaspari Bavini (95/197) – Primula auricola L. = Auricula Ursi omnium maxima Catalogi Horti Pisani (15/35) – Ruscus sp. = Ruscus latifolius, sive Lau- rus Alexandrina, fructu in medio folia- rum extrapendente Cathalogi Horti Pisani (9/18) – Ruscus sp. = Ruscus latifolius, sive Lau- rus Alexandrina, fructu in medio folia- rum extrapendente Cathalogi Horti Pisani (9/18) – Prunella vulgaris L. = Brunella maior, fo- lio non dissecto Gaspari Bavini, seu Con- solida minor Mattioli (135/261) – Prunella vulgaris L. B s.l. = Papaver orientale hirsutissimum magno flore Thur- nephort corollario (71/159) M M – Malva alcea L. Cfr. = Malva vulgaris flore minore, folio rotundo Iohannis Bavini, seu malva Mattioli (43/106) – Papaver rhoeas L. = Palium primum Mat- tioli (77/172) – Matthiola incana R. Br. s.l. = Leucoium album Mattioli (23/54) – Papaver somniferum L. = Papaver herra- ticum rubrum campestre Ioannis Bavini, seu papaver erraticum Mattioli (69/156) – Matthiola situata (L.) R. Br. = Leucoium purpureum Mattioli (23/55) – Papaver somniferum L. = Papaver horten- se semine albo, sativum Dioscoridi, album Plinio, Gaspari Bavini (69/157) – Medicago arabica (L.) Hudson = Medica maior erectior, floribus purpurascentibus Ioannis Bavini (29/74) – Parietaria cfr. officinalis L. = Parietaria Officinarum et Dioscoridis, Caspar Bahuin (111/229) – Medicago orbicularis (L.) Bartal. = Medica orbiculata Ioannis Bavini (29/76) – Pelargonium sp. Cfr. = Geranium Batra- chioides odoratum Gaspari Bavini (63/145) – Medicago sativa L. = Medica glabra annua folio cordato, maculato, fructu compresso, echinis longioribus et tenuioribus erectis Cathalogi Horti Pisani (29/75) – Pelargonium sp. Cfr. = Geranium pober- tianum primum Rubens Gaspari Bavini, seu Geranium tertium Mattioli (63/147) – Melilotus italicus (L.) Lam., ma il campio- ne corrisponde ad altra specie non identifi- cata = Melilotus Italica folliculis rotundis Gaspari Bavini (75/162) – Petroselinum crispum (Mill.) Fuss = Apium hortense Mattioli (121/242) – Petroselinum crispum (Mill.) Fuss = Apium hortense Mattioli (121/242) – Petroselinum crispum (Mill.) Fuss = Apium hortense Mattioli (121/242) – Peucedanum ostruthium (L.) Koch = Im- peratoria Mattioli (69/163) – Peucedanum ostruthium (L.) Koch = Im- peratoria Mattioli (69/163) – Peucedanum ostruthium (L.) Koch = Im- peratoria Mattioli (69/163) – Melilotus officinalis (L.) Palla = Melilotus Officinę Germanię Gaspari Bavini seu Lo- tus urbana Mattioli (71/161) – Phillyrea angustifolia L. = Phillirea Mat- tioli (133/260) – Melissa officinalis L. s.l. = Melissa horten- sis Gaspari Bavini, sive Apiastrum Mattio- li (17/38)i – Phyllitis scolopendrium (L.) Newman = Lingua cervina Officinarum Gaspari Bavi- ni seu Phillitis Mattioli (153/286) – Melissa officinalis L. subsp. altissima (Sm.) Arcang. = Melissa Romana molliter irsu- – Plantago lanceolata L. = Plantago longa Mattioli (115/233) Paolo Emilio Tomei e Francesca Malfanti 48 – Plantago major L. s.l. = Plantago latifolia sinuata Gaspari Bavini (115/231) – Ranunculus sp. = Ranunculus montanus subhirsutus, Gerani folio Gaspari Bavini (107/222) – Ranunculus sp. = Ranunculus montanus subhirsutus, Gerani folio Gaspari Bavini (107/222) – Plantago media L. s.l. Elenco delle specie presenti nell’Hortus Siccus Pisanus 49 Iacobea perennis altissima Simonii folio Institutionum Rei Herbarię (65/151) Iacobea perennis altissima Simonii folio Institutionum Rei Herbarię (65/151) – Salvia sclarea L. = Sclarea Mattioli (21/47) – Salvia sclarea L. = Sclarea Mattioli (21/47) – Salvia sp. = Sclarea laciniatis foliis Insti- tutionum Rei Herbarię (21/48) – Senecio gibbosus (Guss.) DC. subsp. cine- raria (DC.) Peruzzi, Passal. & Soldano = Iacobea vulgaris laciniata Gaspari Bavini (67/152) – Salvia sp. = Salvia affricana fruticans, fo- lio subrotundo, glauco flore, magno, aureo, Horti Amstelodamensis (83/183) – Senecio jacobaea L. = Iacobęa sicula Cri- santemi facie Bocconi (65/154) – Sambucus ebulus L. = Sambucus humilis, sive Ebulus Gaspari Bavini, seu Ebulus sive Sambucus erbacea Ioannis Bavini (27/67) – Sambucus ebulus L. = Sambucus humilis, sive Ebulus Gaspari Bavini, seu Ebulus sive Sambucus erbacea Ioannis Bavini (27/67) – Silene latifolia Poir. subsp. alba (Mill.) Greuter & Burdet Cfr. = Lycnis sylvestris, quę Been album vulgo Gaspari Bavini (143/272) – Sambucus nigra L. = Sambucus Mattioli (25/65) – Sambucus nigra L. = Sambucus Mattioli (25/65) – Sanguisorba minor Scop. s.l. = Pimpinella minor Mattioli (25/62) – Sanguisorba minor Scop. s.l. = Pimpinella minor Mattioli (25/62)i – Silene vulgaris (Moench) Garke s.l. = Lyc- nis sylvestris alba, simplex Gaspari Bavini seu Ocymoides Mattioli (141/271) – Sanguisorba officinalis L. = Pimpinella agrimonides odorata Horti Regis Parisien- sis (25/63) – Sanguisorba officinalis L. = Pimpinella agrimonides odorata Horti Regis Parisien- sis (25/63) – Silybum marianum Gaertn. = Cardus Marię non maculatus Morioni (131/254) – Santolina marchii Arrigoni = Sanctoli- na foliis terrestribus Institutionum Rei Herbarię, seu Abrotanum fęmina Mattioli (93/193) – Sisymbrium cfr. orientale L. = Sisimbrium crucęfolio, glabro, luteo Institutionum Rei Herbarię (117/240) – Sisymbrium cfr. orientale L. = Sisimbrium crucęfolio, glabro, luteo Institutionum Rei Herbarię (117/240) – Smyrnium olusatrum L. = Smirnium Mat- tioli (99/208) – Santolina sp. = Sanctolina foliis oscure vi- rentibus, flores sulphurei coloris Institutio- num Rei Herbarię (93/194) – Solanum dulcamara L. = Solanum scan- dens, seu Dulcamara Gaspari Bavini seu vitis sylvestris Mattioli (77/165) – Santolina sp. Cfr. = Abrotanum mas an- gustifolium maximum Gaspari Bavini (82/176) – Solanum sodomaeum L. = Solanum pomi- ferum, frutescens affricanum, spinosum nigricans, borraginis flore foliis profon- de laciniatis Horti Lugduni Batavorum (75/166) – Santolina sp. Cfr. = Abrotanum mas an- gustifolium incanum Gaspari Bavini (82/177) – Saponaria officinalis L. B = Brunella maior, fo- lio non dissecto Gaspari Bavini, seu Con- solida minor Mattioli (135/261) – Pulmonaria officinalis L. = Pulmonaria Italorum ad Buglossum accedens Ioannis Bavini, seu Pulmonaria altera Mattioli (39/95) S – Salvia cfr. pratensis L. s.l. = Sclarea pra- tensis foliis serratis, flore ceruleo, Institu- tionum Rei Herbarię (21/50) R – Salvia officinalis L. = Salvia maior, an Sphacelus Thophrasti Gaspari Bavini, seu Salvia maior Mattioli (85/179)i – Ranunculus ficaria L. s.l. = Ranunculus vernus rotundis foliis minor Institutionum Rei Herbarię, seu Chelidonium minus Mattioli (109/225) – Ranunculus ficaria L. s.l. = Ranunculus vernus rotundis foliis minor Institutionum Rei Herbarię, seu Chelidonium minus Mattioli (109/225) – Salvia officinalis L. = Salvia maior folio ex viridi, et luteo variegato Horti regis pari- siensis (83/180) (Fig. 13) – Salvia officinalis L. = Salvia maior folio ex viridi, et luteo variegato Horti regis pari- siensis (83/180) (Fig. 13) – Ranunculus sceleratus L. = Ranunculus primis Mattioli (107/219) – Salvia officinalis L. = Salvia maior, an Sphacelus Teophrasti floribus albis Gaspa- ri Bavini (85/181)i – Ranunculus sp. = Ranunculus pratensis, repens, hirsutus Gaspari Bavini (107/220) – Ranunculus sp. = Ranunculus fibrata radi- ce foliis maculatis Horti Pisani (109/221) – Ranunculus sp. = Ranunculus pratensis, repens, hirsutus Gaspari Bavini (107/220)i – Salvia officinalis L. (varietà) = Salvia lati- folia serrata Gaspari Bavini (83/182) – Ranunculus sp. = Ranunculus fibrata radi- ce, foliis maculatis Horti Pisani (109/221) Elenco delle specie presenti nell’Hortus Siccus Pisanus Elenco delle specie presenti nell’Hortus Siccus Pisanus = Lycnis sylve- stris, quę Saponaria vulgo Institutionum Rei Herbarię (141/270) – Spirea hypericifolia L. Cfr. = Spirea Hi- pericifolio non crenato Institutionum Rei Herbarię, seu Pruno Silvestri affinis cana- densis Gaspari Bavini (27/70)i – Scandix pecten-veneris L. s.l. = Scandia semine rostrato vulgaris Gaspari Bavini, seu pecten Veneris Mattioli (51/125) – Scandix pecten-veneris L. s.l. = Scandia semine rostrato vulgaris Gaspari Bavini, seu pecten Veneris Mattioli (51/125) – Stachys officinalis (L.) Trevisan = Betoni- ca Mattioli (23/58) – Staphylea pinnata L. = Staphiledendron Mattioli, seu Nux vesicaria Dodonei (3/5) – Scrophularia sp. = Scrophularia foliis Urticę Gaspari Bavini (15/34) – Scrophularia sp. = Scrophularia foliis Urticę Gaspari Bavini (15/34) – Scrophularia sp. Cfr. = Scrophularia sco- rodonifoliis Morioni (15/32) – Stellaria media (L.) Vill. s. l., il campione è così identificabile, ma la descrizione è quella di Stellaria nemorum L. s. l. = Alsine altissima nemorum Gaspari Bavini (113/228) – Scrophularia sp. Cfr. = Scrophularia flore luteo Gaspari Bavini (15/33) – Scrophularia umbrosa Dumort Cfr. = Scrophularia maxima radice fibrosa Ioan- nis Bavini (15/31) – Symphytum officinale L. s.l. = Symphytum maius Mattioli, seu Symphitum, seu Con- solida maior Gaspari Bavini (7/12) – Senecio gibbosus (Guss.) DC. subsp. cine- raria (DC.) Peruzzi, Passal. & Soldano = Thlapsi vulgatius Ioannis Bavini seu Th- laspi primum Mattioli (65/149) – Symphytum tuberosum L. = Symphytum minus tuberosa radice Gaspari Bavini (7/13)l – Senecio gibbosus (Guss.) DC. subsp. cine- raria (DC.) Peruzzi, Passal. & Soldano = – Syringa vulgaris L. = Lilac flore albo In- stitutionum Rei Herbarię (33/87) 50 Paolo Emilio Tomei e Francesca Malfanti – Thalictrum cfr. flavum L. = Thalictrum maius siliqua angulosa aut striata Gaspari Bavini (139/268) – Thalictrum simplex L. subsp. galioides (DC.) Korsh. = Thalictrum pratense angustissimo folio Gaspari Bavini (139/269) – Thapsia garganica L. = Thapsia sive Turbith Garga- nicum semine latissimo Iohannis Bavini (145/274) (Fig. 14) – Tussilago fanfara L. = Tussilago Mattioli (17/36) U – Umbilicus rupestris (Salisb.) Dandy = Cotiledon maior Gaspari Bavini seu Umbilicus Veneris Mattioli (99/209) – Umbilicus rupestris (Salisb.) Dandy = Cotiledon maior Gaspari Bavini seu Umbilicus Veneris Mattioli (99/209) V – Valeriana dioica L. = Marrubium album vulgare Ga- spari Bavini seu Marrubium Mattioli, seu Prasium Of- ficinarum (53/127) – Valeriana dioica L. = Marrubium album vulgare Ga- spari Bavini seu Marrubium Mattioli, seu Prasium Of- ficinarum (53/127) – Valeriana phu L. Elenco delle specie presenti nell’Hortus Siccus Pisanus = Valeriana optima Cęsalpini, seu Phu maius Mattioli (53/128) – Valerianella dioica L. = Valerianella arvensis pręcox humilis semine compresso Morisoni, seu Locusta Her- ba Ioannis Bavini (59/139) (Fig. 15) – Verbascum sinuatum L. = Verbascum aliud Mattioli (11/21) – Verbascum thapsus L. s.l. = Verbascum primum Mattio- li (9/20) – Viburnum lantana L. = Viburnum Mattioli (49/120) – Viburnum opulus L. = Opulus Rovelli, seu Sambucu aquatica flore simplici Gaspari Bavini (27/69) – Viburnum tinus L. = Tinus prior Clusi (101/210) – Vinca major L. = Pervinca vulgaris angustifolia flore cęruleo Institutionum Rei Herbarię, seu Clematis pri- ma Mattioli (43/107) Fig. 15 Campione di Valerianella dioica L. (n. 139). – Vinca major L. = Pervinca angustifolia variegata Istitu- tionum Rei Herbarię (45/109)l – Vinca major L. = Pervinca vulgaris latifolia flore ceru- leo Istitutionum Rei Herbarię (45/110)l T – Teucrium chamaedrys L. s.l. = Camedrus maior repens Gaspari Bavini seu Trissago, seu Camedrus Mattioli (75/167) – Vinca major L. = Pervinca vulgaris latifolia flore albo Istitutionum Rei Herbarię (45/111) – Vinca minor L., forma a foglie screziate di bianco = Pervinca vulgaris angustifolia flore albo Institutionum Rei Herbarię (43/108) – Teucrium chamaedrys L. s.l. = Camedrus altera Mattio- li (73/168) – Teucrium chamaedrys L. s.l. = Camedrus altera Mattio- li (73/168) – Teucrium scordium L. s.l. = Camedrus palustris cane- scens seu Scordium Mattioli (75/169) – Vincetoxicum hirundinaria Medicus s.l. = Asclepias albo flore Gaspari Bavini seu Vincitoxicum Mattioli (79/170) – Teucrium scordium L. s.l. = Camedrus palustris cane- scens seu Scordium Mattioli (75/169) – Thalictrum aquilegifolium L. = Thalictrum alpinum Aquilegię foliis, forum staminibus purpurascentibus institutionum Rei Herbarię (139/267) – Thalictrum aquilegifolium L. = Thalictrum alpinum Aquilegię foliis, forum staminibus purpurascentibus institutionum Rei Herbarię (139/267) (79/170) – Vincetoxicum nigrum (L.) Moench = Ascle flore Caspar Bahuin (79/171) – Vincetoxicum nigrum (L.) Moench = Asclepias nigro flore Caspar Bahuin (79/171) Aquilegię foliis, forum staminibus purpurascentibus institutionum Rei Herbarię (139/267) – Vincetoxicum nigrum (L.) Moench = Asclepias nigro flore Caspar Bahuin (79/171) Appunti sulla Società Botanica di Cortona e su Mattia Moneti* Bruno Gialluca V ivida attestazione del forte interesse per la botanica nutrito dalla società colta di Cortona fin dai primi decenni del Settecento è il resoconto di Giovanni Targioni Tozzetti del sopralluogo effettuato a Cortona nell’au- tunno 1732 insieme a Pier Antonio Micheli, suo maestro1, che costituisce una testimo- nianza preziosa dei rapporti stretti della an- cora giovane Accademia Etrusca (nata nel 1728) con le altre istituzioni culturali della Toscana e, insieme, dello spazio che nella vita del giovane sodalizio cortonese avevano gli interessi naturalistici e scientifici. Il fine principale che aveva mosso Micheli da Fi- renze non era infatti quello di effettuare una ‘erborizzazione’ nel territorio di Cortona, che pure ebbe luogo, ma quello di riordinare la sezione naturalistica delle collezioni dell’Ac- cademia Etrusca: maturazione il «processo culturale che aveva decretato la lenta e inevitabile agonia delle Rariter- und Wunderkammern a favore di un collezionismo più specializzato e metodico»4. La vivace descrizione di Targioni Tozzetti, completa di riferimenti topografici precisi (il «terzo armario»), restituisce lo spessore ed il peso che nelle collezioni e nella libre- ria accademiche ebbero fin dall’inizio libri, materiali e strumenti delle scienze fisiche e naturali e racconta della presenza nell’Ac- cademia Etrusca di interessi che abbraccia- vano discipline diverse dall’archeologia, alle quali il sodalizio assegnava la stessa dignità della ricerca antiquaria: V Niente dirò della sontuosa libreria, piena di raris- simi ed utilissimi libri in ogni scienza, né di tanti istrumenti ottici e meccanici, che nel terzo armario si conservano. Niente pur dirò delle cose naturali che sono in detto museo, e nel disporre le quali fu occu- pato il sig. Micheli, perché si spera vederne un co- pioso ed esatto catalogo, annesso a quello del museo e libreria, negli atti della non mai abbastanza lodata Accademia Etrusca5. La mattina de’ 24, 25 e 26 ottobre fu occupato il sig. Micheli in disporre le cose naturali del Museo del- l’Accademia Etrusca, per il qual fine principalmente si era fatto questo viaggio2. * Abbreviazioni: B.C.A.E.C. Biblioteca del Comune e dell’Accademia Etrusca, Cortona; B.N.F. Biblioteca Nazionale Centrale, Firenze; B.C.A. Biblioteca Consortile di Arezzo. 1 Relazione d’un viaggio fatto in ricerca di cose naturali per il Valdarno di sopra fino a Cortona da Giovanni Targioni Tozzetti nel mese d’ottobre 1732 in compagnia del celebre Pier’Antonio Micheli suo maestro, e botanico di S.A.R. in: G. Targioni Tozzetti, Relazioni d’alcuni viaggi fatti in diverse parti della Toscana, per osservare le produzioni naturali, e gli antichi monumenti di essa dal dottor Giovanni Targioni Tozzetti, 6 voll., in Firenze nella Stamperia imperiale, 1751-1754, vol. 5, pp. 349-416. Ivi, p. 370. 2 Ibidem. 3 Il legato Baldelli venne acquisito dall’Accademia Etrusca (allora Accademia degli Occulti) nel 1727. 4 A. Tosi, «Biblioteche della natura». Collezioni naturalistiche nella Toscana del primo Settecento, «Annali della Scuola Normale Superiore di Pisa», Classe di Lettere e Filosofia, Serie III, Vol. XIX, 3, Pisa 1989. 5 Targioni Tozzetti, op. cit., p. 368. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Fig. 1 Frontespizio del 1° Volume dell’erbario Moneti 3 Il legato Baldelli venne acquisito dall’Accademia Etrusca (allora Accademia degli Occulti) nel 1727. 4 A. Tosi, «Biblioteche della natura». Collezioni naturalistiche nella Toscana del primo Settecento, «Annali della Scuola Normale Superiore di Pisa», Classe di Lettere e Filosofia, Serie III, Vol. XIX, 3, Pisa 1989. 5 Targioni Tozzetti, op. cit., p. 368. 6 Sul ruolo di Micheli nella fondazione della Società Botanica Fiorentina si veda P. Baccarini, Notizie intorno ad alcuni documenti della Società Botanica Fiorentina del 1716-1783 ed alle sue vicende, «Annali di Botanica», I, 1904, pp. 225-254. Al momento del sopralluogo a Cortona Pier Antonio Micheli aveva già pubblicato la sua opera principale: P.A. Micheli, Nova plantarum genera iuxta Tournefortii methodum disposita quibus plantae 1900 recensentur, scilicet fere 1400 nondum observatae, reliquae suis sedibus restitutae; quarum vero figuram exhibere visum fuit, eae ad 550 aeneis tabulis 108 graphice expressae sunt; adnotationibus, atque observationibus, praecipue fungorum, mucorum, affiniumque plantarum sationem, ortum, & incrementum spectantibus, interdum adiectis ... Auctore Petro Antonio Michelio Flor., Florentiae, typis Bernardi Paperinii, 1729. 7 Ridolfino dal 1730 si era trasferito a Roma e Filippo, il minore, all’epoca risiedeva a Cortona. 9 Di questo rapporto istituzionale resta traccia nelle stesse Relazioni di Giovanni Targioni Tozzetti. Nella prima edizione in sei tomi, comparsi a Firenze tra il 1751 e il 1754, ciascun tomo era dedicato a una importante accademia. Il quinto, comparso nel 1752, nel quale compare il resoconto del sopralluogo a Cortona, era appunto dedicato all’Accademia Etrusca: «All’inclita Accademia Etrusca di Cortona la quale indefessamente e con tanta sua gloria si occupa in togliere dall’oblio e mettere in luminosa vista gli antichi monumenti principalmente della Toscana Gio. Targioni Tozzetti consagra il Fig. 2 Uno degli indici presenti nell’erbario Moneti Appunti sulla Società Botanica di Cortona e su Mattia Moneti* È noto che il nucleo iniziale delle raccolte dell’Accademia Etrusca, il legato Baldelli3, messo insieme negli anni a Roma da Ono- frio Baldelli, comprese accanto a una se- zione archeologica e a una biblioteca anche una sezione scientifico-naturalistica succes- sivamente andata dispersa, espressione di un collezionismo ancora eclettico, nonostan- te già alla fine del Seicento fosse arrivato a L’Accademia Etrusca progettava dunque di descrivere le raccolte accademiche in un catalogo tripartito, uno per i materiali archeo- logici, uno per i naturalia, costituito dalla ca- talogazione del Micheli, uno per la libreria. I cataloghi così concepiti avrebbero dovuto tro- vare esito editoriale nei Saggi di dissertazioni accademiche pubblicamente lette nella nobile Accademia Etrusca dell’antichissima città di 54 Bruno Gialluca Fig. 2 Uno degli indici presenti nell’erbario Moneti Il resoconto di Giovanni Targioni Tozzetti mostra in particolare come in Filippo Venu- ti già nel 1732 gli interessi botanici fossero particolarmente vivi: proprio al Micheli e al Targioni Tozzetti Filippo Venuti espone- va il proposito di stampare nel primo tomo dei Saggi di Dissertazioni un catalogo delle piante spontanee del territorio da lui stesso realizzato10. Al progetto di Filippo Venuti non dovette essere estraneo il lavoro di descrizio- ne e di catalogazione della flora spontanea di Cortona, già da qualche anno avviato dal sa- cerdote Mattia Moneti, consegnato ai volumi manoscritti dell’Ars naturam fingens11 (Figg. 1 e 2) dei quali proprio nel 1732 il religio- so aveva completato il primo volume, come risulta dalla data apposta sul frontespizio. Proprio il rapporto avviato dal giovane Ve- nuti con Micheli costituì la ragione prossima di un avvicinamento del giovane cofondatore e figura di spicco dell’Accademia Etrusca a Mattia Moneti: l’ormai maturo religioso, forte dell’esperienza maturata, avrebbe realizzato i disegni delle piante via via inviate dal Ve- nuti a Micheli e ad altri studiosi. Il rapporto di collaborazione dovette sfociare anche in amicizia se, completato il secondo volume nel 173812, Moneti lo dedicava all’assai più giovane Filippo Venuti.i 8 L’Accademia Etrusca e la Società Botanica di Firenze avevano uno stesso presidente, Filippo Buonarroti. L’attività preparatoria effettuata da Antonio Cocchi e Marcello Venuti si desume dalle lettere di Marcello a Filippo conservate nella Biblioteca Marucelliana (N.A. 368.29). Antonio Cocchi era in contatto diretto anche con Filippo Venuti. Si vedano nelle sue Effemeridi, conservate nella Biblioteca Medica dell’Ospedale di Careggi, le seguenti annotazioni contenute nel quadernetto n. 13, riferite all’anno 1732: «7 ottobre, martedì. […] Lettera dell’abate Venuti.»; «Primo novembre, sabato. […] Il giorno venne da me Micheli ritornato da Cortona. Mi portò alcune cose naturali dono del sig. abate Venuti». Nello stesso quadernetto, nell’anno 1733 Cocchi registrava: «23 maggio, sabato. Lettere scritte: All’abate Venuti. Cortona». Lettere scritte: All’abate Venuti. Cortona». 14 Joseph Pitton Tournefort ... Institutiones rei herbariae ..., editio altera, gallica longe autior, quingentis circiter tabulis aeneis adornata, Parisiis e typographia Regia, 1700, 3 voll. Una seconda edizione in latino, pure in tre volumi, comparve sempre a Parig nel 1719. Fig. 3 Carta n. 32 raffigurante la Linaria lutea con riferimento a Tournefort (I.R.H.) 15 J. P. Tournefort, Elémens de botanique, ou Methode pour connaître les plantes. A Paris de l’Imprimerie royale, 1694. p 12 Anche il secondo volume dell’Ars naturam fingens reca l’anno del completamento annotato nel frontespizio. 13 Narciso Fabbrini, Miscellanea cortonese, B.C.A.E.C., ms. 708, ad vocem. 13 Narciso Fabbrini, Miscellanea cortonese, B.C.A.E.C., ms. 708, ad vocem. presente volume come attestato di sua ossequiosa riconoscenza per l’onor ricevuto di essere ammesso tra’ suoi socj e come saggio delle premure da lui usate in illustrare le antichità fisiche di essa Toscana». Degli altri tomi, il t. I (1751) era dedicato alla «celebratissima Società Bottanica Fiorentina»; il t. II (1751), alla «inclita Società Colombaria Fiorentina»; il t. III (1751), alla «Accademia Imperiale de’ Curiosi della Natura»; il t. IV (1752), alla «Accademia della Crusca»; il t. VI (1754), all’«Accademia Imperiale Petropolitana». presente volume come attestato di sua ossequiosa riconoscenza per l’onor ricevuto di essere ammesso tra’ suoi socj e come saggio delle premure da lui usate in illustrare le antichità fisiche di essa Toscana». Degli altri tomi, il t. I (1751) era dedicato alla «celebratissima Società Bottanica Fiorentina»; il t. II (1751), alla «inclita Società Colombaria Fiorentina»; il t. III (1751), alla «Accademia Imperiale de’ Curiosi della Natura»; il t. IV (1752), alla «Accademia della Crusca»; il t. VI (1754), all’«Accademia Imperiale Petropolitana». 10 Il l di Fili V i presente volume come attestato di sua ossequiosa riconoscenza per l’onor ricevuto di essere ammesso tra’ suoi socj e come saggio delle premure da lui usate in illustrare le antichità fisiche di essa Toscana». Degli altri tomi, il t. I (1751) era dedicato alla «celebratissima Società Bottanica Fiorentina»; il t. II (1751), alla «inclita Società Colombaria Fiorentina»; il t. III (1751), alla «Accademia Imperiale de’ Curiosi della Natura»; il t. IV (1752), alla «Accademia della Crusca»; il t. VI (1754), all’«Accademia Imperiale Petropolitana». 10 Il lavoro di Filippo Venuti, come detto poco sopra, non sarebbe comparso a stampa, né nei Saggi di dissertazioni accademiche né altrove. Il manoscritto (Filippo Venuti, Sylloge plantarum quae in agro Cortoniensi sponte nascuntur, tournefortiana methodo disposta), annotato da Lodovico Coltellini, è conservato in B.N.F., Fondo Targioni, n. 119. Del proposito di Filippo Venuti si ha seguente notizia: «La sera giungemmo alla villa di Catrosso, e vi ci trattenemmo tutta la mattina seguente, affinché il signor Micheli potesse finire di nominare le piante, che nei giorni passati si erano trovate intorno a Cortona, siccome ancora alcune altre, che molto prima il Signor Proposto Filippo Venuti aveva osservato nascere spontanee nella campagna di Cortona, e delle quali pensava farne un catalogo, e stamparlo nell’atti dell’Accademia Etrusca.». Targioni Tozzetti, Relazione d’un viaggio, cit., p. 371. 11 Mattia Moneti, Ars naturam fingens seu nonnullarum plantarum ac stirpium quae in agro cortoniensi sponte nascuntur catalogus… (Vol. I, 1732, cc. 143, tavv. 123; Vol. II, 1738, cc. 95, tavv. 95; Vol. III, 1755, cc. 179, tavv. 84), Cortona, B.C.A.E.C., Mss. 400, 401, 402. Su Mattia Moneti si veda A. Tosi, Mattia Moneti, in L’Accademia Etrusca, Cortona, Palazzo Casali, mostra maggio-ottobre 1985, catalogo a cura di P. Barocchi e D. Gallo, Milano, Electa, 1985, pp. 199-200, con bibliografia precedente. 12 Anche il secondo volume dell’Ars naturam fingens reca l’anno del completamento Cortona. Le cose poi andarono diversamente, poiché fu presto chiaro che i Saggi di Disser- tazioni Accademiche, il cui primo tomo ebbe una genesi travagliata, non erano la sede giu- sta e i cataloghi non vi comparvero. All’epo- ca però, siamo nel 1732, le idee sui Saggi di Dissertazioni Accademiche non erano ancora chiarissime: il primo tomo sarebbe venuto alla luce solo nel dicembre avanzato del 1735. L’Ars naturam fingens costituisce il pro- dotto più maturo dell’attività botanica a Cor- tona nella prima metà del XVIII secolo e converrà pertanto spendere qualche parola su di essa e sul suo autore. Le poche notizie sulla vita del sacerdote Mattia Moneti (Corto- na, 1687-1758) si desumono dalla omonima voce contenuta nella Miscellanea cortonese: D’altra parte Pier Antonio Micheli6, pro- tagonista insieme a Filippo Buonarroti della fondazione, nel 1717, della Società Botanica Fiorentina, certo non effettuò il sopralluogo a Cortona a titolo personale. La visita del Mi- cheli e del Targioni Tozzetti a Cortona, dove soggiornarono nella bella villa suburbana di Catrosse di proprietà dei fratelli Marcello, Ridolfino e Filippo Venuti, i tre principali fondatori dell’Accademia Etrusca, era stata preparata da Marcello Venuti (il maggio- re dei fratelli, allora a Firenze)7 ed Antonio Cocchi, all’epoca segretario della Società Botanica di Firenze8, nel quadro di quello che sembra essere una collaborazione istitu- zionale tra l’Accademia Etrusca e la Società Botanica Fiorentina9, attivamente impegna- ta a convogliare i forti interessi scientifici e naturalistici presenti non solo a Firenze ma anche in centri minori. Mattia Vincenzo Moneti nacque in Cortona il 22 gen- naio 1687, da Antonio di Mattia Moneti e donna Eli- sabetta di Pier Antonio Santini, ambedue di famiglia popolare e civile di Cortona. Compiti i studj di uma- nità e teologia morale, si ordinò sacerdote nel 1711. Visse con fama di ottimo ecclesiastico. Amante della vita campestre e portato dal genio allo studio della bo- tanica e cognizione dei semplici, vacando (…) la par- rocchia di S. Angelo del Succhio in villa di Mitigliano, l’anno 1735 gli fu conferita dal Capitolo della Catte- drale, dove concluse la sua vita il 3 dicembre175813. Le succinte notizie biografiche insistono sulla sua passione per la botanica, certamen- te coltivata e praticata assai precocemente, 55 Appunti sulla Società Botanica di Cortona e su Mattia Moneti presente volume come attestato di sua ossequiosa riconoscenza per l’onor ricevuto di essere ammesso tra’ suoi socj e come saggio delle premure da lui usate in illustrare le antichità fisiche di essa Toscana». Degli altri tomi, il t. I (1751) era dedicato alla «celebratissima Società Bottanica Fiorentina»; il t. II (1751), alla «inclita Società Colombaria Fiorentina»; il t. III (1751), alla «Accademia Imperiale de’ Curiosi della Natura»; il t. IV (1752), alla «Accademia della Crusca»; il t. VI (1754), all’«Accademia Imperiale Petropolitana». almeno dai tempi della ordinazione sacerdo- tale, nulla però sappiamo dei suoi studi e del modo in cui venne in contatto con la classifi- cazione delle piante introdotta dal naturali- sta francese Joseph Pitton de Tournefort. pianta riprodotta in una tavola acquerellata era apposto in calce il nome volgare, quel- lo attribuito dal Tournefort e, assai spesso, l’acronimo «IRH» seguito da un numero car- dinale (Fig. 3). L’acronimo sta per Institutio- nes rei herbariae14, la traduzione latina degli Elémens de botanique15, l’opera principale L’Ars naturam fingens costituisce un cata- logo della flora spontanea di Cortona: a ogni 56 Bruno Gialluca studi botanici. In data 4 novembre 1737 l’aba- te cortonese annotava i seguenti pensieri: Nel sopraddetto libro alla pag. 28119 vi è un esame di diversi monumenti su quali vi è delle piante, che gl’antiquarij confondono col Loto d’Egitto. È certo che l’istoria antica ricerca il soccorso della bottani- ca. Monsieur Mahudel, che intendeva l’una, e l’altra rigetta l’opinion di coloro, che credono doversi con- sultare Teofrasto, Plinio, Dioscoride per spiegar le piante, che si vedono ne’ monumenti antichi, perché le loro descrizioni essendo imperfette, di cose da loro non vedute, non vi si possono applicare. Cinque sorte di piante sono state sin qui confuse, ed incognite nelle cose egizie, Isidi, Serapidi, Arpocrati etc. Il loto. La Faba egiziana; la Colocasia; la Persia e la Musa: queste possono dirsi piante teologiche fra gl’Egizzij, perché oltre l’esser nutritizie, erano ancora misteriose nella religione. Teofrasto ed Esischio convengono che molte diverse piante avevano il nome di loto. I moderni scrittori, e viaggiatori, non trovando erba più comune intorno al Nilo della ninfea, hanno detto esser quella il vero loto degl’antichi. 20 Ibn al-Baytār (Malaga 1197 - Damasco 1248), conosciuto in Occidente come Embitar, medico, farmacologo e botanico spagnolo, di lingua araba. Scrisse in greco un trattato De limonibus, che ebbe una versione latina di Andrea Alpago, bellunese, pubblicata per la prima volta a Venezia nel 1583 (Ebenbitar arabs De limonibus, iamdiu per Andraeam Bellunensem in Latinum translatus, nunc primum in lucem editus, Venetiis, apud Oratium de Gobbis, 1583) e successivamente a Parigi nel 1602 (De limonibus, tractatus Embitar arabis, per Andream Bellunensem latinitate donatus, Parisiis apud Petrum Cheualier impensis Gasparis Bindonij bibliopolae Bononiensis, 1602). Così Abanditar20, Prospero Alpi- no21, m.r Lippi22, e l’autor dell’Orto malabarico23, che ne riporta la figura, che si ritrova molto conforme alle parti, che ce ne danno i marmi e medaglie. Il loto fu consacrato al sole e però posto sopra Osiride e gl’altri dei e sacerdoti egiziani. I re loro che affettavano di- vinità se ne formarono delle corone; e nelle monete si vede ora col gambo, col fiore, e col fructo. La Fava d’Egitto è la seconda pianta che si trova ne’ monumenti antichi. I moderni botanici la pongono sotto altra specie di Ninfea, Erodoto (lib. 2, cap 32) parla di una specie di giglio bianco ed altro color di rosa, che nasce frequente intorno al Nilo. È di co- lor rosso, dice Ateneo che era il fior presentato ad Adriano col nome di Loto d’Antinoo. Nelle medaglie alle volte, e nelle gemme serve di sedile a un putto: Plutarco in Iside dice, che tale espressione significa il crepuscolo. di Tournefort, e il numero designa la tavo- la tournefortiana corrispondente alla ripro- duzione del Moneti. L’adozione del modello sistematico elaborato da Tournefort16 e le tavole, assai essenziali e immediate (Moneti delineava con un contorno deciso a inchio- stro ciascuna pianta riprodotta e la colorava a tempera in maniera sommaria), collocano pienamente l’opera nel primo Settecento, di- staccandola nettamente dalla iconografia e sistematica botanica cinque-seicentesca17. La Colocasia ha il fiore simile all’Aro del quale si crede una specie da Fab(rizio) Colonna; delle radi- che ne facevano que’ popoli pane. Questo fiore fatto a foggia d’orecchia sta spesso in capo d’Arpocrate, e de segni pantei. La Persea è arboscello che nasce nelle vicinanze del Cairo; ha le foglie simili all’alloro, e il frutto alla pera. È sempreverde, ed aromatico; ed ha le foglie fatte a foggia di lingua. Il frutto contiene dentro una mandorla; e questo si vede in capo ad Iside, alle vol- te intero, alle volte smezzato per farne vedere il di dentro. L’interesse di Filippo Venuti per la bota- nica, pure propiziato dalla frequentazione e dall’esempio del Moneti, nasceva senza dub- bio soprattutto da forti motivazioni personali ed era strettamente connesso con i suoi stu- di antiquari. C’è un passo del diario da lui tenuto dal novembre 1737 all’agosto 173818, rivelatore della connessione strettissima che Filippo Venuti istituiva tra studi antiquari e La Musa. Fig. 4 Tavola a c. 7r. del primo volume della Miscellanea erudita e amena di Filippo Venuti (B.C.A.E., ms. 446, 2 voll.). 16 Aix in Provence, 5 giugno 1656 - Parigi, 28 dicembre 1708. 17 Tosi, op. cit., p. 200. 18 Filippo Venuti, Miscellanea erudita e amena, B.C.A., ms. 446. 19 Histoire de l’Académie Royale des Inscriptions & belles lettres, t. II, Paris 1736. 16 Aix in Provence, 5 giugno 1656 - Parigi, 28 dicembre 1708. 17 Tosi, op. cit., p. 200. 18 Filippo Venuti, Miscellanea erudita e amena, B.C.A., ms. 446. 19 Histoire de l’Académie Royale des Inscriptions & belles lettres, t. II, Paris 1736. 20 Ibn al-Baytār (Malaga 1197 - Damasco 1248), conosciuto in Occidente come Embitar, medico, farmacologo e botanico spagnolo, di lingua araba. Scrisse in greco un trattato De limonibus, che ebbe una versione latina di Andrea Alpago, bellunese, pubblicata per la prima volta a Venezia nel 1583 (Ebenbitar arabs De limonibus, iamdiu per Andraeam Bellunensem in Latinum translatus, nunc primum in lucem editus, Venetiis, apud Oratium de Gobbis, 1583) e successivamente a Parigi nel 1602 (De limonibus, tractatus Embitar arabis, per Andream Bellunensem latinitate donatus, Parisiis apud Petrum Cheualier impensis Gasparis Bindonij bibliopolae Bononiensis, 1602). Per ciascun associato era prevista una patente, recante dentro una cornice floreale un giardino, una fontana e la scritta Flores mei fructus honoris et ho- nestatis (Fig. 5). Il contenuto della patente chiarisce assai bene i propositi della Società Botanica di Cortona: una vivace testimonianza proprio da parte di Tommaso Coltellini, che scriveva a Livorno a Filippo Venuti28 pregandolo di voler invia- re qualche pianta dell’orto botanico di Pisa «per formare questo nostro orticciuolo», poi- ché erano stati smarriti i semi che Filippo aveva procurato. Lo informava inoltre del fat- to che «la nostra società si va giornalmente aumentando e ultimamente si ascrisse alla medesima il signor Saverio Manetti custode del giardino imperiale di Firenze» che aveva promesso «di farci parte a suo tempo delle piante più rare». Per marzo quindi si pensa- va di «far l’apertura del medesimo orto e di tenerci in tal congiuntura una specie di acca- demia letteraria, leggendovi cioè alcune dis- sertazioni sopra cui attualmente si lavora da noi». Lo statuto, approvato il 15 luglio 175429, rende immediatamente conto delle finalità: oltre alle consuete figure di presidente, se- gretario, tesoriere ecc., lo statuto prevedeva un lettore: «si deputerà uno dei soci, il qua- le dovrà leggere e spiegare agli altri soci le istituzioni botaniche nel nostro giardino in quelle stagioni che lo permetteranno». Nel- la Società Botanica nata con il fine di inse- gnare a conoscere le piante medicinali e di controllarne la raccolta, l’altra figura chiave era il custode dell’orto, il cui compito sarebbe stato quello di «tener fornito di piante l’orto botanico» provvedendo a «tenere un catalo- go esatto di tutte le medesime piante e de- scriverle opportunamente». L’attività sociale principale era la lettura di una dissertazione, una volta al mese, nell’orto botanico o in un posto riparato durante la cattiva stagione. La Società Botanica aveva un’insegna, un sigillo raffigurante una palma intorno alla quale era scritto In edito stat admirabilis. Botanophi- les Cortonae institutis. Per ciascun associato era prevista una patente, recante dentro una cornice floreale un giardino, una fontana e la scritta Flores mei fructus honoris et ho- nestatis (Fig. 5). 25 Sulla Società Botanica di Cortona si veda L. Tongiorgi Tomasi, A. Tosi, Ars naturam fingens. L’Accademia Etrusca di Cortona tra interessi scientifici, reperti da collezione, illustrazioni naturalistiche e giardini, in L’Accademia Etrusca, cit., pp. 190- 199, con bibliografia precedente. 26 «Novelle Letterarie», num. 44, Firenze, 4 novembre 1757, coll. 691-697. 27 Tommaso Coltellini, abate ancora nel 1755, quando lesse l’elogio di Marcello Venuti, successivamente avrebbe esercitato il notariato. Nell’Archivio di Stato di Firenze sono conservate cinque filze di suoi protocolli notarili (A.S.F., Notarile Moderno, nn. 29279- 29283), che coprono gli anni 1769-1798. Sposò una Caterina Fabbrini, avendone un figlio, Agostino. Tommaso fu a lungo cancelliere della Curia vescovile. Il 2 dicembre 1798, ammalato,fece il suo ultimo testamento (A.S.F., Notarile Moderno, notaio Vincenzo Lupi, n. 28253, cc. 41-42) e poco dopo morì. Per Agostino Coltellini, vedi C. Nepi in questo stesso volume. 28 B.C.A.E.C., ms. 572, c. 28, lettera del 26 giugno 1754. 29 B C A E C ms 602 c 74 sgg 24 La tavola forse è stata realizzata da Marco Tuscher, al quale si devono molti dei disegni del primo tomo della Miscellanea erudita e amena, come si apprende dallo stesso Filippo Venuti. 22 Auguste Lippi (1678-1704), medico e naturalista francese, viaggiò a lungo in Egitto e morì assai giovane, assassinato in Etiopia. 23 Hendrik Adriaan Van Reede Tot Drakenstein, Hortus Indicus Malabaricus, continens regni Malabarici apud Indos celeberrimi omnis generis plantas rariores, Latinis, Malabaricis, Arabicis, & Bramanum characteribus nominibusque expressas, una cum floribus, fructibus & seminibus, naturali magnitudine a peritissimis pictoribus delineatas, & ad vivum exhibitas. Addita insuper accurata earundem descriptione, ... Adornatus per Henricum van Rheede, van Draakenstein, ... et Johannem Casearium, ... Notis adauxit, & commentariis illustravit Arnoldus Syen, Amstelodami, sumptibus Joannis van Someren, et Joannis van Dyck, 1678-1703, 12 voll. Dal vol. 2 a cura di Johannes Commelin. È pianta pelusiaca, che ha una canna lunga con foglie larghe, e lunghe, e ottuse simili alla palma, delle quali hanno in capo alcuna le statue, e figure egizie. Il suo frutto si mangia, fatto a foggia di piccoli cocomerini dorati, con scorza aromatica e polpa di mele. Si vede ornata Iside nel corpo di questi frutti. Darò nell’annessa le figure di queste piante, che ri- porta il detto accademico alla meglio. Appunti sulla Società Botanica di Cortona e su Mattia Moneti 57 La botanica diventava così uno strumento prezioso per lo studio degli antichi monumen- ti, come mostra la tavola24 (Fig. 4) associata nel manoscritto al testo di sopra, efficace il- lustrazione del pensiero di Filippo Venuti. una vivace testimonianza proprio da parte di Tommaso Coltellini, che scriveva a Livorno a Filippo Venuti28 pregandolo di voler invia- re qualche pianta dell’orto botanico di Pisa «per formare questo nostro orticciuolo», poi- ché erano stati smarriti i semi che Filippo aveva procurato. Lo informava inoltre del fat- to che «la nostra società si va giornalmente aumentando e ultimamente si ascrisse alla medesima il signor Saverio Manetti custode del giardino imperiale di Firenze» che aveva promesso «di farci parte a suo tempo delle piante più rare». Per marzo quindi si pensa- va di «far l’apertura del medesimo orto e di tenerci in tal congiuntura una specie di acca- demia letteraria, leggendovi cioè alcune dis- sertazioni sopra cui attualmente si lavora da noi». Lo statuto, approvato il 15 luglio 175429, rende immediatamente conto delle finalità: oltre alle consuete figure di presidente, se- gretario, tesoriere ecc., lo statuto prevedeva un lettore: «si deputerà uno dei soci, il qua- le dovrà leggere e spiegare agli altri soci le istituzioni botaniche nel nostro giardino in quelle stagioni che lo permetteranno». Nel- la Società Botanica nata con il fine di inse- gnare a conoscere le piante medicinali e di controllarne la raccolta, l’altra figura chiave era il custode dell’orto, il cui compito sarebbe stato quello di «tener fornito di piante l’orto botanico» provvedendo a «tenere un catalo- go esatto di tutte le medesime piante e de- scriverle opportunamente». L’attività sociale principale era la lettura di una dissertazione, una volta al mese, nell’orto botanico o in un posto riparato durante la cattiva stagione. La Società Botanica aveva un’insegna, un sigillo raffigurante una palma intorno alla quale era scritto In edito stat admirabilis. Botanophi- les Cortonae institutis. 21 Prospero Alpini (Marostica, 23 novembre 1553 - Padova, 6 febbraio 1617), medico, botanico e scienziato italiano. Filippo Venuti fa qui riferimento all’opera Prosperi Alpini De plantis Aegypti liber. In quo non pauci, qui circa herbarum materiam irrepserunt, errores, deprehenduntur, qourum causa hactenus multa medicamenta ad vsum medicine admodum expetenda, plerisque medicorum, non sine artis iactura, occulta, atque obsoleta iacuerunt. ... Accessit etiam liber de balsamo alias editus, Venetiis: apud Franciscum de Franciscis Senensem, 1592. 22 Auguste Lippi (1678-1704), medico e naturalista francese, viaggiò a lungo in Egitto e morì assai giovane, assassinato in Etiopia. 23 Hendrik Adriaan Van Reede Tot Drakenstein, Hortus Indicus Malabaricus, continens regni Malabarici apud Indos celeberrimi omnis generis plantas rariores, Latinis, Malabaricis, Arabicis, & Bramanum characteribus nominibusque expressas, una cum floribus, fructibus & seminibus, naturali magnitudine a peritissimis pictoribus delineatas, & ad vivum exhibitas. Addita insuper accurata earundem descriptione, ... Adornatus per Henricum van Rheede, van Draakenstein, ... et Johannem Casearium, ... Notis adauxit, & commentariis illustravit Arnoldus Syen, Amstelodami, sumptibus Joannis van Someren, et Joannis van Dyck, 1678-1703, 12 voll. Dal vol. 2 a cura di Johannes Commelin. 24 La tavola forse è stata realizzata da Marco Tuscher, al quale si devono molti dei disegni del primo tomo della Miscellanea erudita e amena, come si apprende dallo stesso Filippo Venuti. 25 Sulla Società Botanica di Cortona si veda L. Tongiorgi Tomasi, A. Tosi, Ars naturam fingens. L’Accademia Etrusca di Cortona tra interessi scientifici, reperti da collezione, illustrazioni naturalistiche e giardini, in L’Accademia Etrusca, cit., pp. 190- 199, con bibliografia precedente. 26 «Novelle Letterarie», num. 44, Firenze, 4 novembre 1757, coll. 691-697. 27 Tommaso Coltellini, abate ancora nel 1755, quando lesse l’elogio di Marcello Venuti, successivamente avrebbe esercitato il notariato. Nell’Archivio di Stato di Firenze sono conservate cinque filze di suoi protocolli notarili (A.S.F., Notarile Moderno, nn. 29279- 29283), che coprono gli anni 1769-1798. Sposò una Caterina Fabbrini, avendone un figlio, Agostino. Tommaso fu a lungo cancelliere della Curia vescovile. Il 2 dicembre 1798, ammalato,fece 28 B.C.A.E.C., ms. 572, c. 28, lettera del 26 giugno 1754. 29 B.C.A.E.C., ms. 602, c. 74 sgg. 28 B.C.A.E.C., ms. 572, c. 28, lettera del 26 giugno 1754. 21 Prospero Alpini (Marostica, 23 novembre 1553 - Padova, 6 febbraio 1617), medico, botanico e scienziato italiano. Filippo Venuti fa qui riferimento all’opera Prosperi Alpini De plantis Aegypti liber. In quo non pauci, qui circa herbarum materiam irrepserunt, errores, deprehenduntur, qourum causa hactenus multa medicamenta ad vsum medicine admodum expetenda, plerisque medicorum, non sine artis iactura, occulta, atque obsoleta iacuerunt. ... Accessit etiam liber de balsamo alias editus, Venetiis: apud Franciscum de Franciscis Senensem, 1592. 22 Auguste Lippi (1678-1704), medico e naturalista francese, viaggiò a lungo in Egitto e morì assai giovane, assassinato in Etiopia. 23 Hendrik Adriaan Van Reede Tot Drakenstein, Hortus Indicus Malabaricus, continens regni Malabarici apud Indos celeberrimi omnis generis plantas rariores, Latinis, Malabaricis, Arabicis, & Bramanum characteribus nominibusque expressas, una cum floribus, fructibus & seminibus, naturali magnitudine a peritissimis pictoribus delineatas, & ad vivum exhibitas. Addita insuper accurata earundem descriptione, ... Adornatus per Henricum van Rheede, van Draakenstein, ... et Johannem Casearium, ... Notis adauxit, & commentariis illustravit Arnoldus Syen, Amstelodami, sumptibus Joannis van Someren, et Joannis van Dyck, 1678-1703, 12 voll. Dal vol. 2 a cura di Johannes Commelin. 24 La tavola forse è stata realizzata da Marco Tuscher, al quale si devono molti dei disegni del primo tomo della Miscellanea erudita e amena, come si 21 Prospero Alpini (Marostica, 23 novembre 1553 - Padova, 6 febbraio 1617), medico, botanico e scienziato italiano. Filippo Venuti fa qui riferimento all’opera Prosperi Alpini De plantis Aegypti liber. In quo non pauci, qui circa herbarum materiam irrepserunt, errores, deprehenduntur, qourum causa hactenus multa medicamenta ad vsum medicine admodum expetenda, plerisque medicorum, non sine artis iactura, occulta, atque obsoleta iacuerunt. ... Accessit etiam liber de balsamo alias editus, Venetiis: apud Franciscum de Franciscis Senensem, 1592. Il contenuto della patente chiarisce assai bene i propositi della Società Botanica di Cortona: Nell’instituire che facemmo, la Società Botanica, nella La partenza da Cortona di Filippo Venuti per Clerac, nell’estate 1738, non interrompe- va la ricerca e l’interesse per la botanica in sede locale: l’Accademia Etrusca all’inter- no della sua attività culturale continuava ad alimentare anche un vivace filone botanico. Esito ultimo di interessi naturalistici vivi e presenti nella società colta di Cortona fu in- fatti l’Accademia o Società Botanica e d’Isto- ria Naturale di Cortona25, nata il 30 giugno 1754, con ogni probabilità anche sull’impulso della nascita, l’anno precedente, dell’Accade- mia dei Georgofili, prima società pubblica in Europa di studi agrari, a sua volta nata dalla necessità di migliorare la produzione agrico- la attraverso un uso razionale del suolo. Prendiamo come data ufficiale di nascita l’approvazione dello statuto, tuttavia già dal 1753 era stata avviata a Cortona una intensa attività preparatoria propedeutica alla costi- tuzione formale della Società Botanica, alla quale partecipò anche il Moneti, socio del- l’Accademia dei Georgofili. Tra i primissimi animatori, o se non addi- rittura il primo, Tommaso Coltellini, fratello del più noto Lodovico, come lui traduttore dal francese: Lodovico Coltellini il 4 ottobre 1757 inviava una lettera a Giovanni Lami, che quest’ultimo provvide a pubblicare in- tegralmente sulle Novelle Letterarie26, nella quale lo informava dei progressi della Socie- tà Botanica di Cortona, della quale ricordava gli esordi nel 1753: Nel mese di settembre dell’anno 1753 fu da alcuni eruditi soggetti, tanto ecclesiastici che secolari di questa patria, ad insinuazione di Tommaso mio fra- tello, ideato, e risoluto d’instituire in Cortona l’Ac- cademia suddetta, che avesse per iscopo lo studio dell’erbe, e delle piante, e delle cose naturali. A tale effetto si ridussero quei Sigg. in corpo ed elessero un giardino, che in appresso riempirono di piante, ed erbe tanto indigene, che forestiere. Furono, in ap- presso, fissate le leggi della nascente Società27. Nell’instituire che facemmo, la Società Botanica, nella città nostra, abbiamo specialmente avuto in mira tre oggetti, vale a dire, lo studio delle piante, l’istoria na- turale, ed alcuna parte dell’agricoltura, acciocché il dilettevole coll’utile, convenientemente unito ne fosse. Della intensa attività preparatoria che pre- cedette la nascita ufficiale dell’Accademia, volta ad apprestare l’orto botanico, abbiamo 58 Bruno Gialluca 30 Tongiorgi Tomasi, Tosi, op. cit., p. 197. 31 Il resoconto della cerimonia registra che «[…] pubblicamente riscosse degli applausi il medesimo donatore [Moneti], per averci ora regalato questo terzo volume delle sue opere botaniche, quali contengono le piante già colorite al naturale, co’ loro fiori, disegnate e dipinte dal medesimo, che si trovano nell’agro cortonese. Nella disposizione di tali piante si serve del Metodo di Tournefort, per uniformarsi a’ più celebri botanici di questo secolo». (B.C.A.E.C. ms. 450, Atti, cc. 212 segg.). 32 Il volume andò successivamente disperso. Fig. 5 Immagine della patente della Società Botanica Cortonese (foto di G. Poccetti - Fotomaster). 30 Tongiorgi Tomasi, Tosi, op. cit., p. 197. Fu la grande stagione di Lodo- vico Coltellini, figura sempre più importan- te e rappresentativa della Società Botanica, che curava personalmente le associazioni, le pubbliche relazioni e la promozione della giovane società, presente nel dibattito scien- tifico, grazie anche allo spazio del quale po- teva disporre nelle Novelle Letterarie, alle quali aveva agevole accesso, per il lungo so- dalizio con Giovanni Lami, anche lui asso- ciato al sodalizio cortonese. Nonostante il fervore dell’attività promozionale del Coltel- lini, a dieci anni dalla fondazione la Società Botanica non ancora aveva pubblicato una raccolta di atti, segno di una inadeguatezza della produzione scientifica. Nel 1764, ri- spondendo a una lettera di Francesco Gri- selini (collaboratore dell’Accademia dei Georgofili, fresco autore presso Bonducci della Nuova maniera di seminare e coltivare il grano), che chiedeva se la Società Botani- ca avesse dato alle stampe qualche volume di dissertazioni, Coltellini ammetteva «che non per anco quest’Accademia Bottanica e d’Istoria naturale di Cortona ha dato qual- che raccolta d’atti alla luce. Veramente tra gli altri disegni e propositi vi è ancor que- sto, ma prima di venire alla effettuazione si reputa necessario e tempo, e buon criterio nella scelta»39. Anche in seguito, la Società Botanica di Cortona non avrebbe mai pub- blicato un contributo a stampa e già negli anni Settanta si perdono le testimonianze della sua attività. Il sodalizio soccombe- va, oltre che per debolezze proprie, anche per la mutata situazione generale: la poli- tica culturale di Pietro Leopoldo accordava notavano le Notti Coritane, nel verbale della riunione della sera del 10 novembre: «Si fa ricordo in questa sera qualmente nella nostra città, ove si è aumentato il buon gusto della erudizione e delle scienze, si è ultimamente stabilita una società per lo studio dei sempli- ci e della botanica il quale studio può facil- mente servire di strada all’universale della storia naturale»33. Poco dopo, nel 1755, l’orto botanico della società fu ufficialmente aperto al pubblico con una grande cerimonia. Per l’occasione venne stampato un sonetto, che la Società dedicava a Filippo Venuti, cele- brato sia come Presidente di essa che come prima dignità della chiesa livornese.34 Nel- l’autunno dello stesso anno moriva Marcello Venuti, perdita gravissima per la cultura e la società cittadine. Molte furono le cerimo- nie di commemorazione tenute in città, la più imponente e spettacolare ebbe luogo proprio nel giardino della Società Botanica e di essa, oltre alla descrizione necessariamente più compendiaria inserita nei Pietosi officij35, ci resta un minuzioso resoconto inviato per lettera a Filippo Venuti, che non aveva par- tecipato alle esequie poiché, proposto della cattedrale di Livorno, non aveva potuto la- sciare la città labronica. redatto nel 1758 testimonia di una intensa e riuscita campagna volta ad ascrivere alla neonata Società Botanica nomi di grande prestigio: Laura Bassi (Laura Maria Cate- rina Bassi Veratri, 1711-1778, professore di fisica e matematica all’università di Pado- va, la prima donna in Italia titolare di una cattedra universitaria), Francesco Zanotti dell’Accademia delle Scienze di Bologna, Giovanni Bianchi (Iano Planco) di Rimini, Antonio Vallisnieri, professore di storia na- turale a Padova, Angelo Attilio Tilli prefet- to del giardino botanico pisano, Antonio e Raimondo Cocchi, Saverio Manetti, Ubaldo Montelatici38. Fu la grande stagione di Lodo- vico Coltellini, figura sempre più importan- te e rappresentativa della Società Botanica, che curava personalmente le associazioni, le pubbliche relazioni e la promozione della giovane società, presente nel dibattito scien- tifico, grazie anche allo spazio del quale po- teva disporre nelle Novelle Letterarie, alle quali aveva agevole accesso, per il lungo so- dalizio con Giovanni Lami, anche lui asso- ciato al sodalizio cortonese. Nonostante il fervore dell’attività promozionale del Coltel- lini, a dieci anni dalla fondazione la Società Botanica non ancora aveva pubblicato una raccolta di atti, segno di una inadeguatezza della produzione scientifica. Nel 1764, ri- spondendo a una lettera di Francesco Gri- selini (collaboratore dell’Accademia dei Georgofili, fresco autore presso Bonducci della Nuova maniera di seminare e coltivare il grano), che chiedeva se la Società Botani- ca avesse dato alle stampe qualche volume di dissertazioni, Coltellini ammetteva «che non per anco quest’Accademia Bottanica e d’Istoria naturale di Cortona ha dato qual- che raccolta d’atti alla luce. Veramente tra gli altri disegni e propositi vi è ancor que- sto, ma prima di venire alla effettuazione si reputa necessario e tempo, e buon criterio nella scelta»39. 33 B.C.A.E.C., ms. 443, c. 240. 34 Tongiorgi, Tosi, op. cit., p. 197. 35 O. Maccari, Pietosi officii prestati a Cortona alla memoria del Marchese Cavalier Marcello de’ Venuti, Livorno 1755, p. 37. 36 Piis · Manibus · Nic. · Marcelli · Venuti · (...) · Eclogium · funebre · lapidarium · quod · propositum · est · arbitratu · conlegi · botanophilorum · CORTONENSIUM · QVVM · PVBLICA · EIDEM · PARENTALIA · APUD · URBEM · SUAM · CELEBRARENT · V · KAL. · SEPTEMBR. · ANNO · AB · CHRISTO · N · M·CC·LV·, Florentiae, ex Typographia Ioannelliana s.d. Il resoconto si trova in una lettera di Carlo Antonioli, lettore nell’Università di Pisa e Gian Filippo Malevoli, ambedue Scolopi, inviata da Cortona a Filippo Venuti il 2 settembre 1755 (B.C.A., ms 447, vol. II, pp. 255 e segg.). 37 Tongiorgi Tomasi, Tosi, op. cit., p. 198. 38 Per le associazioni alla Società Botanica di Cortona si veda Tongiorgi Tomasi, Tosi, op. cit., p. 198. 39 F. Griselini, Nuova maniera di seminare e coltivare il grano opera utilissima rivista corretta ed arricchita di nuove tavole e d’altri discorsi ed esperimenti fatti da vari Accademici Georgofili di Firenze, Firenze nella stamperia di Andrea Bonducci, 1764. 39 F. Griselini, Nuova maniera di seminare e coltivare il grano opera utilissima rivista corretta ed arricchita di nuove tavole e d’altri discorsi ed esperimenti fatti da vari Accademici Georgofili di Firenze, Firenze nella stamperia di Andrea Bonducci, 1764. 38 Per le associazioni alla Società Botanica di Cortona si veda Tongiorgi Tomasi, Tosi, op. cit., p. 198. 37 Tongiorgi Tomasi, Tosi, op. cit., p. 198. lute, avrebbe potuto portare e costituì una implicita valutazione dello spessore della sua opera, che, superata la metà del secolo, dovette apparire modesta, sia in riferimento al valore scientifico, sia alla resa artistica, che non superava un onesto livello documen- tario30. Nello stesso anno un altro importan- te riconoscimento premiava Mattia Moneti: nella seduta del 14 aprile 1755 l’Accademia Etrusca, alla quale aveva donato nel 1751 i primi due volumi della Ars naturam fin- gens, lo nominava accademico. Poco dopo, il 31 luglio 1755, Moneti, nel corso di una pubblica cerimonia31, presentava e donava all’Accademia anche il terzo volume, appena terminato, e l’anno successivo, completato il quarto volume, lo donava ancora all’Accade- mia Etrusca32. La morte, nel 1758, poneva fine alla sua attività. La società Botanica di Cortona nasceva dall’esigenza di affrontare in un’ottica ampia il mondo naturale, con un riferimento al- l’agricoltura, che sembra in qualche misura recepire quelle esigenze che avevano dato vita nel 1753 all’Accademia dei Georgofili. Primo presidente fu Filippo Venuti, che, tor- nato dalla Francia, era ormai ormai propo- sto a Livorno; direttore, un incarico del tutto onorario, Mattia Moneti, premiato in questo modo per la sua lunga e continuativa atti- vità come illustratore e catalogatore della flora spontanea del territorio, alla quale ave- va continuato ad attendere in silenzio e in posizione appartata con il terzo e il quarto volume dell’Ars naturam fingens (completati, rispettivamente, nel 1755 e nel 1756). L’in- carico conferito all’ormai anziano botanico, solo onorifico e sostanzialmente marginale, fu la conseguenza di una realistica valuta- zione del contributo che il sacerdote, in età avanzata e in non buone condizioni di sa- Verso la fine del 1754 l’Accademia o Società Botanica era stata presentata uffi- cialmente all’Accademia Etrusca, come an- Appunti sulla Società Botanica di Cortona e su Mattia Moneti 59 Appunti sulla Società Botanica di Corton redatto nel 1758 testimonia di una intensa e riuscita campagna volta ad ascrivere alla neonata Società Botanica nomi di grande prestigio: Laura Bassi (Laura Maria Cate- rina Bassi Veratri, 1711-1778, professore di fisica e matematica all’università di Pado- va, la prima donna in Italia titolare di una cattedra universitaria), Francesco Zanotti dell’Accademia delle Scienze di Bologna, Giovanni Bianchi (Iano Planco) di Rimini, Antonio Vallisnieri, professore di storia na- turale a Padova, Angelo Attilio Tilli prefet- to del giardino botanico pisano, Antonio e Raimondo Cocchi, Saverio Manetti, Ubaldo Montelatici38. vol. II, pp. 255 e segg.). Anche in seguito, la Società Botanica di Cortona non avrebbe mai pub- blicato un contributo a stampa e già negli Nello scritto viene citata una iscrizione apposta dentro una nicchia nel giardino, at- tribuita al solo Tommaso Coltellini, ma fir- mata anche da Lodovico Coltellini: un lungo elogio funebre indirizzato a Marcello Venuti dalla Società Botanica36. Come è stato osser- vato, nell’apparato funebre in onore di Mar- cello Venuti descritto dal lungo e minuzioso resoconto, il giardino botanico perdeva la sua tradizionale connotazione scientifica e diventava un luogo teatrale37. Nel 1756 fu ripetuta la solenne cerimonia di apertura del giardino botanico di Cortona, in occasione della quale Filippo Venuti dedi- cava un sonetto al vescovo Giuseppe Ippoliti (il futuro presule di Pistoia, antecessore di Scipione de’ Ricci) a sua volta socio della So- cietà Botanica. Le funzioni del segretario assai pre- sto vennero assunte da Lodovico Coltelli- ni, che compariva in questo ruolo già nel 1757 e sarebbe stato a lungo l’anima infa- ticabile del sodalizio. Il catalogo dei soci 60 Bruno Gialluca Fig. 6 Appunti sulla Società Botanica di Cortona e su Mattia Moneti 61 Fig. 7 62 Bruno Gialluca uno spazio sempre più asfittico alle istitu- zioni accademiche minori. nel centro dell’etruscheria, che mai diranno se sta- bilmente si vedesse in Cortona la Società Bottanica Georgofila, giacchè il nostro territorio è certamente l’Esperia toscana. Destiamoci dunque e spero mi ri- troverà prontissimo ai suoi cenni […]40. Della sostanziale sterilità della Società Botanica una parte del ceto colto di Cortona fu pienamente avvertita e cercò, inutilmente, di correre ai ripari, in particolare la persona- lità allora di maggiore spicco e autorevolezza a Cortona, il marchese Benvenuto Giuseppe Venuti, figlio primogenito di Marcello Venu- ti. Nato nel 1741, dopo aver prodotto qualche saggio erudito – come all’epoca era presso- ché d’obbligo a Cortona per i rampolli del- le principali famiglie – ben presto avvertiva con qualche insofferenza la inadeguatezza della tradizionale impostazione antiqua- ria ed erudita e si accostava ad interessi di agronomia e di economia, testimoniati dal suo carteggio. Benvenuto Giuseppe Venuti, già socio dal 1757 della Società Botanica di Cortona (la patente gli era stata rimessa nello stesso anno con una cerimoniosissima lettera da Lodovico Coltellini), ben inserito a corte e perfettamente consapevole della nuova poli- tica culturale, avvertiva lo isterilimento del- la Società Botanica di Cortona e poneva la questione di un suo aggiornamento. 40 B.C.A.E., ms. 273, cc. 80r.v. 41 B.C.A.E.C., ms. 596, cc. 24-25. L’erbario dipinto di Mattia Moneti: note botaniche Maria Adele Signorini con la collaborazione di Laura Vivona R acconta il naturalista toscano Giovanni Targioni Tozzetti1 (1776) che nel 1732 accompagnò il botanico Pier Antonio Miche- li (1679-1737) in un’escursione a Cortona ed ebbe occasione di osservare numerose piante di quel territorio, di cui riporta una diligente lista. Ma scopo – o pretesto – dell’escursio- ne non era solo la conoscenza della flora del territorio: Micheli si recava a Cortona per ordinare la collezione di oggetti naturali del- l’Accademia Etrusca, un’accolita di studiosi sorta pochi anni prima in quella città per iniziativa di un gruppo di cortonesi, tra cui il religioso e naturalista Filippo Venuti. L’Acca- demia nasceva nel solco di una consuetudine che nel XVII e XVIII secolo ispirò l’istituzio- ne di un gran numero di queste aggregazioni di intellettuali e volenterosi che coltivavano interessi come le belle lettere, l’archeologia o i vari aspetti di quella che allora veniva chiamata la Filosofia naturale. R botaniche anche il prete botanofilo – come lo definì Targioni Tozzetti – Mattia Moneti, su cui si trovano notizie negli scritti di Dragone Testi, Buresti e Gialluca3 e nei documenti da loro citati, molti dei quali manoscritti. q Della vita di Moneti è presto detto: nato a Cortona nel 1687, nel 1711 viene ordina- to sacerdote e nel 1735 gli viene affidata la parrocchia in campagna di Sant’Angelo in Metelliano. Nel 1754, al momento della fon- dazione della Società Botanica cortonese di cui è fatto presidente Filippo Venuti, Moneti ne viene nominato direttore; l’anno succes- sivo è ascritto all’Accademia Etrusca. Muo- re nel 1758 a 72 anni a Sant’Angelo, dove è sepolto. Si sa da Targioni Tozzetti che, come Venuti, anche Moneti erborizzò assiduamen- te nel territorio cortonese, ma le piante da lui raccolte – al pari di quelle di Venuti – non ci sono arrivate. Di lui non si sarebbe dunque conservata particolare memoria se, oltre alla lapide sulla tomba che ne loda le qualità di religioso e di botanico, non fossero rimasti anche i volumi di immagini di piante da lui dipinte, conservati nella Biblioteca di Corto- na (Fig. 8). 1 G. Targioni Tozzetti, Relazioni d’alcuni viaggi fatti in diverse parti della Toscana (II ed.), vol. VIII, Cambiagi, Firenze 1776, pp. 464-480 (copia anastatica dell’originale, Forni, Bologna 1972). Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 1972). 2 Cfr. anche quanto riportato da G. Dragone Testi, Un ignorato Centro di Studi Scientifici del secolo XVIII: la Società Botanica Cortonese e l’opera di Mattia Moneti, Annuario Accad. Etrusca di Cortona (1936-1937), 1938, p. 6 («Certo è che il Micheli conobbe l’abate Mattia Moneti, ed andò anche ad erborizzare accanto ad una cappella da quest’ultimo goduta»); p. 14 («[Moneti] fu anche incoraggiato [allo studio della botanica] dallo stesso Micheli, che conosciutolo in occasione del viaggio a Cortona lo ebbe quale assai diligente e premuroso compagno di escursioni in ogni territorio»). 3 Dragone Testi, op. cit.; D. Buresti, Una Società Botanica a Cortona nel ‘700 seconda in Italia dopo quella di Firenze, «L’Universo», 59(2), 1979, pp. 401-434; B. Gialluca, L’Erbario di Mattia Moneti, in AA.VV., Da Andrea Cesalpino ai nostri giorni. Erbari aretini in mostra, Arezzo, Museo d’Arte Medievale e Moderna, 4 marzo - 27 maggio 2005, 2005, oltre che B. Gialluca in questo stesso volume. Scriveva infatti a Lodovico Coltellini, segretario della Società Botanica, il 27 febbraio 1769: È evidente nella lettera la consapevolezza degli orientamenti sovrani in materia di po- litica culturale e la precisa cognizione della inadeguatezza della Accademia Botanica e della necessità della sua associazione all’Ac- cademia dei Georgofili, condizione perché Cortona potesse reinserirsi nella più dinami- ca realtà della Toscana di Pietro Leopoldo. p L’Accademia Botanica non ebbe la forza di raccogliere le indicazioni e il 30 giugno 1775 Benvenuto Venuti presentava all’Acca- demia Etrusca (prova questa della cessata operatività della Società Botanica) il progetto per la fondazione di «una Accademia di agri- coltura, bottanica e commercio» che avrebbe dovuto mantenere uno stretto rapporto con le altre istituzioni scientifiche toscane, pri- ma fra tutte l’Accademia dei Georgofili (della quale, era detto esplicitamente nel progetto, in sostanza sarebbe stata ‘colonia’) e la atti- vità avrebbe dovuto essere finalizzata all’in- cremento del commercio, contribuendo così alla pubblica utilità e alla felicità maggiore dell’intera Toscana41. Il fatto che Benvenuto Venuti proponesse di assegnare alla proget- tata nuova società l’impresa e il sigillo che erano appartenuti alla Società Botanica è chiara dimostrazione che essa, all’epoca, aveva ormai cessato di vivere. Il nuovo so- dalizio, comunque, anche per la prematura morte del Venuti (1780), non avrebbe mai visto la luce. g , ella non può mai supporsi quanto mi sia a cuore la nostra società botanica. Quando dunque faremo una sessione per trattar seriamente della medesima? Quando scriveremo a Firenze per dichiararci consoci della Fiorentina Regia Georgofila? Dei soci utili non ne mancheranno, gli onorari saranno infiniti, l’arte è lunga, la vita è breve. Se l’Accademia Etrusca fu da tutti dichiarata meritamente e giustamente instituita Fig. 8 Carta del 1° Volume dell’erbario Moneti elegantemente illustrata e recante la collocazione all’interno della biblioteca comunale. Fig. 9 Carta raffigurante un gambero di fiume, con riportato il luogo del suo rinvenimento. Fig. 10 Carta raffigurante “Chelidonium maius, sive Hirundinaria”. 4 Citato in Buresti, op. cit. 5 Devo all’amica Chiara Nepi, che ringrazio, tutte le notizie sugli esemplari cortonesi dell’erbario Micheli. 6 A differenza degli altri, il terzo volume non è datato, anche se risulta che venne donato all’Accademia Etrusca nel 1755, vedi Gialluca, op. cit., ed in questo stesso volume. L’erbario dipinto di Mattia Moneti: note botaniche Nelle sue giornate cortonesi, che dovette- ro certo lasciare una forte impronta nella vita culturale della cittadina, Micheli si dedicò dunque ad ordinare le collezioni dell’Acca- demia, a erborizzare e a identificare le sue piante, e anche le molte che erano state rac- colte in precedenza dallo stesso Venuti con lo scopo di dare alle stampe un catalogo del- la flora di Cortona, che sfortunatamente, per quanto si sa, non vide mai la luce. È verosi- mile2 che, oltre a Venuti, Micheli abbia avuto per compagno di escursioni e di dissertazioni Il carattere dell’uomo emerge con chia- rezza dagli aggettivi usati negli scritti dei biografi: la Dragone Testi lo definisce umile, mite, buono, ma più frequentemente modesto; Buresti usa i termini studioso, umile, riser- vato, ignoto, onesto, laborioso, paziente, ma soprattutto mite; il contemporaneo Lodovico Fig. 8 Fig. 8 Fig. 8 Fig. 8 L’erbario dipinto di Mattia Moneti 65 Fig. 10 ig. 9 Fig. 10 Fig. 10 Fig. 9 interessarsi ad essa, visto che nel suo erbario si trova un esemplare di Polypodium vulgare con l’etichetta «Nel bosco de Cappuccini di Cortona copioso, 1732» e l’indicazione «It. Corton. n. 100», indicazione che farebbe supporre un buon numero di campioni rac- colti nella stessa escursione. Invece, a fronte delle quasi cinquanta specie del citato elen- co di Targioni Tozzetti e del numero d’ordine riportato nell’etichetta dell’esemplare miche- liano, attualmente nell’erbario Micheli (con- servato nella Sezione Botanica del Museo di Storia Naturale di Firenze) si trovano solo 10 esemplari provenienti da Cortona, 4 dei qua- li con la dicitura «It. Corton.» e numeri da 71 a 1125. Coltellini4 lo pianse alla morte come schivo, piacevole, modesto, caritativo; in Gialluca si trovano gli aggettivi marginale, apparta- to. Dunque, un semplice e bonario curato di campagna, pittore dilettante con la passione delle piante. Passione che pare fosse nata perché, avendogli il medico prescritto per certi suoi mali un Polypodium che cresceva nel bosco dei Cappuccini, Moneti andò per- sonalmente a cercarlo, appassionandosi alle erborizzazioni. Racconta Targioni Tozzetti che nel 1732 in quello stesso bosco Micheli raccolse un «Polypodium (an) majus, acuto folio, viterbiense … copioso». E aggiunge: «Il Sig. Micheli dubita se il Polipodio che trovammo noi sia l’accennato del Barrelliero, e dal Boccone… e nel suo Libro Rariorum ne farà diligente esame». 6 A differenza degli altri, il terzo volume non è datato, anche se risulta che venne donato all’Accademia Etrusca nel 1755, vedi Gialluca, op. cit., ed in questo stesso volume. 4 Citato in Buresti, op. cit. 5 Devo all’amica Chiara Nepi, che ringrazio, tutte le notizie sugli esemplari cortonesi dell’erbario Micheli. 6 A differenza degli altri il terzo 7 Un elenco delle specie dell’erbario Moneti secondo la nomenclatura della prima metà del ’900 si trova in Dragone Testi, op. cit. Non tutte le sue identificazioni appaiono tuttavia completamente condivisibili. 8 Qui e in seguito le figure sono identificate con il numero del volume, seguito da quello del foglio. g Fig. 14 Carta raffigurante “Asplenium, sive Ceterach”. Fig. 15 Carta raffigurante Orobanche maior “Orobanche maior Caryophyllum olens”. Fig. 16 Carta raffigurante “Nymphea alba maior, sive Nenuphar”. Fig. 11 Carta raffigurante “Scrophularia nodosa fetida”. Fig. 12 Carta raffigurante “Ranunculus sylvaticus, sive sylvestris” (= Anemone apennina L.). i Fig. 11 Carta raffigurante “Scrophularia nodosa fetida”. Fig. 12 Carta raffigurante “Ranunculus sylvaticus, sive sylvestris” (= Anemone apennina L.). Fig. 13 Carta raffigurante funghi eduli. Fig. 14 Carta raffigurante “Asplenium, sive Ceterach”. Fig. 15 Carta raffigurante Orobanche maior “Orobanche maior Caryophyllum olens”. Fig. 16 Carta raffigurante “Nymphea alba maior, sive Nenuphar”. L’erbario dipinto di Mattia Moneti: note botaniche Inevitabile in questa discussione immaginare accanto a Micheli la presenza di Moneti, a dibattere dell’identità proprio di quella pianta cui doveva l’interes- se per la botanica. E anche Micheli dovette Come risulta dagli scritti citati, l’erbario dipinto di Moneti conservato a Cortona consta di 3 volumi rilegati, eseguiti rispettivamente nel 1732 (l’anno dell’escursione di Micheli), 1738 e 17556. A questi si aggiunge un ulte- riore volume, donato all’Accademia Etrusca 66 Maria Adele Signorini 2 11 Fig. 12 Fig. 1 g. 14 ig. 13 4 Fig. 14 Fig. 14 Fig. 13 L’erbario dipinto di Mattia Moneti 67 Fig. 15 Fig. 16 Fig. 15 dopo la morte di Moneti e costituito da una trentina di fogli con numerazione a sbalzi, su cui si trovano figure in parte incomplete accompagnate dal solo nome volgare in ita- liano. Si tratta evidentemente di materiale in forma provvisoria, ancora da sistemare. Sem- bra che Moneti avesse invece completato e donato all’Accademia Etrusca già nel 1756 un quarto volume del suo erbario dipinto, che sarebbe oggi perduto. scibili, anche grazie a dettagli che rivelano l’occhio botanico allenato dell’autore. Tra i caratteri diagnostici in evidenza ve ne sono di sottili, come il lattice aranciato sulle super- fici di taglio del Chelidonium majus (1:48)8 (Fig. 10) o i rizomi nodosi della Scrophularia (2:19) (Fig. 11) in rari casi sono riportati det- tagli separati (vedi Anemone apennina, 1:70) (Fig. 12). L’immagine occupa di norma tutto il foglio, indipendentemente dalle dimensio- ni reali delle piante, che quindi non sono in scala tra loro. In pochi casi in una stessa ta- vola compaiono più specie, come le due Aju- ga (1:46) o i funghi eduli (2:95) (Fig. 13). Radici e parti ipogee sono spesso lasciate bianche o appena accennate. Nelle immagini di qualche pianta di rupe o muro (come Cete- rach 1:138 e Cymbalaria 2:17) o di ambiente umido (come Typha latifolia 2:88) si vede un cenno dell’ambiente di crescita (Fig. 14). scibili, anche grazie a dettagli che rivelano l’occhio botanico allenato dell’autore. Tra i caratteri diagnostici in evidenza ve ne sono di sottili, come il lattice aranciato sulle super- fici di taglio del Chelidonium majus (1:48)8 (Fig. 10) o i rizomi nodosi della Scrophularia (2:19) (Fig. 11) in rari casi sono riportati det- tagli separati (vedi Anemone apennina, 1:70) (Fig. 12). p ) Fig. 13 Carta raffigurante funghi eduli. L’erbario dipinto di Mattia Moneti: note botaniche 15 «… essendo che la salute pubblica, che molte volte dipende dall’uso dei Semplici, veniva affidata ad alcuni vili uomini mercenari, che privi affatto di quelle cognizioni a ciò necessarie, erano incaricati dai farmacopoli di rintracciarli… per ovviare al continuo evidente pericolo dei miei cari concittadini mi accinsi all’impresa… sperando, che delle mie fatiche suddette ne avrei potuto ritrar due vantaggi nel tempo stesso, con precludere cioè la strada all’ignoranza, e impostura dei nostri pseudo bottanici, … e con animare altrui sull’esempio mio a coltivare lo studio di questa bella parte della Fisica e Medicina…» Dunque, le figure del suo erbario dipinto dovevano servire soprattutto per il riconoscimento delle piante medicinali che, per la mancanza di bravi e onesti erboristi, al tempo non davano garanzie di corretta identificazione. zona. Non mancano però piante di ambien- te umido o acquatiche come Nymphaea alba (3:63) (Fig. 16), oggi verosimilmente rare- fatte o scomparse in seguito alla bonifica, e antiche infestanti come il fiordaliso (3:102), rese anch’esse rare, qui come altrove, dalle mutate tecniche agronomiche. Infine, nume- rose entità raffigurate dovevano anche allora non essere comuni, come Erithronium dens- canis (1:88), Nepeta cataria (2:22), Physalis alkekengi (2:7); altre si troverebbero a Corto- na al limite dell’areale, come Hymenocarpus circinnatus (2:29), Hypericum perfoliatum (2:40) e Tetragonolobus purpureus (3:81), quest’ultima non citata oggi per la Toscana nella Flora d’Italia14. linomia per la quale sono citati con sigle gli autori di riferimento: per lo più Tournefort, ma anche i Bauhin e più raramente altri pre- linneani, a cominciare da Mattioli9. Un’al- tra mano ha aggiunto quasi sempre la sigla I.R.H. e un numero, con evidente riferimento alle Institutiones Rei Herbariae di Tournefort (1700). Tournefortiano è anche l’ordinamen- to sistematico seguito. In tre casi si trovano riferimenti a Micheli, tutti però in diversa grafia10. In genere è riportata anche la di- stribuzione, talvolta con aggiunte successive di altra mano: si va da indicazioni generi- che come Nasce da p. tutto, Nasce p. li sodi a descrizioni dettagliate come Questo nasce molto di là dalla V.a di Pergo nella strada quando si va p. andare a Valecchie (3:91). In un caso, quello dell’Orobanche (1:31) (Fig. 15), è aggiunto un commento pratico: Questa pianta si trova di rado ed è bene. L’orobanche è anche l’unica corredata del nome in volgare Succiamele. 9 J.P. Tournefort (1656-1708), J. Bauhin (1541-1613) e C. Bauhin (1560-1624), P. A. Mattioli (1500?-1577). Nel terzo volume dell’erbario i riferimenti agli autori mancano, a parte poche eccezioni in cui è citato Mattioli. 10 Quinquefolium montanum folio inferne canescente Michel. H. Pis. (2: 45). Parthenium minus foliis tenuissimis achillaea. Coesuris Michel. (2: 82). Persicaria non maculosa urens, spicis longis strigosis I. R. H. Michel. (2: 89). (In neretto le aggiunte di grafia diversa). 11 Cfr. P.A. Micheli, Relazione dell’erba detta da’ Bottanici Orobanche, e volgarmente Succiamele, fiamma, e mald’occhio, ecc. Tartini e Franchi, Firenze 1723. 12 Vedi ad esempio 1: 77, 1: 100. 13 Vedi 2: 15, 2: 54, 2: 57. 14 Cfr. S. Pignatti, Flora d’Italia, voll. 1-3, Edagricole, Bologna 1982. 15 «… essendo che la salute pubblica, che molte volte dipende dall’uso dei Semplici, veniva affidata ad alcuni vili uomini mercenari, che privi affatto di quelle cognizioni a ciò necessarie, erano incaricati dai farmacopoli di rintracciarli… per ovviare al continuo evidente pericolo dei miei cari concittadini mi accinsi all’impresa… sperando, che delle mie fatiche suddette ne avrei potuto ritrar due vantaggi nel tempo stesso, con precludere cioè la strada all’ignoranza, e impostura dei nostri pseudo bottanici, … e con animare altrui sull’esempio mio a coltivare lo studio di questa bella parte della Fisica e Medicina…» Dunque, le figure del suo erbario dipinto dovevano servire soprattutto per il riconoscimento delle piante medicinali che, per la mancanza di bravi e onesti erboristi, al tempo non davano garanzie di corretta identificazione. 11 Cfr. P.A. Micheli, Relazione dell’erba detta da’ Bottanici Orobanche, e volgarmente Succiamele, fiamma, e mald’occhio, ecc. Tartini e Franchi, Firenze 1723. 12 Vedi ad esempio 1: 77, 1: 100. 13 Vedi 2: 15, 2: 54, 2: 57. 14 Cfr. S. Pignatti, Flora d’Italia, voll. 1-3, Edagricole, Bologna 1982. 15 «… essendo che la salute pubblica, che molte volte dipende dall’uso dei Semplici, veniva affidata ad alcuni vili uomini mercenari, che privi affatto di quelle cognizioni a ciò necessarie, erano incaricati dai farmacopoli di rintracciarli… per ovviare al continuo evidente pericolo dei miei cari concittadini mi accinsi all’impresa… sperando, che delle mie fatiche suddette ne avrei potuto ritrar due vantaggi nel tempo stesso, con precludere cioè la strada all’ignoranza, e impostura dei nostri pseudo bottanici, … e con animare altrui sull’esempio mio a coltivare lo studio di questa bella parte della Fisica e Medicina…» Dunque, le figure del suo erbario dipinto dovevano servire soprattutto per il riconoscimento delle piante medicinali che, per la mancanza di bravi e onesti erboristi, al tempo non davano garanzie di corretta identificazione. L’erbario dipinto di Mattia Moneti: note botaniche L’immagine occupa di norma tutto il foglio, indipendentemente dalle dimensio- ni reali delle piante, che quindi non sono in scala tra loro. In pochi casi in una stessa ta- vola compaiono più specie, come le due Aju- ga (1:46) o i funghi eduli (2:95) (Fig. 13). Radici e parti ipogee sono spesso lasciate bianche o appena accennate. Nelle immagini di qualche pianta di rupe o muro (come Cete- rach 1:138 e Cymbalaria 2:17) o di ambiente umido (come Typha latifolia 2:88) si vede un cenno dell’ambiente di crescita (Fig. 14). I tre volumi che ci sono pervenuti com- prendono in tutto 338 figure relative a 313 specie, (25 compaiono due volte): oltre a un animale (Fig. 9), vi si trovano 6 specie di fun- ghi, 6 di licheni, 1 briofita, 6 pteridofite, 293 angiosperme7. Le piante, ritratte a tempera, mostrano un aspetto alquanto rigido e inge- nuo che ricorda quasi quello di certe raffigu- razioni cinquecentesche più che le ben più fedeli iconografie botaniche coeve. Tuttavia le immagini non sono prive di una loro piace- volezza e le piante sono quasi sempre ricono- Sotto all’immagine è riportato il nome della pianta, secondo una nomenclatura po- 68 Maria Adele Signorini 9 J.P. Tournefort (1656-1708), J. Bauhin (1541-1613) e C. Bauhin (1560-1624), P. A. Mattioli (1500?-1577). Nel terzo volume dell’erbario i riferimenti agli autori mancano, a parte poche eccezioni in cui è citato Mattioli. 10 Quinquefolium montanum folio inferne canescente Michel. H. Pis. (2: 45). Parthenium minus foliis tenuissimis achillaea. Coesuris Michel. (2: 82). Persicaria non maculosa urens, spicis longis strigosis I. R. H. Michel. (2: 89). (In neretto le aggiunte di grafia diversa). 11 Cfr. P.A. Micheli, Relazione dell’erba detta da’ Bottanici Orobanche, e volgarmente Succiamele, fiamma, e mald’occhio, ecc. Tartini e Franchi, Firenze 1723. 12 Vedi ad esempio 1: 77, 1: 100. 13 Vedi 2: 15, 2: 54, 2: 57. 14 Cfr. S. Pignatti, Flora d’Italia, voll. 1-3, Edagricole, Bologna 1982. L’erbario dipinto di Mattia Moneti: note botaniche E chissà se anche la particola- re attenzione rivolta a questa parassita non sia un’eco dei contatti tra Moneti e Micheli, che l’aveva approfonditamente indagata11. In pochi casi si trovano notazioni sulla rarità della specie12 o sul fatto che sia sativa, cioè coltivata13. linomia per la quale sono citati con sigle gli autori di riferimento: per lo più Tournefort, ma anche i Bauhin e più raramente altri pre- linneani, a cominciare da Mattioli9. Un’al- tra mano ha aggiunto quasi sempre la sigla I.R.H. e un numero, con evidente riferimento alle Institutiones Rei Herbariae di Tournefort (1700). Tournefortiano è anche l’ordinamen- to sistematico seguito. In tre casi si trovano riferimenti a Micheli, tutti però in diversa grafia10. In genere è riportata anche la di- stribuzione, talvolta con aggiunte successive di altra mano: si va da indicazioni generi- che come Nasce da p. tutto, Nasce p. li sodi a descrizioni dettagliate come Questo nasce molto di là dalla V.a di Pergo nella strada quando si va p. andare a Valecchie (3:91). In un caso, quello dell’Orobanche (1:31) (Fig. 15), è aggiunto un commento pratico: Questa pianta si trova di rado ed è bene. L’orobanche è anche l’unica corredata del nome in volgare Succiamele. E chissà se anche la particola- re attenzione rivolta a questa parassita non sia un’eco dei contatti tra Moneti e Micheli, che l’aveva approfonditamente indagata11. In pochi casi si trovano notazioni sulla rarità della specie12 o sul fatto che sia sativa, cioè coltivata13. Di tutte sarebbe interessante verificare la presenza attuale nel cortonese, meglio se dopo una preliminare e accurata verifica delle identificazioni di Dragone Testi. Alla luce di quanto scrive Moneti all’inizio del III volume a proposito dello scopo della sua opera15, desta infine qualche stupore la pre- senza tra le piante raffigurate non già di sole specie medicinali, ma di ogni genere di ve- getali, compresi molti di cui non si conosce alcun uso officinale. Evidentemente, al di là delle intenzioni iniziali, nel corso dell’opera la passione per le piante doveva aver preso la mano al volenteroso abate cortonese. Le piante raffigurate sono in massima parte comuni e certo tuttora presenti nella L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi fatti come, almeno ancora agli inizi del ’900, l’erbario fosse formato da tre parti distinte: la prima, terminante a pagina 130, in cui le piante appaiono ben determinate secondo il Durante; la seconda costituita da alcuni fa- scicoli sciolti, con numerazione diversa dalla precedente (da pagina 1 a pagina 114, ma con lacune) e intercalativi successivamente, con piante meno ben conservate, peggio de- terminate o mancanti di denominazione; la terza costituita da pochi fascicoli legati in- sieme a quelli della prima, ma con numera- zione diversa (da pagina 21 a 77 ma senza interruzioni), con saltuarie citazioni del testo del Durante. Secondo Baccarini, quindi, gli autori dell’erbario potrebbero essere diversi ma forse uno di essi è identifi cabile attra- verso una notazione posta in fondo a pagina 130 (Fig. 3). Qui infatti si legge: «Semplici appiccicati in questo libro numero 323» a cui segue un breve elenco di piante presenti nell’Orto di uno (probabilmente il primo) dei compilatori dell’erbario. Si noti che insieme all’elenco di piante compaiono dei riferimen- ti mal interpretabili (neanche Baccarini era riuscito a decifrarli) che potrebbero tutta- via indicare la localizzazione delle piante nell’Orto. All’elenco segue la notazione (in grafi a differente) «Adesso non ve ne sono di queste se non che tre ……ma io Gio. Iaco- pelli spero di rimetterle con molte altre, come che sono geniale di simili piante». È dunque plausibile, come riporta Baccarini, ritenere fatti come, almeno ancora agli inizi del ’900, l’erbario fosse formato da tre parti distinte: la prima, terminante a pagina 130, in cui le piante appaiono ben determinate secondo il Durante; la seconda costituita da alcuni fa- scicoli sciolti, con numerazione diversa dalla precedente (da pagina 1 a pagina 114, ma con lacune) e intercalativi successivamente, con piante meno ben conservate, peggio de- terminate o mancanti di denominazione; la terza costituita da pochi fascicoli legati in- sieme a quelli della prima, ma con numera- zione diversa (da pagina 21 a 77 ma senza interruzioni), con saltuarie citazioni del testo del Durante. Secondo Baccarini, quindi, gli autori dell’erbario potrebbero essere diversi ma forse uno di essi è identifi cabile attra- verso una notazione posta in fondo a pagina 130 (Fig. 3). Qui infatti si legge: «Semplici appiccicati in questo libro numero 323» a cui segue un breve elenco di piante presenti nell’Orto di uno (probabilmente il primo) dei compilatori dell’erbario. Fig. 1 la bella bacheca della Biblioteca dove viene esposto l’erbario insieme ad altri preziosi volumi (foto di l. lastrucci). 1 P. Baccarini, Sopra un antico erbarietto conservato nella biblioteca comunale di Poppi, «Bull. Soc. Bot. Ital.», 7, 1910, pp. 102-106. 2 C. durante Herbario nuovo, con fi gure che rappresentano le viue Piante, che nascono in tutta Europa, & nell’Indie Orientali, & Occidentali, con versi latini... Presso Gian Giacomo hertz, Venezia 1684. 3 G. Cipriani, Inventario dei manoscritti della Biblioteca comunale di Poppi, luigi Bordandini, Forlì 1896. L’erbario della Biblioteca Rilliana di Poppi Fig. 1 Fig. 1 Fig. 1 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 1 P. Baccarini, Sopra un antico erbarietto conservato nella biblioteca comunale di Poppi, «Bull. Soc. Bot. Ital.», 7, 1910, pp. 102-106. 2 C. durante Herbario nuovo, con fi gure che rappresentano le viue Piante, che nascono in tutta Europa, & nell’Indie Orientali, & Occidentali, con versi latini... Presso Gian Giacomo hertz, Venezia 1684. 3 G. Cipriani, Inventario dei manoscritti della Biblioteca comunale di Poppi, luigi Bordandini, Forlì 1896. L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi Si noti che insieme all’elenco di piante compaiono dei riferimen- ti mal interpretabili (neanche Baccarini era riuscito a decifrarli) che potrebbero tutta- via indicare la localizzazione delle piante nell’Orto. All’elenco segue la notazione (in grafi a differente) «Adesso non ve ne sono di queste se non che tre ……ma io Gio. Iaco- pelli spero di rimetterle con molte altre, come che sono geniale di simili piante». È dunque plausibile, come riporta Baccarini, ritenere L ’erbario della Biblioteca Comunale «Ril- liana» di Poppi (Figg. 1 e 2) si presenta, sotto alcuni aspetti, piuttosto misterioso, sia per quanto riguarda la sua precisa datazione sia per ciò che concerne l’autore o gli autori. Esso è stato già oggetto di attenzione dell’in- signe botanico P. Baccarini che nel 1910 ri- portava alcuni commenti nel Bollettino della Società Botanica Italiana1. Sul come questo erbario sia giunto a Poppi, è sicuro che esso sia pervenuto per una donazione ad opera del Conte Rilli Orsini, «il quale a sua vol- ta aveva ereditato il titolo nobiliare ed i libri da un suo congiunto appartenente a quella celebre famiglia Romana». Nel 1825 infatti il Conte Fabrizio Rilli Orsini fece dono alla Comunità di Poppi di manoscritti, incunabo- li e altre pubblicazioni dei secoli XII-XVIII. Per quanto riguarda la datazione dell’erba- rio, pur non esistendo alcuna indicazione relativa a date precise, può esser di aiuto il fatto che, all’interno di esso, si trovano rife- rimenti al testo del Durante Herbario nuovo nell’Edizione Herziana del 16842; è naturale quindi ritenere l’erbario composto dopo tale data, e collocarlo almeno alla fi ne del XVII secolo. A conferma di ciò si cita l’Inventario dei manoscritti della Biblioteca comunale di Poppi redatto nel 1896 da G. Cipriani in cui compare: «Erbario colle pianticelle naturali. Sec. XVII in 4°»3. Incertezze vi sono anche sull’autore, o forse sarebbe meglio dire sugli autori dell’erbario; Baccarini sottolinea in- L 72 Lorenzo Lastrucci e Alessandro Brezzi che l’erbario sia stato realizzato da un autore che coltivava le piante in un piccolo Orto allo scopo probabilmente di renderne più facile l’identificazione. È probabile poi che l’Orto, passato alla cura di altre persone meno dili- genti, sia andato via via decadendo finché il nuovo curatore, tale Iacopelli, si ripromise di rimetterlo in auge. Secondo Baccarini, Iaco- pelli potrebbe essere dunque l’autore della seconda e terza parte dell’erbario (o almeno di una delle due). Fig. 2 Il castello dei Conti Guidi a Poppi, in cui si trova la Biblioteca Rilliana che ospita l’erbario (foto di L. Lastrucci). Fig. 3 Pagina 130: si legge il numero di piante contenuto nell’erbario (323), l’elenco di piante presenti nell’Orto al tempo dell’autore e la notazione successiva di Iacopelli. Fig. 4 Pagina 5: in calce a tre dei quattro campioni è presente il riferimento all’utilizzo galenico della pianta e la citazione della pagina del testo del Durante. L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi Resta tuttavia il mistero su chi abbia composto la prima parte del- l’erbario che, come riportava Baccarini, ap- pariva la più curata, sia per l’esattezza delle determinazioni che per i continui riferimenti al testo del Durante, mancanti invece nella seconda e sporadici nella terza parte. Ciò che invece sembra abbastanza certo è l’intento professionale e galenico per cui è stato costi- tuito l’erbario, dal momento che esso ospita piante utilizzate a scopo medicinale (“sem- plici”) e, come già accennato, molte di esse portano anche indicazioni sulle loro proprie- tà medicamentose tratte dall’opera di Du- rante sopra menzionata (Figg. 4, 5 e 6). Tali “semplici” dovevano, almeno per la maggior parte di loro, essere stati coltivati nell’Orto suddetto, di cui resta comunque misteriosa la localizzazione. L’erbario come ci appare oggi, si presenta in un unico volume legato in pergamena che non presenta alcuna indicazione all’esterno; rispetto a quanto riportato da Baccarini non compaiono fascicoli sciolti. Va subito detto, peraltro, che l’erbario è stato oggetto pochi anni fa di una profonda opera di restauro da L’erbario della Biblioteca Rilliana di Poppi 73 Fig. 3 Pagina 130: si legge il numero di piante contenuto nell’erbario (323), l’elenco di piante presenti nell’Orto al tempo dell’autore e la notazione successiva di Iacopelli. parte del Laboratorio di Restauro del Libro delle suore benedettine dell’Abbazia della SS. Annunziata di Rosano. Gli interventi di restauro hanno riguardato, tra le altre cose, la ricomposizione e cucitura dei fascicoli, che naturalmente è stata fatta in conformità con l’originale pervenuto al Laboratorio. Sono inoltre stati restaurati alcuni campioni di cui è stato consolidato l’ancoraggio al supporto; in calce all’erbario sono presenti alcuni fo- gli di nuova apposizione in cui i restauratori hanno incollato alcuni campioni evidente- mente staccatisi nel tempo o forse mai incol- lati nell’erbario e di cui non è stato possibile risalire alla posizione originaria. I fogli dell’erbario sono numerati a fronte e retro, con la pagina dispari che porta la numerazione in alto a destra e la pari in alto a sinistra; in alcune, tuttavia, sono presenti altri numeri il che potrebbe far pensare ad una sorta di numerazione anche dei campio- ni oppure ad un riferimento circa la loro po- sizione nell’Orto dell’autore. I fogli sono più o meno nettamente divisi in due colonne ed i 74 Lorenzo Lastrucci e Alessandro Brezzi totale e sono concentrate nelle prime sedici pagine dell’erbario. Fig. 6 Particolare delle proprietà curative di una pianta (Sena, colonna destra di Pagina 4) con riferimento alla pagina dell’opera di Durante. Fig. 5 Pagina 4: ancora campioni recanti riferimenti al testo del Durante in calce. L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi Alcune piante da Bac- carini citate come esempi delle terza parte dell’erbario quali Therriacaria del Cesalpi- no, a pagina 43 o Camumilla con odore di melo Appio, a pagina 44, appaiono interca- late all’interno del volume; risulta pertanto difficile capire come si sia giunti all’attuale sequenza dei fogli dell’erbario e sapere se siano venute a mancare alcune parti di esso; campioni sono incollati, a seconda delle loro dimensioni, uno per colonna o due per co- lonna (per un totale di quattro campioni per foglio); non mancano fogli con tre campioni così come, più raramente, compaiono fogli con un unico campione. Le piante che por- tano particolareggiati riferimenti al Durante (che quindi appartengono alla prima parte dell’erbario come lo aveva descritto il Bac- carini) sono relativamente poche rispetto al L’erbario della Biblioteca Rilliana di Poppi 75 non bisogna dimenticare che tra la visita di Baccarini alla biblioteca comunale di Poppi ed oggi sono passati molti anni e soprattutto due guerre per cui diventa arduo ricostruire la storia dell’erbario almeno fino agli anni ’70. È da notare poi che nel 1937 muore l’ul- timo bibliotecario comunale e da quella data manca una figura ufficiale addetta alla ge- stione della Biblioteca che viene affidata al preposto. Da questa data mancano dunque documenti che registrino tutti i movimenti relativi al materiale della biblioteca e que- sta situazione si è protratta fino agli Ottanta, allorché la gestione della biblioteca torna di Fig. 5 Pagina 4: ancora campioni recanti riferimenti al testo del Durante in calce. Fig. 6 Particolare delle proprietà curative di una pianta (Sena, colonna destra di Pagina 4) con riferimento alla pagina dell’opera di Durante. non bisogna dimenticare che tra la visita di Baccarini alla biblioteca comunale di Poppi ed oggi sono passati molti anni e soprattutto due guerre per cui diventa arduo ricostruire la storia dell’erbario almeno fino agli anni ’70. È da notare poi che nel 1937 muore l’ul- timo bibliotecario comunale e da quella data manca una figura ufficiale addetta alla ge- stione della Biblioteca che viene affidata al preposto. L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi Da questa data mancano dunque documenti che registrino tutti i movimenti relativi al materiale della biblioteca e que- sta situazione si è protratta fino agli Ottanta, allorché la gestione della biblioteca torna di 76 76 Lorenzo Lastrucci e Alessandro Brezzi ni Tozzetti4 si ricava che Lunaria minore era utilizzato (non in modo univoco, peraltro) per designare Botrychium lunaria (L.) Swartz, che corrisponde in effetti al campione presen- te nel foglio. In calce al campione compare la notazione «Quata: fredda et secca» che altro non è che la citazione di quanto riportato a pagina 253 del Durante in cui si legge per Lunaria maggiore: «Qualità. E’ frigida, & secca». Allo stesso modo a pagina 8 in bas- so a destra compare in testa al campione la scritta Podagraria (Aegopodium podagraria L.). In calce si legge la scritta «Qualita calda et secca et a le medesime faculta del Ebulo» che è poi una sintesi di quanto in effetti ripor- tato a pagina 351 del Durante. Da pagina 16 in poi mancano particolareggiati riferimenti all’Herbario nuovo anche se talvolta compaio- no commenti relativi a qualche campione. I nomi delle piante in tutto l’erbario sono ge- neralmente in italiano, (es. Agliaria, Guado, Sicembro aquatico). Talvolta sono presenti ci- tazioni di botanici cinquecenteschi o seicen- teschi quali Cesalpino, a proposito della già citata Therriacaria (è un’Ononis) che in effet- ti compare come Theriacaria a pagina 238 dell’opera di Cesalpino5. Tra gli altri autori citati troviamo Mattioli (1500-1577), Taber- naemontanus (Dietrich Jacob, 1520?-1590) e Colonna (1576-1650) mentre compaiono an- che citazioni dell’autore latino Plinio: è plau- sibile che oltre al testo del Durante chi aveva costituito l’erbario dovesse conoscere anche le opere di questi illustri botanici. nuovo alle competenze di un dipendente co- munale, il Sig. Alessandro Brezzi. Al di là delle vicende storiche che lo hanno interessa- to, l’erbario appare complessivamente in di- screto stato, anche se alcuni esemplari in esso contenuti sono parzialmente o profondamente danneggiati e molti fogli appaiono macchia- ti, nonostante il restauro recente. I campioni presenti sono in gran parte Dicotiledoni (in misura minore ci sono anche alcune Pterido- fite e Monocotiledoni) e, come già accennato, almeno nelle prima parte dell’erbario, com- paiono, per molte piante, precisi riferimenti al testo del Durante relativi alle loro proprietà medicamentose. A pagina 5 (si veda Fig. 4 O. Targioni Tozzetti, Dizionario botanico italiano che comprende i nomi volgari italiani specialmente toscani e vernacoli delle piante..., II ed., parte prima, Firenze 1858. 5 A. Cesalpino, De Plantis Libri XVI. Florentiae 1583. Fig. 7 Pagina 122 con esemplare di Erba saetta (Sagittaria sagittifolia L., colonna sinistra) con la curiosa notazione in calce. L’erbario della biblioteca Rilliana di Poppi Lorenzo Lastrucci e Alessandro Brezzi 4) in alto a destra, ad esempio, compare la nota- zione Lunaria minore, evidentemente sovra- scritta su una precedente Lunaria maggiore (che è poi il nome presente in Durante per designare la pianta in questione): da Targio- Particolarmente curiose risultano alcune locuzioni che accompagnano diverse piante a cui l’Autore (o, forse, qualcuno degli Auto- ri) non sapeva attribuire il nome o di cui non conosceva la provenienza: a pagina 100, ad esempio, sotto un campione di Amaranthus si legge «Questa non si sa che sia» mentre a pa- gina 122 la frase che accompagna il campione di Erba saetta (Sagittaria sagittifolia L.) reci- ta: «dove si trova questa erba curiosa?» (Fig. 7). Resta anche in questo caso da capire se l’Autore intendesse sapere dove crescesse la pianta in natura oppure in che posto dell’Orto essa fosse coltivata. Anche questi interrogati- vi, tuttavia, stando ai dati attualmente dispo- nibili sulla storia dell’erbario e del suo o dei suoi Autori, sono destinati a restare irrisolti. Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna Secondino Gatta È evidente qui che san Francesco d’Assi- si non sottovaluta l’importanza della predi- cazione fra gl’infedeli, alla quale egli stesso si era dedicato, ma semplicemente vuole im- partire al suo compagno un insegnamento di umiltà. Del resto il Santo era fermamente convinto che «frate corpo» fosse creato in funzione dell’assolvimento degli atti spiri- tuali, cioè della preghiera, e che, pertanto, andasse curato e mantenuto sano. L’uomo, egli diceva, Andando un poco più oltre, santo Francesco chia- mava ancora forte: “O frate Lione, pecorella di Dio, benché il frate Minore parli con lingua d’Agnolo, e sappia i corsi delle stelle e le virtù delle erbe, e fus- songli rivelati tutti li tesori della terra, e conoscesse le virtù degli uccelli e de’ pesci e di tutti gli animali e delle pietre e delle acque; iscrivi che non è in ciò perfetta letizia1. L L e Fonti Francescane ci testimoniano che già al tempo di Francesco d’Assisi le conoscenze di erboristeria e medicinali non erano estranee alla missione dei frati. I Fioretti di san Francesco, infatti, benché siano una fonte tarda – si tratta di una com- pilazione di ambito toscano del ’300 ma che raccoglie importanti tradizioni orali coeve al Santo – riportano questo interessante dialo- go tra il Santo e Frate Leone. Il passo, che a prima vista sembrerebbe sottintendere un giudizio negativo, in realtà si inserisce in un discorso più ampio sulla ‘perfetta letizia’: la conoscenza delle virtù terapeutiche degli elementi naturali non deve essere motivo di orgoglio personale giacché è una delle Gra- zie che Dio concede ai suoi frati per farne un uso caritatevole. Che in questo insegnamento del Santo non vi sia un intento di condanna delle conoscenze medicinali da parte dei fra- ti lo si evince chiaramente dall’esortazione che segue immediatamente dopo: deve provvedere con discrezione al suo fisico, in ma- niera che fratello corpo non abbia a protestare3. deve provvedere con discrezione al suo fisico, in ma- niera che fratello corpo non abbia a protestare3. È del resto una testimonianza molto si- gnificativa di questa attenzione dell’As- sisiate il dato che i miracoli attribuiti alla sua intercessione siano in gran parte volti a sanare: a ricostituire, cioè, l’equilibrio tra corpo e anima. 1 I fioretti di san Francesco, Cap. VIII, FF, 1836. 2 Ibidem. 3 Specchio di perfezione, FF. 1796. Fig. 1 Farmacia Antica, sec. XVI- XVIII, Santuario della Verna (foto di A. Ferrini, Archivio fotografico della Verna). L’erbario Venturini del Santuario della Verna Fig. 1 Fig. 1 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 1 I fioretti di san Francesco, Cap. VIII, FF, 1836. Fig. 3 Alambicco del XVII secolo, Museo della Verna (foto di A. Ferrini, Archivio fotografico della Verna). Fig. 4 Libri di medicina negli scaffali della farmacia (foto di S. Gatta). Fig. 5 Albarelli per medicamenti, sec. XVIII. Museo della Verna (foto di A. Ferrini, Archivio fotografico della Verna). Fig. 3 Alambicco del XVII secolo, Museo della Verna (foto di A. Ferrini, Archivio fotografico della Verna). Fig. 4 Libri di medicina negli scaffali della farmacia (foto di S. Gatta). Fig. 5 Albarelli per medicamenti, sec. XVIII. Museo della Verna (foto di A. Ferrini, Archivio fotografico della Verna). Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna Importante è la notizia risalente al 1723-1724 della costruzione di un nuovo «stillatoio», con tanto di nuovi distillatori (Fig. 3). È dell’Ottocento l’arredo della bel- lissima farmacia – ancora oggi visitabile –, mentre i locali dei laboratori hanno subito solo nel secolo scorso un sostanziale cam- biamento di destinazione. Lo Speziale della Verna era general- mente un Frate Laico al quale l’Ordine dei Fig. 2 Veduta del convento de La Verna (foto di S. Gatta). mondo, i frati residenti allacciano intensi rapporti con il territorio circostante. No- nostante la loro presenza offra soprattutto assistenza religiosa e spirituale, essa si ca- ratterizza fin dall’inizio anche come atten- zione ai bisogni materiali delle popolazioni locali. È il caso degli aiuti alimentari di- stribuiti nei periodi di carestia ma è anche, soprattutto, la costante disponibilità ad ac- cogliere, a ricoverare, a curare ed aiutare malati ed infermi. Per l’attività terapeutica, che è sicuramente presente fin dagli inizi, abbiamo documentazione già dal ’400. La presenza nel convento di una spezieria è te- stimoniata nel 1462 attraverso una cronaca che segnala la morte di fra’ Pietro Francio- si, citato come infermiere. Lo stesso frate è raffigurato in un dipinto seicentesco, ora al Museo della Verna, e un’iscrizione lo defini- sce aromatario. Nel 1478 sono ricordati un fra’ Tommaso, anch’egli definito infermiere e la struttura in cui operava: l’infermeria. Due annotazioni amministrative del Libro dei Conti degli anni dal 1481 al 1518 ci ragguagliano su lavori di sistemazione ed arredo per l’infermeria, nel 1497 e poi nel 1515. In quest’ultimo caso troviamo la de- finizione di «infermeria nova», segno che ve n’era un’altra ormai non più funzionale. Altri lavori nel Cinquecento ci confermano che l’infermeria doveva essere una struttura ben organizzata se era in grado di ottene- re addirittura finanziamenti dalla duchessa Eleonora di Toledo, moglie di Cosimo I de’ Medici. Nei secoli XVI e XVII è attestata in numerosi documenti una «spetiaria» o «aromataria» con tanto di laboratori per le lavorazioni officinali. Nel Settecento ven- gono fatti importanti lavori di ammoder- namento, ristrutturazione e rinnovo delle dotazioni. Importante è la notizia risalente al 1723-1724 della costruzione di un nuovo «stillatoio», con tanto di nuovi distillatori (Fig. 3). È dell’Ottocento l’arredo della bel- lissima farmacia – ancora oggi visitabile –, mentre i locali dei laboratori hanno subito solo nel secolo scorso un sostanziale cam- biamento di destinazione. Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna La presenza alla Verna di un erbario come quello del Venturini non deve quindi mera- vigliare giacché la conoscenza delle piante medicinali era un requisito essenziale per i frati addetti alla ben documentata Spezieria del Santuario (Fig. 1). Fondato, come è noto, attorno agli anni 1213-1214, a seguito del dono del monte da parte del Conte Orlando di Chiusi, il Convento diviene ben presto, oltre che un centro di spiritualità e contem- plazione per i frati, un luogo di accoglienza ed assistenza caritatevole di viandanti e pel- O frate Lione, benchè ‘l frate Minore sapesse si’ bene predicare, che convertisse tutti gl’infedeli alla Fede di Cristo, iscrivi che non è ivi perfetta letizia2. 80 Secondino Gatta mondo, i frati residenti allacciano intensi rapporti con il territorio circostante. No- nostante la loro presenza offra soprattutto assistenza religiosa e spirituale, essa si ca- ratterizza fin dall’inizio anche come atten- zione ai bisogni materiali delle popolazioni locali. È il caso degli aiuti alimentari di- stribuiti nei periodi di carestia ma è anche, soprattutto, la costante disponibilità ad ac- cogliere, a ricoverare, a curare ed aiutare malati ed infermi. Per l’attività terapeutica, che è sicuramente presente fin dagli inizi, abbiamo documentazione già dal ’400. La presenza nel convento di una spezieria è te- stimoniata nel 1462 attraverso una cronaca che segnala la morte di fra’ Pietro Francio- si, citato come infermiere. Lo stesso frate è raffigurato in un dipinto seicentesco, ora al Museo della Verna, e un’iscrizione lo defini- sce aromatario. Nel 1478 sono ricordati un fra’ Tommaso, anch’egli definito infermiere e la struttura in cui operava: l’infermeria. Due annotazioni amministrative del Libro dei Conti degli anni dal 1481 al 1518 ci ragguagliano su lavori di sistemazione ed arredo per l’infermeria, nel 1497 e poi nel 1515. In quest’ultimo caso troviamo la de- finizione di «infermeria nova», segno che ve n’era un’altra ormai non più funzionale. Altri lavori nel Cinquecento ci confermano che l’infermeria doveva essere una struttura ben organizzata se era in grado di ottene- re addirittura finanziamenti dalla duchessa Eleonora di Toledo, moglie di Cosimo I de’ Medici. Nei secoli XVI e XVII è attestata in numerosi documenti una «spetiaria» o «aromataria» con tanto di laboratori per le lavorazioni officinali. Nel Settecento ven- gono fatti importanti lavori di ammoder- namento, ristrutturazione e rinnovo delle dotazioni. Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna Lo Speziale della Verna era general- t F t L i l l l’O di d i Fig. 2 Veduta del convento de La Verna (foto di S. Gatta). Fig. 2 Veduta del convento de La Verna (foto di S. Gatta). legrini (Fig. 2). Questi ultimi, che le raccol- te di miracoli già ci testimoniano numerosi in concomitanza con le presenze del Santo, aumenteranno di numero dopo la sua mor- te, quando verrà reso noto il prodigio delle Stimmate (settembre 1224). Benché scelto come romitorio per fuggire i clamori del Lo Speziale della Verna era general- mente un Frate Laico al quale l’Ordine dei Frati Minori aveva fatto fare solidi studi. I 81 Spezieria ed erbari tra scienza e carità Fig. 4 numerosi trattati scientifici che vanno dal XVI al XX secolo – conservati nella Biblio- teca del Convento (citiamo per tutti i volumi della enciclopedia naturalistica di Ulisse Aldrovandi) o ancora disposti negli scaffali della farmacia (Fig. 4) – danno piena testi- monianza di questo impegno di studio. Con un’adeguata formazione botanica, medica e farmaceutica, dunque, il frate speziale pre- parava le ricette magistrali per i confratelli ma anche per i pazienti che salivano al Con- vento. Non raramente si recava egli stesso al capezzale del malati per recare i medica- menti necessari (Fig. 5) o per interventi di assistenza sanitaria o di piccola chirurgia. Tutto questo in modo sempre gratuito. La vacchetta per ricette di spezieria, un quader- no manoscritto con coperta in pergamena in uso dal 1691 al 1807, testimonia mira- bilmente questa attività (Fig. 6). Vi sono annotati, insieme al medicamento e alla posologia, il destinatario della prescrizione – il paziente, diremmo oggi – il cui nome è generalmente segnato in testa alla ricetta: vi si trovano insieme nomi di frati e nomi di persone comuni, abitanti del territorio cir- costante, le cui famiglie, talvolta, sono an- cora oggi identificabili. Il servizio offerto al territorio dal frate farmacista, in particolare a favore degli abitanti delle isolate frazioni montane del Casentino e della Valtiberina, è stato più volte riconosciuto di fondamen- tale importanza sia dalle autorità locali che dai medici condotti, ed è rimasto attivo fino alla prima metà del secolo scorso. Fig. 5 Per le necessità delle preparazioni il frate speziale seguiva la coltivazione delle piante medicinali in un apposito orto dei semplici, la cui ubicazione è ancor oggi localizzabile. Fig. 6 Vacchetta per le ricette di Spezieria, secc. XVII-XVIII, Museo della Verna (foto di A. Ferrini, Archivio fotografi co della Verna). Fig. 7 I Discorsi di Pietro Andrea Mattioli, Venezia 1557, Museo della Verna (foto di Tekne restauro, Arezzo). L’erbario Venturini: note botaniche Lorenzo Lastrucci e Guido Moggi Lorenzo Lastrucci e Guido Moggi L ’erbario Venturini conservato al Conven- to dei frati francescani alla Verna si pre- senta come un unico volume in pergamena di dimensioni 28 x 22 e rilegato a fili, composto da 391 carte tutte numerate con numerazio- ne posta in alto a destra solo sul recto. Nel frontespizio si trova la scritta «Semplici in Natura / Racolti dal Dottor’ / Francesco Ma- ria / Venturini / In Fiorenza l’anno del Sig: / M.DCC.XI». Sempre nel frontespizio si tro- vano, in calce, due annotazioni manoscritte di mano e inchiostro diversi, datate rispetti- vamente 1711 e 1807 (Fig. 8). L (da c.1r a c.134v) (Fig. 9) ed una seconda, che parrebbe quasi da completare (da c.135r alla fine), con campioni per lo più non incol- lati e con la presenza di etichettine rettan- golari di carta sciolte, in cui viene riportato il nome della pianta; la grafia presente nelle etichettine appare coeva all’erbario mentre alcune notazioni sui fogli appaiono posteriori (grafia settecentesca, ottocentesca e alcune note risalenti probabilmente al 1928). I campioni presenti nell’erbario appaiono generalmente in discrete condizioni anche se sono presenti talvolta esemplari danneggiati e, soprattutto nella prima parte, nei fogli ap- paiono evidenti alcune macchie (in qualche caso anche molto estese) probabilmente do- vute all’umidità. I campioni essiccati sono collocati per lo più solo sul recto. Dalla c.1r alla c.134r i campioni sono numerati ed il loro numero complessivo risulta di 258; dalla c. 135r non sono più numerati. Fra la c. 135r e la c. 391r sono collocati circa 150 esemplari, tutti non incollati (escluso alcuni fra c.135r e 144r); molte carte sono prive di campioni e l’ultimo è collocato alla c.389r. Quasi tutte le piante sono incollate sulla carta e recano il nome scritto direttamente sul foglio; dalla c.142r il nome della pianta è scritto quasi sempre su una piccola etichetta volante. Alla c. 230r compare la notazione «Anno VI della Rivo- luzione / fascista / luglio 1928 / Ciampelli» dopo cui compaiono molti fogli vuoti, con qualche campione sparso, non incollato. I campioni presenti nell’erbario sono disposti di norma in numero di due o tre per foglio, anche se non mancano esempla- ri singoli. I campioni sono disposti in modo casuale (né in ordine alfabetico né sistema- tico); essi sono spesso parziali (un rametto, una foglia singola, infiorescenze) ed ap- paiono generalmente in buone condizioni. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Spezieria ed erbari tra scienza e carità: il Santuario francescano della Verna A queste produzioni medicinali si aggiunge- vano, quando arrivava il «tempo balsamico», la raccolta dei prodotti vegetali spontanei. Erbari come quello Venturini, ma prima ancora erbari a stampa, erano gli essenziali strumenti di conoscenza di questa attività. Un pregiato esemplare del 1557 de I Discorsi (...) nei sei libri della materia medicinale di Pedacio Dioscoride, del senese Pietro An- drea Mattioli, oggi esposto nel Museo (Fig. 7), mostra i segni di un lungo uso, sia nel laboratorio di spezieria che all’aria aperta: Fig. 5 82 SECoNdINo GATTA gocce di pioggia cadute sulle pagine durante un’erborizzazione nella foresta l’hanno in- fatti signifi cativamente segnato, regalandoci un’immagine vivissima della sua utilizzazio- ne sul campo. Un’ultima curiosità ci testimo- nia la storica familiarità dei religiosi con le erbe offi cinali: alcuni ampi locali nel sotto- tetto del Convento – oggi adibiti a magazzi- no – segnalano inequivocabilmente la loro antica funzione di essiccatoi per le essenze vegetali. Essi sono ancora chiamati con il nome della pianta principale che vi veniva lavorata: la belladonna. Fig. 6 Fig. 6 Fig. 7 Fig. 7 Fig. 7 L’erbario Venturini: note botaniche Lorenzo Lastrucci e Guido Moggi Essi sono costituiti per la quasi totalità da dicotiledoni (tra le eccezioni, un esemplare di Iris florentina, s. n., a c.323r) e consta- no di erbe e piante officinali di cui tutta- via non si riesce a risalire alle località di raccolta visto che mancano quasi ovunque riferimenti geografici anche approssimativi. Dalle poche località citate (Argentario, S. In sostanza l’erbario appare suddiviso al- meno in due parti: una prima, accurata, con campioni incollati, numerati e determinati Lorenzo Lastrucci e Guido Moggi 84 Fig. 8 Fig. 8 raccolta di erbe utilizzate prevalentemente a scopi galenici. I campioni presenti sono generalmente identificati con nomi italiani (es. Erba di S. Giovanni) e latini pre-linneani (in generale con uno o due termini) talvolta seguiti da al- cune note di commento. Le prime 11 carte portano sul verso il nome (o i nomi) attribuiti al campione presente sul recto; successiva- mente i nomi sono scritti per lo più sul recto, cioè sulla stessa carta dove si trova il relativo campione. In diversi casi i nomi sono cor- redati di citazioni di autori cinquecenteschi o seicenteschi che testimoniano da parte del realizzatore dell’erbario una certa conoscen- za delle opere classiche dei secoli XVI-XVII, anche se forse piuttosto sommaria. Gli auto- ri menzionati più frequentemente sono P.A. Mattioli e G. Bauhin; evidentemente per il primo l’autore dell’erbario si riferisce alle nu- merose edizioni dei Discorsi e dei Commen- tarii su Dioscoride (dal 1544 in poi) e per il secondo all’opera Pinax Theatri Botanici (del 1623). Tuttavia numerosi altri autori vengono citati: fra quelli cinquecenteschi ad es. D. Dodoens (Dodonaeus), M. de L’Obel (Lobe- lius), O. Brunfels (citato erroneamente «Bru- sf.»), L. Fuchs (citato erroneamente «Fusc.»), ecc. e fra i secenteschi C. de L’Ecluse (Clu- sius), mentre non si fa mai riferimento a J.P. de Tournefort, il cui lavoro Institutiones rei herbariae (1700) risultò a partire dal ’700 l’opera fondamentale di botanica per tutti gli studiosi. Da rilevare anche qualche citazione riferita a Plinio, Dioscoride e Teofrasto. Fig. 8 Giuliano) si può desumere che, almeno per parte delle piante contenute nell’erbario, la provenienza sia la Toscana ma risulta co- munque difficile avventurarsi in ipotesi più dettagliate. Allo stesso modo, se si esclude la data di realizzazione dell’erbario ripor- tata nell’intestazione dello stesso, non sono presenti ulteriori riferimenti cronologici per cui non è possibile desumere le date di rac- colta dei campioni. Fig. 10 La c. 45r, contenente tra gli altri un campione di Teucrium (n. 94) con una curiosa dicitura. Fig. 8 Frontespizio dell’Erbario Venturini, con la data di allestimento (1711). Fig. 9 Alcune delle pagine iniziali dell’Erbario. A sinistra la c. 1v (con i nomi delle piante presenti sulla carta precedente; da notare le numerose citazioni di autori) e a destra la c. 2r con un esemplare di amaranto (n. 4) e una foglia di asaro (n. 5). A destra la traccia di un campione di Caltha non più esistente (n. 6). Fig 10 La c 45r contenente tra gli Fig. 8 Frontespizio dell’Erbario Venturini, con la data di allestimento (1711). Fig. 9 Alcune delle pagine iniziali dell’Erbario. A sinistra la c. 1v (con i nomi delle piante presenti sulla carta precedente; da notare le numerose citazioni di autori) e a destra la c. 2r con un esemplare di amaranto (n. 4) e una foglia di asaro (n. 5). A destra la traccia di un campione di Caltha non più esistente (n. 6). Fig 10 La c 45r contenente tra gli L’erbario Venturini: note botaniche Lorenzo Lastrucci e Guido Moggi La stessa tipologia di piante contenute nell’erbario, costituite da specie erbacee non tutte strettamente offici- nali (come potrebbe apparire dal titolo del frontespizio), rende difficile desumere con certezza le finalità dell’erbario, anche se è plausibile che esso potesse servire come Fra le pagine più significative dell’erbario si possono segnalare anzitutto le prime 16 carte (contenenti una quarantina di campio- ni), sulle quali sono conservati gli esemplari in migliori condizioni e forniti di nomi spes- so dettagliati e chiari riferimenti ad autori cinque-seicenteschi. Fig. 8 Frontespizio dell’Erbario Venturini, con la data di allestimento (1711). Un campione piuttosto curioso è il n. 94 situato a c.45r (Fig. 10): si tratta di un ra- metto di un Teucrium (Labiatae) che porta la curiosa dicitura Pollius montanus / isopo del Coloña con / cui fù abeverato Giesù Xpo. Probabilmente l’autore fa riferimento al ramo d’issopo (Isopo del Coloña) che al momen- to della crocifissione fu usato da un soldato per avvolgervi la spugna impregnata di ace- L’erbario Venturini: note botaniche Fig. 10 Fig. 9 L’erbario Venturini: note botaniche 85 Fig. 9 Fig. 10 Fig. 10 Lorenzo Lastrucci e Guido Moggi 86 Lorenzo Lastrucci e Guido Moggi Fig. 12 Fig. 11 Fig. 11 Fig. 12 Fig. 12 87 L’erbario Venturini: note botaniche Fig. 13 Fig. 14 88 Lorenzo Lastrucci e Guido Moggi frontespizio; in merito a questa non vi sono però altri riferimenti. La seconda è il 1928 che è citata alla c.230r. Qui infatti, come si è detto, è posta la dicitura Anno VI della Rivoluzione / fascista / luglio 1928 / Ciam- pelli che farebbe pensare che in quella data l’erbario sia stato esaminato e controllato, ma non vi sono altri elementi per approfon- dire il problema (quali interventi si ebbero in quell’anno? Chi era Giuseppe Ciampelli? Cfr. anche alla c.218r) (Fig. 15). Come si deduce quindi molti restano gli interrogativi intorno a questo erbario, dalle località alle date di raccolta dei campioni in esso contenuti, dalle finalità alle vicende che lo hanno interessato nel 1807 e nel 1928. I punti fermi che si deducono dall’intestazione sono invece la data di composizione del vo- lume (1711) e il nome dell’autore, Francesco Maria Venturini. Resta anche il problema dell’identificazione di questo autore, di cui non si sono potute reperire notizie. 1 P. Luzzi, F. Fabbri, I tre Orti Botanici di Firenze, in S. Ferri, F. Vannozzi (a cura di), I Giardini dei Semplici e gli Orti Botanici della Toscana, Quattroemme, Giunta Reg. Toscana. Firenze 1993, pp. 49-68. Fig. 11 Un campione di Lotus (n. 214) a c. 98r per il quale sono indicate le proprietà curative . Fig. 12 Sulla c. 7r sono presenti un campione di Calendula (n. 12) ed uno parziale di Aquilegia (n. 13), rappresentato da una foglia e da un fiore. Fig. 11 Un campione di Lotus (n. 214) a c. 98r per il quale sono indicate le proprietà curative . Fig. 12 Sulla c. 7r sono presenti un campione di Calendula (n. 12) ed uno parziale di Aquilegia (n. 13), rappresentato da una foglia e da un fiore. Fig. 13 Un bel campione di Bagolaro o Celtis australis (n. 252 a c. 128r) definito Lotus arbor con riferimento a Mattioli. Fino al XVIII secolo il Bagolaro era infatti chiamato anche Loto. Fig. 14 Un campione di Erba lombrica (Scorpiurus muricatus L.) senza numero a c. 166r. Si noti a sinistra l’etichetta volante col nome e il riferimento all’autore (G.B. = Gaspar Bauhin). Fig 15 La c 230r con la notazione Fig. 14 Un campione di Erba lombrica (Scorpiurus muricatus L.) senza numero a c. 166r. Si noti a sinistra l’etichetta volante col nome e il riferimento all’autore (G.B. = Gaspar Bauhin). F 15 L 230 l i Fig. 15 La c. 230r con la notazione del 1928 (cfr. testo). Fig. 13 Un bel campione di Bagolaro o Celtis australis (n. 252 a c. 128r) definito Lotus arbor con riferimento a Mattioli. Fino al XVIII secolo il Bagolaro era infatti chiamato anche Loto. L’erbario Venturini: note botaniche Lorenzo Lastrucci e Guido Moggi Il fatto che sia definito «Dottor» farebbe pensare che si trattasse di un medico o di uno spezia- le, probabilmente operante a Firenze nella prima metà del ’700. Un’ipotesi che potrebbe essere avanzata riguarda eventuali collega- menti con l’Ospedale di S. Maria Nuova (o Ospedale di S. Egidio), esistente in Firenze fin dal XIII secolo ed in pieno esercizio nel XVIII. Da segnalare che alla metà del ’700 nell’area dell’Ospedale nel centro di Firenze era attivo anche un Giardino dei Semplici, dedicato alle erbe medicinali (evidentemente ad uso dei malati dell’ospedale), ricco di più di 1000 specie1. Fig. 15 to con cui fu dissetato Gesù (cfr. Vangelo di Giovanni, 19.29). È da notare tuttavia che il campione presente nell’erbario non è issopo (Hyssopus officinalis L.), ma – come si è detto – un Teucrium (molto probabilmente T. po- lium L.). Un secondo problema tuttora aperto ri- guarda l’anno e le modalità di arrivo del- l’erbario al Convento della Verna: attraverso quali vie infatti l’erbario è pervenuto alla Verna e in che periodo? Fu un dono dell’au- tore ai frati oppure fu un’opera commissio- nata? Quale collegamento può esservi stato fra questa piccola collezione – evidentemen- te incompleta – e la spezieria del convento, sicuramente a quell’epoca molto attiva? Allo stato attuale questi interrogativi restano senza risposta; non disperiamo tuttavia che un’analisi attenta dei documenti conservati nell’archivio del Convento della Verna possa risolvere almeno alcuni di questi dubbi. Anche se molte piante fra quelle presenti nell’erbario sono identificabili come specie medicinali, una sola volta si fa riferimento a proprietà curative, e cioè alla c.98r. Infatti qui il campione n. 214 porta la dicitura Loto Maggiore sive / Lotus pentafillos siliquosus et villosus / contra vermes experta quocumq. modo assumpta (Fig. 11, 12, 13 e 14). Due date poste nell’erbario farebbero pensare a controlli effettuati dopo il 1711, data che risulta nel frontespizio e che do- vrebbe essere l’anno di raccolta delle piante. La prima è il 1807 che figura sul L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio Paolo Emilio Tomei e Lucia Amadei L ’interesse degli studiosi italiani per la fl ora egiziana risale al XVI secolo quan- do il botanico Prospero Alpino1 si recò al Cairo dove condusse numerosi studi e ricer- che. Pubblicò poi diverse opere fra cui De plantis Aegypti, il primo testo sulla fl ora di questo paese. Successivamente il naturali- sta Vitaliano Donati fu in Egitto dal 1759 al ’60, colà inviato da Carlo Emanuele III per raccogliere materiale che doveva andare ad arricchire il Museo di Torino2. L È da qui che iniziò quel fl usso di personaggi che, dal nostro paese, andarono ad occupare numerosi punti chiave dell’amministrazione egiziana5. Forni condusse con sé Giambattista Broc- chi che raccolse numerose specie vegetali esplorando anche il Libano e il Sennar6. Successivamente il botanico fi orentino Giuseppe Raddi fu in Egitto al seguito della spedizione franco-toscana guidata da Fran- cois Champollion7. Ma il più eminente fra i naturalisti di questo periodo fu Antonio Fi- gari (1804-1870); recatosi colà come farma- cista, fu assunto alle dipendenze governative dapprima come Ispettore Farmacista e Pro- fessore della Scuola farmaceutica e botani- ca del Cairo, successivamente con l’incarico di eseguire ricerche minerarie nell’intero territorio egiziano. Per più di 40 anni poté così percorrere il paese raccogliendo reper- ti minerali, zoologici e vegetali. Genovese di nascita, fu allievo del botanico Domenico Viviani e per questa sua formazione, durante le sue attività eminentemente orientate ver- so uno scopo pratico, non trascurò mai l’os- servazione delle piante. Il Figari dette alle stampe un interessantissimo lavoro sulla Geografi a botanica dell’Egitto, dove illustra- va le diverse zone fl oristiche della regione e la loro suddivisione; quest’opera può essere considerata – come quella dell’Alpino per la fl ora – il primo contributo alla conoscenza fi togeografi a del territorio egiziano8 (Fig. 1). A partire dall’800 le presenze italiane nella terra del Nilo divennero costanti e nu- merose; infatti proprio in questo periodo il paese venne nuovamente scoperto dall’Eu- ropa3. Il primo luglio 1789 Napoleone era sbarcato ad Alessandria e oltre al suo eserci- to aveva portato anche un gruppo di 166 stu- diosi: fra questi fi guravano geologi, zoologi e botanici, dal lavoro dei quali prese forma e fu data alle stampe la Description de l’Egypte, dove questo paese veniva illustrato da tutti i punti di vista4. Fig. 1 Villaggio nel delta del Nilo - da G. Ebers, L’Egitto antico e moderno, Edoardo Perino editore–tipografo, roma 1893. 3 Cfr. l.A. Balboni, Gli italiani nella civiltà egiziana del secolo XIX. Alessandria d’Egitto, Stabilimento Tipo-litografi co V. Penasson, 1906; A. Siliotti, La scoperta dell’antico Egitto, White Star, Vercelli 1999. 4 Cfr. J.C. herold, Bonaparte in Egitto, Einaudi, Torino 1962; AA.VV., Description de l’Egypte. de l’imprimerie imperiale, Parigi 1809; E. Bresciani, Il richiamo della piramide. J.-F. Champollion e I. Rosellini in Egitto, in La Piramide e la Torre. Due secoli di archeologia egiziana, a cura di E. Bresciani, Cassa di risparmio di Pisa, 2000. 5 Cfr. Balboni, op. cit. 6 Giambattista Brocchi (1772- 1826). Studioso di geologia e paleontologia nato a Bassano, soggiornò in Egitto dal 1822 al Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 3 Cfr. l.A. Balboni, Gli italiani nella civiltà egiziana del secolo XIX. Alessandria d’Egitto, Stabilimento Tipo-litografi co V. Penasson, 1906; A. Siliotti, La scoperta dell’antico Egitto, White Star, Vercelli 1999. 1 Prospero Alpino (1553-1617). Nacque a Marostica, fu medico e botanico; dopo il soggiorno in Egitto divenne professore presso l’Università degli Studi di Padova. Il suo lavoro è rimasto per circa duecento anni l’unico contributo alla conoscenza della fl ora dell’Egitto in quanto sarà solo nel 1775 che comparirà la successiva Flora aegyptiaco arabica dello svedese Peter Forskal; r. Almagià, L’opera degli italiani per la conoscenza dell’Egitto e per il suo risorgimento civile ed economico, Provveditorato generale dello Stato, roma 1926; P. Alpino, De Plantis Aegypti, Venezia 1592. 2 Vitaliano donati (1717-1762). Padovano, fu professore di botanica nella reale Università degli Studi di Torino. la maggior parte delle piante raccolte dal donati – che perse la vita in quel viaggio – inizialmente stivate in quattro casse, fu guastata dagli insetti. In seguito fu pubblicato un elenco delle 84 che erano rimaste; Almagià, op. cit.; G. Bonino, Biografi a medica piemontese, 2, Torino 1825, pp. 169-171. L’erbario egiziano di Jacob Corinaldi Fig. 1 1 Prospero Alpino (1553-1617). Nacque a Marostica, fu medico e botanico; dopo il soggiorno in Egitto divenne professore presso l’Università degli Studi di Padova. Il suo lavoro è rimasto per circa duecento anni l’unico contributo alla conoscenza della fl ora dell’Egitto in quanto sarà solo nel 1775 che comparirà la successiva Flora aegyptiaco arabica dello svedese Peter Forskal; r. Almagià, L’opera degli italiani per la conoscenza dell’Egitto e per il suo risorgimento civile ed economico, Provveditorato generale dello Stato, roma 1926; P. Alpino, De Plantis Aegypti, Venezia 1592. 2 Vitaliano donati (1717-1762). Padovano, fu professore di botanica nella reale Università degli Studi di Torino. la maggior parte delle piante raccolte dal donati – che perse la vita in quel viaggio – inizialmente stivate in quattro casse, fu guastata dagli insetti. In seguito fu pubblicato un elenco delle 84 che erano rimaste; Almagià, op. cit.; G. Bonino, Biografi a medica piemontese, 2, Torino 1825, pp. 169-171. 3 Cfr. l.A. Balboni, Gli italiani nella L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio Dopo la presenza francese, il vicerè d’Egitto Mohammed Ali aprì le porte agli eu- ropei non escludendo gli italiani; il milanese Giuseppe Forni, che dal 1815 aveva esplo- rato queste contrade alla ricerca di sostanze nitrose, nel 1821 ritornò in Italia su ordine del viceré per l’ingaggio di collaboratori, in particolare «ingegneri minerari» per la ri- cerca e lo sfruttamento delle miniere, nonchè medici e farmacisti per gli ospedali militari. 92 Paolo Emilio Tomei e Lucia Amadei Anche durante il soggiorno in quel paese non abbandonò il suo interesse per la botanica, approfittando di raccogliere preziosi reperti da studiare. Collezionò infatti diverse pian- te, in particolare durante l’anno 1826, con le quali preparò un erbario di flora egizia- na. Dopo 5 anni dal suo ritorno in Italia, nel 1831, Corinaldi donò all’Accademia Valdar- nese una prima collezione di 42 piante rac- colte in Egitto. Nel Rapporto delle adunanze tenute dall’Accademia valdarnese nell’anno 1831 si legge infatti: (è stato donato) «dallo zelantissimo Socio Sig. Jacob Corinaldi un fascicolo contenente n.42 piante egizie, con la descrizione annessa a ciascheduna specie contenente le notizie da lui medesimo pre- se sul posto. L’intiera collezione poi che egli promette all’Accademia, sarà composta di 150 specie, ed a questa unirà i semi di tutte quelle medesime piante egizie da esso con successo coltivate in Pisa, acciò ne sia pure tentata la cultura nelle nostre campagne». È in questo ampio contesto storico-culturale che si inserisce la figura di Jacob Corinaldi. È in questo ampio contesto storico-culturale che si inserisce la figura di Jacob Corinaldi. Egli nacque a Ferrara il 15 dicembre 1782, ma trascorse gran parte della sua vita in To- scana e morì a Pisa, il 23 marzo 1847. Il poco che si conosce di lui riguarda la sua passione per le scienze naturali, principalmente la bo- tanica: amava soprattutto occuparsi di alghe, che andava a ricercare lungo le coste toscane, in particolare nel livornese. Le sue indagini su questo gruppo di vegetali e i ritrovamenti che egli fece contribuirono notevolmente a in- tegrare le conoscenze di allora sull’argomen- to. Per questa sua passione si trovò in contatto con noti algologi del tempo, ai quali era solito fornire materiale da studiare, frutto delle sue erborizzazioni (Fig. 2). Fu membro di varie accademie italiane ed estere, in particolare dell’Accademia Valdar- nese del Poggio, nella quale ricoprì l’incari- co di conservatore del Museo e presidente. Fig. 2 Lettera di Corinaldi al noto algologo G. Meneghini, conservata nell’archivio del Museo Botanico dell’Università di Pisa. 1826 con l’incarico di curare l’attivazione di nuove miniere metallifere. Durante questo soggiorno viaggiò molto nel territorio egiziano riportando in un «Giornale di viaggio» le sue dettagliate osservazioni riguardo ai luoghi, alla natura e ai costumi delle popolazioni. Con la piante che raccolse compilò un Erbario, conservato oggi a Bassano; G. Busnardo, Gli erbari Brocchi Montini Parolini riordinati da Giuseppe Marchente, «Boll. Mus. Civ. Bassano» 3/6 (1987-88), 1990, pp. 83-94; S. Pernigotti, L’avventura egiziana di Giambattista Brocchi (1772-1826), Atti del Convegno. Bassano del Grappa 9-10 novembre, 1985, pp. 103-124. 7 Giuseppe Raddi (1770-1829). Le piante raccolte da Raddi in Egitto – così come quelle riportate da una precedente spedizione in Brasile – sono conservate nel Museo Botanico di Pisa e nella Sezione Botanica del Museo di Storia Naturale di Firenze; E. Francini Corti, Giuseppe Raddi (1770-1829), in G. Raddi, Flora Brasiliana, Roma 1976; R.E.G. Pichi Sermolli, M.P. Bizzarri, A revision of Raddi’s pteridological collection from Brazil (1817-1818), «Webbia», 60(1), 2005, pp. 1-393; P.E. Tomei, Le raccolte botaniche di Giuseppe Raddi in Egitto, Atti del Convegno «Ippolito Rosellini: passato e presente di una disciplina», Pisa, 30-31 maggio 1982, suppl. a «Evo», 3, 1982, pp. 25-31; P.E. Tomei, R. M. Baldini, L. Amadei, S. Maccioni, Le raccolte egiziane conservate nell’Herbarium Horti Pisani, «Museologia scientifica», 20 (2), 2005, pp. 235-333. 7 Giuseppe Raddi (1770-1829). Le piante raccolte da Raddi in Egitto – così come quelle riportate da una precedente spedizione in Brasile – sono conservate nel Museo Botanico di Pisa e nella Sezione Botanica del Museo di Storia Naturale di Firenze; E. Francini Corti, Giuseppe Raddi (1770-1829), in G. Raddi, Flora Brasiliana, Roma 1976; R.E.G. Pichi Sermolli, M.P. Bizzarri, A revision of Raddi’s pteridological collection from Brazil (1817-1818), «Webbia», 60(1), 2005, pp. 1-393; P.E. Tomei, Le raccolte botaniche di Giuseppe Raddi in Egitto, Atti del Convegno «Ippolito Rosellini: passato e presente di una disciplina», Pisa, 30-31 maggio 1982, suppl. a «Evo», 3, 1982, pp. 25-31; P.E. Tomei, R. M. Baldini, L. Amadei, S. Maccioni, Le raccolte egiziane conservate nell’Herbarium Horti Pisani, «Museologia scientifica», 20 (2), 2005, pp. 235-333. 1826 con l’incarico di curare l’attivazione di nuove miniere metallifere. Durante questo soggiorno viaggiò molto nel territorio egiziano riportando in un «Giornale di viaggio» le sue dettagliate osservazioni riguardo ai luoghi, alla natura e ai costumi delle popolazioni. Con la piante che raccolse compilò un Erbario, conservato oggi a Bassano; G. Busnardo, Gli erbari Brocchi Montini Parolini riordinati da Giuseppe Marchente, «Boll. Mus. Civ. Bassano» 3/6 (1987-88), 1990, pp. 83-94; S. Pernigotti, L’avventura egiziana di Giambattista Brocchi (1772-1826), Atti del Convegno. Bassano del Grappa 9-10 novembre, 1985, pp. 103-124. Cfr. AA.VV., op. cit. 14 Le piante nominate da Corinaldi come nuove per la flora dell’Egitto sono le seguenti: Agrostis stolonifera L., Panicum zonale Guss., Trigonella laciniata L., Lasiopogon muscoides DC. fra le fanerogame; Sargassum diversifolium Ag., Cystosira ericoides Ag., Sphaerococcus confervoides Ag., Cystosira abrotanifolia Ag., Cystosira discors Ag. fra le alghe. Cfr. AA.VV., op. cit. 12 Cfr. J. Corinaldi, Piante egiziane raccolte dal dott. Jacob Corinaldi l’anno 1826, e dal medesimo donate all’Accademia Valdarnese del Poggio nell’Ottobre 1831, «Memorie Valdarnesi», 3, 1842, pp. 73-97. 11 Cfr. F. Parlatore, Notizie botaniche, «Giorn. Bot. Ital.», 1(3), 1847 pp. 89-91. 8 A. Figari, Studi scientifici sull’Egitto e le sue adiacenze, compresa la penisola dell’Arabia Petrea, Tip. G. Giusti, Lucca 1865. 9 Cfr. AA.VV., Atti della terza riunione degli scienziati italiani tenuta in Firenze nel settembre del 1841, Firenze, 1841; P.A. Saccardo, La botanica in Italia, Bologna, 1895; I. Cantù, L’Italia scientifica contemporanea. Notizie sugli italiani ascritti ai cinque primi Congressi, attinte alle fonti più autentiche, Stella, Milano 1844. 13 Cfr. AA.VV., op. cit. 8 A. Figari, Studi scientifici sull’Egitt e le sue adiacenze, compresa la penisola dell’Arabia Petrea, Tip. G. Giusti, Lucca 1865. 9 Cfr. AA.VV., Atti della terza riunione degli scienziati italiani tenuta in Firenze nel settembre del 1841, Firenze, 1841; P.A. Saccardo La botanica in Italia, Bologna, 1895; I. Cantù, L’Italia scientifica contemporanea. Notizie sugli italian ascritti ai cinque primi Congressi, attinte alle fonti più autentiche, Stella, Milano 1844. 10 Cfr. G. Branchi, Risposta del Chiarissimo sig. Dottore Giuseppe Branchi, Professore di Chimica nell’I. e R. Università di Pisa, al Dottore Iacob Corinaldi, membro di varie Accademie scientifiche, Pisa, 29 ottobre 1832; G. Branchi, Lettera del chiarissimo sig. Professor Giuseppe Branchi al Dott. Jacob Corinaldi, Pisa, 6 luglio 1841; J. Corinaldi, Descrizione di alcune crittogame trovate nel valdarno di sopra, Pisa 1818; J. Corinaldi, Lettera al Sig. Antonio Bottari Dottore in Farmacia a Serravezza, Pisa, 5 giugno 1821; J. Corinaldi, Sull’Anzarut. Lettera del Dottore Iacob Corinaldi, al Chiarissimo sig. Dottore Giuseppe Branchi, Professore di Chimica nell’I. e R. Università di Pisa, Pisa, 18 ottobre 1832; J. Corinaldi, Elenco di alcune alghe del mare Labronico, Pisa 1839; J. J. Corinaldi, Sulla Polysyphonia parasitica, Atti Congr. Scienz. Ital. in Lucca, 1843; G. Meneghini, Lettera del professore Giuseppe Meneghini al dottore Jacob Corinaldi, Padova, 9 novembre 1840, «Giorn. Toscano di Scienze Mediche, Fisiche e Naturali», 2, 1840a, pp. 1-4; G. Meneghini, Di alcune nuove specie di Alghe. Lettera del Prof. Giuseppe Meneghini al Dott. Jacob Corinaldi a Pisa, Padova, 23 maggio 1840, Tipografia Prosperi, Pisa 1840b; G. Meneghini, Alghe mediterranee italiane (con lettera al sig. Dott. Jacob Corinaldi. Padova, 15 gennaio 1841), Tipografia Nistri, Pisa 1841; G. Meneghini, Quattro nuove specie di alghe trovate dal dottore Jacob Corinaldi ai bagni di S. Giuliano di Pisa e pubblicate dal professore Giuseppe Meneghini di Padova, s. d. 11 Cfr. F. Parlatore, Notizie botaniche «Giorn Bot Ital » 1(3) «Memorie Valdarnesi», 3, 1842, pp. 73-97. L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio In veste di deputato per l’Accademia stessa, partecipò alla prima e alla terza Riunione degli scienziati italiani9. Come cultore di storia naturale e come socio dell’Accademia rimase in contatto epistolare con diversi na- turalisti suoi contemporanei, pubblicandone talvolta la corrispondenza10. Dopo questa donazione ne seguì una se- conda fino ad arrivare però a un totale di 83 specie e non 150 come era stato preannuncia- to. Nel 1842 infine Corinaldi pubblicò un ca- talogo in cui enumerava tutte le piante donate all’Accademia e ne illustrava le caratteristiche, commentando per ciascuna di esse eventuali utilizzi, luoghi di ritrovamento e curiosità12. Negli anni 1825-26 si recò in Egitto, dove insegnò al Cairo come pubblico professore11. 93 L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio Fig. 3 Campione di Agrostis stolonifera L. Fig. 3 Campione di Agrostis stolonifera L. La collezione fu presentata, insieme al catalogo, nel corso della terza Riunione de- gli Scienziati italiani in Firenze, alla quale Corinaldi partecipò come presidente dell’Ac- cademia Valdarnese13. Durante questo inter- vento egli mise in evidenza un certo numero di specie spontanee da lui ritrovate in Egitto e all’epoca non ancora indicate per la flora di quel paese14 (Figg. 3, 4, 5, 6). Dopo di ciò, sempre a nome dell’Acca- demia Valdarnese del Poggio, mostrò e di- stribuì in dono ai membri della «Sezione di Botanica e Fisiologia vegetabile» presenti al convegno esemplari secchi di tali specie. Dopo di ciò, sempre a nome dell’Acca- demia Valdarnese del Poggio, mostrò e di- stribuì in dono ai membri della «Sezione di Botanica e Fisiologia vegetabile» presenti al convegno esemplari secchi di tali specie. Fig. 5 Fig. 6 Fig. 8 Fig. 9 F 6 Fig 5 Fig. 4 Fig. 5 Fig. 4 Fig. 6 Fig. 9 Fig. 8 Fig. 8 Fig. 9 Fig. 9 Fig. 11 Fig. 12 Fig. 12 L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio 95 Fig. 13 Fig. 14 Fig. 14 Fig. 13 Questi campioni si trovano oggi conservati in diversi erbari italiani tra cui Firenze, Torino, Roma e Pisa15. trova un piccolo cartellino con il numero pro- gressivo stampato in caratteri assai grandi. In terza pagina è fissato il campione, accom- pagnato da un’etichetta stampata con l’indi- cazione del nome della pianta e della località di raccolta. Corinaldi si occupò anche di indagini sulla mirra, cercando di individuare l’albero dal quale la medesima veniva ricavata, es- sendo allora questo un problema di difficile interpretazione16. I fogli sono quindi tutti raccolti in una camicia che ha impresso a stampa il titolo Piante egiziane raccolte nel MDCCCXXVI dal dottore Jacob Corinaldi di Pisa e dal me- desimo donate all’Accademia Valdarnese del Poggio nell’ottobre MDCCCXXXI (Fig. 10). Descrisse inoltre alcuni frutti da lui ac- quistati nelle drogherie del Cairo e donati al Museo dell’Accademia17: tra questi ricordava un frutto di Hyphane tebaica che indicava come proveniente dalla necropoli di Sakka- ra. Infine si occupò anche della sostanza in- dicata dagli arabi con il nome di «anzarut», che giudicò essere sarcocolla (Penaea mu- cronata e P. sarcocolla)18. Il tutto è infine contenuto in una cartella di cartone che ha sulla copertina un’etichetta con la scritta «Erbario Egiziano Raccolto e Dichiarato dal M. On. Dr. Jacob Corinaldi di Pisa»19 (Fig. 11). I luoghi delle erborizzazioni di Corinaldi si trovavano tutti nell’area che andava dalla costa marittima nei pressi di Alessandria fino al Cairo: in particolare egli raccolse piante sulle coste e sulle scogliere del por- to di Alessandria, lungo le rive del Nilo e di canali suoi affluenti nei pressi del Cai- ro, nelle campagne del Cairo e sull’isola di Roda. Corinaldi preparò anche due Erbari di flo- ra valdarnese che donò all’Accademia dove sono tuttora conservati (Figg. 7, 8, 9). 15 G. Forneris, “Spigolature” nelle collezioni dell’Erbario torinese (TO), «Webbia», 48, 1993, pp. 267-285; M. Iberite, P. Marchi, A. Millozza, Museo dell’Erbario di Roma, in I Musei dell’Università “La Sapienza” a cura di M. Barbanera e I. Venafro, Istituto Poligrafico e Zecca dello Stato, Roma 1993, pp. 77-91; Tomei, Baldini, Amadei, Maccioni, Fig. 13 Campione di Gnaphalium luteo-album L. 19 Recentemente l’erbario è stato restaurato a cura dell’Accademia e reso noto e accessibile il suo contenuto. A questo proposito cfr.: P.E. Tomei, L’erbario egiziano 16 Cfr. J. Corinaldi, Osservazioni sulla mirra, «Memorie Valdarnesi», 1, Pisa 1835a, pp. 68-72. Fig. 4 Campioni di alghe raccolti nel porto di Alessandria. Fig. 5 Campione di Lasiopogon muscoides DC. Fig. 6 Campione di Panicum zonale Guss. Fig. 7 Frontespizio di uno dei due erbari della flora del Valdarno superiore donati all’Accademia Valdarnese del Poggio. Fig. 8 Campione di felce appartenente alla flora del Valdarno. Fig. 9 Campioni di muschi appartenenti alla flora del Valdarno. Fig. 10 Frontespizio dell’Erbario egizio. Fig. 11 Copertina dell’Erbario egizio. Fig. 12 Campione di Caulerpa prolifera Lamk. Fig. 13 Campione di Gnaphalium luteo-album L. Fig. 14 Campione di Convolvulus arvensis L 17 Cfr. J. Corinaldi, Cenni sopra alcuni frutti trovati nelle drogherie del Cairo, «Memorie Valdarnesi», 1, Pisa 1835b, pp. 72-77. Fig. 14 Campione di Convolvulus arvensis L. Tomei, Baldini, Amadei, Maccioni, op. cit. Fig. 12 Campione di Caulerpa prolifera Lamk. 15 G. Forneris, “Spigolature” nelle collezioni dell’Erbario torinese (TO), «Webbia», 48, 1993, pp. 267-285; M. Iberite, P. Marchi, A. Millozza, Museo dell’Erbario di Roma, in I Musei dell’Università “La Sapienza” a cura di M. Barbanera e I. Venafro, Istituto Poligrafico e Zecca dello Stato, Roma 1993, pp. 77-91; Tomei, Baldini, Amadei, Maccioni, op. cit. 16 Cfr. J. Corinaldi, Osservazioni sulla mirra, «Memorie Valdarnesi», 1, Pisa 1835a, pp. 68-72. 17 Cfr. J. Corinaldi, Cenni sopra alcuni frutti trovati nelle drogherie del Cairo, «Memorie Valdarnesi», 1, Pisa 1835b, pp. 72-77. 18 Cfr. Corinaldi, Sull’Anzarut , cit. 19 Recentemente l’erbario è stato restaurato a cura dell’Accademia e reso noto e accessibile il suo contenuto. A questo proposito cfr.: P.E. Tomei, L’erbario egiziano Corinaldi a Montevarchi, in: E. Gusmeroli, L. Lastrucci (a cura di) Atti del Convegno Evoluzione delle conoscenze botaniche e problematiche della conservazione in provincia di Arezzo da Andrea Cesalpino ad oggi, Provincia di Arezzo, Università degli Studi di Firenze, Arti grafiche Cianferoni, Stia, 2006; P.E. Tomei, L. Amadei, Notizie sull’erbario egiziano di Jacob Corinaldi, Atti dell’Accademia Valdarnese del Poggio, Memorie Valdarnesi, anno 163, serie VII, fascicolo XI:119-132, 1998. In Appendice viene presentata una tabella che riporta le specie presenti nella raccolta, messe a confronto con quelle illustrate nel catalogo. 16 Cfr. J. Corinaldi, Osservazioni sulla mirra «MemorieValdarnesi» 18 Cfr. Corinaldi, Sull’Anzarut , cit. 19 , g Corinaldi a Montevarchi, in: E. Gusmeroli, L. Lastrucci (a cura di) Atti del Convegno Evoluzione delle conoscenze botaniche e problematiche della conservazione in provincia di Arezzo da Andrea Cesalpino ad oggi, Provincia di Arezzo, Università degli Studi di Firenze, Arti grafiche Cianferoni, Stia, 2006; P.E. Tomei, L. Amadei, Notizie sull’erbario egiziano di Jacob Corinaldi, Atti dell’Accademia Valdarnese del Poggio, Memorie Valdarnesi, anno 163, serie VII, fascicolo XI:119-132, 1998. In Appendice viene presentata una tabella che riporta le specie presenti nella raccolta, messe a confronto con quelle illustrate nel catalogo. L’erbario L’erbario è costituito attualmente da 67 cam- pioni vegetali, numerati e fissati ai fogli con sottile filo di refe. I fogli sono di carta filigra- na, delle dimensioni di cm 50x36, ripiegati a metà. Sulla prima pagina in alto a destra, si Anche dopo essere tornato in Italia, ri- mase in contatto epistolare con amici in Egitto che raccoglievano per lui campioni ve- Paolo Emilio Tomei e Lucia Amadei 96 Fig. 16 Fig. 17 Fig. 17 Fig. 16 Fig. 17 Fig. 15 Fig. 16 Fig. 17 Fig. 15 Fig. 15 Fig. 16 Fig. 15 getali, in particolare alghe, e glieli inviavano (Figg. da 12 a 18). nale – a sinistra – e quella attuale – a de- stra. Quando la revisione ha confermato l’identificazione e la nomenclatura di Co- rinaldi niente è stato aggiunto. Il segno = indica conferma dell’identificazione e ag- giornamento della nomenclatura, l’assenza di questo segno indica invece che il nome è di nuova identificazione. Nelle colonne di destra sono riportate rispettivamente la nu- merazione dei campioni nell’ambito dell’er- bario e del catalogo. Nell’introduzione al Catalogo delle piante donate all’Accademia, che Corinaldi aveva pubblicato nel 1842, egli specificava come il territorio egiziano fosse per lo più costituito da aree desertiche in cui i vegetali scarseggiava- no, tanto che «il botanico pochi può sperare di raccoglierne anche dietro lunghe e penose pe- regrinazioni». Le aree irrigate dal Nilo, sotto- poste ad intensa coltivazione, mostravano una flora che era il risultato della mescolanza fra le piante indigene e le esotiche, introdottevi in epoche diverse e gradualmente naturalizzate. Questi motivi «mi determinarono di erborizza- re nelle terre incolte, giudicando che in quelle avrei trovato le vere piante spontanee». Il primo gruppo di piante, numerate da 1 a 9, è costituito da alghe marine, mentre per il resto si tratta di fanerogame. Le entità vegetali del catalogo sono nu- merate (1-83); mettendo a confronto le due liste si nota come, tranne in pochi casi, non esista una esatta corrispondenza fra le spe- cie dell’erbario e quelle del catalogo: anche se le specie nominate sono per lo più le stes- se, le rispettive posizioni sono generalmente diverse. Fig. 15 Campione di Cyperus olivaris Targ. Tozz. Fig. 16 Campione di Poa cynosuroides Willd. Fig. 17 Campione di Cystosira amentacea Bory L’erbario La collezione di piante egiziane prepara- te da Corinaldi e conservata presso l’Acca- demia Valdarnese, nonostante il numero di campioni estremamente esiguo che raccoglie, è comunque interessante per la presenza di alcune specie allora nuove ed è indicativa di quell’interesse per la storia naturale del- l’Egitto che attraverso i secoli gli italiani hanno sempre dimostrato. Per ciascuna di esse viene indicato il nome dato da Corinaldi, il nuovo nome con cui è stato identificato il campione nel corso della revisione sistematica e la numerazione nell’ambito dell’erbario e del catalogo (Tab. 1). Da ciò compare immediatamente che, della serie completa illustrata nel catalogo, mancano all’erbario almeno 16 campioni (n. 12, 13, 16, 17, 20, 21, 24, 29, 32, 77, 78, 79, 80, 81, 82, 83). Fig. 18 Campione di Francoeuria crispa Cass. Appendice Le specie presenti nella raccolta sono elencate secondo la nomenclatura origi- 97 97 L’erbario egiziano di Jacob Corinaldi dell’Accademia Valdarnese del Poggio SPECIE erb. cat. Sargassum vulgare Ag. 1 1 Sargassum vulgare var. confertum Ag. 2 Sargassum diversifolium Ag. - 2 Cystosira ericoides Ag. - 3 Cystosira amentacea Bory. 3 Cystosira abrotanifolia Ag. 4 4 Cystosira abrotanifolia Ag. var. Boryana Menegh. 5 4 bis Cystosira discors Ag. - 5 Caulerpa prolifera Lmx. 9 6 Rytiphlaea tinctoria Ag. 7 7 Sphaerococcus confervoides Ag. 6 8 Chondria obtusa Ag. 8 9 Fimbristylis dichotomum Vahl. = Fimbristylis bis-umbellata (Forsk.) Bub. 10 10 Cyperus olivaris Targ. Tozzett. = Cyperus rotundus L. 11 11 Polypogon monospeliensis Rom et Schult. - 12 Agrostis stolonifera L. Agrostis semiverticillata (Forsk. ) Christens. - 13 Crypsis aculeata Willd. 14 14 Aristida pungens Desf. Aristida acutiflora Trin. et Rupr. 15 15 Phalaris paradoxa L. - 16 Saccarum cylindricum Lamk. - 17 Saccarum aegyptiacum Willd. - 18 Panicum zonale Guss. Echinochloa colonum (L.) Link. 18 19 Panicum leiogonum Delil. Echinochloa colonum (L.) Link. 19 20 Lolium perenne L. - 21 Poa aegyptiaca Willd. - 22 Poa cynosuroides Willd. = Desmostachya bipinnata (L.) Stapf 22 - Avena fatua L. Avena alba Vahl. 23 - Poa cynosuroides Willd. 23 Arundo phragmitis L. 24 Halochnemum strobilaceum M. Br. (1) Salicornia glauca Delile (2) = Arthrocnemum macrostachyum (Moric.) Moris (2) 25 25 Atriplex coriacea Forsk 26 27 Fig. 18 Campione di Francoeuria crispa Cass. Fig. 18 C crispa C SPECIE erb. cat. Sargassum vulgare Ag. 1 1 Sargassum vulgare var. confertum Ag. 2 Sargassum diversifolium Ag. - 2 Cystosira ericoides Ag. - 3 Cystosira amentacea Bory. 3 Cystosira abrotanifolia Ag. 4 4 Cystosira abrotanifolia Ag. var. Boryana Menegh. 5 4 bis Cystosira discors Ag. - 5 Caulerpa prolifera Lmx. 9 6 Rytiphlaea tinctoria Ag. 7 7 Sphaerococcus confervoides Ag. 6 8 Chondria obtusa Ag. 8 9 Fimbristylis dichotomum Vahl. = Fimbristylis bis-umbellata (Forsk.) Bub. 10 10 Cyperus olivaris Targ. Tozzett. = Cyperus rotundus L. 11 11 Polypogon monospeliensis Rom et Schult. - 12 Agrostis stolonifera L. Agrostis semiverticillata (Forsk. ) Christens. - 13 Crypsis aculeata Willd. 14 14 Aristida pungens Desf. Aristida acutiflora Trin. et Rupr. 15 15 Phalaris paradoxa L. - 16 Saccarum cylindricum Lamk. - 17 Saccarum aegyptiacum Willd. - 18 Panicum zonale Guss. Echinochloa colonum (L.) Link. 18 19 Panicum leiogonum Delil. Echinochloa colonum (L.) Link. 19 20 Lolium perenne L. Appendice - 21 Poa aegyptiaca Willd. - 22 Poa cynosuroides Willd. = Desmostachya bipinnata (L.) Stapf 22 - Avena fatua L. Avena alba Vahl. 23 - Poa cynosuroides Willd. 23 Arundo phragmitis L. 24 Halochnemum strobilaceum M. Br. (1) Salicornia glauca Delile (2) = Arthrocnemum macrostachyum (Moric.) Moris (2) 25 25 Atriplex coriacea Forsk. 26 27 Salsola glomerulata Delil. - 26 Paolo Emilio Tomei e Lucia Amadei 98 Polygonum maritimum L. 27 28 Passerina hirsuta L. Reaumuria vermiculata L. 28 29 Ficus sycomorus L. - 30 Croton obliquifolium Vis. 31 Statice monopetala L. = Limoniastrum monopetalum (L.) Boiss. 30 32 Plantago major L. 31 33 Hyosciamus albus L. - 34 Solanum esculentum Dum. = Solanum melongena L. 33 35 Heliotropium europaeum L. 34 36 Lithospermum callosum 37 Convolvulus arvensis var. ? Convolvulus arvensis L. 35 38 Cressa cretica L. 36 39 Sesamum orientale L. Sesamum indicum L. 37 40 Cynanchum monspeliacum Willd.. = Cynanchum acutum L 38 41 Zollikoferia chondrilloides DC. = Launaea resedifolia O.Kuntze 39 42 Sonchus ciliatus Lamk. = Sonchus oleraceus L. 40 43 Gnaphalium luteo-album L. 41 44 Gnaphalium luteo-album L. var. Gnaphalium luteo-album L. 42 - Gnaphalium niliaceum Spreng. = Gnaphalium indicum L. 43 45 Lasiopogon muscoides DC. Filago desertorum Pomel 44 46 Trichogyne cauliflora DC. - 47 Gnaphalium trifidum Thumb. ? = Gnaphalium luteo-album L. 45 - Conyza aegyptiaca (L.) Ait. 46 48 Pulicaria arabica Cass. - 49 Francoeuria crispa Cass. = Francoeuria crispa (Forsk.) Cass. 47 50 Anthemis cairica Vis. - 51 Senecio arabicus L. Senecio aegyptius L. 48 52 Senecio aegyptius L. 49 53 Ceruana pratensis Forsk. 50 54 Cotula anthemoides L. 51 55 Ambrosia maritima L. 52 56 Ammi visnaga Lamk = Ammi visnaga .(L.) Lam. 53 57 Anethum graveolens L. non identificabile 54 58 Coriandrum sativum L. non identificabile 55 59 Cucumis colocynthis L. = Citrullus colocynthis (L.) Schrad. 56 60 s.n. non identificabile 57 - Luffa aegyptiaca Mill. - 61 Potentilla supina L. non identificabile 58 62 Glinus lotoides L. - 63 Lawsonia alba Lamk. Lawsonia inermis L. 59 64 Tamarix senegalensis DC. non identificabile 60 65 Cassia fistula L. 61 66 Cassia sophera L. 67 Cassia senna L. Cassia italica (Mill.) Lam. ex Steud. 62 68 Cassia absus L. 63 69 Acacia lebbek Willd. 64 70 Parkinsonia aculeata L. 65 71 Melilotus parviflora Desf. = Melilotus indica (L.) All. 66 72 Vicia sativa L. var. floribus ... Appendice Visiani Vicia sativa L. 67 73 Sesbania aegyptiaca Pers. = Sesbania sesban (L.) Merrill 68 74 Trigonella laciniata L. 69 75 Trigonella foenum-graecum L. 70 76 Alhagi maurorum Tournef. = Alhagi maurorum Medic. 71 77 Nigella sativa L. 72 78 Enarthrocarpus lyratus DC. 73 79 Sisymbrium irio L. 74 80 Frankenia revoluta Forsk. 75 81 Alsine succulenta Delil. - 82 Arenaria media L. non identificabile 76 83 Reaumuria vermiculata L. Herbarium M. Padulae Michele Padula cenosi del Monte Falco è stata censita con un rilevamento fl oristico successivamente pubblicato1. La formazione dell’erbario risale al 1965. La raccolta di piante vascolari ha inte- ressato soprattutto l’Appennino tosco-ro- magnolo, in particolare il territorio delle foreste demaniali forlivesi e casentinesi; peraltro, inizialmente, sono stati raccolti anche campioni di varie regioni d’Italia in occasione di escursioni. La raccolta era fi - nalizzata alla conoscenza e allo studio della fl ora e del relativo ambiente in cui si trova- no le piante, per ricavarne utili indicazio- ni anche ai fi ni della gestione delle foreste stesse. Si è realizzato così, in un periodo di quasi 30 anni, un piccolo erbario di oltre 2000 campioni. Fra questi sono numerose le specie di un certo interesse fi togeogra- fi co o rare quali Lycopodium clavatum L., Tozzia alpina L., Cotoneaster integerrimus Medicus, Staphylea pinnata L., e altre che vivono in ambienti rupestri e circoscritti come quelli del versante Nord occidentale del Monte Falco ai confi ni di due regioni (Emilia-Romagna e Toscana) e tre province (Forlì, Firenze e Arezzo). Qui sono state rac- colte, fra altre, Saxifraga moschata Wulf., Saxifraga latina (Terracc.) Hayek, Linum alpinum Jacq., Seseli libanotis (L.) Koch., Senecio doronicum (L.) L., Rhynchosinapis cheiranthos (Vill.) Dandy, Galium austria- cum Jacq., Hypericum richeri Vill. La fi to- Queste specie hanno notevole interesse scientifi co, non soltanto per la loro limitata diffusione e per lo studio della vegetazione di questa zona, ma anche per la loro impor- tanza nell’ecosistema di cui fanno parte, co- stituendo un tipo ambientale della fl ora del territorio con potere diversifi cante e quindi con valore biologico. Ha fatto parte dell’erbario Padula anche una raccolta di circa 1500 campioni della fl ora del Parco Nazionale del Circeo. Questi, unitamente ai campioni raccolti da altri ri- cercatori, hanno costituito la base documen- taria, per la redazione della Flora Vascolare del Parco Nazionale del Circeo2. I campioni di Padula sono stati donati alla direzione del Parco del Circeo. Con l’inizio dell’anno 2000 è stata estesa e concentrata la raccolta, la preparazione e lo studio delle piante del Casentino, bacino idrografi co della prima parte dell’Arno, ab- bastanza ben delimitato topografi camente e le cui conoscenze sulla fl ora sono scarse. La superfi cie della valle oggetto di studio è di circa 71000 ettari. Attualmente l’erbario è costituito, complessivamente, da circa 4500 campioni. 1 M. Padula e G. Crudele, Descrizione naturalistica delle foreste demaniali casentinesi di Campigna-Lama nell’Appennino tosco-romagnolo, regione Emilia- romagna, 1988. 2 B. Anzalone, E. lattanzi, F. lucchese, M. Padula, Flora Vascolare del Parco Nazionale del Circeo, «Webbia» 51(2), 1997. Fig. 1 Campione di Aconitum lycoctonum l. subsp. neapolitanum (Ten.) Nyman dell’Herbarium M. Padulae (foto di A. Alterini). erbari di oggi in provincia di Arezzo Michele Padula e Vincenzo Gonnelli Herbarium M. Padulae Michele Padula Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Fig. 1 1 M. Padula e G. Crudele, Descrizione naturalistica delle foreste demaniali casentinesi di Campigna-Lama nell’Appennino tosco-romagnolo, regione Emilia- romagna, 1988. romagna, 1988. 2 B. Anzalone, E. lattanzi, F. lucchese, M. Padula, Flora Vascolare del Parco Nazionale del Circeo, «Webbia» 51(2), 1997. Fig. 2 Campione di Epipactis flaminia Savelli et Alessandrini dall’Herbarium M. Padulae (foto di A. Alterini). Fig. 3 Esempio di cartellino utilizzato per l’Herbarium M. Padulae, contenente indicazioni sulla stazione di rinvenimento del campione (foto di A. Alterini). Herbarium M. Padulae Michele Padula Gli esemplari raccolti in Casentino sono oltre 2000 e sono in continuo aumento; dalla loro identifi cazione sono risultati oltre 800 taxa specifi ci. 102 Michele Padula e Vincenzo Gonnelli Fig. 2 Fig. 3 Fig. 3 Numerosi sono i campioni d’erbario di specie esotiche, naturalizzate o non, che ve- getano in Casentino. Varie specie arboree sono state introdotte con i rimboschimenti sia in complessi puri o misti con altre spe- cie, sia come rinfoltimenti di boschi cedui degradati, sia per consolidamento di frane e scarpate. Così vi sono campioni di specie dei generi Cedrus, Pinus e Abies, oltre a Pseu- dotsuga, Chamaecyparis, Thuja, Cupressus e Robinia. Di queste, alcune si riproducono naturalmente. Ma varie altre piante erbacee e arbustive si sono diffuse in misura notevole o fanno la loro comparsa nella zona per la prima volta: Reynoutria x bohemica Chrtek et Chrtkova, primo reperto in Italia, Soli- dago gigantea Aiton, Abutilon theophrasti Medicus, Vitis riparia Michx., Fallopia bal- dschuanica (Regel) Holub, Impatiens balfou- rii Hook., Matricaria discoidea DC., primo reperto per la Toscana. Fig. 2 Fra le tante specie di interesse fitogeogra- fico, o selvicolturale, o rare si può indicare: Aconitum lycoctonum L. subsp. neapolita- num (Ten.) Nyman (Fig. 1), Betula pendula Roth, Alnus incana (L.) Moench, Epipactis flaminia Savelli et Alessandrini (Fig. 2), ra- rissima in Toscana, Euonymus latifolius (L.) Mill., Crataegus laciniata Ucria. Di que- st’ultima specie, esclusiva della Sicilia, sono stati trovati diversi esemplari nella zona di Badia Prataglia – Val della Meta (Arezzo); è in corso uno studio di carattere morfolo- gico ed ecologico. Interessanti sono anche i reperti di specie del genere Rosa delle quali il Casentino è molto ricco. Finora sono stati raccolti vari campioni relativi a 16 taxa. p p Un erbario della flora locale, riferito ad una data area, costituisce la documentazio- ne essenziale di base, duratura nel tempo se ben conservato, per la conoscenza stori- ca, naturalistica ed ecologica del territorio preso in esame. Pertanto nel cartellino d’ac- compagnamento di tutti i singoli esemplari sono riportate diverse sintetiche informazio- ni, soprattutto di carattere ecologico (Fig. 3). Fig. 4 Quadro realizzato con un campione di Quercus macrolepis Kotschy del Museo Forestale «Carlo Siemoni» (foto di A. Zoccola). 3 G. Moggi (a cura di), Guida agli erbari della Toscana, Giunta Regionale Toscana, 1994. Herbarium M. Padulae Michele Padula Così, oltre alla nomenclatura identificativa del taxon, quanto più possibile aggiornata e con i sinonimi di più frequente uso cui si può fare riferimento ed eventuali note criti- che di carattere sistematico, è stata riportata la località di raccolta, in modo più preciso possibile, con almeno le coordinate UTM del reticolo di 1 Kmq (100 ettari) del reperto. Si aggiungono alcune informazioni quali la quota; l’esposizione del versante; il substrato Si indicano per la loro rarità: Rosa ande- gavensis Bastard., Rosa serafinii Viv., Rosa agrestis Savi, Rosa dumalis Bechst, Rosa ru- brifolia Vill. 103 Erbari di oggi in provincia di Arezzo Fig. 4 Quadro realizzato con un campione di Quercus macrolepis Kotschy del Museo Forestale «Carlo Siemoni» (foto di A. Zoccola). litologico e se possibile il tipo di suolo e la misura del pH del terreno dove l’esemplare è stato raccolto; il tipo di vegetazione o comun- que l’ambiente dove si trova, ossia l’habitat; la fase fenologica, peraltro limitata alla fiori- tura e fruttificazione alla data della raccolta; nonché un’indicazione della sua diffusione in quell’ambiente. Insomma delle informazioni abbastanza oggettive, relative al taxon censi- to, nelle condizioni reali in cui si trova. Erbario del Museo Forestale «Carlo Siemoni» di Badia Prataglia (Poppi, Arezzo) Michele Padula L’erbario conservato nel museo forestale di Badia Prataglia è di proprietà del Corpo Fo- restale dello Stato, Ufficio Territoriale per la Biodiversità di Pratovecchio (Arezzo). L’at- tuale collocazione deriva dalla fusione di due erbari esistenti nel museo forestale di Camal- doli dal 1978 e in quello di Badia Prataglia dal 1989. I due erbari, singolarmente, sono stati riportati nel volume Guida agli erbari della Toscana3 edito dalla Giunta Regionale Toscana nell’aprile 1994. glitore. In una figurina è rappresentata la diffusione della specie in Italia. Sempre nel cartellino è indicato l’areale della specie e vengono date diverse sintetiche informazioni di carattere ecologico e colturale. L’erbario ha finalità didattiche e divul- gative. Si presume che, almeno a livello di grandi specie, tutte le arboree indigene della flora forestale italiana siano rappresentate. Tra le varie piante dell’erbario si possono ci- tare Pinus cembra L. dell’arco alpino, Acer lobelii Ten. e Pinus leucodemis Ant., ende- mismi dell’Italia meridionale, Quercus troia- na Webb e Quercus macrolepis Kotschy (Fig. 4), esclusive della Puglia, e le esotiche Thuja plicata D. Don e Pseudotsuga menziesii (Mir- bel) Franco. La collezione riguarda piante arboree, e anche alcune arbustive, di flora forestale, spontanee in Italia e alcune tra le più diffuse specie esotiche o naturalizzate, introdotte in Italia per scopi selvicolturali. L’erbario è stato realizzato da Michele Padula. I campioni, in numero di 161, sono costituiti da rametti con fiori e/o frutti, o stro- bili e semi e sono inquadrati in una cornice di legno con vetro. L’allestimento nel quadro è stato curato da Guido Crudele. Numerose specie sono ripetute, peraltro raccolte in epo- che e località diverse. In quasi tutti i campio- ni sono riportate due date di raccolta, quella dell’esemplare in fiore e quella dell’esempla- re in frutto. La raccolta è avvenuta in varie località italiane, in occasione di escursioni dell’autore e di vari collaboratori che hanno successivamente inviato il materiale fresco che è stato revisionato ed essiccato. Il car- tellino di accompagnamento di ogni specie riporta i dati stazionali della località di rac- colta, la relativa data e il nome del racco- Fig. 5 Campione di Ribes rubrum L. dell’Herbarium Alvernae (foto di N. Siemoni). Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) Michele Padula L’erbario della Verna è sorto per una inizia- tiva comune tra la Provincia Toscana di S. Francesco stimmatizzato e la Comunità Mon- tana del Casentino, con finanziamenti della Regione Toscana. Esso è stato realizzato da 104 Michele Padula e Vincenzo Gonnelli Geologicamente l’area è costituita dal- l’imponente formazione calcarea della Ver- na (calciruditi e calcareniti) che poggia su una coltre di argille scagliose del complesso caotico, che si estendono marginalmente. La zona è largamente boscata con varie tipo- logie forestali, naturali o seminaturali e di impianto, ma ne comprende anche altre, in misura minore, quali prati, arbusteti, incolti, rupi boscate, nonché aree calanchive, frano- se e detritiche. Nicola Siemoni e Carlo Ricceri nel periodo 1989-1996, con la raccolta, identificazione Fig. 5 Campione di Ribes rubrum L. dell’Herbarium Alvernae (foto di N. Siemoni). In questa area, relativamente piccola, ma con habitat differenti e ben caratteriz- zata morfologicamente e geologicamente, si riscontra una flora ampiamente diversificata. I taxa che costituiscono l’erbario sono circa 550 (solo pochissimi campioni sono ripetuti). Il censimento non è certo completo, ché varie altre specie sono state raccolte e osservate nella zona, comunque la ricchezza floristi- ca rilevata è notevole. Di recente l’erbario è stato di supporto per la individuazione delle tipologie forestali del Piano di gestione della foresta della Verna. Le specie presenti nell’erbario rappre- sentano un buon campionario della flora del Monte della Verna. Di particolare interesse naturalistico sono gli arbusti e piccoli albe- ri che caratterizzano gli spazi, più o meno aperti, di rupi, spesso fessurate e fratturate e dei detriti calcarei, quali Daphne alpina L., Cotoneaster integerrimus Medicus, Berberis vulgaris L., Amelanchier ovalis Medicus, Rhamnus alpina L., Ribes multiflorum Kit., Ribes rubrum L. (Fig. 5), Ribes alpinum L.. Si possono segnalare inoltre per la loro rarità in Toscana, Opopanax chironium (L.) Koch, Digitalis ferruginea L., Trochiscanthes nodi- flora (All.) Koch, e Pyrus magiarica Terpò. Quest’ultimo taxon è un endemismo dell’Un- gheria che è stato trovato di recente anche in Toscana, o meglio identificato come tale e distinto dal comune e somigliante Pyrus pi- raster Burgsd. Nicola Siemoni e Carlo Ricceri nel periodo 1989-1996, con la raccolta, identificazione e allestimento di 551 campioni di piante. A questi si aggiungono una dozzina di cam- pioni raccolti successivamente da Stefania Gualazzi e Laura Piaggi nel 1998. L’erbario è conservato nel Convento della Verna. Fig. 6 Campione dell’erbario Gonnelli di Laserpitium gallicum L., specie rara in Toscana, segnalata per il Monte Rondinaio in provincia di Lucca e in Provincia di Arezzo dove vegeta sull’Alpe della Luna nel Comune di Badia Tedalda e a Montenero nel Comune di Pieve S. Stefano (foto di V. Gonnelli). Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) Michele Padula Nicola Siemoni e Carlo Ricceri nel periodo 1989-1996, con la raccolta, identificazione e allestimento di 551 campioni di piante. A questi si aggiungono una dozzina di cam- pioni raccolti successivamente da Stefania Gualazzi e Laura Piaggi nel 1998. L’erbario è conservato nel Convento della Verna. Nicola Siemoni e Carlo Ricceri nel periodo 1989-1996, con la raccolta, identificazione e allestimento di 551 campioni di piante. A questi si aggiungono una dozzina di cam- pioni raccolti successivamente da Stefania Gualazzi e Laura Piaggi nel 1998. L’erbario è conservato nel Convento della Verna. Fig. 5 Campione di Ribes rubrum L. dell’Herbarium Alvernae (foto di N. Siemoni). Il censimento floristico ha interessato tutta la proprietà forestale dei frati france- scani, di una superficie di circa 200 ettari ed è stato esteso anche alle zone perimetra- li dove la foresta, soprattutto verso Ovest e Sud, si disperde nelle praterie e nei primi abitati di Chiusi della Verna. La superficie esplorata è, approssimativamente di 260 ettari. L’altimetria dell’area censita va da circa metri 1000 ai 1283 della cima del Monte Penna. A corredo di questa sintetica descrizione dell’«Herbarium Alvernae» segnalo, sempre relativamente alla flora della Verna, un altro erbario e cioè quello fotografico di fra’ Gi- nepro. Esso rappresenta un tipo di erbario dei tempi moderni, ma che si rifà agli an- tichi erbari figurati del passato; questi sono 105 Erbari di oggi in provincia di Arezzo costituiti da disegni di piante, di solito a colori, con indicazione identificativa e, tal- volta, descrizione dell’esemplare raffigurato. Il francescano fra’ Ginepro (al secolo Sante Giacomelli), studioso di Scienze naturali, di cui fu professore nell’Istituto Pontificio di Santa Chiara a Napoli, si dedicò allo studio della flora del monte della Verna, illustrando le piante che via via censiva con fotografie, senza raccogliere alcun esemplare. Egli ese- guì tra il 1979 e il 1996 oltre un migliaio di fotografie a colori di piante della Verna; per ciascuna è riportato il nome scientifico della specie e la data di esecuzione della stessa. Da questa raccolta sono state selezionate da Erminio Ferrarini e Rodolfo E. G. Pichi Ser- molli 181 fotografie di piante, alcune ripe- tute per metterne in evidenza dei particolari quali fiore o frutto e pubblicate in un volume: La Verna. Cantico della Creazione. I fiori del Monte di Francesco visti da Fra’ Ginepro4. 4 E. Ferrarini, La Verna. Cantico della Creazione. I fiori del Monte di Francesco visti da Fra’ Ginepro, a cura di R. E. G. Pichi Sermolli, Edizioni La Verna, 1998. Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) Michele Padula Il volume comprende alcuni brevi scritti di correligionari a ricordo di fra’ Ginepro, oltre ad alcune riflessioni dello stesso fra’ Ginepro sul “Cantico delle Creature” di San France- sco, pure riportato nell’Incipit. Le foto sono molto belle e curate nei par- ticolari. Nel volume sono state ordinate, di solito due per pagina, non in ordine sistema- tico, né alfabetico, ma in base alla stagione di fioritura. Così l’esposizione iconografica inizia dalla fine dell’inverno con Crocus bi- florus Mill. (17.3.1995) per terminare nel- l’autunno con Cyclamen hederifolium Aiton (16.10.1983). A prima vista si rimane colpiti dall’affasci- nante bellezza delle foto, che quasi distoglie l’attenzione dall’accuratezza scientifica del- l’identificazione dell’esemplare, che è stato ripreso, spesso, nel suo ambiente di vegeta- zione più tipico e con i particolari diagnostici più rappresentativi. Fra le specie illustrate si possono indicare, per l’interesse fitogeografi- co, gli endemismi appenninici Hypochoeris robertia Fiori (sinonimo di Robertia taraxa- coides (Loisel) DC. e Sedum monregalense Balb. e, per la loro rarità, Moneses uniflora (L.) A. Gray, e Carduus crispus L. Infine ri- cordo Corydalis cava (L.) Schweigg. et Korte, che è l’ultima foto scattata da fra’ Ginepro il 20 aprile 1996, pochi momenti prima della sua caduta dalla rupe del Sasso Spicco, che lo condusse alla morte il 26 aprile 1996. L’esposizione è corredata dei contributi di Ferrarini e Pichi Sermolli, relativi all’ambien- te e alla vegetazione della Verna, considera- zioni fitogeografiche sul popolamento vegetale e un’accurata storia dell’esplorazione floristica della Verna da Cesalpino a fra’ Ginepro. Erbario Gonnelli Vincenzo Gonnelli L’erbario, iniziato negli anni ’80 del secolo scorso, con raccolte di piante vascolari (Pte- ridophyta e Magnoliophyta), è formato per lo 106 Michele Padula e Vincenzo Gonnelli Fig. 7 Alcuni campioni sono conservati presso l’Erbario Centrale Italiano di Firenze e presso l’Erbario dell’Università degli Studi di Siena. Sono presenti campioni delle più impor- tanti specie dal punto di vista fitogeografico che vegetano nell’Appennino tosco-romagno- lo, fra le quali si citano: Daphne alpina L., Hordelymus europaeus (L.) Harz, Staphylea pinnata L., Berberis vulgaris L., Cardamine enneaphyllos (L.) Crantz, Laserpitium galli- cum L. (Fig. 6); Stipa tirsa Steven, Teucrium siculum Rafin., Centaurea arrigonii Greuter, Hieracium prenanthoides Vill., Moneses uni- flora (L.) A. Gray ecc. Altre specie importanti come Cirsium alpis-lunae Br.-Catt. et Gubell. (Fig. 7), Iso- pyrum thalictroides L., Ribes alpinum L., Ribes multiflorum Kit., Heracleum sphondylium L. subsp. ternatum (Velen.) Brummit, Ononis masquillierii Bertol., Lembotropis nigricans (L.) Griseb., ecc. Fig. 7 Campione dell’erbario Gonnelli di Cirsium alpis-lunae Br.-Catt. & Gubell., pianta endemica dell’Alpe della Luna e di Montenero in Valtiberina, è inserita nella lista rossa nazionale delle specie a rischio di estinzione in Italia (foto di V. Gonnelli). Fig. 8 Campione di Rhitisma acerinum (Pers.) Fr. «Croste nere dell’acero contenuto nell’erbario dell’IPSAA di Pieve S. Stefano», della sezione «erbario patologico». Si tratta di una malattia abbastanza diffusa sugli aceri causata da un fungo ascomicete; in genere nei boschi non costituisce un problema. In questo caso è su acero montano (Acer pseudopatanus L.) (foto di V. Gonnelli). Herbarium Alvernae (Convento della Verna, Chiusi della Verna, Arezzo) Michele Padula hanno, in provincia di Arez- zo (Valtiberina e Casentino) le uniche stazio- ni di vegetazione in Toscana. Sono presenti inoltre campioni delle principali specie che vegetano nelle ofioliti della Valtiberina. Alcuni campioni provengono da nuove stazioni di vegetazione in ambito regionale (Toscana ed Emilia-Romagna) che sono sta- ti, nel tempo, oggetto di segnalazioni floristi- che come ad esempio Arceuthobium oxycedri (DC.) Bieb, Moneses uniflora (L.) A. Gray, Corydalis pumila (Host) Rchb., Typha mini- ma Hoppe ecc. Per alcune specie, partico- larmente rare in Valtiberina, non sono stati raccolti campioni d’erbario, ma ci siamo li- mitati ad una documentazione fotografica. Fig. 7 più da campioni provenienti dall’Appennino tosco-romagnolo marchigiano (Valtiberina e Casentino). Sono presenti tuttavia anche esemplari provenienti da altre Regioni. Fig. 7 Campione dell’erbario Gonnelli di Cirsium alpis-lunae Br.-Catt. & Gubell., pianta endemica dell’Alpe della Luna e di Montenero in Valtiberina, è inserita nella lista rossa nazionale delle specie a rischio di estinzione in Italia (foto di V. Gonnelli). Fig. 8 Campione di Rhitisma acerinum (Pers.) Fr. «Croste nere dell’acero contenuto nell’erbario dell’IPSAA di Pieve S. Stefano», della sezione «erbario patologico». Si tratta di una malattia abbastanza diffusa sugli aceri causata da un fungo ascomicete; in genere nei boschi non costituisce un problema. In questo caso è su acero montano (Acer pseudopatanus L.) (foto di V. Gonnelli). È presente anche una piccola collezione di licheni per lo più provenienti dalla Valti- berina con 41 campioni per 24 specie fra le più diffuse. Comprende raccolte personali di Vincen- zo Gonnelli con oltre 2.000 campioni, dei quali 1.312 già montati, numerati, muniti di cartellino ed inseriti in un database per fa- cilitare la ricerca e la consultazione. I cam- pioni – riuniti in pacchi – sono conservati in scatole di plastica. Nell’Erbario è in sostanza presente qua- si tutta la flora della Riserva Naturale del Sasso di Simone, e la flora pteridologica del Parco Nazionale delle Foreste Casentinesi, Monte Falterona e Campigna oltre, come già ricordato, la presenza di quasi tutte le specie endemiche e di interesse fitogeografico del- l’Appennino Tosco-Romagnolo. L’erbario è suddiviso in due sezioni, una generale ed una specifica per la flora della Riserva Naturale del Sasso di Simone che comprende esclusivamente campioni prove- nienti dalla Riserva. Nell’erbario, che è consultabile su richie- sta solo per motivi di studio, sono periodica- mente intercalati nuovi esemplari. Fig. 8 108 MIChElE PAdUlA E VINCENzo GoNNEllI erbario Istituto Professionale di stato per l’Agricoltura e l’Ambiente «A.M. Camaiti» (Pieve di s. stefano) Vincenzo Gonnelli erbario, nel campione è inserita una scheda con la descrizione della patologia presente. erbario Istituto Professionale di stato per l’Agricoltura e l’Ambiente «A.M. Camaiti» (Pieve di s. stefano) Vincenzo Gonnelli L’Istituto conserva anche l’erbario Mazzo- li, costituito da 130 campioni in prevalenza di specie forestali. L’erbario è completato da una spermoteca (collezione di semi) con 113 campioni di semi di specie forestali (arboree ed arbustive), una raccolta di frutti e di strobili (carpoteca) di quasi tutte le specie forestali Italiane e di alcune esotiche utilizzate in epoche passate nei rimboschimenti della Provincia. L’Istituto Professionale di Stato per l’Agricol- tura e l’Ambiente «A.M. Camaiti» di Pieve S. Stefano è una scuola ad indirizzo forestale ed agrario. Fin dagli anni ’80 ha iniziato la costi- tuzione di un erbario con fi nalità didattiche e scientifi che per lo studio della fl ora forestale italiana e delle principali specie foraggere. In totale nella collezione sono presenti 50 campioni di frutti delle angiosperme e 55 campioni di strobili (pigne) delle gimno- sperme, oltre ad alcuni campioni di pomacee poste sotto liquido conservativo. Attualmente, nell’erbario dell’Istituto, sono conservati complessivamente 457 cam- pioni di specie agrarie e forestali (molti pro- venienti dalla Provincia di Arezzo). È pre- sente in erbario quasi tutta la fl ora forestale italiana e alcuni campioni di specie esoti- che introdotte nei boschi della Provincia. Per le principali specie dell’Appennino Aretino sono conservati in una piccola xilo- teca anche campioni di legno. L’erbario, pur rispondendo rigorosamente a criteri scientifi ci, ha prevalente funzione didattica. Per questa ragione e per cercare di evitare il danneggiamento durante la con- sultazione, i campioni sono conservati dentro buste di plastica riunite in raccoglitori. Di particolare importanza è la presenza di campioni di vecchie cultivar di melo, di pero e di alcune orticole della Valtiberina Tosca- na che complessivamente sono 46 campioni per 35 cultivar. Questi campioni fanno parte del progetto di conservazione del germopla- sma della Valtiberina Toscana. L’erbario e le altre collezioni sono con- sultabili su appuntamento per gli studiosi e le scuole che desiderano approfondire la conoscenza della fl ora forestale della nostra Provincia. È stato realizzato anche un erbario fi to- patologico costituito da 44 campioni di pian- te forestali attaccate da patologie specifi che (Fig. 8). In questo caso, oltre al cartellino di Traduzione a cura di Leonardo Magionami D alla prefazione all’opera De Plantis di Andrea Cesalpino si comprende che egli allestì l’Hortus Siccus dietro insistenza fortissima e commissione del Granduca di Toscana Cosimo I; di questo un esemplare fu donato e offerto al Granduca, l’altro al suo sostenitore e patrono Alfonso Tornabuoni ve- scovo di Sansepolcro. Del primo esemplare invero si deve lamentare la perdita nonostan- te siano state effettuate delle ricerche in di- versi periodi e non se ne ha traccia in nessun luogo. D mani del suo allievo Giovanni Targioni, che tra il 1737 e il 1738 la riprese dal maestro, arricchendola con annotazioni notevoli. Successivamente l’Hortus Siccus del Ce- salpino rimase dimenticato fino all’anno 1818; anno in cui fu ritrovato in casa del- la famiglia Nencini erede dei Pandolfini da Ottaviano Targioni figlio di Giovanni; e poco dopo fu acquisito dalla biblioteca Palatina del Granduca di Toscana. Nello stesso anno, il Brocchi, cultore di storia naturale che al- lora risiedeva a Firenze pubblicò una breve notizia del ritrovamento dell’erbario nella Biblioteca Italica (vol. X pag. 203): l’anno successivo il Professor Bertoloni della scuola bolognese pubblicò un commento sull’erba- rio dal titolo Memoria sopra l’Erbario e una lettera del Cesalpino. Nel frattempo Otta- viano Targioni, che allora possedeva l’ope- ra manoscritta del Micheli, si apprestò alla pubblicazione di quella dopo aver definito una nuova suddivisione delle piante secondo la nomenclatura di Linneo, in modo tale che la descrizione del Micheli fosse ulteriormen- te aggiornata secondo i nuovi studi ma l’ope- ra prodotta si fermò ad un primo abbozzo. g L’altro esemplare ha avuto una sorte mi- gliore poiché, confluito con l’eredità Tor- nabuoni nei beni della famiglia fiorentina Pandolfini, è rimasto dimenticato nella loro biblioteca domestica, finchè Pietro Antonio Micheli, dopo aver svolto molte ricerche, gra- zie ai suggerimenti di William Sherard, lega- to di Smirne per la Regina d’Inghilterra, lo ritrovò nell’anno 1717 (come lo stesso Micheli tramanda negli appunti lasciati manoscritti). Micheli, venuto a conoscenza del valore no- tevole dell’erbario felicemente riscoperto si preoccupò che le piante fossero classificate in modo più accurato secondo la allora vi- gente nomenclatura Tournefortiana, e in bre- ve tempo ne definì la classificazione secondo la tassonomia da lui proposta. * Si premette all’edizione anastatica dell’Illustratio la traduzione dell’introduzione del 1858. Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Traduzione a cura di Leonardo Magionami Rivendica un primo posto rispetto al nostro probabilmente un Hortus Siccus di un certo Greault, chirurgo di Lione, conservato prima da Jussieu nella sua collezione e ora a Parigi nell’Hortus Re- gius e attribuito all’anno 15581. È posteriore quello di Leida detto Rauwolfiano, preparato grazie alle piante raccolte da Rauwolfio nei suoi viaggi effettuati dall’anno 1573 al 1575. L’Aldrovandiano che si trova a Bologna, si può asserire che sia quasi dello stesso pe- riodo del Cesalpino, è noto a tutti che Aldro- vandi è stato contemporaneo e condiscepolo del Cesalpino, ma mancano argomenti per affermarlo2. Segue, certamente alquanto posteriore, l’Hortus Siccus di Bauhin che è conservato a Basilea, preparato tra gli anni 1576 e 16233. Nell’anno 1844 quando passò dalla Bi- blioteca Palatina al Museo di storia natura- le, l’Hortus Siccus del Cesalpino era raccolto in un unico volume rilegato in pergamena. In quella occasione, alle piante già rovina- te dalle tarme furono evitati ulteriori danni grazie alla disinfestazione voluta dal Profes- sor Parlatore e l’erbario dietro sua indicazio- ne fu interfoliato con fogli cartacei posti tra una pagina e l’altra, e rispettata la sequenza originale fu rilegato in tre volumi distinti ri- coperti in pelle rossa. E sebbene l’erbario sia talmente rovinato per le questioni che sopra ho ricordato, le piante sono in uno stato tale da essere riconoscibili per la maggior par- te. Le più piccole sono integre fatta eccezio- ne per la radice o altre parti sotterranee; di quelle più grandi è presente soltanto la parte superiore, fiorifera o fruttifera, spesso con qualche foglia o inferiore o radicale; qual- che volta ma raramente è presente soltanto il fiore. Alcuni esemplari, pochi certamente rispetto agli altri, e cioè quelli arborei o frut- tiferi che si riconoscono per questo motivo, mancano del fiore e del frutto. Se si presta Anche io, nell’illustrazione dell’Hortus Siccus di Cesalpino ho mantenuto questo criterio, ho riportato fedelmente la lettera e gli indici per quanto lo consente la stampa mantenendo l’ordine, l’ortografia e le note distintive del testo autografo. Passando all’il- lustrazione delle singole piante ho riportato secondo il numero quelle che sono elencate nell’erbario. Inoltre alle piante conservate oltre al numero progressivo, ho aggiunto il numero della carta in cui compaiono. Ho tra- scritto i nomi delle piante, anche se inesatti mantenendomi fedele all’esemplare. Traduzione a cura di Leonardo Magionami Ma l’opera, che l’uomo insigne si era proposto di pubbli- care, non so per quali ragioni, allora non fu data alla stampa, e dopo la morte di Micheli passò, insieme a tutta la sua collezione nelle Finalmente nell’anno 1844 l’Hortus Sic- cus del Cesalpino fu spostato dalla Biblio- teca Palatina al Regio Museo fiorentino di storia naturale su richiesta del Professore Filippo Parlatore, che prefetto della facoltà di Botanica era allora passato alla direzione del Museo. Egli pensò subito di riprendere la Gli erbari aretini da Andrea Cesalpino ai giorni nostri 112 descrizione dell’Hortus Siccus del Cesalpino ma, distratto da altre incombenze non poté realizzare questo desiderio. attenzione ci si accorge subito che le piante con le quali è stato allestito l’erbario proven- gono da diversi territori della Toscana e al- cune anche dagli orti. L’erbario è composto da 260 carte di grande formato detto comu- nemente in folio, tutte numerate, e le piante sono 767. Gli esemplari sono incollati sulle carte o singolarmente o a gruppi a seconda della grandezza. Fatta eccezione per quat- tro casi, vicino ad ogni pianta è riportato il nome di mano del Cesalpino in greco, latino e in italiano. Una serie di numeri progressivi risulta apposta da un’altra mano certamente posteriore, come credo, forse con probabilità proprio la mano del Micheli. L’ordine della numerazione non è sempre corretto e non se- gue neppure l’ordine pensato da Cesalpino. L’erbario vero e proprio è preceduto da una lettera in italiano di grande valore scientifico che Cesalpino scrive di suo pugno al vescovo Alfonso Tornabuoni. In questa lettera spie- ga come abbia cominciato a pensare ad una classificazione delle piante, abbozzando con questo erbario le prime linee del suo meto- do di classificazione, metodo che dopo venti anni di studio si concretizza nel fondamen- tale testo De Plantis; di seguito chiarisce la ragione della distribuzione in classi e del- l’attribuzione dei nomi. La lettera è seguita da due indici uno in greco, uno in latino e italiano, i quali rimandano alla numerazione delle carte dell’erbario. A questo punto disquisendo riguardo la sua antichità l’Hortus Siccus del Cesalpino, risulta al momento certamente il più antico tra quelli conosciuti se non addirittura il più antico in assoluto poiché questo è ascri- vibile all’anno 1563. 1 Dottor Puel, Parigi, in litt. 2 Prof. G. Bertoloni, Bologna, in litt. 3 Prof. Meisner, Basilea, in litt. Traduzione a cura di Leonardo Magionami Ho citato i passi del De Plantis nei quali con certezza o grande probabilità è descritta la pianta e ho aggiunto i nomi delle piante come ivi com- paiono. Ho messo in evidenza in quale ma- niera si presentano gli esemplari, la parte del fiore, quella del frutto o la pianta per intero. Introduzione all’edizione di Teodoro Caruel (1858) Introduzione all’edizione di Teodoro Caruel (1858) 113 Infine ho aggiunto i nomi più recenti delle specie, li ho resi più chiari con i loro sinonimi o dove possibile con noterelle o citazioni. arricchirle. Ritenendo che questo mio lavoro potrà essere utile sia per quelli che attraver- so lo studio della geografia botanica si sono domandati quale fosse la flora in toscana ai tempi di Cesalpino, e a coloro i quali si sono dedicati alla storia delle specie e soprattutto a coloro che si sono scontrati con la difficile lettura del De Plantis senza questa spiega- zione non mi dispiace di aver profuso tanta fatica in quest’opera. Di questo sono soprat- tutto lieto, di poter pagare un tributo di reve- renza a Micheli, insigne studioso primo tra i botanici, tributo negato dalla sorte sia a lui che ad altri importanti studiosi. Ora la mia opera vede la luce, opera a cui ho atteso non senza fatica e con la con- sapevolezza di non portare un vantaggio alla scienza poiché soprattutto si impiega così tanto tempo che gli studi di storia botanica diventano subito obsoleti e pochissimi si de- dicano alla riscoperta delle tracce lasciate dai padri su quella via che segnarono per i posteri o a cercare di scoprire in che modo essi si siano impadroniti di quelle conoscenze che ci hanno tramandato per conservarle ed 114 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI edizione Anastatica Theodori Caruelii illustratio in hortum siccum Andreae Caesalpini Florentiae Typis Le Monnier MDCCCLVIII XII, 128 p. ; 8° Copia riprodotta: Biblioteca di Scienze – Sezione Botanica, Università di Firenze N.° I.° 398 (Coll. 39 F 22) Theodori Caruelii illustratio in hortum siccum Andreae Caesalpini Florentiae Typis Le Monnier MDCCCLVIII XII, 128 p. ; 8° Copia riprodotta: Biblioteca di Scienze – Sezione Botanica, Università di Firenze N.° I.° 398 (Coll. Introduzione all’edizione di Teodoro Caruel (1858) 39 F 22) edizione Anastatica edizione Anastatica Theodori Caruelii illustratio in hortum siccum Andreae Caesalpini Florentiae ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 115 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 116 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 117 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 118 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 119 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 120 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 121 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 122 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 123 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 124 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 125 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 126 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 127 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 128 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 129 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 130 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 131 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 132 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 133 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 134 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 135 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 136 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 137 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 138 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 139 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 140 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 141 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 142 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 143 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 144 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 145 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 146 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 147 148 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 149 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 150 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 151 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 152 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 153 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 154 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 155 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 156 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 157 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 158 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 159 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 160 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 161 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 162 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 163 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 164 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 165 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 166 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 167 168 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 169 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 170 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 171 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 172 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 173 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 174 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 175 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 176 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 177 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 178 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 179 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 180 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 181 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 182 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 183 ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 184 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI ILLUSTRATIO IN HORTUM SICCUM ANDREAE CAESALPINI 185 GlI ErBArI ArETINI dA ANdrEA CESAlPINo AI GIorNI NoSTrI 186 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press Gli erbari aretini da Andrea Cesalpino ai giorni nostri Gli erbari aretini da Andrea Cesalpino ai giorni nostri 188 Studi di Siena e insegna codicologia presso la Facoltà di Lettere e Filosofia di Arezzo. La sua attività scientifica e di ricerca si è concentrata prevalentemente sullo studio e catalogazione del manoscritto medievale e moderno. Ha al suo attivo pubblicazioni ine- renti alla storia del libro manoscritto e delle scritture esposte e su questi argomenti ha te- nuto conferenze in Italia e all’estero. Studi di Siena e insegna codicologia presso la Facoltà di Lettere e Filosofia di Arezzo. La sua attività scientifica e di ricerca si è concentrata prevalentemente sullo studio e catalogazione del manoscritto medievale e moderno. Ha al suo attivo pubblicazioni ine- renti alla storia del libro manoscritto e delle scritture esposte e su questi argomenti ha te- nuto conferenze in Italia e all’estero. Michele Padula – Laureato in Scienze Fo- restali, libero docente di Botanica forestale all’Università di Firenze e assistente volonta- rio presso la cattedra di botanica della Facol- tà di Agraria. Ufficiale del Corpo Forestale dello Stato, ha svolto vari incarichi tra cui quello di amministratore delle Foreste De- maniali di Corniolo (FC) e Pratovecchio (AR) e di coordinatore regionale del CFS della To- scana. Ha svolto ricerche e studi di carattere floristico, ecologico e selvicolturale. L’attività di ricerca prosegue tuttora a titolo personale ed in particolare con lo studio della flora del Casentino. Francesca Malfanti – Laureata in Scienze Biologiche, si occupa di museologia scien- tifica e di ricerche floristiche sul campo. In particolare ha esplorato diverse aree del ter- ritorio pisano e pubblicato alcuni lavori in proposito. È curatore dell’Orto Botanico dei Frignoli in Lunigiana. Maria Adele Signorini – Laureata in Scienze Forestali, è ricercatrice presso la Facoltà di Agraria di Firenze, dove insegna Botanica sistematica. Per diversi anni re- sponsabile dell’erbario FIAF (erbario dei Laboratori di Agraria e forestale del Dipar- timento di Biologia vegetale), è autrice di numerose pubblicazioni su argomenti legati alla iconografia botanica, agli erbari e alla storia della botanica. Guido Moggi – Già professore ordinario di Botanica e direttore del Museo Botanico del- l’Università di Firenze dal 1974 al 1998. Si è occupato di flora dell’Italia meridionale e della Somalia, di giardini e parchi storici, di erbari e collezioni botaniche. È stato presi- dente dell’Associazione Nazionale dei Musei Scientifici e di O.P.T.I.M.A., organizzazione internazionale che studia la flora del Medi- terraneo. Note sugli autori Lucia Amadei – Laureata in Scienze Bio- logiche, è conservatore del Museo botanico dell’Università di Pisa. Si occupa in partico- lare di indagini sugli erbari antichi, sia di- pinti che costituiti da exsiccata, sui quali ha prodotto numerosi contributi. Vincenzo Gonnelli – Laureato in Scien- ze e Tecnologie Agrarie, è insegnante presso l’Istituto Professionale di Stato per l’Agricol- tura e l’Ambiente «A.M. Camaiti» di Pieve S. Stefano (AR). Si occupa di studi botanici e della conservazione della biodiversità in particolare nella gestione e conservazione di ecosistemi naturali, su cui ha pubblicato nu- merosi lavori. Alessandro Brezzi – Laureato in Scienze Politiche, è Responsabile della Biblioteca Comunale Rilli-Vettori di Poppi e Respon- sabile dell’Ufficio Cultura del Comune di Poppi. Ha coordinato il restauro conservativo dell’erbario anonimo della Rilliana e ne ha curato l’esposizione al pubblico. Enrico Gusmeroli – Laureato in Scienze Naturali, è Responsabile dell’U.O. Reti Eco- logiche del Servizio Conservazione della Na- tura della Provincia di Arezzo e si occupa di tutela della biodiversità e gestione della rete ecologica provinciale. È stato co-curatore della mostra «Da Andrea Cesalpino ai nostri giorni. Erbari aretini in mostra». Secondino Gatta – Laureato in Lettere ad indirizzo etno-antropologico, Dottore di Ri- cerca in Storia della Chiesa, attualmente è Direttore del Museo della Verna. Ha colla- borato alla realizzazione degli allestimenti e della comunicazione di diversi Musei, tra cui il Museo delle Erbe di Aboca. Ha inoltre col- laborato a mostre sia di ambito ecclesiastico che storico sulla cultura materiale. Lorenzo Lastrucci – Laureato in Scienze Biologiche e Dottore di Ricerca in Biosiste- matica ed Ecologia Vegetale, sta svolgendo attività di ricerca presso il Dipartimento di Biologia Vegetale dell’Università di Firenze occupandosi prevalentemente di flora e vege- tazione delle zone umide. Molti dei suoi stu- di hanno riguardato il territorio provinciale di Arezzo. Bruno Gialluca – Laureato in Filosofia presso l’Università degli Studi di Firenze, è direttore del Servizio per i beni culturali del Comune di Cortona. I suoi interessi e le sue pubblicazioni vertono sull’antiquaria tosca- na, in particolare di Cortona, al tempo degli ultimi Medici. Leonardo Magionami – Laureato in Con- servazione dei beni culturali, è ricercatore di Paleografia latina presso l’Università degli Gli erbari aretini da Andrea Cesalpino ai giorni nostri Si è dedicato recentemente alla storia della botanica, pubblicando ricerche su A. Cesalpino, su P.A. Micheli e sulla sto- ria della botanica in Italia. Paolo Emilio Tomei – Laureato in Scienze Naturali è Professore associato di Fitogeogra- fia presso la Facoltà di Scienze dell’Universi- tà di Pisa. Persegue diverse linee di ricerca, nel cui ambito rivestono un ruolo non tra- scurabile gli studi relativi alla storia della botanica; su questa tematica ha pubblicato numerosi lavori. Chiara Nepi – Laureata in Scienze Agrarie, Dottore di Ricerca in Biosistematica ed Eco- logia Vegetale, è conservatrice della Sezione Botanica del Museo di Storia Naturale del- l’Università di Firenze. Si occupa in partico- lare dello studio e della conservazione delle collezioni storiche (erbari, modelli botanici, dipinti, ecc.) ivi conservate, sulle quali ha pubblicato diversi contributi. Laura Vivona – Laureata in Scienze Biolo- giche, lavora presso il Dipartimento di Bio- logia Vegetale (Sez. di Botanica Forestale e Ambientale) dell’Università di Firenze. Ha notevoli esperienze professionali come dise- gnatrice naturalistica. Indice dei nomi Abate G. 41 Adriano, imperatore romano 56 Aldrovandi U. 4, 9, 31, 33, 81, 112 Alessandrini A. 41 Almagià R. 91 Alpago A. 56 Alpini P. 56-57, 91, 103 Amadei L. 91-92, 95, 187 Anderson F. J. 3 Anzalone B. 101 Apuleius Platonicus 3 Arber A. 3-4 Aristotele 6-7 Baccarini P. 54, 71-72, 74-75 Balboni L.A. 91 Baldelli O. 53 Baldini R.M. 92, 95 Barocchi P. 55 Bassi L.M. C. 59 Bauhin G. (o C.) 15, 33, 68, 84, 88, 112 Bauhin J. 68 Barrelier J. 65 Bertoloni A. 10, 111-112 Bianchi G. 59 Bigazzi A. ix, xiii, 4, 13 Bizzarri M.P. 92 Blasi C. 41 Boccone P. 15, 65 Boeravio H. 33 Bonino G. 91 Boutroue M.E. 31 Branchi G. 93 Breccia Fratadocchi M. 31 Bremekamp C.E.B. 7-8 Bresciani E. 91 Brezzi A. 71, 76, 187 Brocchi G.B. 10, 12-13, 15, 17, 91-92, 111 Brunfels O. 8, 84 Buonarroti F. 54 Buresti D. 63, 65 Busnardo G. 92 Buttò S. 31 Calzolari F. 4 Carlo Emanuele III Savoia, re di Sardegna 91 Caruel T. xiv, 7, 15-17, 111 Cesalpino A. ix, xi, xiii-xiv, 1, 3-18, 23, 31, 33, 63, 74, 76, 95, 105, 111-113, 187-188 Champollion F. 91 Ciampelli G. 83, 88 Cibo G. 3, 9 Cipriani G. 71 Clemente VIII, papa 5 Cocchi A. 25, 54 Cocchi R. 59 Collins M. 3 Colonna F. 56 Coltellini A. xiv, 23, 25, 28, 57 Coltellini L. 25, 28, 55, 57, 59, 62, 65 Coltellini T. 25, 28, 57, 59 Abate G. 41 Adriano, imperatore romano 56 Aldrovandi U. 4, 9, 31, 33, 81, 112 Alessandrini A. 41 Almagià R. 91 Alpago A. 56 Alpini P. 56-57, 91, 103 Amadei L. 91-92, 95, 187 Anderson F. J. 3 Anzalone B. 101 Apuleius Platonicus 3 Arber A. 3-4 Aristotele 6-7 Baccarini P. 54, 71-72, 74-75 Balboni L.A. 91 Baldelli O. 53 Baldini R.M. 92, 95 Barocchi P. 55 Bassi L.M. C. 59 Bauhin G. (o C.) 15, 33, 68, 84, 88, 112 Bauhin J. 68 Barrelier J. 65 Bertoloni A. 10, 111-112 Bianchi G. 59 Bigazzi A. ix, xiii, 4, 13 Bizzarri M.P. 92 Blasi C. 41 Boccone P. 15, 65 Boeravio H. 33 Bonino G. 91 Boutroue M.E. 31 Branchi G. 93 Breccia Fratadocchi M. 31 Bremekamp C.E.B. 7-8 Bresciani E. 91 Brezzi A. 71, 76, 187 Brocchi G.B. 10, 12-13, 15, 17, 91-92, 111 Brunfels O. 8, 84 Buonarroti F. Indice dei nomi 54 Buresti D. 63, 65 Busnardo G. 92 Buttò S. 31 Boutroue M.E. 31 Branchi G. 93 Breccia Fratadocchi M. 31 Bremekamp C.E.B. 7-8 Bresciani E. 91 Brezzi A. 71, 76, 187 Brocchi G.B. 10, 12-13, 15, 17, 91-92, 111 Brunfels O. 8, 84 Buonarroti F. 54 Buresti D. 63, 65 Busnardo G. 92 Buttò S. 31 Buresti D. 63, 65 Busnardo G. 92 Buttò S. 31 Calzolari F. 4 Calzolari F. 4 Baccarini P. 54, 71-72, 74-75 Carlo Emanuele III Savoia, re di Sardegna 91 Caruel T. xiv, 7, 15-17, 111 Cesalpino A. ix, xi, xiii-xiv, 1, 3-18, 23, 31, 33, 63, 74, 76, 95, 105, 111-113, 187-188 Cesalpino A. ix, xi, xiii-xiv, 1, 3-18, 23, 31, 33, 63, 74, 76, 95, 105, 111-113, 187-188 Champollion F. 91 Bauhin G. (o C.) 15, 33, 68, 84, 88, 112 Ciampelli G. 83, 88 Cibo G. 3, 9 Cipriani G. 71 Barrelier J. 65 Bertoloni A. 10, 111-112 Clemente VIII, papa 5 Cocchi A. 25, 54 Cocchi R. 59 Chiara Nepi, Enrico Gusmeroli ( a cura di ), Gli erbari aretini da Andrea Cesalpino ai giorni nostri, ISBN 978-88-8453-765-2 (print), ISBN 978-88-8453-803-1 (online) © 2008 Firenze University Press 190 Gli erbari aretini da Andrea Cesalpino ai giorni nostri Andrea Cesalpino ai giorni nostri Commelin J. 33, 57 Conti F. 41 Corinaldi J. xiv, 89, 91-93, 95-96 Cosimo I dei Medici, granduca di Toscana 8, 10, 80, 111 Cristofolini G. 4 Crudele G. 101, 103 Dioscoride 3-4, 6, 13, 18, 33, 56, 81, 84 Dodoens D. 84 Donati V. 91 Dragone Testi G. 63, 67-68 Durante C. 71-72, 74-76 Eleonora di Toledo, duchessa 80 Erodoto 56 Fabbri F. 88 Fabbrini C. 28, 57 Fabbrini N. 55 Ferdinando III di Asburgo-Lorena, granduca di Toscana 10, 15 Ferrarini E. 6, 105 Ferri S. 4, 88 Figari A. 91, 93 Forneris G. 95 Forni G. 91 Francesco d’Assisi, santo 79, 103, 105 Francini Corti E. 92 Franciosi P., frate 80 Fuchs L. 8, 84 Fusi P. 31, 38 Galeno 6 Gallo D. 55 Garbari F. 4, 6 Gatta D. 79-80, 187 Gesù 88 Ghini L. 4-5, 9, 31 Ghizzi G. 38-39 Ghizzi L. 39 Giacomelli S. (fra’ Ginepro) 6, 18, 104-105 Gialluca B. 24-25, 28, 53, 63, 65, 187 Ginori 28 Girault J. 4 Gnagnoni E. 38 Gnagnoni G. 38 Gonnelli V. 101, 105-106, 108, 187 Greault 112 Griselini F. 59 Gualazzi S. Napoleone 91 Targioni Tozzetti A. 9, 15, 24 Targioni Tozzetti G. 5-6, 8-10, 14-17, 24, 53-55, 63, 65, 76, 111 Targioni Tozzetti O. 15-17, 111 Teofrasto 6, 8, 13, 18, 56, 84 Indice dei nomi 59 Morelli Timpanaro M.A. 28 Moretti G. 10 Moriani A. 31, 38 Morton A.G. 6-7 Rosselli S. 9, 14 Rosselli S. 9, 14 Saccardo P.A. 93 Sachs J. 7 Santini E. 54 Sherard W. 9, 111 Siemoni N. 103-104 Signorini M.A. 63, 188 Siliotti A. 91 Sorbelli A. 38 Montacchini F. 3 Siemoni N. 103-104 Signorini M.A. 63, 188 Siliotti A. 91 Moriani A. 31, 38 Moriani A. 31, 38 Morton A.G. 6-7 Morton A.G. 6-7 Tabernaemontanus D.J. 76 Tabernaemontanus D.J. 76 Tilli A.A. 59 Tilli A.A. 59 Osiride 56 Indice dei nomi 104 Gusmeroli E. xiii, 4, 13, 95, 187 Herold J. C. 91 Hoffmann C. 33 Iacopelli G. 71-73 Iberite M. 95 Ibn al-Baytār 56 Ippoliti G. 59 Iside 56 Jonquet D. 33 Jussieu 112 Lais G. 5 Lami G. 25, 57, 59 Lastrucci L. 4, 13, 71-72, 83, 95, 187 Lattanzi E. 101 L’Ecluse C. de 84 Leonardo da Vinci 31 Leone, frate 79 Leopoldo II di Asburgo-Lorena, granduca di Toscana 10 Linneo C. xi, 18, 23, 28, 111 Lippi A. 57 L’Obel M. de 33, 84 Lucchese F. 101 Lupi V. 57 Luzzi P. 88 Maccari O. 59 Maccioni S. 92, 95 Magionami L. 31-32, 111, 187 Mahudel, monsieur 56 Manetti S. 25, 57, 59 Maniero F. 24 Marchi P. 95 Marescotto G. 9 Mattioli P.A. 4, 8, 18, 33, 38, 68, 76, 81-82, 84, 88 Mazzoli 108 Meisner 112 Meneghini G. 92-93 Merini M. 3-4, 9, 23 Micheli P.A. 8-10, 13-15, 17, 23, 25, 53-55, 63, 65, 68, 111-113, 188 Francini Corti E. 92 Manetti S. 25, 57, 59 Gesù 88 Ghini L. 4-5, 9, 31 Ghizzi G. 38-39 Ghizzi L. 39 Giacomelli S. (fra’ Ginepro) 6, 18, 104-105 Gialluca B. 24-25, 28, 53, 63, 65, 187 Mazzoli 108 Ginori 28 Meisner 112 Girault J. 4 Gnagnoni E. 38 Gnagnoni G. 38 191 Indice dei nomi Millozza A. 95 Moggi G. 3-8, 31, 83, 103, 188 Mohammed Ali 91 Moneti A. 54 Moneti M. xiv, 51, 53-58, 63, 65, 67-68 Montacchini F. 3 Montelatici U. 59 Morelli Timpanaro M.A. 28 Moretti G. 10 Moriani A. 31, 38 Morton A.G. 6-7 Napoleone 91 Nencini, famiglia 10, 15, 111 Nepi C. xiii, 10-11, 23, 57, 65, 188 Nunziati A. 38 Orlando di Chiusi, conte 79 Osiride 56 Padula M. 101, 103, 188 Pandolfini P. 10, 111 Paolo III, papa 12, 41 Parlatore F. ix, xi, xiv, 9-11, 93, 111-112 Pazzagli R. 5-6 Pernigotti S. 92 Petrollini F. 4, 9 Piaggi L. 104 Pichi Sermolli R.E.G. 6, 8-9, 17, 92, 105 Pietro Leopoldo di Asburgo-Lorena, granduc di Toscana 23, 25, 28, 59, 62 Pignatti S. 41, 68 Plinio 6, 8, 18, 33, 56, 76, 84 Plutarco 56 Puel 112 Raddi G. 91-92 Ragazzini S. 14-15 Ratzenberger C. 4 R lf L 112 Millozza A. 95 Moggi G. 3-8, 31, 83, 103, 188 Mohammed Ali 91 Moneti A. 54 Moneti M. xiv, 51, 53-58, 63, 65, 67-68 Montacchini F. 3 Montelatici U. Osiride 56 Tomei P.E. 41, 91-92, 95, 188 Tommasini A.C. 39 Tommasini G.T. 39 Tommasini I.V. 39 Tommasini L. 38-39 Tommasini L. 38-39 Parlatore F. ix, xi, xiv, 9-11, 93, 111-112 Tommasini R.G. 39 Tommasini R.G. 39 Pazzagli R. 5-6 Pazzagli R. 5-6 Tommasini T.R. 39 Pernigotti S. 92 Pernigotti S. 92 Tongiorgi Tomasi L. 6, 57-59 Petrollini F. 4, 9 Petrollini F. 4, 9 Tornabuoni A. 9-10, 12, 14, 16-18, 111-112 Tornabuoni A. 9-10, 12, 14, 16-18, 111-112 Piaggi L. 104 Tosi A. 6, 53, 55-59 Pichi Sermolli R.E.G. 6, 8-9, 17, 92, 105 Pichi Sermolli R.E.G. 6, 8-9, 17, 92, 105 Tournefort J.P. de 15, 24, 33, 45, 55-56, 58 68 84 Pietro Leopoldo di Asburgo-Lorena, granduca di Toscana 23, 25, 28, 59, 62 Pietro Leopoldo di Asburgo-Lorena, granduca di Toscana 23, 25, 28, 59, 62 Touwaide A. 31 Pignatti S. 41, 68 Turner W. 4 Turner W. 4 Tuscher M. 57 Venturini F.M. 77, 79, 81, 83-84, 88 Venuti B.G., marchese 62-63 Ragazzini S. 14-15 Venuti F. 54-59, 63 Venuti M. 54, 59, 62 Venuti R. 54 Viviani D. 91 Finito di stampare presso la tipografia editrice Polistampa Titoli pubblicati 1. P. Dolara, G. Fiorini (a cura di), La collezione storica di farmaci dell’Università di Firenze 2. L. Borrelli, F. Gherardi, G. Fiorito, A Catalogue of Body Patterning in Cephalopoda 3. M. Gasperini, Arch Cube 4. C. Nepi, E. Gusmeroli (a cura di), Gli erbari aretini da Andrea Cesalpino ai giorni nostri Di prossima pubblicazione G. Chelazzi, G. Barsanti (a cura di), Il Museo di Storia Naturale dell’Università di Firenze. La Specola e le collezioni zoologiche A. M. Jasink, L. Bombardieri (a cura di), Le collezioni egee del Museo archeologico nazionale di Firenze M. Raffaelli (a cura di), Il Museo di Storia Naturale dell’Università di Firenze. Le collezioni botaniche 1. P. Dolara, G. Fiorini (a cura di), La collezione storica di farmaci dell’Università di Firenze 2. L. Borrelli, F. Gherardi, G. Fiorito, A Catalogue of Body Patterning in Cephalopoda 3. M. Gasperini, Arch Cube 4. C. Nepi, E. Gusmeroli (a cura di), Gli erbari aretini da Andrea Cesalpino ai giorni nostri M. Raffaelli (a cura di), Il Museo di Storia Naturale dell’Università di Firenze. Le collezioni botaniche Finito di stampare presso la tipografia editrice Polistampa
https://openalex.org/W4377291409	https://www.frontiersin.org/articles/10.3389/fpsyg.2023.1148391/pdf	English	null	Hotspots, trends, and advice: a 10-year visualization-based analysis of painting therapy from a scientometric perspective	Frontiers in psychology	2,023	cc-by	11,859	TYPE Review PUBLISHED 22 May 2023 DOI 10.3389/fpsyg.2023.1148391 TYPE Review PUBLISHED 22 May 2023 DOI 10.3389/fpsyg.2023.1148391 OPEN ACCESS OPEN ACCESS EDITED BY Mingwei Lin, Fujian Normal University, China REVIEWED BY Dejian Yu, Nanjing Audit University, China Lili Zhang, Fujian Jiangxia University, China Cynthia Whissell, Laurentian University, Canada CORRESPONDENCE Hongzhong Qiu hzqiu@163.com Qianrong Liang 1,2†, Jiarong Ye 3†, Yingyin Lu 1, Junjie Dong 4,2, Heyong Shen 2,5† and Hongzhong Qiu 6† 1 School of Finance, Guangdong University of Foreign Studies, Guangdong, China, 2 Institute of Analytical Psychology, Faculty of Humanities and Social Sciences, City University of Macau, Macau, China, 3 The Second Clinical College of Guangzhou University of Chinese Medicine (The Second Affiliated Hospital of Guangzhou University of Chinese Medicine), Guangzhou, China, 4 Department of Social Work, Kunming Children Hospital, Kunming, China, 5 School of Psychology, South China Normal University, Guangzhou, China, 6 School of Public Health and Management, Guangzhou University of Chinese Medicine, Guangzhou, China †These authors have contributed equally to this work RECEIVED 23 January 2023 ACCEPTED 04 May 2023 PUBLISHED 22 May 2023 CITATION Liang Q, Ye J, Lu Y, Dong J, Shen H and Qiu H (2023) Hotspots, trends, and advice: a 10-year visualization-based analysis of painting therapy from a scientometric perspective. Front. Psychol. 14:1148391. doi: 10.3389/fpsyg.2023.1148391 RECEIVED 23 January 2023 ACCEPTED 04 May 2023 PUBLISHED 22 May 2023 CITATION Liang Q, Ye J, Lu Y, Dong J, Shen H and Qiu H (2023) Hotspots, trends, and advice: a 10-year visualization-based analysis of painting therapy from a scientometric perspective. Front. Psychol. 14:1148391. doi: 10.3389/fpsyg.2023.1148391 Purpose: Research on painting therapy is available worldwide and painting therapy is widely applied as a psychological therapy in different fields with diverse clients. As an evidence-based psychotherapy, previous studies have revealed that painting therapy has favorable therapeutic effects. However, limited studies on painting therapy used universal data to assemble in-depth evidence to propose a better recommendation on it for the future use. Large-scale retrospective studies that used bibliometric methodology are lacking. Therefore, this study presented a broad view of painting therapy and provided an intensively analytical insight into the structure of knowledge regarding painting therapy employing bibliometric analysis of articles. CiteSpace software was used to evaluate scientific research on painting therapy globally published from January 2011 to July 2022. COPYRIGHT © 2023 Liang, Ye, Lu, Dong, Shen and Qiu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). frontiersin.org Frontiers in Psychology 1. Introduction CiteSpace evaluates co-citations, co-authors, and co-occurrence keywords and analyzes the data using procedural steps: time slicing, thresholding, modeling, pruning, merging, and mapping (Chen, 2004). Analysis using CiteSpace is beneficial for users to obtain key information regarding the research field under study and recognize the trends in a certain area. Due to the lack of bibliometric analysis of painting therapy, quantitative results on the use, collaboration, exchange and research of painting therapy internationally, as well as on the future direction of research development are unclear. The Citespace-based bibliometric analysis may address this issue, providing a comprehensive quantitative overview, display a broad panorama and visual analysis on the current state and the hotspot of painting therapy. Therefore, this study presented a broad view of painting therapy and provided an intensively analytical insight into the structure of knowledge regarding painting therapy employing bibliometric analysis of articles.h This paper used bibliometric analysis and social network analysis to analyse the publications on the painting therapy during 2011–2022. To enhance understanding of the field in particular, the main contributions of this paper are as follows: (1) Discussions on the published articles were conducted at the national/regional, institutional and author levels, followed by a cluster analysis based on the research literature on painting therapy in order to understand the topical issues in this field. (2) Social network analysis was used to quantify the extent of national and institutional collaboration quantitatively. (3) Analysis of the content of key literature qualitatively provides deep insight into issues of close interest to the authors, as well as future research trends. (4) In-depth discussions were held from the perspective of current burning issues, future trends and challenges, and limitations. Bibliometrics has become an influential method for in-depth quantitative analysis and is a scientific tool that can be used to investigate hotspots and reveal trends in a certain research area (Durieux and Gevenois, 2010; Huertas González-Serrano et al., 2020). This approach refers to the study of the distribution, quantitative relationships, and change patterns of a field of study by taking the measurement characteristics of the literature as the object of study and then exploring the structure, characteristics, and patterns in a subject area through the filtering and processing of massive amounts of information. It can also reveal the evolution of a journal (Chen et al., 2022; Envelope et al., 2022). 1. Introduction economics and management (Zhong and Lin, 2022; Zhang et al., 2023). In the early 1980s, psychotherapy with various art forms such as painting, music, literature, crafts, and drama became known as “expressive art therapy” at Leiris University, and subsequently developed into an independent field of study. The American Art Therapy Association defines art therapy as an integrative mental health and human services profession that enriches the lives of individuals, families, and communities through active art-making, creative processes, applied psychological theory, and human experience in a psychotherapeutic relationship (Association TAAT, n.d.). As a form of art therapy, painting therapy can capture psychological information that is difficult to obtain from conventional questionnaires, thereby circumventing the limitations of language (Qiu, 2004; Hu et al., 2021). It also helps to break through patients’ psychological resistance and open their hearts, and allows them to share their feelings, views, and emotions with others. Painting therapy has psychological therapeutic effects such as cultivating emotional resilience, resolving distress, and fostering self-confidence (Qiu, 2004; Puetz et al., 2013; Hu et al., 2021). It is widely applied as a psychological therapy in different fields with diverse clients, including individuals with emotional problems (Qiu et al., 2017), children or adolescences with special needs (Klop, 2017; Durrani, 2019), and patients with mental illnesses (MacIntosh, 2017). In addition, painting therapy can address the psychological needs of patients with cancer (Czamanski- Cohen et al., 2019; Sw et al., 2019; Thyme et al., 2022), Parkinson’s disease (Schofield, 2019), diabetes (Zamanifard et al., 2022), functional gastrointestinal diseases (Kai et al., 2022), and other physical illnesses. Although research on this evidence-based psychotherapy is available worldwide and the amount of its publications have emerged, limited studies on painting therapy have used universal data. Large-scale retrospective studies that used bibliometric techniques are also lacking. It is significantly essential to summarize the current studies on painting therapy by bibliometric approach. CiteSpace, developed by Chen (2004), is a Java application software for data analysis and visualization based on the theory of co-citation analysis. It is a unique and leading application for information visualization analysis using three major concepts: burst detection, betweenness centrality, and heterogeneous networks. These three concepts can recognize the character of research frontiers, mark keywords, and identify emerging trends and sudden changes over time, respectively (Chen and CiteSpace, 2006). OPEN ACCESS The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Methods: Publications related to painting therapy from 2011 to 2022 were searched using the Web of Science database. This study employed bibliometric techniques to perform co-citation analysis of authors, visualize collaborations between countries/regions as network maps, and analyze keywords and subjects relevant to painting therapy by using CiteSpace software. Results: In total, 871 articles met the inclusion criteria. We found that the number of painting therapy publications generally trended incrementally. The United States and United Kingdom made the most contributions to painting therapy research and had the greatest impact on the practical application in other countries. Arts in Psychotherapy and Frontiers in Psychology occupied key publishing positions in this research field. The application groups were mainly children, adolescents, and females, and Western countries paid high attention to painting therapy. The main areas of application of painting therapy were Alzheimer’s disease and other psychosomatic disease fields. Identified research priorities for painting therapy were emotion regulation and mood disorder treatment, personality disorder treatment, personal self-esteem enhancement, and medical humanistic care. Three keywords, “depression,” “women,” and “recovery,” had the strongest citation bursts, which emphasized the research trends. Conclusion: The general trend for painting therapy research is positive. Our findings provide useful information for researchers on painting therapy to determine new directions in relate to popular issues, collaborators, and research frontiers. Painting therapy holds a promising future, and further studies could explore the clinical implications of this therapy in terms of mechanisms and criteria for assessing efficacy. Frontiers in Psychology 01 frontiersin.org Liang et al. 10.3389/fpsyg.2023.1148391 KEYWORDS KEYWORDS painting therapy, bibliometrics, CiteSpace, Web of Science, psychosomatic health painting therapy, bibliometrics, CiteSpace, Web of Science, psychosomatic health painting therapy, bibliometrics, CiteSpace, Web of Science, psychosomatic health Frontiers in Psychology frontiersin.org 1. Introduction It applies mathematics, statistics and computer science to all sorts of documentary data with the competence to identify trends in the development of various disciplines and dig out the potential value of knowledge (Yu et al., 2018). It has been widely used in the field of engineering and technology (Yu et al., 2018, 2020; Lin et al., 2021; Chen et al., 2022; Envelope et al., 2022), humanities and social sciences (Tal and Gordon, 2018), medicine (Chen et al., 2022; He et al., 2022), 2.1. Search strategy The data used in this study were obtained from Web of Science (WoS), a database developed by Thomson Scientific and has been used in tens of thousands of academic studies. It is one of the most trustworthy international citation datasets, covering a large amount of literature published from 1900 to the present and providing journals with high quality and information about literature worldwide in details (Chen et al., 2022). It is not only applied as a research tool to 02 frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. 10.3389/fpsyg.2023.1148391 widely used evaluation index that reflects the overall quality of the journal, but it fail to measure the impact of a specific article or research scholar. The H-index, on the other hand, can reflect the quality of most articles published in the field of painting therapy rather than identifying individual papers with more citations, which is more accurate and targeted than the IF in an analysis of a specific field. The H-index is a delicate balance between the number of papers and their citation rate, thus reducing the ‘over-rating’ of smaller journals. In fact, the H-index and the IF are like a complementary relationship, and the two together make for more reliable journal evaluations. In scientific research literature, keywords often reflect the main research content of the literature. Therefore, clustering analysis of keywords can aid in understanding the development lineage, main research contents, and research frontier hotspots of a particular research field. What’s more, the keyword co-occurrence network can reflect current research hotspots as well as hotspots that were previously generated in the field. Centrality is a key indicator for analyzing the importance of keywords. The greater the centrality, the greater the importance and influence of the node in the study. The combination of node size, centrality and frequency of keyword occurrence reveals the focus and hotspots in the research field. support a variety of scientific tasks in various domains of knowledge, but also as a statistic set for large-scale data-intensive scientific research. gi Art therapy was chosen as the search term since painting therapy belongs to it. However, painting therapy may be expressed in a variety of ways. It can also be named after different forms of painting. 2.4. Analysis methods The goal of bibliometrics can be depicted as the performance of research that contributes to the development of the field of knowledge by inferring and interpreting relevant analyses (Castanha and Grácio, 2014; Merigó et al., 2019; Mulet-Forteza et al., 2021). As a bibliometric software, CiteSpace produces information to visualize data better. This software was used to visualize scientific maps for publications on painting therapy from 2011 to 2022 with years per slice (slice length = 1), author co-citation networks, inter-country and inter- regional collaboration networks and network maps, co-cited journals, and co-occurring keywords for top painting therapy research (timeline view). Co-citations occur when two articles receive citations from the same third study (Henry, 1973; Merigó et al., 2019). Every option in the terminology source was selected, and a single node type was chosen at a time based on particular conditions. The “top 25 levels” were used as thresholds for the most frequently cited or referenced articles in the corresponding time slice. Nodes and links were essential elements to develop a visualization knowledge figure. A node represents an element (e.g., an author, a keyword, a region, or a country) with the node’s size proportional to the occurrence or citation frequency, and the node’s color indicates the year. Furthermore, each node is represented by a series of concentric circles or “tree rings” on a time slice. The concentric circles’ size indicates the publications’ number, and different colored circles from the inside to the outside of a node indicate the years (2011–2022). Links between two nodes represent collaboration, co-occurrence, or co-citation relationships. The green ring represents the mediated centrality of the literature (Chen et al., 2009), whereas the red indicates the time slice where sudden rise in citations or detectable citation bursts happened (Kleinberg, 2002). 2.2. Inclusion criteria The title field was painting therapy, and “articles” was the only document type selected in the advanced search. Other document types, such as editorial material, letters, and book reviews, were excluded. 1,297 articles were returned in this advanced search process. Irrelevant documents were removed by manually filtering each article when downloading the data. These irrelevant documents mainly included documents with the search terms in the subject but that were not about painting therapy (e.g., combination antiretroviral therapy in the medical field) and articles on non-painting therapy such as music therapy, dance therapy, art exhibitions, art professions, art taste, and other unrelated articles. Removal of duplicates was performed using CiteSpace. Finally, 871 valid documents were identified as the data source for this study. The title field was painting therapy, and “articles” was the only document type selected in the advanced search. Other document types, such as editorial material, letters, and book reviews, were excluded. 1,297 articles were returned in this advanced search process.i Irrelevant documents were removed by manually filtering each article when downloading the data. These irrelevant documents mainly included documents with the search terms in the subject but that were not about painting therapy (e.g., combination antiretroviral therapy in the medical field) and articles on non-painting therapy such as music therapy, dance therapy, art exhibitions, art professions, art taste, and other unrelated articles. Removal of duplicates was performed using CiteSpace. Finally, 871 valid documents were identified as the data source for this study. 2.1. Search strategy In order to collect more comprehensive literature and cover a wider range of painting therapy modalities to analyze the current state of painting therapy, various forms of painting were included as search terms in this article. Combination Antiretroviral Therapy (cART) and Assisted Reproductive Technology (ART), as the abbreviations are the same as ART, have been added to the exclusion terms to exclude literature that is not relevant to painting therapy. In addition, painting therapy is not belonged to art education, and articles on art education were also excluded. Therefore, we performed data acquisition on July 26, 2022 using the following search terms: title = (((((TS = (Art Therapy OR art therapy OR arts therapies)) OR (((TI = (Art OR arts))OR (TI = (artists))))) OR ((((((((((((((TI = (draw OR paint))) OR (TI = (sketch))) OR (TI = (coloring))) OR (TI = (doodle))) OR (TI = (doodling))) OR (TI = (collage))) OR (TI = (craft))))) OR (TI = (expressive*))) OR (TI = (Tracing))) OR (TI = (still life))) OR (TI = (Mandala)))) NOT (AB = (cART OR combination antiretroviral therapy OR art education OR HIV OR Assisted Reproductive Technology)) NOT (TS = (cART OR combination antiretroviral therapy OR art education OR HIV OR Assisted Reproductive Technology))) NOT ALL = (cART OR combination antiretroviral therapy OR art education OR HIV OR Assisted Reproductive Technology)) and time span = 2011–2022. Frontiers in Psychology frontiersin.org 2.3. Data extraction The authors extracted publications and used Microsoft Excel 2016 and EndNote to analyze the publications obtained from the WoS database. In addition, basic information for each publication were downloaded and recorded as bibliometric indicators, including citation frequency, authors’ country or region, and journal. The Hirsch index (H-index) is used to measure the citation frequency of research in academic journals or researchers (Wang et al., 2019). The IF and the H-index were used to evaluate journal quality and scientific research influence. The IF is a measure of the number of citations to articles published in a journal over a 2 year period. It is a widely recognized and Scientific maps are usually featured as a group of points and lines to indicate collaboration among publications (Chen and CiteSpace, 2006). A point is used to signify an author, country/region, journal, or keyword, whereas a line represents a link between points, with wider lines indicating stronger links (Zheng and Wang, 2019). Moreover, a 03 frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. manuscripts on painting therapy, which were published in 61 countries and regions. All articles are written in English. CiteSpace was used to analyze the distribution of these articles in terms of which countries or regions they were published in. The node type was selected as “country,” g-index (k = 25) as the threshold, and slice = 1. The top 10 countries or regions are highlighted in Table 1. The average article citations, total citations on WoS, and H-index for these top 10 countries or regions are also presented. The H-index is used to measure the influence of authors’ scientific achievements. The United States had the most published manuscripts (222 studies), together with the most citations on WoS (2,157 citations), and the highest H-index value (H = 38). Its average number of article citations (9.72 citations) was ranked second. The United Kingdom had the second largest number of publications (102 studies) and a total of 999 citations on WoS. It had the highest average number of citations (9.79 citations). These findings indicated that the United States and the United Kingdom, as the countries with the earliest application of painting therapy, made the most contribution to the field of painting therapy research and had the greatest impact on the practical application of this therapy in other countries. This laid a research foundation for painting therapy development. 3.1. Publication outputs and time trends As can be seen in Figure 1, the number of articles on painting therapy was relatively stable between 2011 and 2017, with about 40–60 publications per year, and then steadily grew from year to year. This showed that increasing attention has been paid to painting therapy since 2017. There was a more rapid increase in the number of articles between 2019 and 2020, and in 2020, the number of articles on painting therapy exceeded 100 for the first time. The average annual number of publications in the last 2 years has been in excess of 130. Overall, the general trend in the number of articles published on painting therapy was incremental. In 2022, however, as of July 2022, only 59 papers were published in this approximately six-month period, which is less than the number of articles in 2020 and 2021. This needs to be followed up further. Figure 2 displays the collaboration networks among countries or regions. In total, 61 nodes and 124 links were contained in the collaboration networks. The size of the circle represents the amount of publications in a certain country. The largest node was located in the United States with the maximum publication count on painting therapy. In addition, a lighter color indicates a study was published earlier. Thicker lines between nodes reflect more cooperative publications. The shorter the distance between two nodes, the stronger cooperation between the two countries or regions. We observed that in the early stages of research on painting therapy, there was cooperation between European countries, with other countries around the world engaging in a greater level of cooperation in research as the field deepened and developed. The darkest purple circle was shown for Spain (H = 0.5), meaning that Spain had the greatest amount of collaborations with other countries or regions in this research field. 2.3. Data extraction scientific map includes nodes that represent the frequency of citations for certain topics. Burst nodes in a central red circle form indicate increasing co-occurrence or number of citations over time. The purple nodes represent centrality and suggest essential knowledge is shown by the data (Chen and CiteSpace, 2006; Chen et al., 2012; Zheng and Wang, 2019). Occurrence bursts in the map represent the topics’ frequency, whereas citation bursts indicate the references’ frequency (Chen and CiteSpace, 2006). Citation bursts for keywords and references can be used to explore trends and indicate whether particular authors were gaining considerable attention in the field (Chen and CiteSpace, 2006). These maps build understanding of emerging trends and help to track hot topics through burst detection analysis (Liang et al., 2017; Miao et al., 2017). 3. Results 3.2. Distribution by country or region Studies published from 2011 to 2022 in the WoS database were selected and analyzed with a time slice of 1 year. This included 871 FIGURE 1 Time sequence of relevant articles on painting therapy published from 2011 to 2022 in Web of Science. FIGURE 1 Time sequence of relevant articles on painting therapy published from 2011 to 2022 in Web of Science. FIGURE 1 Time sequence of relevant articles on painting therapy published from 2011 to 2022 in Web of Science. 04 04 Frontiers in Psychology frontiersin.org Liang et al. 10.3389/fpsyg.2023.1148391 TABLE 1 Top ten countries or regions that published articles related to painting therapy. p g p p g py Country Publications Frequency H_index Total citations Citations per paper USA 222 0.255 38 2,157 9.72 United Kingdom 102 0.117 32 999 9.79 Israel 87 0.1 13 437 5.02 Germany 60 0.069 21 476 7.93 Canada 56 0.064 25 456 8.14 China 47 0.054 15 213 4.53 Australia 39 0.045 17 153 3.92 Korea 33 0.038 12 255 7.73 Netherlands 26 0.03 15 150 5.77 Italy 25 0.029 9 123 4.92 FIGURE 2 The collaboration networks among countries or regions. In this map, the circle represents a country or a region, whose size reflects the amount of publications in a certain country. Thicker lines between nodes indicate more cooperative publications. The shorter the distance between two nodes, the stronger cooperation between the two countries or regions. FIGURE 2 The collaboration networks among countries or regions. In this map, the circle represents a country or a region, whose size reflects the amount of publications in a certain country. Thicker lines between nodes indicate more cooperative publications. The shorter the distance between two nodes, the stronger cooperation between the two countries or regions. FIGURE 2 The collaboration networks among countries or regions. In this map, the circle represents a country or a region, whose size reflects the amount of publications in a certain country. Thicker lines between nodes indicate more cooperative publications. The shorter the distance between two nodes, the stronger cooperation between the two countries or regions. Frontiers in Psychology frontiersin.org 3.3. Distribution by journals The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. Journal IF (2022) H_index Publications Arts in Psychotherapy 1.847 20 222 Frontiers in Psychology 4.232 8 83 Psychology of Aesthetics Creativity and the Arts 2.325 5 13 Disability and Rehabilitation 2.439 8 10 Dementia-International Journal of Social Research and Practice 2.624 5 8 British Journal of Guidance and Counselling 1.125 3 8 Australian and New Zealand Journal of Family Therapy 1.037 2 8 European Journal of Oncology Nursing 2.588 4 7 BMJ Open 3.006 4 6 Family Process 3.497 4 6 FIGURE 3 The map of the co-citation journals. In this map, the node represents a journal. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. FIGURE 3 The map of the co-citation journals. In this map, the node represents a journal. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. citations), and PLOS One (140 citations). Thus, based on the publications and co-citation counts analysis, Arts in Psychotherapy and Frontiers in Psychology held key positions in the painting therapy field. distance between two nodes reflects more cooperation between authors. The color of the circle highlighted the author of the same cluster. Blue circles mean earlier outputs, whereas yellow circles represent more recent publications. The large-sized node that represented authors Braun V, Malchiodi CA, and Gantt L, indicated that these authors were the most co-cited authors, which can be regarded a remarkable progress in the field of painting therapy. 3.3. Distribution by journals decade, and had the highest H-index (H = 20), followed by Frontiers in Psychology (83 publications, H = 8), and Psychology of Aesthetics Creativity and the Arts (13 publications, 2022 IF: 2.325, H = 5). Figure 3 presents the map of the co-citation journals, which included 534 nodes and 1,862 links. A journal is represented by a node in this co-citation map. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. A high co-citation count identifies journals with the greatest academic impact that occupy the key position in the research field. Arts in Psychotherapy had the greatest number of co-citation count at 698, followed by Frontiers in Psychology (158 Table 2 lists the top 10 journals publishing manuscripts related to painting therapy, along with the journals’ impact factors (IFs) and H-index. It shows that articles were mostly published in Arts in Psychotherapy, Frontiers in Psychology, Psychology of Aesthetics Creativity and the Arts, Disability and Rehabilitation, and Dementia- International Journal of Social Research. The IFs of these journals varied between 1.037 and 4.232 (average IF: 2.472). Seven of these top 10 journals had an IF > 2, of which Frontiers in Psychology had the highest IF for 2022 (4.232). In addition, Arts in Psychotherapy (2022 IF: 1.847) published 222 articles on painting therapy in the past 05 10.3389/fpsyg.2023.1148391 Liang et al. TABLE 2 Top 10 journals publishing manuscripts related to painting therapy. TABLE 2 Top 10 journals publishing manuscripts related to painting therapy. TABLE 2 Top 10 journals publishing manuscripts related to painting therapy. Journal IF (2022) H_index Publications Arts in Psychotherapy 1.847 20 222 Frontiers in Psychology 4.232 8 83 Psychology of Aesthetics Creativity and the Arts 2.325 5 13 Disability and Rehabilitation 2.439 8 10 Dementia-International Journal of Social Research and Practice 2.624 5 8 British Journal of Guidance and Counselling 1.125 3 8 Australian and New Zealand Journal of Family Therapy 1.037 2 8 European Journal of Oncology Nursing 2.588 4 7 BMJ Open 3.006 4 6 Family Process 3.497 4 6 FIGURE 3 The map of the co-citation journals. In this map, the node represents a journal. The larger the node, the more publications in that journal. 3.4. Distribution by authors A total of 711 authors had contributed to the research in the field of painting therapy. Table 3 presents the top 10 authors publishing manuscripts studies related to painting therapy according to the output amount and citations by others. However, the overall publication counts were not large. Author Levwiesel R published the most manuscripts (10 publications), followed by Snir S (nine publications) and Kaimal G (nine publications). For co-citation counts, Braun V (65 citations) and Malchiodi CA had the most co-citations, followed by Gantt L (61 citations), McNiff S (51 citations), and Csikszentmihalyi M (38 citations). Figure 4 shows the co-citation network. These nodes represent the number of articles published by each author: a larger node indicates more publications, and a shorter Frontiers in Psychology frontiersin.org 3.5. Analysis of keywords The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. FIGURE 4 The map of the co-citation authors. In this map, the node represents a journal. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. FIGURE 4 The map of the co-citation authors. In this map, the node represents a journal. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. as the index to judge research hotspots. The keyword “Alzheimer’s disease/dementia” had the highest centrality (0.17/0.13). In this network map, nodes corresponded to keywords, and the size of the nodes indicated the frequency of co-occurring keywords. Colored lines represented by time zone that appeared between items: the red can be considered as the newest, and the light grey the oldest. The top 30 most cited or co-occurring keywords were selected from each slice. The highest frequency was for “depression” at 94, followed by “children,” “intervention,” “mental health,” and “quality of life.” Most of the nodes marked with purple circles represented good betweenness centrality, and suggested that these items were essential. In other words, these nodes with the strongest bursts reflected emerging tendency in painting therapy field. Figure 6 was sorted by the chronological order to obtain Figure 7, which displays the chronological view of the painting research field. The historical progression of clusters reflected the development of research on painting therapy over a specific period of time and the thematic concentration of each cluster. Figure 5 presents the top 15 keywords with the strongest citation bursts. “Depression” was one of the strongest keywords bursts (4.05) and highlighted the emerging trend of painting therapy studies in the recent decade. The words “women” and “recovery” were similar to “depression,” with bursts of 3.21 and 2.94, respectively. Keywords that received major attention in 2022 and formed the focus of current painting research were “virtual reality” (4.31), “anxiety” (3.24), “COVID-19” (3.24), “resilience” (3.01), and “creative arts therapy” (2.77). Co-occurring keywords reflected research hotspots in painting therapy field. Frontiers in Psychology 3.5. Analysis of keywords The keywords analysis showed the research hotspots and provided a method to identify further research topics. Table 4 displays the top 30 keywords that had the most frequent use (e.g., “Alzheimer’s disease/dementia”), the method or function of painting therapy (e.g., “psychotherapy,” “drawing,” and “experience”), and the applied groups (e.g., “adolescent,” “women,” and “children”). As the mediated centrality can also reflect research hotspots to some extent, this study comprehensively considered the centrality and frequency of keywords 06 10.3389/fpsyg.2023.1148391 Liang et al. TABLE 3 The top 10 authors publishing manuscripts studies related to painting therapy according to the output amount and citations by others. h TABLE 3 The top 10 authors publishing manuscripts studies related to painting therapy according to the output amount and citations by others. Number Author Co-citation counts Author Publications 1 Braun V 65 RACHEL LEVWIESEL 10 2 Malchiodi CA 65 SHARON SNIR 9 3 Gantt L 61 GIRIJA KAIMAL 9 4 McNiff S 51 LIMOR GOLDNER 8 5 Csikszentmihalyi M 38 DAFNA REGEV 8 6 WINNICOTT DW 36 SUZANNE HAEYEN 8 7 Hass-Cohen N 35 EPHRAT HUSS 7 8 Van Lith T 35 VICKY KARKOU 6 9 Reynolds F 34 HOD ORKIBI 5 10 Collie K 33 MICHAL BAT OR 4 FIGURE 4 The map of the co-citation authors. In this map, the node represents a journal. The larger the node, the more publications in that journal. The link between two nodes represents the frequency of co-citation, and a thicker purple circle suggests greater impact on the field of painting therapy. 10 authors publishing manuscripts studies related to painting therapy according to the output amount and citations by others. TABLE 3 The top 10 authors publishing manuscripts studies related to painting therapy according to the output amount and c Number Author Co-citation counts Author Publications 1 Braun V 65 RACHEL LEVWIESEL 10 2 Malchiodi CA 65 SHARON SNIR 9 3 Gantt L 61 GIRIJA KAIMAL 9 4 McNiff S 51 LIMOR GOLDNER 8 5 Csikszentmihalyi M 38 DAFNA REGEV 8 6 WINNICOTT DW 36 SUZANNE HAEYEN 8 7 Hass-Cohen N 35 EPHRAT HUSS 7 8 Van Lith T 35 VICKY KARKOU 6 9 Reynolds F 34 HOD ORKIBI 5 10 Collie K 33 MICHAL BAT OR 4 FIGURE 4 The map of the co-citation authors. In this map, the node represents a journal. The larger the node, the more publications in that journal. frontiersin.org 3.6. Analysis of subjects A set of closely related keywords formed a certain subject. The subjects were clustered using the log-likelihood ratio algorithm and the results were labeled according to the importance level to obtain the subject clustering, as shown in Figure 8. Two indicators, modularity Q and the contour value S, were used to evaluate clusters. Q values >0.3 suggest that the network is clear and essential, S values >0.5 mean that the clustering results are ideal and > 0.7 imply the results are reliable (Jin et al., 2022). In this analysis, the modularity Q was 0.8306 and S was 0.9293, reflecting that the results were reliable and meaningful. The smaller the number of the cluster, the more keywords it contained, and the larger the corresponding area. Moreover, the corresponding time was distinguished by the cluster’s color, with cooler colors signifying earlier years and warmer colors indicating more recent ones. In total, 14 clusters were found, and 13 clusters could be clearly presented. The details corresponding to each cluster are presented in Table 5, with the tag words highlighted in the cited subjects. The top 10 subjects are listed in order of the keyword count. The keywords contained in each cluster may be duplicated, and the duplication is reflected in the overlap area of the cluster mapping. The top-ranked subject by counts was “qualitative content analysis” (year: 2010) in Cluster 0 at 39, followed by “emotion regulation” (year: 2011) in Cluster 1 at 34, “oncology” (year: 2008) in Cluster 2 with a count of 33, and “grounded theory” (year: 2012) in Cluster 3 with a count of 33. (#3) and “well-being” (#6) continued to be the popular topics in last few years, indicating that they were classical topics for painting therapy. In addition, Clusters 0 and 6 had more keywords appearing in 2019–2020 than other clusters of classic topics, with “personality disorder” in Cluster 0 and “self-esteem” in Cluster 6. The horizontal line in “compassion” (#9) was lighter in color and had recent keywords such as “medical humanity” and “emotion regulation.” The recent keywords in the above clusters suggested that researchers recently investigated areas related to these keywords, meaning these clusters can be considered the frontiers of painting therapy research. 3.5. Analysis of keywords With “Keywords” as the node type, the g-index (k = 25) as the threshold, slice = 1, pathfinder as the network trimming method, and other parameters as default values, 405 nodes and 1,377 inter- node links were obtained. After the keyword “art therapy” and the words belonging to that phrase (“art” and “therapy”) were hidden, the keyword co-occurrence network map was obtained as shown in Figure 6. “Depression” formed the largest cluster (#0), followed by “learning strategies” (#1) and “qualitative research” (#2). “Depression,” “arts” 07 frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. TABLE 4 The top 30 keywords with the most frequency and centrality in painting therapy field. FIGURE 5 The top 15 keywords with the strongest citation bursts. The blue measures reflect infrequent citation of keywords while the red measures represent frequent citation of keywords. FIGURE 5 The top 15 keywords with the strongest citation bursts. The blue measures reflect infrequent citation of keywords while the red measures represent frequent citation of keywords. painting therapy field. Number Publications Centrality Keywords 1 92 0.07 Depression 2 82 0.09 Children 3 72 0.05 Intervention 4 62 0.13 Mental health 5 56 0.03 Quality of life 6 55 0.06 Anxiety 7 52 0.11 Psychotherapy 8 50 0.15 Adolescent 9 39 0.03 Health 10 39 0.03 Symptom 11 38 0.01 Trauma 12 37 0.10 Women 13 34 0.00 People 14 34 0.03 Model 15 34 0.10 Drawing 16 34 0.10 Experience 17 34 0.13 Dementia 18 33 0.10 Life 19 33 0.13 Cancer 20 32 0.07 Scale 21 31 0.05 Creativity 22 30 0.03 Program 23 30 0.05 Disorder 24 30 0.10 Breast cancer 25 29 0.03 Impact 26 28 0.02 Care 27 27 0.03 Mindfulness 28 25 0.11 Schizophrenia 29 23 0.03 Emotion 30 23 0.17 Alzheimers disease 3.6. Analysis of subjects Therefore, it can be predicted that painting therapy for emotion regulation and mood disorder treatment, personality disorder treatment, personal self-esteem enhancement, and medical humanity care will be the major focus of further research in the painting therapy field internationally. Frontiers in Psychology frontiersin.org 4.1. Global trends in painting therapy research This study performed a bibliometric analysis of painting therapy studies over the past decade. The findings reflected that painting therapy research has been held worldwide and provided scholars with suggestions for further research. In terms of the overall analysis of 08 Frontiers in Psychology frontiersin.org Liang et al. 10.3389/fpsyg.2023.1148391 published articles, the characteristics, and countries or regions are summarized below. point as it was the first time that more than 100 articles were published. Overall, research on painting therapy received increasing attention from scholars between 2011 and 2022. First, the number of published articles per year has increased in the last decade. From 2011 to 2022, the annual volume of publications in painting therapy was divided into the following three phases: the beginning, second and third phases. The beginning phase was from 2011 to 2016 with <60 publications annually. The second phase was 2017–2019, which showed a steady increase in publication volume. The third phase was 2019–2022, with 2020 being a significant turning Second, the articles covered 61 countries or regions, and The United States had the maximum publication outputs, followed by Europe. According to the amount of citations on WoS, the citation counts of each study, and the H-index, publications on painting therapy from developed Western countries (e.g., the United States and the United Kingdom) had a greater impact than those from other countries or regions. Furthermore, China, as a leading example of developing countries, published 47 manuscripts, and was ranked the sixth in the productive countries. FIGURE 6 The keyword co-occurrence network map. In this map, the node represents a keyword and the link represents the co-occurrence frequency. The larger the node, the more publications of a certain keyword. A thicker purple circle suggests greater impact in this field. Colored lines represented by time zone that appeared between items: the red can be considered as the newest, and the light grey the oldest. FIGURE 6 Third, based on the data for the authors’ total amount of citations, publications counts, and the top 10 authors cited, the amount of publications for a certain author was generally low, and even the research by the author with the highest count of studies (Levwiesel R) accounted for only 1.15% of publications, suggesting a lack of leading scholars to conduct intensive and systematic research in this field. 4.2. Research focus on the research frontier and hot topics Painting therapy had a range of research objects, and focused on special groups in recent years. The keyword co-occurrence mapping and keyword frequency and centrality indicated that the objects of painting therapy research were mainly women, adolescents, and children. The research encompassed a wide range of psychological problems, psychological disease groups, and physical disease groups. Research focused on individuals with psychological problems and diseases included groups of psychiatric diseases such as eating disorders, substance abuse, personality disorders, and posttraumatic The keyword co-occurrence network map. In this map, the node represents a keyword and the link represents the co-occurrence frequency. The larger the node, the more publications of a certain keyword. A thicker purple circle suggests greater impact in this field. Colored lines represented by time zone that appeared between items: the red can be considered as the newest, and the light grey the oldest. The keyword co-occurrence network map. In this map, the node represents a keyword and the link represents the co-occurrence frequency. The larger the node, the more publications of a certain keyword. A thicker purple circle suggests greater impact in this field. Colored lines represented by time zone that appeared between items: the red can be considered as the newest, and the light grey the oldest. FIGURE 7 Recurring research on painting therapy after Figure 6 data are sorted into chronological order. In this map, nodes corresponded to keywords, and the size of the nodes indicated the frequency of co-occurring keywords. The nodes on the same line belong to the same cluster. FIGURE 7 Recurring research on painting therapy after Figure 6 data are sorted into chronological order. In this map, nodes corresponded to keywords, and the size of the nodes indicated the frequency of co-occurring keywords. The nodes on the same line belong to the same cluster. 09 Frontiers in Psychology frontiersin.org Liang et al. 10.3389/fpsyg.2023.1148391 10.3389/fpsyg.2023.1148391 FIGURE 8 The subjects clustered using the log-likelihood ratio algorithm. In this map, each cluster contained a set of closely related keywords which formed a certain subject. The smaller the number of the cluster, the more keywords it contained, and the larger the corresponding area. The corresponding time was distinguished by the cluster’s color, with cooler colors signifying earlier years and warmer colors indicating more recent ones. 4.2. Research focus on the research frontier and hot topics In addition, with the development of modern technology, painting therapy is not limited to drawing on paper, but can be produced online using electronic applications and virtual drawing. This keyword may also relate to the dramatic rise in psychological problems during the COVID-19 pandemic and the increased difficulty of language-based and venue-required counseling. In this analysis, cluster 11 (e-mental health) was identified, which may offer an important direction for further exploring painting therapy. stress disorder, in addition to depression (MacIntosh, 2017; Qiu et al., 2017; Hirano and Hosaka, 2018). For physical diseases, Cluster 2 (oncology) explored the application of painting therapy in oncology, with painting therapy mainly applied with patients with breast cancer (Czamanski-Cohen et al., 2019; Thyme et al., 2022), gynecological cancers (Wiswell et al., 2019), and other tumors (Puetz et al., 2013; Bozcuk et al., 2017; Elimimian et al., 2020). Painting therapy was also used with patients with Parkinson’s disease (Schofield, 2019), diabetes (Zamanifard et al., 2022), and children with gastrointestinal disorders (Kai et al., 2022). Alzheimer’s disease (Pongan et al., 2017) was the main focus of applicable diseases, with the highest keyword word frequency centrality; this also formed a cluster (Cluster #4) as the object of study on painting therapy. These results reflected the focus of painting therapy research on psychosomatic disorders. Studies also increasingly paid attention to special groups, including prisoners (Qiu et al., 2017; Gonzalez Barajas and Ho, 2020), pregnant women (Wahlbeck et al., 2020), children with autism (Durrani, 2019), people with low education and literacy levels (Crombie et al., 2022), and older people (Beauchet et al., 2020), reflecting medical humanism and community care. What’s more, it is practical to conduct research on painting therapy in hospitals and colleges. With the shift from traditional medical systems to biopsychosocial medical networks, the role of psychological factors in the etiology and prognosis of physical illnesses has received more attention. On the one hand, physical illnesses can lead to psychological problems and even psychological disorders. For example, the diagnosis and treatment of cancer is often accompanied by changes in physical status and function, unpleasant side effects, reduced quality of life and impaired social relationships (Caruso et al., 2017); the prevalence of psychiatric disorders is much higher in cancer patients, with depression and anxiety being the two most common psychiatric symptoms (Niedzwiedz et al., 2019; Pilevarzadeh et al., 2019; Hashemi et al., 2020). 4.2. Research focus on the research frontier and hot topics In terms of the outcome indicators, the efficacy of painting therapy was evaluated by most researchers using psychological scales for pre- and post-intervention efficacy assessment, and few studies applied physiological indicators as an efficacy assessment indicator. Because of the lack of reliability and validity of efficacy assessment tools based on painting elements (e.g., line, composition, color, prototype imagery), few studies used painting elements for efficacy assessment. Most studies focused on the changes in mental health indicators, and the cluster analysis also suggested that the frontier research hotspots included key words such as medical humanities, indicating that the evaluation of the painting therapy effectiveness covered changes in psychological scale indicators and the improvement of psychological conditions, as well as humanistic care for community rehabilitation, special populations, and clinical patients. For example, attention was paid to the improvement of patients’ quality of life (Bozcuk et al., 2017), reduction of pain (Stinley et al., 2015), improvement of interpersonal relationships (Durrani, 2019), improvement of psychological resilience (Huss et al., 2012), and increased self-esteem, self-efficacy, and happiness (Beauregard, 2014). A study (Versluis et al., 2022) that investigated the hair loss experience and feelings of patients with cancer receiving chemotherapy based on negative emotions caused by hair loss through painting therapy suggested that medical practitioners should pay attention to and take measures against hair loss in these patients. FIGURE 8 The subjects clustered using the log-likelihood ratio algorithm. In this map, each cluster contained a set of closely related keywords which formed a certain subject. The smaller the number of the cluster, the more keywords it contained, and the larger the corresponding area. The corresponding time was distinguished by the cluster’s color, with cooler colors signifying earlier years and warmer colors indicating more recent ones. FIGURE 8 The subjects clustered using the log-likelihood ratio algorithm. In this map, each cluster contained a set of closely related keywords which formed a certain subject. The smaller the number of the cluster, the more keywords it contained, and the larger the corresponding area. The corresponding time was distinguished by the cluster’s color, with cooler colors signifying earlier years and warmer colors indicating more recent ones. It is noteworthy that the term “COVID-19” was found in the top 15 keywords with the strongest citation bursts, suggesting that painting therapy can also be applied to the psychological problems and diseases associated with the recent pandemic. Frontiers in Psychology frontiersin.org 4.2. Research focus on the research frontier and hot topics On the other hand, mental health status is crucial to patient recovery, as it is associated with reduced quality of life, impaired social relationships, prolonged recovery time, poor treatment compliance and abnormal behaviour, as well as potentially shorter survival (Riba and Grassi, 2012). Painting therapy has relevance in hospital clinical practice as an affordable, easy-to-use and highly participatory psychotherapeutic tool that can improve patients’ psychological well-being and compliance. In addition, the dramatic increase in mental health problems among university students under the COVID-19 pandemic has exerted enormous pressure on the In terms of research interventions, painting therapy can be applied alone, or combined with other therapies such as music therapy (Keidar et al., 2021), dance therapy (Kennedy et al., 2014), narrative therapy (Kruger and Swanepoel, 2017), mindful training (Jang et al., 2016), and cognitive behavioral therapy (Iosa et al., 2021), as well as other non-pharmacological treatment modalities. For patients with mental or physical illness, most interventions were pharmacotherapy combined with painting therapy with a view to exploring the effects of painting therapy in a safe and ethical manner. There were also studies (Feniger-Schaal et al., 2022) that explored the possibility of online/remote painting therapy, using virtual reality drawing to explore changes in patients’ mental impairment, physical functioning, and cognitive abilities (Iosa et al., 2021), which offered an innovative form of painting therapy intervention. 10 10.3389/fpsyg.2023.1148391 Liang et al. TABLE 5 The largest 13 clusters of painting therapy document co-citation, identified by subject headings. 4.2. Research focus on the research frontier and hot topics Cluster ID Size Silhouette Mean (year) Top terms (LLR) 0 39 0.904 2010 Qualitative content analysis (9.14, 0.005); evaluation (4.55, 0.05); amplifications (4.55, 0.05); relaxation (4.55, 0.05); social change (4.55, 0.05); yoga (4.55, 0.05); eating disorder (4.55, 0.05); mind-body therapy (4.55, 0.05); incest survivors (4.55, 0.05); adults (4.55, 0.05); eurythmy (4.55, 0.05) 1 34 0.874 2016 Emotion regulation (8.38, 0.005); compassion (4.17, 0.05); spirituality (4.17, 0.05); symptoms (4.17, 0.05); compassion focused therapy (4.17, 0.05); maternal representation (4.17, 0.05); joint drawing (4.17, 0.05); art-based evaluation (4.17, 0.05); internal representations (4.17, 0.05); assessment (4.17, 0.05); bias (4.17, 0.05) 2 33 0.994 2008 Oncology (6.64, 0.01); arts-focused intervention (6.64, 0.01); museum object (6.64, 0.01); heritage- focused intervention (6.64, 0.01); health services research (6.64, 0.01); psychosis (6.64, 0.01); psychosocial (6.64, 0.01); occupational therapy (6.64, 0.01); art (6.07, 0.05); well-being (3.94, 0.05); rehabilitation (3.94, 0.05) 3 33 0.904 2012 Grounded theory (7.59, 0.01); therapeutic process (7.59, 0.01); adult mental health (4.1, 0.05); clinical practice (4.1, 0.05); survey (4.1, 0.05); depression (4.1, 0.05); transformation in parent–child relations (3.78, 0.1); counselling psychology (3.78, 0.1); major depression (3.78, 0.1); trauma (3.78, 0.1); medical art therapy (3.78, 0.1) 4 31 0.893 2015 Dementia (10.89, 0.001); visual art (10.09, 0.005); cognition (10.09, 0.005); art (7.39, 0.01); neurodegeneration (5.01, 0.05); acceptability (5.01, 0.05); music (5.01, 0.05); dance (5.01, 0.05); neuropsychology (5.01, 0.05); non-pharmacological intervention (5.01, 0.05); longitudinal (5.01, 0.05); 5 30 0.902 2016 Military (9.83, 0.005); traumatic brain injury (9.83, 0.005); veterans (9.83, 0.005); art therapy (7.97, 0.005); self-awareness (4.89, 0.05); trauma therapy (4.89, 0.05); suicide survivors (4.89, 0.05); avoidance (4.89, 0.05); bereavement (4.89, 0.05); military trauma (4.89, 0.05); drawings (4.89, 0.05) 6 23 0.953 2014 Pictures as signs of a neuroception of safety (6.37, 0.05); pictorial semiotics (6.37, 0.05); creating alongside (6.37, 0.05); art-based research (6.37, 0.05); assessment tool (6.37, 0.05); art-therapy (6.37, 0.05); art therapy and trauma (6.37, 0.05); trauma-therapy (6.37, 0.05); signs of dissociation and recovery (6.37, 0.05); altered state of consciousness and trauma (6.37, 0.05); increased positive affect and states (6.37, 0.05) 7 21 0.896 2017 Virtual reality (18.07, 1.0E-4); arts-based research (8.94, 0.005); psychomotor therapy (8.94, 0.005); psychotherapy (5.34, 0.05); poverty (4.45, 0.05); digital art therapy (4.45, 0.05); autonomic nervous system (4.45, 0.05); skin conductance (4.45, 0.05); school-based art therapy (4.45, 0.05); vr therapeutic setting (4.45, 0.05); anxiety (4.45, 0.05) 8 18 0.984 2018 Creative arts therapies (8.64, 0.005); nursing practice (6.13, 0.05); corona covid-19 (6.13, 0.05); tele- creative arts therapies (6.13, 0.05); tele-psychotherapy (6.13, 0.05); narrative gerontology (6.13, 0.05); tele-arts therapy (6.13, 0.05); nursing research (6.13, 0.05); photocollage (6.13, 0.05); creative self- efficacy (6.13, 0.05); older adults (6.13, 0.05) 9 15 0.946 2017 Cancer (12.22, 0.001); pediatric (10.98, 0.001); creative arts therapy (10.98, 0.001); quality of life (7.29, 0.01); cortisol (5.45, 0.05); il-6 (5.45, 0.05); posture (5.45, 0.05); psychosocial well-being (5.45, 0.05); complementary alternative therapy (5.45, 0.05); coloring (5.45, 0.05); survivorship care (5.45, 0.05) 10 13 1 2018 Intersectionality (11.52, 0.001); ghostly matters (5.71, 0.05); white supremacy (5.71, 0.05); resilience (5.71, 0.05); white antiracism (5.71, 0.05); museum (5.71, 0.05); intergenerational relations (5.71, 0.05); loneliness (5.71, 0.05); creative process (5.71, 0.05); self-reflexivity (5.71, 0.05); radical care (5.71, 0.05) 11 13 0.999 2017 e-mental health (7.15, 0.01); videoconferencing psychotherapy (7.15, 0.01); expressive arts therapy (7.15, 0.01); psychotherapists (7.15, 0.01); telemental health (7.15, 0.01); psychological wellbeing (7.15, 0.01); cognitive function (7.15, 0.01); telehealth (7.15, 0.01); mild cognitive impairment (7.15, 0.01); covid-19 (4.43, 0.05); psychotherapy (3.44, 0.1) 12 12 0.961 2017 Psychiatry (7.35, 0.01); manual (7.35, 0.01); expectancy (7.35, 0.01); patient preferences (7.35, 0.01); practice (7.35, 0.01); group arts therapies (7.35, 0.01); group (7.35, 0.01); intervention development (4.63, 0.05); arts therapies (3.63, 0.1); mental health (2.54, 0.5); art therapy (1.11, 0.5) 13 12 0.896 2012 Joint attention (7.15, 0.01); relational processes (7.15, 0.01); work-related stress (7.15, 0.01); care home (7.15, 0.01); older people (7.15, 0.01); staff groups (7.15, 0.01); arts for health (7.15, 0.01); art-viewing (7.15, 0.01); community (4.43, 0.05); art-making (4.43, 0.05); dementia (2.81, 0.1) TABLE 5 The largest 13 clusters of painting therapy document co-citation, identified by subject headings. Conflict of interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. 4.2. Research focus on the research frontier and hot topics 11 Frontiers in Psychology frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. further studies could focus on the clinical impact of painting therapy for human psychosomatic health. school counselling system (Charles et al., 2021; Mauer et al., 2022). Financial constraints and staff shortages have made face-to-face case consultations no longer an urgent solution for the currently large group of university students with psychological conditions, while painting therapy, an affordable, accessible and rapid assessment and treatment, may address this problem. Therefore, it is important to explore painting therapy in colleges and hospitals for practical reasons. Further exploration of the neuroendocrine mechanisms of painting therapy’s effect on mental health in the future may help to clarify the mechanisms of painting therapy and its scope of application. Finally, several studies demonstrated the effectiveness of painting therapy in improving physical and mental health. Furthermore, painting therapy can provide benefit to a range of individuals, especially women, adolescents, and patients with Alzheimer’s disease. It also has significant effects on certain special groups, such as prisoners, pregnant women, people with low education and literacy levels, or people with certain physical diseases such as cancer. Meanwhile, there are some limitations to this article, such as the omission of certain publications led by the single database source and the specific search settings, a bias in reference frequency resulted from the time constraints of the search, information complication because of the inability of CiteSpace to perform full and fractional count systems. Therefore, further research in the future may consider additional databases and more loosely defined terms for data collection, perform quantitative exploration of the relationship between citation rates, themes, and publications, use both quantitative and qualitative indicators to analyze the development of painting therapies in specific journals, or employ with more advanced methods for comprehensive analysis, such as machine learning, text mining, and other advanced approaches. Mechanisms and criteria for assessing the efficacy of painting therapy are necessary for further research to advance the field, and researchers could develop painting therapy criteria to measure clinical practice. 5. Conclusion This work is supported by Research Elite Funds for Institute of Analytical Psychology, Faculty of Humanities and Social Sciences, City University of Macau. (CUM: RE202201). This analysis provides information of publications on trends in the painting therapy area from 2011 to 2022. First, it suggests several theoretical implications regarding publications to enable researchers to further promote their research. The annual publication volume of painting therapy research increased significantly over the past decade, with the overall trend in publication volume nearly tripling from 54 in 2011 to 154 in 2021. In 2022, however, as of July 2022, publication rate seems to have decreased, which needs to be followed up further. This study also enhances the holistic understanding of the global research structure in this field. A considerable degree of collaboration among different countries or regions and authors in painting therapy research was observed, which may expand the knowledge of painting therapy. Second, this study has several other practical implications. Among the keywords citation bursts, the keywords “depression,” “women,” and “recovery” highlight the areas where painting therapy may be helpful in alleviating depression, benefiting women’s groups, and facilitating rehabilitation. The development of painting therapy is promising, and This analysis provides information of publications on trends in the painting therapy area from 2011 to 2022. First, it suggests several theoretical implications regarding publications to enable researchers to further promote their research. The annual publication volume of painting therapy research increased significantly over the past decade, with the overall trend in publication volume nearly tripling from 54 in 2011 to 154 in 2021. In 2022, however, as of July 2022, publication rate seems to have decreased, which needs to be followed up further. This study also enhances the holistic understanding of the global research structure in this field. A considerable degree of collaboration among different countries or regions and authors in painting therapy research was observed, which may expand the knowledge of painting therapy. 4.3. Strengths and limitations This study analyzed large-scale data from painting therapy publications over the past 10 years using CiteSpace. More comprehensive results, instead of simply reviewing articles and studies can be detected by CiteSpace. What’s more, bibliometric, as an in-depth analysis approach, enabled us to detect emerging trends and collaborations among authors and countries or regions.h This bibliometric study had some limitations. First, we only analyzed the publications in WoS; further reviews could select the articles from other canonical databases such as Scopus and PubMed. Second, the time constraints of the search might have led to bias in reference frequency. For instance, some studies were published recently, and they might not yet be cited as frequently as older articles. The search was conducted in July 2022, so full year data for 2022 cannot be presented. Third, quantitative analysis methods were used, and only limited qualitative analysis was conducted. Furthermore, CiteSpace was used to bibliometricly analyzed the publications, but the full count and fractional count system cannot be performed in this software to complicate the information. Therefore, further studies could use both quantitative and qualitative metrics to analyze the development of painting therapy in particular journals, and explore the relationship between citation rates, themes, and publications. Frontiers in Psychology Author contributions QL, HQ, and HS designed the paper. JY, YL, and JD edited the English text of a draft of this manuscript. QL and HQ revised the paper. HS and QL supervised the project. QL, JY, YL, JD, HS, and HQ participated in drafting and reviewing. All authors contributed to the article and approved the submitted version. frontiersin.org References F., Sela, N., Caspi, O., and Weihs, K. (2019). The role of emotional processing in art therapy (REPAT) for breast cancer patients. J. Psychosoc. Oncol. 37, 586–598. doi: 10.1080/07347332.2019.1590491 Lin, M., Chen, Y., and Chen, R. (2021). Bibliometric analysis on Pythagorean fuzzy sets during 2013-2020. Int. J. Intell. Comput. Cybernet. 14, 104–121. doi: 10.1108/ IJICC-06-2020-0067 Durieux, V., and Gevenois, P. A. (2010). Bibliometric indicators: quality measurements of scientific publication. Radiology 255, 342–351. doi: 10.1148/radiol.09090626 MacIntosh, H. B. (2017). A bridge across silent trauma: enactment, art, and emergence in the treatment of a traumatized adolescence. Psychoanal. Dial. 27, 433–453. doi: 10.1080/10481885.2017.1328188 Durrani, H. (2019). Art Therapy's scope to address impaired attachment in children with ASD and comorbid SID. Art Ther. 37, 131–138. doi: 10.1080/07421656.2019.1677063 Mauer, V. A., Littleton, H., Lim, S., Sall, K. E., Siller, L., and Edwards, K. M. (2022). Fear of COVID-19, anxiety, and social support among college students. J. Am. Coll. Heal. 1-8, 1–8. doi: 10.1080/07448481.2022.2053689 Elimimian, E. B., Elson, L., Stone, E., Butler, R. S., Doll, M., Roshon, S., et al. (2020). A pilot study of improved psychological distress with art therapy in patients with cancer undergoing chemotherapy. BMC Cancer 20:899. doi: 10.1186/s12885-020-07380-5 Merigó, J. M., Mulet-Forteza, C., Valencia, C., and Lew, A. A. (2019). Twenty years of tourism geographies: a bibliometric overview. Tour. Geogr. 21, 881–910. doi: 10.1080/14616688.2019.1666913 Envelope, L. Z. A., Envelope, J. L. B., and Envelope, M. L. B. C. P. (2022). Artificial intelligence in renewable energy: a comprehensive bibliometric analysis. Energy Rep. 8, 14072–14088. doi: 10.1016/j.egyr.2022.10.347 Miao, Y., Xu, S.-Y., Chen, L.-S., Liang, G.-Y., Pu, Y.-P., and Yin, L.-H. (2017). Trends of long noncoding RNA research from 2007 to 2016: a bibliometric analysis. Oncotarget 8, 83114–83127. doi: 10.18632/oncotarget.20851 Feniger-Schaal, R., Orkibi, H., Keisari, S., Sajnani, N. L., and Butler, J. D. (2022). Shifting to tele-creative arts therapies during the COVID-19 pandemic: an international study on helpful and challenging factors. Arts Psychother. 78:101898. doi: 10.1016/j. aip.2022.101898 Mulet-Forteza, C., Lunn, E., Merigó, J. M., and Horrach, P. (2021). Research progress in tourism, leisure and hospitality in Europe (1969–2018). Int. J. Contemp. Hosp. Manag. 33, 48–74. doi: 10.1108/IJCHM-06-2020-0521 Gonzalez Barajas, A. K., and Ho, R. T. H. (2020). Drawings from a man at a penitentiary Centre in Mexico. Epidemiol. Psychiatr. Sci. 29:e171. doi: 10.1017/ S2045796020000852 Niedzwiedz, C. L., Knifton, L., Robb, K. A., Katikireddi, S. V., and Smith, D. J. (2019). References psychres.2021.113706 Kai, H., Cao, L., Li, Y., Zhu, J., Zhu, D., and Qi, L. (2022). Observation on the effect of painting therapy on bad emotion in children with functional gastrointestinal diseases. Minerva Pediatr. (Torino) 74, 245–247. doi: 10.23736/S2724-5276.21.06515-0 Keidar, L., Regev, D., and Snir, S. (2021). Non-Haredi arts Therapists' perceptions of therapy with ultra-orthodox children. Front. Psychol. 12:599872. doi: 10.3389/ fpsyg.2021.599872 Chen, C. (2004). Searching for intellectual turning points: progressive knowledge domain visualization. Proc. Natl. Acad. Sci. U. S. A. 101, 5303–5310. doi: 10.1073/ pnas.0307513100 Kennedy, H., Reed, K., and Wamboldt, M. Z. (2014). Staff perceptions of complementary and alternative therapy integration into a child and adolescent psychiatry program. Arts Psychother. 41, 21–26. doi: 10.1016/j.aip.2013.10.007 Chen, C., and CiteSpace, I. I. (2006). Detecting and visualizing emerging trends. J. Am. Soc. Informat. Sci. Technol. 57, 359–377. doi: 10.1002/asi.20317 Chen, C., Hu, Z., Liu, S., and Tseng, H. (2012). Emerging trends in regenerative medicine: a scientometric analysis in CiteSpace. Expert. Opin. Biol. Ther. 12, 593–608. doi: 10.1517/14712598.2012.674507 Kleinberg, I. (2002). A mixed-bacteria ecological approach to understanding the role of the oral bacteria in dental caries causation: an alternative to Streptococcus mutans and the specific-plaque hypothesis. Crit. Rev. Oral Biol. Med. 13, 108–125. doi: 10.1177/154411130201300202 Chen, Y., Lin, M., and Zhuang, D. (2022). Wastewater treatment and emerging contaminants: Bibliometric analysis. Chemosphere 297:133932. doi: 10.1016/j. chemosphere.2022.133932 Klop, S. (2017). Sometimes words just Ain't enough – enhancing the contribution of children in therapy through creative expression. Aust. N. Z. J. Fam. Ther. 38, 283–294. doi: 10.1002/anzf.1218 Chen, C., Yue, C., Horowitz, M., Hou, H., Liu, Z., and Pellegrino, D. (2009). Towards an explanatory and computational theory of scientific discovery. J. Informet. 3, 191–209. doi: 10.1016/j.joi.2009.03.004 Kruger, D., and Swanepoel, M. (2017). Gluing the pieces together: female adolescents' construction of meaning through digital metaphoric imagery in trauma therapy. Arts Psychother. 54, 92–104. doi: 10.1016/j.aip.2017.04.011 Crombie, M., Dutt, A., Dey, P., Nandi, R., and Evans, J. (2022). Examination of the validity of the 'Papadum test': an alternative to the clock drawing test for people with low levels of education. Clin. Neuropsychol. 1-18, 1–18. doi: 10.1080/13854046.2022.2047789 Liang, Y. D., Li, Y., Zhao, J., Wang, X. Y., Zhu, H. Z., and Chen, X. H. (2017). Study of acupuncture for low back pain in recent 20 years: a bibliometric analysis via CiteSpace. J. Pain Res. 10, 951–964. doi: 10.2147/JPR.S132808 Czamanski-Cohen, J., Wiley, J. Publisher’s note Second, this study has several other practical implications. Among the keywords citation bursts, the keywords “depression,” “women,” and “recovery” highlight the areas where painting therapy may be helpful in alleviating depression, benefiting women’s groups, and facilitating rehabilitation. The development of painting therapy is promising, and All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. 12 frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. References Hu, J., Zhang, J., Hu, L., Yu, H., and Xu, J. (2021). Art therapy: a complementary treatment for mental disorders. Front. Psychol. 12:686005. doi: 10.3389/ fpsyg.2021.686005 Association TAAT. About art therapy. Available at: https://arttherapy.org/about-art- therapy/ (Accessed January 18, 2023). Beauchet, O., Lafleur, L., Remondière, S., Galery, K., Vilcocq, C., and Launay, C. P. (2020). Effects of participatory art-based painting workshops in geriatric inpatients: results of a non-randomized open label trial. Aging Clin. Exp. Res. 32, 2687–2693. doi: 10.1007/s40520-020-01675-0 Huertas González-Serrano, M., Jones, P., and Llanos-Contrera, O. (2020). An overview of sport entrepreneurship field: a bibliometric analysis of the articles published in the web of science. Sport Soc. 23, 296–314. doi: 10.1080/17430437.2019.1607307 Huss, E., Nuttman-Shwartze, O., and Altman, A. (2012). The role of collective symbols as enhancing resilience in children's art. Arts Psychother. 39, 52–59. doi: 10.1016/j. aip.2011.11.005 Beauregard, C. (2014). Effects of classroom-based creative expression programs on children's well-being. Arts Psychother. 41, 269–277. doi: 10.1016/j.aip.2014.04.003 Bozcuk, H., Ozcan, K., Erdogan, C., Mutlu, H., Demir, M., and Coskun, S. (2017). A comparative study of art therapy in cancer patients receiving chemotherapy and improvement in quality of life by watercolor painting. Complement. Ther. Med. 30, 67–72. doi: 10.1016/j.ctim.2016.11.006 Iosa, M., Aydin, M., Candelise, C., Coda, N., Morone, G., Antonucci, G., et al. (2021). The Michelangelo effect: art improves the performance in a virtual reality task developed for upper limb Neurorehabilitation. Front. Psychol. 11:611956. doi: 10.3389/ fpsyg.2020.611956 Caruso, R., Nanni, M. G., Riba, M. B., Sabato, S., and Grassi, L. (2017). The burden of psychosocial morbidity related to cancer: patient and family issues. Int. Rev. Psychiatry 29, 389–402. doi: 10.1080/09540261.2017.1288090 Jang, S. H., Kang, S. Y., Lee, H. J., and Lee, S. Y. (2016). Beneficial effect of mindfulness- based art therapy in patients with breast Cancer-a randomized controlled trial. Explore (N.Y.) 12, 333–340. doi: 10.1016/j.explore.2016.06.003 Castanha, R. C. G., and Grácio, M. C. C. (2014). Bibliometrics contribution to the Metatheoretical and domain analysis studies. Knowl. Organ. 41, 171–174. doi: 10.5771/0943-7444-2014-2-171 Jin, L., Zegang, X., and Hui, C. (2022). Research review of rural public space in China based on CiteSpace (in Chinese). South Architect. 2, 11–21. doi: 10.3969/j.issn.1000-0232.2022.02.002 Charles, N. E., Strong, S. J., Burns, L. C., Bullerjahn, M. R., and Serafine, K. M. (2021). Increased mood disorder symptoms, perceived stress, and alcohol use among college students during the COVID-19 pandemic. Psychiatry Res. 296:113706. doi: 10.1016/j. Frontiers in Psychology References Depression and anxiety among people living with and beyond cancer: a growing clinical and research priority. BMC Cancer 19:943. doi: 10.1186/s12885-019-6181-4 Hashemi, S.-M., Rafiemanesh, H., Aghamohammadi, T., Badakhsh, M., Amirshahi, M., Sari, M., et al. (2020). Prevalence of anxiety among breast cancer patients: a systematic review and meta-analysis. Breast Cancer 27, 166–178. doi: 10.1007/s12282-019-01031-9 Pilevarzadeh, M., Amirshahi, M., Afsargharehbagh, R., Rafiemanesh, H., Hashemi, S.- M., and Balouchi, A. (2019). Global prevalence of depression among breast cancer patients: a systematic review and meta-analysis. Breast Cancer Res. Treat. 176, 519–533. doi: 10.1007/s10549-019-05271-3 He, Z., Lin, M., Xu, Z., Yao, Z., Chen, H., Alhudhaif, A., et al. (2022). Deconv- transformer (DecT): a histopathological image classification model for breast cancer based on color deconvolution and transformer architecture. Inf. Sci. 608, 1093–1112. doi: 10.1016/j.ins.2022.06.091 Pongan, E., Tillmann, B., Leveque, Y., Trombert, B., Getenet, J. C., Auguste, N., et al. (2017). Can musical or painting interventions improve chronic pain, mood, quality of life, and cognition in patients with mild Alzheimer's disease? Evidence from a randomized controlled trial. J. Alzheimers Dis. 60, 663–677. doi: 10.3233/JAD-170410 Henry, S. (1973). Co-citation in the scientific literature: a new measure of the relationship between two documents. J. Am. Soc. Inf. Sci. 24, 265–269. doi: 10.1002/ asi.4630240406 Puetz, T. W., Morley, C. A., and Herring, M. P. (2013). Effects of creative arts therapies on psychological symptoms and quality of life in patients with cancer. JAMA Intern. Med. 173, 960–969. doi: 10.1001/jamainternmed.2013.836 Hirano, Y., and Hosaka, K. (2018). Creativity and arts in psychiatry. Psychiatry Clin. Neurosci. 72:1. doi: 10.1111/pcn.12617 13 Frontiers in Psychology frontiersin.org frontiersin.org 10.3389/fpsyg.2023.1148391 Liang et al. Wahlbeck, H., Kvist, L. J., and Landgren, K. (2020). Art therapy and counseling for fear of childbirth: a randomized controlled trial. Art Ther. 37, 123–130. doi: 10.1080/07421656.2020.1721399 Qiu, H. Medicine and human culture (in Chinese). Guangzhou: Guangdong Higher Education Press; (2004). Qiu, H. Z., Ye, Z. J., Liang, M. Z., Huang, Y. Q., Liu, W., and Lu, Z. D. (2017). Effect of an art brut therapy program called go beyond the schizophrenia (GBTS) on prison inmates with schizophrenia in mainland China-a randomized, longitudinal, and controlled trial. Clin. Psychol. Psychother. 24, 1069–1078. doi: 10.1002/cpp.2069 Wang, X. Q., Peng, M. S., Weng, L. M., Zheng, Y. L., Zhang, Z. J., and Chen, P. J. (2019). Bibliometric study of the comorbidity of pain and depression research. Neural Plast. 2019, 1–16. doi: 10.1155/2019/1657498 Riba, M., and Grassi, L. References (2012). Clinical psycho-oncology: an international perspective. Hoboken, NJ: John Wiley & Sons. Wiswell, S., Bell, J. G., McHale, J., Elliott, J. O., Rath, K., and Clements, A. (2019). The effect of art therapy on the quality of life in patients with a gynecologic cancer receiving chemotherapy. Gynecol. Oncol. 152, 334–338. doi: 10.1016/j.ygyno.2018.11.026 Schofield, S. (2019). Group art therapy, aesthetic experiences of difference and belonging. Lang. Psychoanal. 8, 30–68. doi: 10.7565/landp.v8i1.1591 Yu, D., Xu, Z., and Hamido, F. (2018). Bibliometric analysis on the evolution of applied intelligence. Appl. Intell. 49, 449–462. doi: 10.1007/s10489-018-1278-z Stinley, N. E., Norris, D. O., and Hinds, P. S. (2015). Creating mandalas for the Management of Acute Pain Symptoms in pediatric patients. Art Ther. 32, 46–53. doi: 10.1080/07421656.2015.1028871 Yu, D., Xu, Z., and Pedrycz, W. (2020). Bibliometric analysis of rough sets research. Appl. Soft Comput. 94:106467. doi: 10.1016/j.asoc.2020.106467 Sw, A., Jgb, B., Mh, C., Joe, D., Kr, B., and Ac, B. (2019). The effect of art therapy on the quality of life in patients with a gynecologic cancer receiving chemotherapy - ScienceDirect. Gynecol. Oncol. 152, 334–338. doi: 10.1016/j.ygyno.2018.11.026 Zamanifard, M., Soltanian, M., Edraki, M., Moravaj, H., and Sharifi, N. (2022). The effects of virtual directed painting therapy on anxiety, depression, and self-efficacy of children with type 1 diabetes: a randomized controlled clinical trial. Int. J. Comm. Based Nurs. Midwifery 10, 210–222. doi: 10.30476/IJCBNM.2022.93400.1927 Tal, D., and Gordon, A. (2018). Women as political leaders: a Bibliometric analysis of the literature. Society 55, 256–261. doi: 10.1007/s12115-018-0248-8 Zhang, L., Ling, J., and Lin, M. (2023). Carbon neutrality: a comprehensive bibliometric analysis. Environ. Sci. Pollut. Res. 30, 45498–45514. doi: 10.1007/ s11356-023-25797-w Thyme, K. E., Tavelin, B., Öster, I., Lindh, J., and Wiberg, B. (2022). Art therapy and self-image: a 5-year follow-up art therapy RCT study of women diagnosed with breast cancer. Palliat. Support. Care 20, 482–490. doi: 10.1017/S1478951521001437 Zheng, K., and Wang, X. (2019). Publications on the association between cognitive function and pain from 2000 to 2018: a Bibliometric analysis using CiteSpace. Med. Sci. Monit. 25, 8940–8951. doi: 10.12659/MSM.917742 Versluis, A., Alphen, K. V., Dercksen, W., Haas, H., Kaptein, A. A., and CVD, H. (2022). Looking bad: female patients drawing their representation of chemotherapy-induced alopecia. J. Health Psychol. 27, 3013–3027. doi: 10.1177/13591053221075503 Zhong, M., and Lin, M. (2022). Bibliometric analysis for economy in COVID-19 pandemic. Heliyon 8:e10757. doi: 10.1016/j.heliyon.2022.e10757 14 frontiersin.org Frontiers in Psychology 14 Frontiers in Psychology
https://openalex.org/W3111604490	https://hal.science/hal-03044240/document	English	null	New geographical and host record of bat ectoparasite Steatonyssus (Steatonyssus) afer Radovsky and Yunker, 1963 (Mesostigmata: Gamasina: Macronyssidae)	Acarologia	2,020	cc-by	3,044	How to cite this article Orlova M. V. et al. (2020), New geographical and host record of bat ectoparasite Steatonyssus (Steatonyssus) afer Radovsky and Yunker, 1963 (Mesostigmata: Gamasina: Macronyssidae). Acarologia 60(4): 951 958; DOI 10.24349/acarologia/20204411 Maria V. Orlovaa,b , Theresa M. Lavertyc , Will K. Reevesd , Elena M. Grattonc , Mallory L. Daviesc , Nikolay V. Anisimova a Tyumen State University, Tyumen, Russia. c Department of Fish, Wildlife, and Conservation Biology, Colorado State University, Fort Collins, Colorado, USA. d Colorado State University, C. P. Gillette Museum of Arthropod Diversity, Fort Collins, Colorado, USA. Acarologia is proudly non-profit, with no page charges and free open access Please help us maintain this system by encouraging your institutes to subscribe to the print version of the journal and by sending us your high quality research on the Acari. Subscriptions: Year 2020 (Volume 60): 450 € http://www1.montpellier.inra.fr/CBGP/acarologia/subscribe.php Previous volumes (2010-2018): 250 € / year (4 issues) Acarologia, CBGP, CS 30016, 34988 MONTFERRIER-sur-LEZ Cedex, France ISSN 0044-586X (print), ISSN 2107-7207 (electronic) Acarologia A quarterly journal of acarology, since 1959 Publishing on all aspects of the Acari All information: http://www1.montpellier.inra.fr/CBGP/acarologia/ acarologia-contact@supagro.fr Acarologia is under free license and distributed under the terms of the Creative Commons-BY-NC-ND which permits unrestricted non-commercial use distribution and The digitalization of Acarologia papers prior to 2000 was supported by Agropolis Fondation under the reference ID 1500-024 through the « Investissements d’avenir » programme (Labex Agro: ANR-10-LABX-0001-01) Acarologia is under free license and distributed under the terms of the Creative Commons-BY-NC-ND which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT Steatonyssus afer is recorded for the first time from Namibia and for the first time from Cistugo seabrae (Chiroptera: Cistugidae). Our finding is the southernmost locality for S. afer, expanding the geographical distribution range of this ectoparasite. We give diagnostic illustrations and measurements of the species. Keywords Steatonyssus afer; Macronyssidae; Namibia; Cistugo seabrae Figure 1 Map of Africa, showing previous (triangles) and new (circle) record sites for the bat ec toparasitic mite Steatonyssus (Steatonyssus) afer. Figure 1 Map of Africa, showing previous (triangles) and new (circle) record sites for the bat ec toparasitic mite Steatonyssus (Steatonyssus) afer. whitewinged serotine Neoromicia tenuipinnis Peters, 1872 (Chiroptera: Vespertilionidae), and tha largeeared slitfaced bat Nycteris macrotis Dobson, 1876 (Chiroptera: Nycteridae)] (Radovsky and Yunker, 1963), in Sierra Leone (host was determined as a “small bat”) (Till, Evans, 1964, and in Afghanistan from the serotine bat Eptesicus serotinus and the mousetailed bat Rhinopoma sp. (Dusbábek, 1970) (Fig. 1). Here, we report S. afer, including a brief description of adult females, in Namibia and associated with the Angolan winggland bat Cistugo seabrae Thomas, 1912 (Chiroptera: Cistugidae), including a brief redescription of female. Introduction Steatonyssus is the largest genus in the family Macronyssidae and has been recorded from every major zoogeographic region (limited in the Australasian to one Australian species) (Radovsky, 1967, 2010). Radovsky (2010) recognized 53 species both in Steatonyssus (52) and Steatonysella (1) subgenera. Mites of genus Steatonyssus parasitize bats (Chiroptera: Vespertilionidae, Molossidae, Rhinolophidae, Hipposideridae, Nycteridae, Emballonuridae, and Pteropodidae), some rodents (Rodentia: Myoxidae), and elephant shrews (Macroscelidea: Macroscelididae). Received 27 January 2020 Accepted 03 December 2020 Published 07 December 2020 Corresponding author Maria V. Orlova: m.v.orlova@utmn.ru Academic editor Roy, Lise DOI 10.24349/acarologia/20204411 ISSN 0044586X (print) ISSN 21077207 (electronic) Copyright Orlova M. V. et al. Distributed under Creative Commons CC-BY 4.0 According to Coffee (1973), for the Afrotropical (Ethiopian) region, 16 species belonging to the genus Steatonyssus have been recorded: Steatonyssus periblepharus Kolenati, 1858, Steatonyssus (Steatonyssus) aelleni Radovsky and Yunker, 1963; Steatonyssus (Steatonyssus) afer Radovsky and Yunker, 1963; Steatonyssus (Steatonyssus) benoiti Till and Evans, 1964; Steatonyssus (Steatonyssus) brucei Lavoipierre, 1956; Steatonyssus (Steatonyssus) calcaratus Radovsky and Yunker, 1963; Steatonyssus (Steatonyssus) crassisetosus Till and Evans, 1964; Steatonyssus (Steatonyssus) eos Zumpt and Till, 1954; Steatonyssus (Steatonyssus) hipposideros Till, 1958; Steatonyssus (Steatonyssus) mutatus Coffee, 1973; Steatonyssus (Steatonyssus) natalensis Zumpt and Patterson, 1951; Steatonyssus (Steatonyssus) nyassae Hirst, 1922; Steatonyssus (Steatonyssus) roeri Coffee, 1973; Steatonyssus (Steatonyssus) sudanensis Hirst, 1923; Steatonyssus (Steatonyssus) tibialis Till and Evans, 1964; and Steatonyssus (Steatonyssus) javensis brevisetosus Till and Evans, 1964. Most of them are known only by the first description and, in some cases, scattered records. Steatonyssus afer is known by the findings in Angola from mixed collection of hosts [(the banana pipistrelle Neoromicia nana Peters, 1852, the by Monadjem et al. (2010) and we recorded the following data for each bat: sex, age (juvenile or adult), forearm length (mm), body mass (g), and reproductive status (nonreproductive, reproductive, pregnant, or lactating). The body (front, back, tail, wings, ears, uropatagium, etc.) of all bats were thoroughly and systematically surveyed in hand and examined for ectoparasites using an LED headlamp. All ectoparasites were removed with forceps, pooled into one sample for each individual bat, and preserved in 95% ethanol before the bats were released. For each ectoparasite sample, we transferred the mites into a new vial containing 70% ethanol and sent all mites to the Institute Xbio of Tyumen State University (Russia) for species identification. After clearing, the mites were mounted on permanent slides with FaureBerlese’s mounting medium (Whitaker, 1988). Morphological identification of mites was done by the first author (MO), according to the keys by Till and Evans (1964). The morphological terminology generally follows Evans and Till (1979). Dorsal and ventral setae were labelled according to the systems of Lindquist and Evans (1965). Photographs were taken with a digital camera AxioCam ICc5 (Zeiss, Germany) via a compound microscope AxioImager A2 (Zeiss, Germany) with a phasecontrast and DIC objectives. All measurements are given in micrometers (μm). The specimens mounted on five slides were deposited at the collection of the Tyumen State University’s Museum of Zoology, Tyumen Province, Russia. Genus Steatonyssus Kolenati, 1858 Type species: Steatonyssus periblepharus Kolenati, 1858, by designation of Till and Evans, 1964 Results and discussion Five specimens of macronyssid mites (Mesotigmata: Gamasina: Macronyssidae) were obtained from adult females of the Angolan winggland bat C. seabrae, in two localities of Namibia: Okongwe (18°59′S 13°9′E) (one bat) and the junction of the Hoanib and Mudorib Rivers (19°19′S 13°14′E) (one bat). Mites (five females) were morphologically identified as S. afer. Measurement information from the present study and previous ones is provided in Table 1. Materials and methods We conducted research on Namib Desert bats and their ectoparasites in the Kunene Region of northwestern Namibia (Fig. 1). This region is comprised of pastoralist communal conservancies and tourism concessions bordering Skeleton Coast Park, and receives ~30 to 100 mm of rainfall, on average, between January and April each year (Jacobson and Jacobson, 2013). From 6 December 2016 to 4 April 2017, we intensively examined captured bats for ectoparasites. We captured bats by deploying mist nets (2 m high by 4 to 12 m in length) above bodies of water for 1 to 3 hr after sunset for a total of 120.37 hr over 37 nights. We removed bats from mist nets within 1 to 5 min of capture and placed them in cloth bags until they could be processed 0 to 15 min later. Bats were morphologically identified using taxonomic descriptions rlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 952 Figures 2–3 Figures 2–3 g Differential diagnosis (female) — Steatonyssus afer is morphologically closest to S. roeri. The anterior margin of the sternal plate can hardly be detected in S. afer (vs. well defined in S. roeri). In S. afer, the peritreme starts ventral and bends over to the dorsal surface for most of its length, while the peritrematal plate is interrupted with a leaflike portion lying over coxae I and II (vs. not interrupted plate in S. roeri). The setae on the opisthonotal and podonotal shields are considerably longer and thicker in S. afer than in S. roeri (5471 vs. 4047, respectively). Brief description of female — Five specimens from Namibia were measured (Table 1). Idiosoma: ovoid, light brown colour. Dorsum (Fig.2AD), podonotal shield clearly reticulated, with 11 pairs of setae, j2 present; setae z2 – 3, s3 – 4 longest; posterior margin of shield flat (Fig. 2A, C); opisthonotal shield concave anteriorly with pronounced reticulate pattern, bearing 7 setal pairs, J1 – 3 long, S5 short, J5 microseta (Figs. 2B, D). Venter, sternal shield, not sharply demarcated from reticulate presternal area, posterior region heavily sclerotized (Fig. 3), with 2 pairs of pores and 3 setal pairs. Epigynial shield triangular, anterior flap has slender sclerotized median projection. Anal shield obovate, with 3 setae. Idiosomal unsclerotized integument with about 102110 pairs of smooth setae (6469 on dorsal side, and 3841 on ventral side); peritremes end over coxa III; peritrematal shield interrupted, separated anterior leaflike part situated over coxa I (Fig. 4). Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 953 Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 Figure 2 Female of Steatonyssus afer. A, C. Podonotal shield (scale bar 100 μm); B, D. Opisthonotal shield (scale bar 100 μm). igure 2 Female of Steatonyssus afer. A, C. Podonotal shield (scale bar 100 μm); B, D. Opisthonotal shield (scale bar 100 μm). Figure 2 Female of Steatonyssus afer. A, C. Podonotal shield (scale bar 100 μm); B, D. Opisthonotal shield (scale bar 100 μm). Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 954 954 Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 Gnathosoma: deutosternum with 7 teeth arranged in single file (Fig. 5). Legs: coxae II bears pronounced anterodorsal spur; in a single specimen from Mudorib River its tip divided into four denticles. Chaetotaxy of the legs typical for the genus (Fig. 6). Material examined — One female sampled from C. seabrae in Okongwe (18°59′S 13°9′E) (4 I 2017; coll. T. Laverty) and four females from C. seabrae close to the Mudorib River (19°19′S 13°14′E) (6 XII 2016; coll. T. Laverty) (Kunene Region, Namibia). Distribution — Angola (type locality) (Radovsky, Yunker, 1963), Sierra Leone (Till, Evans, 1964), Afghanistan (Dusbábek, 1970), Namibia (our data, first record). Hosts — Neoromicia nana (as Pipistrellus nanus – Radovsky, Yunker, 1963), Neoromicia Figure 3 Female of Steatonyssus afer. Sternal shield (scale bar 20 μm). Figure 3 Female of Steatonyssus afer. Sternal shield (scale bar 20 μm). Gnathosoma: deutosternum with 7 teeth arranged in single file (Fig. 5). Legs: coxae II bears pronounced anterodorsal spur; in a single specimen from Mudorib River its tip divided into four denticles. Chaetotaxy of the legs typical for the genus (Fig. 6). Material examined — One female sampled from C. seabrae in Okongwe (18°59′S 13°9′E) (4 I 2017; coll. T. Laverty) and four females from C. seabrae close to the Mudorib River (19°19′S 13°14′E) (6 XII 2016; coll. T. Laverty) (Kunene Region, Namibia). Distribution — Angola (type locality) (Radovsky, Yunker, 1963), Sierra Leone (Till, Evans, 1964), Afghanistan (Dusbábek, 1970), Namibia (our data, first record). Hosts — Neoromicia nana (as Pipistrellus nanus – Radovsky, Yunker, 1963), Neoromicia Figure 4 Female of Steatonyssus afer. Peritreme and peritremal shields (scale bar 100 μm). Figure 4 Female of Steatonyssus afer. Peritreme and peritremal shields (scale bar 100 μm). Figure 4 Female of Steatonyssus afer. Peritreme and peritremal shields (scale bar 100 μm). 955 Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 Table 1 Comparison of metric data of body parts between Steatonyssus afer adult female from Namibia and those described in Angola. All measurements are given in micrometers (μm). Table 1 Comparison of metric data of body parts between Steatonyssus afer adult female from Namibia and those described in Angola. All measurements are given in micrometers (μm). Table 1 Comparison of metric data of body parts between Steatonyssus afer adult female from Namibia and those described in Angola. All measurements are given in micrometers (μm). Table 1 Comparison of metric data of body parts between Steatonyssus afer adult female from Namibia and th measurements are given in micrometers (μm). Steatonyssus afer Steatonyssus afer Steatonyssus roeri Angola Namibia Namibia Body part (Radovsky, Yunker, 1963; Till, Evans, 1964) (this paper) (Coffee, 1973) Idiosoma, length 849 ≈1000 - Idiosoma, width 503 ≈700 - Podonotal shield, length 276 297.0±6.6 (289-304) 288 Podonotal shield, width - 250.8±6.9 (242-258) 278 j 2 - 28.5±2.7 (26-32) 24-26 z 2 – 3, s 3 – 4 67-76 74.9± 3.2 (71-80) - Opistonotal shield, length 308-338 357.2±7.3 (347-367) 365-403 Opistonotal shield, width 180-195 204.4±9.4 (194-213) 202 J 1 – 3 64-71 59.5±5.5 (52-66) 40-47 S 5 - 10.3±0.8 14 Sternal shield, length 57 57.2±3.8 (51-60) 54-59 Sternal shield, width - 155.8±2.7 (153-160) 127-137 St 1 39-43 39 - St 3 67-72 71, 72 - St 1 – St 1 47 61±2.5 (59-64) 61-64 Epigynial shield, length - 313.8±15.0 (296-329) - Epigynial shield, width - 111.8±5.4 (106-119) - Anal shield, length 150-165 186.8±5.2 (181-194) 125-137 Anal shield, width 74-82 90.8±2.2 (88-93) 85-96 Dorsal opistosomal setae, length - 74.5±2.6 (71-78) - Ventral opistosomal setae, length - 50.5±8.0 (39-60) - Tibia I 87 94, 101 - Tarsal I 157-160 177,182 - Tibia IV 80 99, 103 - Tarsal IV 168-180 208 - it would be useful to explore the genetic structure of this taxon by morphological comparisons on a larger and more representative sample and, if possible, by adding a molecular analysis to the morphological analysis. The individuals present the first record of S. afer for Namibia and the southernmost locality for the species to date. Also, C. seabrae is recorded as another host of the parasite. Given the poor knowledge of genus Steatonyssus in Africa, there is a need for further research on Steatonyssus species throughout Africa to establish complete host lists and geographical ranges of these ectoparasites. Figure 5 Female of Steatonyssus afer. Gnathosoma (scale bar 20 μm). Figure 5 Female of Steatonyssus afer. Gnathosoma (scale bar 20 μm). tenuipinnis (Chiroptera: Vespertilionidae) (as Eptesicus tenuipinnis – Radovsky, Yunker, 1963), Nycteris macrotis (Chiroptera: Nycteridae) (Radovsky, Yunker, 1963), Eptesicus serotinus (Chiroptera: Vespertilionidae) (Dusbábek, 1970), Rhinopoma sp. (Chiroptera: Rhinopomatidae) (Dusbábek, 1970), Cistugo seabrae (Chiroptera: Cistugidae) (our data, first record). Remarks — We note that there is a fairly large variation in size among the mites recorded so far (in Namibia in this study and in different regions in the literature; Table 1). Since we were only able to examine five specimens from Namibia (from five bats in two distant localities) and could not standardize our measurements by performing them in parallel on the types, we cannot conclude about possible crypticism within the species S. afer. However, we suggest that Figure 6 Female of Steatonyssus afer. Legs (scale bar 100 μm). Figure 6 Female of Steatonyssus afer. Legs (scale bar 100 μm). Figure 6 Female of Steatonyssus afer. Legs (scale bar 100 μm). Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 956 We are grateful to Dr. B. Halliday for his valuable comments and to the anonymous reviewers, whose advices made our manuscript substantially better. Ethical approval All applicable institutional, national and international guidelines for the care and use of animals were followed. Fieldwork was conducted with in accordance with the guidelines of Colorado State University’s Institutional Animal Care and Use Committee (Protocol #156140A) and the Ministry of Environment and Tourism, Republic of Namibia (research/collecting permits #2122/2016 and #2225/16). Orlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 Acknowledgments This work was supported by a National Science Foundation Graduate Research Fellowship to TML and by the Tomsk State University Competitiveness Improvement Program to MVO. 957 rlova M. V. et al. (2020), Acarologia 60(4): 951-958; DOI 10.24349/acarologia/20204411 g g ( ) Dusbábek F. 1970. Mite parasites (Acarina) of bats from Afghanistan. Folia Parasitolog. 17:6176. References Coffee G. M. 1973. Two new species of Steatonyssus from the Ethiopian Region Acarina Mesostigmata. Zeitschrift fuer Angewandte Zoologie.60 (1):2129. Coffee G. M. 1973. Two new species of Steatonyssus from the Ethiopian Region Acarina Mesostigmata. Zeitschrift fuer Angewandte Zoologie.60 (1):2129. g g ( ) Dusbábek F. 1970. Mite parasites (Acarina) of bats from Afghanistan. Folia Parasitolog. 17:6176. J b P J b K 2013 H d l i l f h i l d l i l i N Jacobson P., Jacobson K. 2013. Hydrologic controls of physical and ecological processes in Namib Desert ephemeral rivers: implications for conservation and management. J. Arid. Environ., 93:8093. doi:10.1016/j.jaridenv.2012.01.010 Monadjem A., Taylor P.J., Cotterill F.P.D., Schoeman M.C. 2010. Bats of Southern and Central Africa: A biogeographic and taxonomic synthesis. Wits University Press: Publisher. pp. 596. Monadjem A., Taylor P.J., Cotterill F.P.D., Schoeman M.C. 2010. Bats of Southern and Central Africa: A biogeographic and taxonomic synthesis. Wits University Press: Publisher. pp. 596. Radovsky F. 1967. The Macronyssidae and Laelapidae (Acarina: Mesotigmata) parasitic on bats. Barkeley, University of California: Publisher. pp. 288. g g p y y pp Radovsky F. 1967. The Macronyssidae and Laelapidae (Acarina: Mesotigmata) parasitic on bats. Barkeley, University of California: Publisher. pp. 288. y y pp Radovsky F. 2010. Revision of Genera of the parasitic mite family Macronyssidae (Mesostigmata: Dermanyssoidea) of the world. Indira Publishing House: Publisher. pp. 170. Dermanyssoidea) of the world. Indira Publishing House: Publisher. pp. 170. Radovsky F. J., Yunker C. E. 1963. Four New Species of Steatonyssus from Africa (Acarina: Dermanys sidae). The Journal of Parasitology, 49 (2): 334339. doi:10.2307/3276009 Radovsky F. J., Yunker C. E. 1963. Four New Species of Steatonyssus from Africa (Acarina: Dermanys sidae). The Journal of Parasitology, 49 (2): 334339. doi:10.2307/3276009 Till M. W., Evans O. G. 1964: The genus Steatonyssus Kolenati (Acari: Mesostigmata). Bulletin of the British Museum (Natural History), Zoology Series. 11: 511582. doi:10.5962/bhl.part.4724 Whi k J O J 1988 C ll i d i i f l i l d E l i l d Till M. W., Evans O. G. 1964: The genus Steatonyssus Kolenati (Acari: Mesostigmata). Bulletin of the British Museum (Natural History), Zoology Series. 11: 511582. doi:10.5962/bhl.part.4724 Whitaker J. O. Jr. 1988. Collecting and preserving ectoparasites for ecological study. References Ecological and B h i l M th d f th St d f B t W hi t S ith i I t P P bli h 459 474 Whitaker J. O. Jr. 1988. Collecting and preserving ectoparasites for ecological study. Ecological and Behavioral Methods for the Study of Bats. Washington: Smithsonian Inst. Press Publisher: 459474. Whitaker J. O. Jr. 1988. Collecting and preserving ectoparasites for ecological study. Ecological and Behavioral Methods for the Study of Bats. Washington: Smithsonian Inst. Press Publisher: 459474. 958
https://openalex.org/W2509773728	https://europepmc.org/articles/pmc5002318?pdf=render	English	null	PREMM: preterm early massage by the mother: protocol of a randomised controlled trial of massage therapy in very preterm infants	BMC pediatrics	2,016	cc-by	12,518	Abstract Background: Preterm infants follow an altered neurodevelopmental trajectory compared to their term born peers as a result of the influence of early birth, and the altered environment. Infant massage in the preterm infant has shown positive effects on weight gain and reduced length of hospital stay. There is however, limited current evidence of improved neurodevelopment or improved attachment, maternal mood or anxiety. The aim of this study is to investigate the effects of infant massage performed by the mother in very preterm (VPT) infants. Effects on the infant will be assessed at the electrophysiological, neuroradiological and clinical levels. Effects on maternal mood, anxiety and mother-infant attachment will also be measured. Methods/Design: A randomised controlled trial to investigate the effect of massage therapy in VPT infants. Sixty VPT infants, born at 28 to 32 weeks and 6 days gestational age, who are stable, off supplemental oxygen therapy and have normal cranial ultrasounds will be recruited and randomised to an intervention (infant massage) group or a control (standard care) group. Ten healthy term born infants will be recruited as a reference comparison group. The intervention group will receive standardised massage therapy administered by the mother from recruitment, until term equivalent age (TEA). The control group will receive care as usual (CAU). Infants and their mothers will be assessed at baseline, TEA, 12 months and 24 months corrected age (CA), with a battery of clinical, neuroimaging and electrophysiological measures, as well as structured questionnaires, psychoanalytic observations and neurodevelopmental assessments. Discussion: Optimising preterm infant neurodevelopment is a key aim of neonatal research, which could substantially improve long-term outcomes and reduce the socio-economic impact of VPT birth. This study has the potential to give insights into the mother-baby relationship and any positive effects of infant massage on neurodevelopment. An early intervention such as massage that is relatively easy to administer and could alter the trajectory of preterm infant brain development, holds potential to improve neurodevelopmental outcomes in this vulnerable population. (Continued on next page) * Correspondence: p.colditz@uq.edu.au †Equal contributors 1Perinatal Research Centre, School of Medicine, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia 2University of Queensland Centre for Clinical Research, Level 4, Bldg 71/918, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia Full list of author information is available at the end of the article © 2016 Lai et al. PREMM: preterm early massage by the mother: protocol of a randomised controlled trial of massage therapy in very preterm infants Melissa M. Lai1,2,3†, Giulia D’Acunto4†, Andrea Guzzetta4, Roslyn N. Boyd5, Stephen E. Rose6, Jurgen Fripp6, Simon Finnigan1,2, Naoni Ngenda3, Penny Love1,2, Koa Whittingham5, Kerstin Pannek6, Robert S. Ware7,8 and Paul B. Colditz1,2,3* * Correspondence: p.colditz@uq.edu.au †Equal contributors 1Perinatal Research Centre, School of Medicine, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia 2University of Queensland Centre for Clinical Research, Level 4, Bldg 71/918, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia Full list of author information is available at the end of the article Lai et al. BMC Pediatrics (2016) 16:146 DOI 10.1186/s12887-016-0678-7 Lai et al. BMC Pediatrics (2016) 16:146 DOI 10.1186/s12887-016-0678-7 Open Access Background total brain volume is increased nearly 3-fold, cortical gray matter volume is increased 4-fold and cerebellar volume is increased 4-fold [7]. Although some differences between term infants and preterm infants at term equivalent age are explained by the effect of complications associated with preterm birth, both the experience of a highly stress- ful ex-utero environment and the lack of the stimulation normally experienced in the womb, exert detrimental ef- fects on the immature brain [6, 9]. The challenges im- posed by the extra-uterine environment, disrupt the above described development of a “neurotypical” brain. Results include numerous primary medical problems experienced by preterm infants such as lung disease of prematurity, physiological instability, asphyxia, suboptimal nutrition, infection, medication side effects and hyperbilirubinaemia, each of which may have its own potentially deleterious impact on brain development [6, 10]. Environmental stressors, which include frequent noxious stimulation, ex- cessive sound and constant light, also adversely affect the normal neurodevelopmental trajectory [6, 10]. Repetitive pain universally experienced by premature infants from frequent invasive procedures is hypothesised to cause ex- cessive activation of central afferent pain pathways and excitotoxic damage to the developing brain [11]. While the impact of these factors and negative influence on the brain are recognised, emergent strategies to lessen harm include environmental enrichment and infant massage. Over the past 50 years, there has been a progressive decrease in preterm infant mortality [1]. A lowering of the age limit of viability and an increase in preterm birth numbers has made preterm birth a significant public health issue [1]. Despite advances in technology, the num- ber of preterm infants with neurodevelopmental com- promise remains high [2]. In recent years the focus of improving preterm outcomes has shifted from increasing survival to minimising morbidity and improving neurode- velopmental outcomes [1, 3]. Long term neurodevelop- mental abnormalities impact up to 50 % of these infants [4] and include motor disability (including cerebral palsy), reduced cognitive performance and behavioural problems [3]. The severity of the deficits is related to the degree of prematurity and the presence of neuroradiological injuries [3], however high rates of more subtle but nonetheless im- portant neurodevelopmental abnormalities also occur in low-risk preterms without obvious brain injury [5, 6]. Even in late preterm infants, the effects of preterm birth on brain development are more significant and long lasting than previously thought [5, 7]. Background The untimely interruption of the developing fetus’ environment is thought to be a major contributor to this phenomenon [6]. Keywords: Preterm, Infant massage, Neurodevelopment, Attachment Abbreviations: AFD, Apparent fibre density; Bayley-III, Bayley Scales of Infant and Toddler Development, Third Edition; BDI, Beck Depression Inventory; BSID-II, Bayley Scales of Infant Development, Second Edition; DASS, Depression Anxiety Stress Scale; dEEG, Dense array electroencephalography; DWI, Diffusion weighted imaging; EEG, Electroencephalography; EPDS, Edinburgh Postnatal Depression Scale; FA, Fractional anisotropy; FOD, Fibre orientation distribution; GMA, General Movements Assessment; HARDI, High angular resolution diffusion imaging; HNNE, Hammersmith Neonatal Neurological Examination; HREC, Human research ethics committee; ITSEA, Infant Toddler Social and Emotional Assessment; MD, Mean diffusivity; MDI, Mental Developmental Index; MIBS, Mother-to-Infant Bonding Scale; MIRI, Maternal Infant Responsiveness Instrument; MPAS, Maternal Postnatal Attachment Scale; MRI, Magnetic Resonance Imaging; MSES, Maternal Self-Efficacy Scale; NVS, Neonatal Vision Scale; PDSS, Postpartum Depression Screening Scale; PMA, Postmenstrual age; PPD, Postpartum depression; SPSS, Statistical Package for the Social Sciences; TEA, Term equivalent age; TSE, Turbo spin echo; VPT, Very preterm infant Disruption of optimal fetal environment The altered neurodevelopmental trajectory observed in preterm infants results from interruption of the intra- uterine environment. The neuromaturation of the cere- bral cortex, initially laid down in cortical layers prior to 20 weeks gestation, is a dynamic process from 30 to 40 weeks gestation. It is during this time, the subplate, a transient population of neurons, guides the development of cortical and thalamocortical connections [8], maturing from its maximal prominence to almost complete re- gression. These connections are fundamental for cortical processing and cognition [8]. Between 29 and 41 weeks, Abstract Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lai et al. BMC Pediatrics (2016) 16:146 Page 2 of 12 (Continued from previous page) Trial registration: Australian New Zealand Clinical Trials Registry: ACTRN12612000335897. Date registered: 22/3/2012. Keywords: Preterm Infant massage Neurodevelopment Attachment Trial registration: Australian New Zealand Clinical Trials Registry: ACTRN12612000335897. Date registered: 22/3/2012. Trial registration: Australian New Zealand Clinical Trials Registry: ACTRN12612000335897. Date reg Keywords: Preterm, Infant massage, Neurodevelopment, Attachment Infant massage Infant massage has been investigated as a potentially effective intervention aimed at providing a form of environmental enrichment [15]. It is often combined with other forms of stimulation such as kinaesthetic stimulation (e.g. passive extension/flexion movements of the arms and legs), talking or eye contact [16]. The pos- sibility that infant massage could provide a form of com- forting touch with positive effects on growth and neurodevelopment is described in a number of studies [16, 17]. Evidence supporting positive effects of massage in preterm infants include increased weight gain [18–20], improved growth and gastrointestinal function [21, 22], im- proved body fat deposition [23], improved neurobeha- vioural outcomes [17, 24–26], pain attenuation [27, 28], reduction of infant stress and stress-related factors [29, 30], reduction of late-onset sepsis [31], improved immune system [32], reduced jaundice [33] and improved heart rate variability [34] as well as a reduction in maternal depression and anxiety [35]. Studies investigating the use of specific oils in massage versus no oil suggest improved weight gain [36–38]. Overall the evidence remains weak, mainly due to small sample sizes, heterogeneity and poor methodology in some studies. The current level of evidence does not sup- port wider use of infant massage without further research [17, 39]. Two key aspects of infant massage as an early intervention have received little attention. The first aspect is the effect of infant massage on direct measures of brain development, such as the maturation of brain electrical ac- tivity, brain structure and the relationship to clinical neuro- behaviour. A recent meta-analysis reported that the association between massage and neurobehavioural devel- opment remained elusive [40]. The second aspect is the po- tential to enhance the efficacy of the intervention by active involvement of the parents; in particular the mother, and what effect of actively massaging her baby may have on the mother-infant relationship. Much like the rest of the brain, considerable develop- ment of the visual system occurs in the third trimester, which increases the potential for long-term visual dysfunc- tion in preterm infants. The Neonatal Vision Scale is an assessment that was originally developed to test vision in full term infants. It has subsequently been applied to preterm infant cohorts to reliably measure the integrity of the visual system [46, 47]. Using this scale could assist in evaluating any measureable effects of massage on the visual system. Environmental enrichment Environmental enrichment and social stimulation induce experience-dependent neuroplasticity in experimental ani- mal models [10]. Neuroplasticity is the capacity of the mammalian brain to use the activity induced by a given experience, enabling the modification of function in neur- onal circuitry [12]. Early interventions based on the ma- nipulation of the extra-uterine environment, among them infant massage therapy, have been used in preterm infants Page 3 of 12 Page 3 of 12 Lai et al. BMC Pediatrics (2016) 16:146 with the aim of optimising the infant’s sensory experience and improving overall functional outcome [13, 14]. as in the appropriate age of administration and scoring systems. This affects the ability to use them in clinical practice or research [42]. The Prechtl’s GMA has the great- est predictive accuracy for the diagnosis of cerebral palsy [41, 43] with a sensitivity of 98 % and specificity of 91 % [44]. The assessment is based on a global visual perception system of the quality and complexity of movements [45]. As such, it is an applicable tool to use evaluating the impact of massage. Infant massage Measuring long-term neurodevelopmental outcomes has been standardised for a number of assessments for use in toddlers and young children. These assessments are particularly useful for assessing high-risk populations such as VPT infants. The Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III) is a structured instrument to assess cognitive and social- emotional development, language and motor abilities [48] and is the most widely used measure to assess neu- rodevelopment of VPT and very low birth weight infants in the first three years [49]. In a recent review, Mental Development Index (MDI) scores were strongly predictive of later cognitive functioning, r = 0.61 (95 % CI = 0.57- 0.64) and motor scale scores were moderately predictive of later motor function, r = 0.34 (95 % CI = 0.26-0.42) [49]. In another recent study [25], 73 very low birthweight pre- term infants who had been randomised to receive massage therapy or not, were followed up at 2 years of age with the Bayley Scales of Infant Development Second edition (BSID-II). Outcomes showed the Mental Development Index (MDI) was higher in the intervention group than the control group indicating better cognitive scores in the group who received massage [25]. In the present study, we will use Bayley-III to review these findings. Clinical neurodevelopmental assessments Traditionally, methods of neurodevelopmental assessment of the preterm infant have included a number of clinical examinations used to evaluate neurological function and neurobehaviour [41]. Those most commonly used in the neonatal period include the Prechtl’s General Movements Assessment (GMA), Hammersmith Neonatal Neurological Examination (HNNE), Amiel-Tison Neurological Assess- ment at term, Neonatal Behavioural Assessment Scale (NBAS), Neurobehavioural Assessment of the Preterm Infant (NAPI) and the Neonatal intensive care unit Network Neurobehavioural Scale (NNNS) [42]. These assessments vary in the time required for training, as well Electroencephalography to assess function Electrophysiological studies have reported significant differences in spectral electroencephalography (EEG) measures between healthy term and preterm infants [64–66]. Maturation of the EEG in preterm infants is characterised by decreases in the total amplitude and delta activity power in quiet and active sleep [67]. A preliminary study examining whether preterm infant massage has a beneficial effect on cerebral function as measured by EEG, found a significant difference in the interburst interval dur- ation and also a difference in maturation of visual function in preterm infants who received massage in comparison to preterm infants who received standard care [68]. A subse- quent study showed a relative increase in global EEG spectral power in delta and beta frequencies in massaged infants when compared to controls, which was interpreted as suggesting that massage therapy in low-risk preterm infants favors a process of maturation of brain electrical activity similar to that observed in term born infants [24]. The Depression Anxiety Stress Scale (DASS) is another widely studied assessment tool in the postnatal period. It is a 42-item questionnaire, completed by the mother, de- signed to measure the magnitude of these three negative emotional states [79, 80]. The Depression scale focuses on reports of low mood, motivation and self-esteem, the Anxiety scale assesses physiological arousal, perceived panic and fear, and the Stress scale measures tension and irritability [79, 80]. Internal consistency for each of the subscales is typically high (e.g. Cronbach’s α of 0.96-0.97, 0.84-0.92, 0.90-0.95 for Depression, Anxiety and Stress re- spectively) [80] and convergent and discriminant validity has been demonstrated [81]. Attachment difficulties can arise from preterm birth disrupting normal physical contact between the mother and infant [82] and withdrawal of the mother from their infant due to distress [83]. This may inhibit the mother’s caregiving behaviour as her desire and ability to provide protection for her preterm infant is interrupted, ultimately impacting the mother’s attachment representations and the child’s attachment patterns [84]. Infant Observation is a method developed by Esther Bick at the Tavistock Clinic in London more than 60 years ago, and has been used to understand the characteristics of the developing relationship between mother and infant [85]. This method has been widely used internationally for training and research in the psychodynamic-psychoanalytic field to capture the rich and complex nature of mother-infant attachment. One of these complexities in particular is that of maternal responsiveness. Magnetic Resonance Imaging to assess structure Magnetic Resonance Imaging to assess structure Magnetic Resonance Imaging (MRI) has become an essen- tial neurodiagnostic tool as it offers high resolution images and can assist prognostication following neonatal brain injury [50]. MRI studies in preterm infants at term- equivalent age have shown that preterm birth alters development of regional brain volume [51], white matter [52–54], cortex [55, 56], deep gray matter [57, 58] and vascular organisation [59]. With diffusion tensor imaging, white matter integrity and white matter maturation can be Lai et al. BMC Pediatrics (2016) 16:146 Page 4 of 12 Page 4 of 12 studied, and white matter pathways can be non-invasively delineated through diffusion tractography [60–62]. Numer- ous studies have evaluated the ability of MRI at term equivalent age to predict neurodevelopmental outcomes at 1 to 9 years and established it as the best imaging tool available for outcome prediction of children born preterm [63]. To our knowledge, MRI studies of infants who have received massage therapy are yet to be described. proactive in the developmental care of their babies. A variety of maternal factors including frequency of mater- nal touch and the degree of postpartum depression (PPD) can influence the neurodevelopmental and cogni- tive skills of the preterm infant [76, 77]. Maternal mood and anxiety have been measured using a number of scales. The most widely researched is the Edinburgh Postnatal Depression Scale [78] which demonstrates moderate to good internal consistency over several studies. Other measures such as the Postpartum Depression Screening Scale (PDSS) and the Beck Depres- sion Inventory (BDI) are also widely used and demonstrate good concurrent validity [78]. Aim The broad aim is to investigate the potential effects of infant massage performed by the mother, in VPT infants born between 28 and 32 weeks and 6 days gestational age. Effects on the infant will be assessed at the electro- physiological, neuroradiological and clinical levels at term equivalent age and 24 months corrected age. In addition, the impact of infant massage implemented by the mother on maternal mood and anxiety and infant attachment will be measured at term equivalent age, 12 months corrected and 24 months corrected age. Inclusion criteria Infants born between 28 and 32 weeks and 6 days gestational age, with a birthweight between the 10th and 90th percentile (for their gestational age and gender), who are clinically stable, off oxygen therapy or respiratory support and have no IVH greater than grade II. The cut-off age for recruitment is 34 weeks and 3 days old, to ensure ample time for infants in the massage group to receive the intervention. Exclusion criteria The presence of abnormalities on brain ultrasound including intraventricular haemorrhage (grade III and IV), periventricular echogenicity or peri- ventricular cysts. Infants with major genetic disorders and malformations will also be excluded. Hypothesis The primary hypothesis to be tested is that infant massage by the mother in VPT infants promotes favourable pro- cesses in brain development, which can be functionally measured with dense array electroencephalography (dEEG). Secondary hypotheses to be tested include the following: p Preterm participants Preterm infants born between 28 weeks and 32 weeks and 6 days gestational age, admitted to the Grantley Stable Neonatal Unit at the Royal Brisbane & Women’s Hospital will be assessed for eligibility. Parents will be approached for consent if a head ultrasound around day 10 after birth confirms the absence of moderate or severe intraventricu- lar haemorrhage (> grade II) [88], periventricular echo- genicity or periventricular cysts. Once written informed consent is obtained, baseline measures will be collected before allocation occurs. Allocation will be performed by opening the next, in sequence, opaque envelope by non- study personnel. In this way infants will be randomised into a massage group or a care as usual group. In this project we will investigate several key aspects of the effect of infant massage on preterm neurodevelop- ment, its neurobiological correlates and the mother-infant relationship. Methods Design coefficient of 0.86 [82]. The Maternal Postnatal Attach- ment Scale (MPAS) is designed to evaluate maternal emotions and cognitions relating to attachment. It is specifically designed for use during the first year of life and focuses on the mother’s subjective experience in relation to her infant. Internal consistency as measured by Cronbach’s coefficient alpha is 0.78. In addition to maternal attachment, this research is also aimed to explore infant mental health. coefficient of 0.86 [82]. The Maternal Postnatal Attach- ment Scale (MPAS) is designed to evaluate maternal emotions and cognitions relating to attachment. It is specifically designed for use during the first year of life and focuses on the mother’s subjective experience in relation to her infant. Internal consistency as measured by Cronbach’s coefficient alpha is 0.78. In addition to maternal attachment, this research is also aimed to explore infant mental health. A randomised controlled trial will be conducted to investi- gate the effect of infant massage. Infants will be recruited and randomised to an intervention group (PREMM), which will receive infant massage by the mother or a control group, which will receive care as usual (CAU). A study coordinator, who is not involved in generating the randomisation sequence, will assess infants for eligibility and obtain written informed consent from the parent(s). Evaluating the infant’s social and emotional functioning can be an indirect measure of the integrity of infant attach- ment. It has been observed that secure infant attachment status is related to lower risk for later peer interaction and behavioral problems [86]. The Infant Toddler Social and Emotional Assessment (ITSEA) is a robust parent-report questionnaire designed to assess a wide array of social- emotional and behavioural problems and competencies [87]. Its psychometric properties are sound with strong test- retest reliability (0.61-0.91, mean = 0.79), with concurrent and discriminant validities (α = 0.69-0.86, mean = 0.76) [87]. Electroencephalography to assess function Parental stress, infant attachment and related assessments The stress and trauma of VPT birth on the parents is well described [69–72]. Preterm delivery has been identi- fied as a risk factor for stress, postpartum posttraumatic stress disorder, postpartum depression and difficulty with initial bonding and attachment [73]. The increased levels of stress, anxiety and depression may negatively influence the already difficult maternal-infant bonding [73]. Historically, a high dependence on technology for life-support, the institutionalisation of preterm infant care and the fear of infection have often resulted in the separation of mother and baby [74]. In the early 1970s, research demonstrated that mothers who were permitted to enter the nursery showed a greater commitment to their infants, were more confident in their mothering abilities and had increased caretaking skills [74]. In response to this and subsequent evidence, parental involvement was encouraged [74]. Maternal responsiveness plays an important role in developing secure attachment and effective bonding patterns between mother and infant. Ineffective maternal responsiveness is directly related to a lack of attachment, low self-esteem and unhealthy growth and development of the child [82]. The Maternal Infant Responsiveness Instrument (MIRI) was designed to measure this concept by appraising the maternal awareness and reflection of the mother’s responsiveness to her infant and her recognition of her infant’s responsiveness to her. It is easy to administer and has an alpha reliability An increase in bonding and attachment behaviours and a decrease in parental depression has been reported in studies where mothers attended massage classes 2 months after birth at term [75]. In the same way, actively providing infant massage could help to mitigate the impact of stress and anxiety in the intensive care nursery environment, by allowing mothers to be more Lai et al. BMC Pediatrics (2016) 16:146 Page 5 of 12 Randomisation (6 times), shoulder to the hand (6 times) left and right, neck to the back to the bottom (6 times), bottom to the feet (6 times) left and right. After returning to “resting hands” position for up to a minute, this sequence will be repeated. During the massage the mother is to breathe deeply and slowly, keeping herself relaxed whilst paying attention to the baby’s stress cues which include crying, skin mottling, hiccups, gagging, apnoea and bradycardia. If signs of stress are evident, the mother will to return to “resting hands” until the baby settles and is ready for fur- ther massage. If the baby doesn’t settle, the mother will stop and re-attempt the massage later. Randomisation will be performed in Statistics Package for the Social Sciences (SPSS) 15.0 (SPSS Inc., Chicago, IL, USA) using random number generation with stratification by gender. The ratio of intervention to care as usual partici- pants will be 1:1. A third party not involved with recruitment will generate the randomisation list and the list will be concealed. After random allocation lists have been generated, the allocation group (intervention or care as usual) of the premature infants will be stored in sequen- tially numbered, sealed, opaque envelopes. These envelopes will be opened by non-study personnel at each study enrol- ment, after baseline measures have been obtained. g For the kinaesthetic phase, the baby is placed in the supine position. Again, the mother will place her palms on the baby’s body to embrace the chest in a “resting hands” position. If the baby’s eyes are open, eye contact and verbal interaction will be encouraged. Kinaesthetic stimulation will follow this sequence: 6 flexions and extensions of the right upper limb, 6 flexions and extensions of the left upper limb, 6 flexions and extensions of the right and left upper limbs together, 6 flexions and extensions of the right lower limb, 6 flexions and extensions of the left lower limb, 6 flexions and extensions of the right and left lower limbs together. After returning to “resting hands” position for up to a minute, this concludes the session. The massage sequence will be taught to the mother by the same paediatric physiotherapist trained in the Field method of massage [89]. Sample size (1)Infant massage by the mother in VPT infants promotes favourable processes in brain development, which can be detected structurally at term equivalent age, and functionally at term equivalent age and 24 months corrected age. Preliminary data from a pilot study, which described global spectral EEG power-associated brain maturation in massaged preterm infants, was used to calculate a sample size [24]. To find a 10 % between-group differ- ence in EEG power in the treatment group [24], with type 1 error rate 0.05, and power of 80 %, we require 20 participants in each group to complete the study. Assuming that two thirds of participants will remain in the study at term equivalent age, we will require a total of 60 preterm infants to be recruited (i.e. 30 per group). (2)Infant massage by the mother in VPT infants reduces maternal depression, anxiety and stress associated with preterm birth. (3)Infant massage by the mother in VPT infants enhances infant attachment. Lai et al. BMC Pediatrics (2016) 16:146 Page 6 of 12 Page 6 of 12 Intervention h The intervention will be introduced to the mother following allocation into the massage group. The massage paradigm is modified from the preterm massage protocol of Field et al. [89] that consists of 2 parts, a tactile phase and a kinaesthetic phase. Mothers will be taught the techniques and encouraged to massage their babies for 15 min, twice a day until term equivalent age. Those allocated to the control group will receive care as usual. Mothers will be given the following instructions both in demonstration and in written form for later reference. The tactile phase will begin with asking “baby’s permis- sion”, in the prone position and enclosing the infant with cupped hands in a “resting hands” position (Fig. 1). A small amount of massage oil (cold-pressed fruit or vegetable oil) will be used. Maintaining a “resting hand” to the back, the mother will slowly stroke the baby with the palm of her other hand with a gentle but firm, rhythmic motion, following this sequence: head to the neck The mothers will be provided with a structured diary to keep a record of the number and duration of each session. Massage sessions will be preferably performed around 60 min before feeding and at least 2 h after the completion of the previous session. Up to three individual teaching sessions will be given to each mother and support will be continued until the mother is considered to be competent and confident. Regular contact with the mother will en- sure that she continues to use the outlined protocol. When the baby is discharged home, the mother will be asked to continue daily massages until TEA. Fig. 1 "Resting hands" prone position in massage protocol Infants of the CAU group will undergo routine nursery care and there will be no specific interventions. Inter- vention and CAU mothers will be assigned to different nursery rooms to prevent contamination. In addition, nursery staff will be briefed on the importance of main- taining the study protocol. Very close monitoring of the data collection process will ensure protocol violations will be identified and action will be undertaken to prevent this, in a sensitive manner. Term participants A term born reference group will be recruited from the postnatal wards or via interested parents for comparison to a normal infant population. Fig. 1 "Resting hands" prone position in massage protocol Page 7 of 12 Lai et al. BMC Pediatrics (2016) 16:146 The term born reference group will be born between 37 and 41 weeks gestation following an uncomplicated pregnancy and delivery, have a birthweight > 10th centile and have not required special care admission. 7. Infant Observations: These are conducted on mother- infant pairs, each of 1 h duration by one Psychoanalytic Psychotherapist trained in the Esther Bick Tavistock model of observation [85]. Each observation will be assessed using the following criteria: amount of touching by the mother, amount of visual checking if the mother is away from the infant or if the infant is asleep, amount of auditory checking, amount of talking about the infant by the mother, the infant's desire for contact with the mother, visual/physical reaching, anxiety apparent in the relationship, reports of illness or difficulties, willingness to let the observer see the infant. This will be the first of four such observations to evaluate the evolving mother-infant relationship. Preterm infants timing of assessments Preterm infants timing of assessments All preterm infants will be asked to return to hospital for assessment at term equivalent age (39 to 42 weeks post menstrual age (PMA)), then further assessments at 12 and 24 months corrected age either at home or at the hospital (Table 1). 3. Hammersmith Neonatal Neurological Examination (HNNE): The HNNE is a discriminative and predictive scale for the neurological examination of the preterm and term newborn [91–93]. It is derived from a combination and adaptation of items taken from Prechtl, Saint Anne Dargassies and Brazelton methods in a simplified and easy to administer protocol [92]. When compared to term born infants, preterm infants at term equivalent age have lower scores when assessed with HNNE [92]. The examination will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist who will score the infant at the same time. 3. Hammersmith Neonatal Neurological Examination (HNNE): The HNNE is a discriminative and predictive scale for the neurological examination of the preterm and term newborn [91–93]. It is derived from a combination and adaptation of items taken from Prechtl, Saint Anne Dargassies and Brazelton methods in a simplified and easy to administer protocol [92]. When compared to term born infants, preterm infants at term equivalent age have lower scores when assessed with HNNE [92]. The examination will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist who will score the infant at the same time. Outcome measures Primary outcome h The primary outcome measure is beta-frequency global power (EEG parameter) in preterm infants, at term equivalent age, who have received massage therapy or care as usual. High-density electroencephalography (dEEG) will be recorded using a 64-electrode sensor net (HydroCel Geodesic Sensor Net, Electrical Geodesics Inc.). Each electrode is a saline sponge, in a geodesic tension structure comprised of silastic threads. The sensor net is prepared by soaking in a normal saline solution and then applied 4. Depression Anxiety Stress Scales (DASS): This self- report measure consists of three 14-item scales assessing depression, anxiety and stress over the past week. Each item has 4 response options ranging from 0 (“Did not apply to me at all”) to 3 (“Applied to me most of the time”) [94]. If questionnaire scores indicate significant emotional stress, psychosocial support will be offered. Mothers will be asked to complete this assessment. Table 1 Outcome measures for preterm infants Measures for Preterm infants Baseline Term Equivalent Age 12mth corrected 24mth corrected Clinical Demographics HNNE GMA HNNE GMA NVS Bayley-III Infant Observation IO IO IO IO Questionnaires DASS MIBS EPDS DASS MIBS EPDS MIRI MPAS DASS ITSEA MSES Radiological MRI Electrophysiological dEEG HNNE = Hammersmith Neonatal Neurological Examination; GMA = General Movements Assessment; NVS = Neonatal Vision Scale; Bayley-III = Bayley Scales of Infant and Toddler Development, 3rd Ed; IO = Infant Observation; DASS = Depression, Anxiety, Stress Scale; MIBS = Mother to Infant Bonding Scale; EPDS = Edinburgh Postnatal Depression Scale; MIRI = Maternal Infant Responsiveness Instrument; MPAS = Maternal Postnatal Attachment Scale; ITSEA = Infant Toddler Social and Emotional Assessment; MSES = Maternal Self Efficacy Scale; MRI = Magnetic Resonance Imaging; dEEG = dense array electroencephalography Table 1 Outcome measures for preterm infants Table 1 Outcome measures for preterm infants Measures for Preterm infants Baseline Term Equivalent Age 12mth corrected 24mth corrected Clinical Demographics HNNE GMA HNNE GMA NVS Bayley-III Infant Observation IO IO IO IO Questionnaires DASS MIBS EPDS DASS MIBS EPDS MIRI MPAS DASS ITSEA MSES Radiological MRI Electrophysiological dEEG 5. Edinburgh Postnatal Depression Scale (EPDS): The EPDS is a 10-item self-administered questionnaire developed for screening postpartum depression in women in outpatient or home visiting settings [95]. Psychosocial support will be offered and appropriate referrals made if the EPDS score is greater than 9. Mothers will be asked to complete this assessment. 5. Measures Baseline measures Baseline measures 1. Demographics: Medical risk factors for outcome using standardised Australian and New Zealand Neonatal Network (ANZNN) data definitions [90] (e.g. gestational age, birthweight) will be collected from the baby’s case notes as baseline demographics. 2. Prechtl’s Qualitative Assessment of General Movement [45]: Video of the baby’s general movements will be collected for general movements assessment performed at a later date by a certified assessor who will be blind to group allocation. Secondary outcomes i. Other EEG parameters at term equivalent age (TEA) Global power in delta, theta and alpha bands, absolute and relative power for each frequency band, interhemispheric coherence and partial directed coherence will be analysed. iii. Preterm infant clinical measures y ii. MRI measures at TEA A brain MRI will be performed using a 3.0-T (Siemens TIM Trio, Erlangen, Germany) and an MRI-compatible incubator with a dedicated head coil (LMT Lammers Medical Technology, Lubeck, Germany). The infants will be fed, fitted with earmuffs to minimise noise exposure (Natus Mini Muffs, Natus Medical Inc., San Carlos, CA) and monitored by pulse oximetry, then wrapped and placed in an MRI-compatible incubator to keep the infant still, warm and supported in the scanner. Scanning will occur for about 45 to 60 min without sedation. The MRI protocol will include T1 turbo spin echo (TSE), T1w MPRage, T2w HASTE and 3 echo T2 map, 30 direction diffusion weighted imaging (DWI), and 64 direction DWI sequences. A neuroradiologist will review clinical sequences and classify any white and grey matter [58, 97] abnormalities. Diffusion images will be acquired using single-shot echo planar multi-direction diffusion-weighted sequence, employing dual bipolar diffusion gradient and double spin echo. This will include the acquisition of a 30 direction DWI protocol (b = 1000s/mm2) and a 64 direction HARDI protocol (b = 2000s /mm2), in which the encoding gradients are uniformly distributed in space (acquisition time 5 min and 11 min respectively). A field map for diffusion data is acquired using two 8. Clinical assessments performed at baseline will also be performed at TEA; Prechtl’s Qualitative General Movements Assessment (GMA) [45] and Hammersmith Neonatal Neurological Examination (HNNE) [92]. All examinations will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist blinded to allocation groups. 8. Clinical assessments performed at baseline will also be performed at TEA; Prechtl’s Qualitative General Movements Assessment (GMA) [45] and Hammersmith Neonatal Neurological Examination (HNNE) [92]. All examinations will be performed by a Paediatric Neurologist, Paediatric Physiotherapist or Neonatologist blinded to allocation groups. ii. MRI measures at TEA A brain MRI will be performed using a 3.0-T (Siemens TIM Trio, Erlangen, Germany) and an MRI-compatible incubator with a dedicated head coil (LMT Lammers Medical Technology, Lubeck, Germany). Outcome measures Primary outcome h Edinburgh Postnatal Depression Scale (EPDS): The EPDS is a 10-item self-administered questionnaire developed for screening postpartum depression in women in outpatient or home visiting settings [95]. Psychosocial support will be offered and appropriate referrals made if the EPDS score is greater than 9. Mothers will be asked to complete this assessment. 6. Mother-to-Infant Bonding Scale (MIBS): The MIBS is designed for use from day one postpartum, offering the mother one-word descriptors of possible emotions towards her new child. It is quick and easy to use and has reasonable reliability (α score 0.71) [96]. Mothers will be asked to complete this assessment. 6. Mother-to-Infant Bonding Scale (MIBS): The MIBS is designed for use from day one postpartum, offering the mother one-word descriptors of possible emotions towards her new child. It is quick and easy to use and has reasonable reliability (α score 0.71) [96]. Mothers will be asked to complete this assessment. Page 8 of 12 Lai et al. BMC Pediatrics (2016) 16:146 distortions [60]. Fractional anisotropy (FA) and mean diffusivity (MD) will be estimated from corrected diffusion data using a diffusion tensor model. Constrained spherical deconvolution implemented in MRtrix will be employed to estimate fibre orientation distribution (FOD) [98]. Whole-brain voxel based analysis of FA and MD will be performed using tract-based spatial statistics optimised for neonates [99]. Whole-brain voxel-based analysis of fibre orientation distributions will be conducted using Apparent Fibre Density (AFD) [100]. Probabilistic tractography will be performed using MRtrix. White matter pathways will be delineated using the multi-regions-of-interest approach. A number of pathways, including corticospinal tract, corpus callosum, superior longitudinal fasciculus and thalamic radiations will be extracted. Summary measures of FA, MD, AFD and T2 within pathways will be calculated and related to EEG and clinical findings. directly onto the head without need for further scalp prep- aration. The infant will be fed, wrapped and placed in an open cot, in a darkened room fitted with a Faraday cage. The recording will be conducted during sleep. EEG signals are amplified by a GES 300 series amplifier (Electrical Geodesics Inc.), digitised (at a sampling rate of 256 Hz) and recorded to the hard drive of a Mac desktop computer via NetStation software (Electrical Geodesics Inc.). The EEG data files will be exported from NetStation software then pre-processed and analysed quantitatively using Curry Scan 7 Neuroimaging Suite signal processing software (Compu- medics Neuroscan™, Compumedics Limited, Australia). Secondary outcomes The infants will be fed, fitted with earmuffs to minimise noise exposure (Natus Mini Muffs, Natus Medical Inc., San Carlos, CA) and monitored by pulse oximetry, then wrapped and placed in an MRI-compatible incubator to keep the infant still, warm and supported in the scanner. Scanning will occur for about 45 to 60 min without sedation. (1)A visual assessment will be performed at TEA. The Neonatal Vision Scale (NVS) [47] is a clinical assessment consisting of a short test battery that explores various aspects of visual function, ranging from the ability to fix and follow a target, to more complex aspects of visual function, such as reaction to a colour target, discrimination of black and white stripes with increasing spatial frequency and attention at a distance [47]. (1)A visual assessment will be performed at TEA. The Neonatal Vision Scale (NVS) [47] is a clinical assessment consisting of a short test battery that explores various aspects of visual function, ranging from the ability to fix and follow a target, to more complex aspects of visual function, such as reaction to a colour target, discrimination of black and white stripes with increasing spatial frequency and attention at a distance [47]. The MRI protocol will include T1 turbo spin echo (TSE), T1w MPRage, T2w HASTE and 3 echo T2 map, 30 direction diffusion weighted imaging (DWI), and 64 direction DWI sequences. A neuroradiologist will review clinical sequences and classify any white and grey matter [58, 97] abnormalities. Diffusion images will be acquired using single-shot echo planar multi-direction diffusion-weighted sequence, employing dual bipolar diffusion gradient and double spin echo. This will include the acquisition of a 30 direction DWI protocol The MRI protocol will include T1 turbo spin echo (TSE), T1w MPRage, T2w HASTE and 3 echo T2 map, 30 direction diffusion weighted imaging (DWI), and 64 direction DWI sequences. A neuroradiologist will review clinical sequences and classify any white and grey matter [58, 97] abnormalities. Diffusion images will be acquired using single-shot echo planar multi-direction diffusion-weighted sequence, employing dual bipolar diffusion gradient and double spin echo. This will include the acquisition of a 30 direction DWI protocol (2)The Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III) will be performed [48] at 24 months corrected age. Children will be assessed in the 5 key developmental domains of cognition, language, social-emotional, motor and adaptive behavior [48]. Secondary outcomes (b = 1000s/mm2) and a 64 direction HARDI protocol (b = 2000s /mm2), in which the encoding gradients are uniformly distributed in space (acquisition time 5 min and 11 min respectively). A field map for diffusion data is acquired using two 2D gradient recalled echo images to assist in correction for residual distortions due to susceptibility in homogeneities (acquisition time 1 min). An extensive pre-processing and quality control procedure will be used to detect and correct image artefacts caused by head movement, cardiac pulsation, and image (3)The Infant Toddler Social and Emotional Assessment (ITSEA) will be measured at 24 months corrected age. The ITSEA is a 136-item parent report questionnaire to assess social and emotional problems and competencies in 4 domains of behavior: behavioural dysregulation; externalising behavior problems; internalising behavior problems and competencies [87]. iv. Maternal measures (1)Infant Observation [85] performed at baseline will be repeated at TEA, 12 months and 24 months Lai et al. BMC Pediatrics (2016) 16:146 Page 9 of 12 Page 9 of 12 Page 9 of 12 Data analysis S Summary statistics will be used to describe demographic and clinical data characteristics at baseline by allocated study treatment. Continuous data will be summarised using either mean and standard deviation, or median and inter- quartile range, depending on the distribution of the variable of interest. Categorical data will be presented as frequencies and percentages. Comparisons between the baseline values of the treatment groups will be conducted to assess the de- gree to which variables are comparable after randomisa- tion. Between group differences will be investigated using linear regression for continuous data and Fisher’s exact test for categorical data. Potential confounding variables for the primary outcome are considered a-priori to be: gestational age at birth, birthweight, parity and maternal education. If any of these variables differ significantly between-groups (at P < 0.005), the identified variable(s) will be controlled for in the main analyses. (4)At 24 months corrected age, the Maternal Self Efficacy Scale (MSES) will be measured. The MSES is a 20-item measure of the mothers’ perceived self-competence of their maternal practice used at 12 months. The measure has good internal consistency (Cronbach’s α = 0.76-0.89) and is a good predictor of maternal competence, with strong concurrent validity with observation [102]. Mothers will be asked to complete this assessment. (4)At 24 months corrected age, the Maternal Self Efficacy Scale (MSES) will be measured. The MSES is a 20-item measure of the mothers’ perceived self-competence of their maternal practice used at 12 months. The measure has good internal consistency (Cronbach’s α = 0.76-0.89) and is a good predictor of maternal competence, with strong concurrent validity with observation [102]. Mothers will be asked to complete this assessment. Potential confounders corrected age. This will be performed by the Psychoanalytic Psychotherapist who is blinded to randomisation. Limitations include the single blinded nature of the study and the risk of potential contamination between interven- tion and CAU groups. Parents whose infants are allocated to the control group may be more likely to withdraw from the study after randomisation, if they perceive there is limited benefit to them. (2)Questionnaires performed at baseline will also be performed at TEA; Depression Anxiety Stress Scales [79], Edinburgh Postnatal Depression Scale [95] and Mother-to-infant Bonding Assessment [96]. Mothers will be asked to complete these assessments. (3)At 12 months corrected age, maternal infant responsiveness and maternal postnatal attachment will be measured. The MIRI, a 22-item scale [82] will be performed. The MPAS which has 19 items, all scored on a 5-point scale, with 1 and 5 indicating low and high attachment respectively, will also be performed [101]. Mothers will be asked to complete this assessment. (3)At 12 months corrected age, maternal infant responsiveness and maternal postnatal attachment will be measured. The MIRI, a 22-item scale [82] will be performed. The MPAS which has 19 items, all scored on a 5-point scale, with 1 and 5 indicating low and high attachment respectively, will also be performed [101]. Mothers will be asked to complete this assessment. Healthy term born infants The primary study outcome is beta frequency global power measured in all VPT infants at term equivalent age. The mean difference between treatment groups will be calculated using linear regression with treatment group entered as the main effect. The corresponding 95 % Wald confidence interval and p-value will be reported. For secondary outcomes, we will present the effect estimate relating to a variable with a continuous outcome as a mean difference, which will be calculated using a linear regression model. If a continuous variable does not meet the assump- tions necessary for linear regression, we will compare groups using the Mann–Whitney U Test. We will present the effect estimate relating to a variable with a binary outcome as an odds ratio, which will be calculated using a logistic regression model. We will present the effect estimate relating to a variable with a count outcome as an incident rate ratio, which will be calculated using a Poisson regression model. For all models the corresponding 95 % Wald confidence interval and p-value will be reported. Once written informed consent has been obtained, an appointment will be arranged for the infants to under- take the same assessments as the preterm cohort, at term equivalent age (39 to 42 weeks postmenstrual age) and infant observations at 12 months of age (Table 2). Consent for publication 20. Diego MA, Field T, Hernandez-Reif M. Preterm infant weight gain is increased by massage therapy and exercise via different underlying mechanisms. Early Hum Dev. 2014;90(3):137–40. Consent to publish the images contained in this document was obtained from the parents or legal guardians. Consent to publish the images contained in this document was obtained from the parents or legal guardians. 21. Choi H, Kim SJ, Oh J, Lee MN, Kim S, Kang KA. The effects of massage therapy on physical growth and gastrointestinal function in premature infants: A pilot study. J Child Health Care. 2016;20(3):394–404. Funding Funding QCPRRC funding project grant Merchant Charitable Foundation Postgraduate Research Scholarship (ML) Funding QCPRRC funding project grant Merchant Charitable Foundation Postgraduate Research Scholarship (ML) g QCPRRC funding project grant Merchant Charitable Foundation Postgraduate Research Scholarship (ML) 14. Lubbe W, Van der Walt CS, Klopper HC. Integrative literature review defining evidence-based neurodevelopmental supportive care of the preterm infant. J Perinatal Neonatal Nurs. 2012;26(3):251–9. Discussion h This protocol paper outlines a randomised controlled trial of infant massage in VPT infants to detect effects on neurodevelopment. To our knowledge this study is the first to directly measure what influences infant mas- sage may have on preterm brain structure and function using EEG, MRI and neurobehavioural assessments, and the mother-infant relationship in a preterm cohort. 7. Baron IS, Litman FR, Ahronovich MD, Baker R. Late preterm birth: a review of medical and neuropsychological childhood outcomes. Neuropsychol Rev. 2012;22(4):438–50. 8. Gatti MG, Becucci E, Fargnoli F, Fagioli M, Aden U, Buonocore G. Functional maturation of neocortex: a base of viability. J Matern Fetal Neonatal Med. 2012;25 Suppl 1:101–3. Availability of data and material 15. Cioni G, D'Acunto G, Guzzetta A. Perinatal brain damage in children: neuroplasticity, early intervention, and molecular mechanisms of recovery. Prog Brain Res. 2011;189:139–54. Data and materials will be made available once the study results have been published but restrictions will apply to prevent compromise of individual privacy. 16. Vickers A, Ohlsson A, Lacy JB, Horsley A. Massage for promoting growth and development of preterm and/or low birth-weight infants. Cochrane Database Syst Rev. 2004 (Issue 2):CD000390. 16. Vickers A, Ohlsson A, Lacy JB, Horsley A. Massage for promoting growth and development of preterm and/or low birth-weight infants. Cochrane Database Syst Rev. 2004 (Issue 2):CD000390. Acknowledgements 10. Perlman JM. Cognitive and behavioral deficits in premature graduates of intensive care. Clin Perinatol. 2002;29(4):779–97. We would like to acknowledge allied health clinicians Sonia Sam, Joanne George, Bernadette Shannon & PhD student Annice Kong for their enthusiastic assistance with assessments and acquisition of the clinical data. We would also like to acknowledge Dr Joel Dulhunty who assisted in generation of the randomisation list. 11. Anand KJ, Scalzo FM. Can adverse neonatal experiences alter brain development and subsequent behavior? Biol Neonate. 2000;77(2):69–82. 12. Kalia M. Brain development: anatomy, connectivity, adaptive plasticity, and toxicity. Metabolism. 2008;57 Suppl 2:S2–5. 13. Symington A, Pinelli J. Developmental care for promoting development and preventing morbidity in preterm infants. Cochrane Database Syst Rev. 2006 (Issue 2):CD001814. Ethics approval and consent to participate The study was approved by the Human Research Ethics Committees (HREC) at the Royal Brisbane & Women’s Hospital (HREC/09/QRBW/296), The Children’s Health Services Queensland (HREC/12/QRCH/40) and The University of Queensland (2014001160). The study was approved by the Human Research Ethics Committees (HREC) at the Royal Brisbane & Women’s Hospital (HREC/09/QRBW/296), The Children’s Health Services Queensland (HREC/12/QRCH/40) and The University of Queensland (2014001160). 22. Tekgunduz KS, Gurol A, Apay SE, Caner I. Effect of abdomen massage for prevention of feeding intolerance in preterm infants. Ital J Pediatr. 2014;40:89. 23. Moyer-Mileur LJ, Haley S, Slater H, Beachy J, Smith SL. Massage improves growth quality by decreasing body fat deposition in male preterm infants. J Pediatr. 2013;162(3):490–5. Blinding Field T, Diego M, Hernandez-Reif M, Dieter JN, Kumar AM, Schanberg S, et al. Insulin and insulin-like growth factor-1 increased in preterm neonates following massage therapy. J Dev Behav Pediatr. 2008;29(6):463–6. 20. Diego MA, Field T, Hernandez-Reif M. Preterm infant weight gain is increased by massage therapy and exercise via different underlying mechanisms. Early Hum Dev. 2014;90(3):137–40. 21. Choi H, Kim SJ, Oh J, Lee MN, Kim S, Kang KA. The effects of massage therapy on physical growth and gastrointestinal function in premature infants: A pilot study. J Child Health Care. 2016;20(3):394–404. 22. Tekgunduz KS, Gurol A, Apay SE, Caner I. Effect of abdomen massage for prevention of feeding intolerance in preterm infants. Ital J Pediatr. 2014;40:89. 23. Moyer-Mileur LJ, Haley S, Slater H, Beachy J, Smith SL. Massage improves growth quality by decreasing body fat deposition in male preterm infants. J Pediatr. 2013;162(3):490–5. 24. Guzzetta A, D'Acunto MG, Carotenuto M, Berardi N, Bancale A, Biagioni E, et al. The effects of preterm infant massage on brain electrical activity. Dev Med Child Neurol. 2011;53 Suppl 4:46–51. 25. Procianoy RS, Mendes EW, Silveira RC. Massage therapy improves neurodevelopment outcome at two years corrected age for very low birth weight infants. Early Hum Dev. 2010;86(1):7–11. 26. Ho YB, Lee RS, Chow CB, Pang MY. Impact of massage therapy on motor outcomes in very low-birthweight infants: randomized controlled pilot study. Pediatr Int. 2010;52(3):378–85. 27. Diego MA, Field T, Hernandez-Reif M. Procedural pain heart rate responses in massaged preterm infants. Infant Behav Dev. 2009;32(2):226–9. 28. Jain S, Kumar P, McMillan DD. Prior leg massage decreases pain responses to heel stick in preterm babies. J Paediatr Child Health. 2006;42(9):505–8. possible massage/received at least 50 % of total possible massage). Any such analyses will be clearly labeled as such and cautiously interpreted as perhaps indicating the maximum theoretical potential of the intervention. Statistical significance will be defined as alpha = 0.05. All tests conducted will be two-tailed. Competing interests Competing interests The authors declare that they have no competing interests. 19. Field T, Diego M, Hernandez-Reif M, Dieter JN, Kumar AM, Schanberg S, et al. Insulin and insulin-like growth factor-1 increased in preterm neonates following massage therapy. J Dev Behav Pediatr. 2008;29(6):463–6. Blinding Clinical assessors will be blinded to the intervention allocation. Likewise MRI and EEG data analyses will be performed in a blinded fashion. Table 2 Outcome measures for term born infants Measures for Term born infants 42 weeks post menstrual age 12 months corrected Clinical HNNE GMA NVS Infant Observation IO IO Questionnaires DASS MIBS EPDS Radiological MRI Electrophysiological dEEG HNNE = Hammersmith Neonatal Neurological Examination; GMA = General Movements Assessment; NVS = Neonatal Vision Scale; IO = Infant Observation; DASS = Depression, Anxiety, Stress Scale; MIBS = Mother to Infant Bonding Scale; EPDS = Edinburgh Postnatal Depression Scale; MRI = Magnetic Resonance Imaging; dEEG = dense array electroencephalography Table 2 Outcome measures for term born infants All analyses will primarily be undertaken using the ‘intention-to-treat’ approach, where all evaluable data is analysed in the treatment group according to which the participant was allocated, regardless of treatment received. Additionally, ‘per protocol’ analyses will be conducted where there have been large deviations from the planned intervention; all individuals with evaluable outcome data will be analysed according to the treat- ment they received (no massage/received < 50 % of total Lai et al. BMC Pediatrics (2016) 16:146 Page 10 of 12 Page 10 of 12 Lai et al. BMC Pediatrics (2016) 16:146 Page 10 of 12 References 1. Allen MC, Cristofalo EA, Kim C. Outcomes of preterm infants: m replaces mortality. Clin Perinatol. 2011;38(3):441–54. 2. Bonifacio SL, Glass HC, Peloquin S, Ferriero DM. A new neurolog neonatal intensive care. Nat Rev Neurol. 2011;7(9):485–94. 3. Saigal S, Doyle LW. An overview of mortality and sequelae of p from infancy to adulthood. Lancet. 2008;371(9608):261–9. 4. Latal B. Prediction of neurodevelopmental outcome after preter Pediatr Neurol. 2009;40(6):413–9. 5. Vohr B. Long-term outcomes of moderately preterm, late preter term infants. Clin Perinatol. 2013;40(4):739–51. 6. Bennet L, Van Den Heuij L, Dean JM, Drury P, Wassink G, Gunn plasticity and the Kennard principle: Does it work for the preter Clin Exp Pharmacol Physiol. 2013;40(11):774–84. 7. Baron IS, Litman FR, Ahronovich MD, Baker R. Late preterm birth medical and neuropsychological childhood outcomes. Neurops 2012;22(4):438–50. 8. Gatti MG, Becucci E, Fargnoli F, Fagioli M, Aden U, Buonocore G maturation of neocortex: a base of viability. J Matern Fetal Neon 2012;25 Suppl 1:101–3. 9. Field T. Massage therapy. Med Clin N Am. 2002;86(1):163–71. 10. Perlman JM. Cognitive and behavioral deficits in premature gra intensive care. Clin Perinatol. 2002;29(4):779–97. 11. Blinding Anand KJ, Scalzo FM. Can adverse neonatal experiences alter br development and subsequent behavior? Biol Neonate. 2000;77( 12. Kalia M. Brain development: anatomy, connectivity, adaptive pla toxicity. Metabolism. 2008;57 Suppl 2:S2–5. 13. Symington A, Pinelli J. Developmental care for promoting deve preventing morbidity in preterm infants. Cochrane Database Sy 2006 (Issue 2):CD001814. 14. Lubbe W, Van der Walt CS, Klopper HC. Integrative literature rev evidence-based neurodevelopmental supportive care of the pre J Perinatal Neonatal Nurs. 2012;26(3):251–9. 15. Cioni G, D'Acunto G, Guzzetta A. Perinatal brain damage in child neuroplasticity, early intervention, and molecular mechanisms o Prog Brain Res. 2011;189:139–54. 16. Vickers A, Ohlsson A, Lacy JB, Horsley A. Massage for promoting development of preterm and/or low birth-weight infants. Coch Database Syst Rev. 2004 (Issue 2):CD000390. 17. Smith JR. Comforting touch in the very preterm hospitalized inf integrative review. Adv Neonatal Care. 2012;12(6):349–65. 18. Massaro AN, Hammad TA, Jazzo B, Aly H. Massage with kinesthetic improves weight gain in preterm infants. J Perinatol. 2009;29(5):35 19. Field T, Diego M, Hernandez-Reif M, Dieter JN, Kumar AM, Scha et al. Insulin and insulin-like growth factor-1 increased in preter following massage therapy. J Dev Behav Pediatr. 2008;29(6):463– 20. Diego MA, Field T, Hernandez-Reif M. Preterm infant weight gai increased by massage therapy and exercise via different underly mechanisms. Early Hum Dev. 2014;90(3):137–40. 21. Choi H, Kim SJ, Oh J, Lee MN, Kim S, Kang KA. The effects of ma therapy on physical growth and gastrointestinal function in pre infants: A pilot study. J Child Health Care. 2016;20(3):394–404. 22. Tekgunduz KS, Gurol A, Apay SE, Caner I. Effect of abdomen massa prevention of feeding intolerance in preterm infants. Ital J Pediatr. 23. Moyer-Mileur LJ, Haley S, Slater H, Beachy J, Smith SL. Massage growth quality by decreasing body fat deposition in male prete J Pediatr. 2013;162(3):490–5. 24. Guzzetta A, D'Acunto MG, Carotenuto M, Berardi N, Bancale A, B et al. The effects of preterm infant massage on brain electrical a Med Child Neurol. 2011;53 Suppl 4:46–51. 25. Procianoy RS, Mendes EW, Silveira RC. Massage therapy improve neurodevelopment outcome at two years corrected age for ver weight infants. Early Hum Dev. 2010;86(1):7–11. 26. Ho YB, Lee RS, Chow CB, Pang MY. Impact of massage therapy outcomes in very low-birthweight infants: randomized controlle study. Pediatr Int. 2010;52(3):378–85. 27. Diego MA, Field T, Hernandez-Reif M. Procedural pain heart rate in massaged preterm infants. Infant Behav Dev. 2009;32(2):226– 28. Received: 26 June 2015 Accepted: 16 August 2016 Blinding Jain S, Kumar P, McMillan DD. Prior leg massage decreases responses to heel stick in preterm babies. J Paediatr Child H 2006;42(9):505–8. References 1. Allen MC, Cristofalo EA, Kim C. Outcomes of preterm infants: morbidity replaces mortality. Clin Perinatol. 2011;38(3):441–54. 2. Bonifacio SL, Glass HC, Peloquin S, Ferriero DM. A new neurological focus in neonatal intensive care. Nat Rev Neurol. 2011;7(9):485–94. 3. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet. 2008;371(9608):261–9. 4. Latal B. Prediction of neurodevelopmental outcome after preterm birth. Pediatr Neurol. 2009;40(6):413–9. 5. Vohr B. Long-term outcomes of moderately preterm, late preterm, and early term infants. Clin Perinatol. 2013;40(4):739–51. 6. Bennet L, Van Den Heuij L, Dean JM, Drury P, Wassink G, Gunn AJ. Neural plasticity and the Kennard principle: Does it work for the preterm brain? Clin Exp Pharmacol Physiol. 2013;40(11):774–84. 7. Baron IS, Litman FR, Ahronovich MD, Baker R. Late preterm birth: a review of medical and neuropsychological childhood outcomes. Neuropsychol Rev. 2012;22(4):438–50. 8. Gatti MG, Becucci E, Fargnoli F, Fagioli M, Aden U, Buonocore G. Functional maturation of neocortex: a base of viability. J Matern Fetal Neonatal Med. 2012;25 Suppl 1:101–3. 9. Field T. Massage therapy. Med Clin N Am. 2002;86(1):163–71. 10. Perlman JM. Cognitive and behavioral deficits in premature graduates of intensive care. Clin Perinatol. 2002;29(4):779–97. 11. Anand KJ, Scalzo FM. Can adverse neonatal experiences alter brain development and subsequent behavior? Biol Neonate. 2000;77(2):69–82. 12. Kalia M. Brain development: anatomy, connectivity, adaptive plasticity, and toxicity. Metabolism. 2008;57 Suppl 2:S2–5. 13. Symington A, Pinelli J. Developmental care for promoting development and preventing morbidity in preterm infants. Cochrane Database Syst Rev. 2006 (Issue 2):CD001814. 14. Lubbe W, Van der Walt CS, Klopper HC. Integrative literature review defining evidence-based neurodevelopmental supportive care of the preterm infant. J Perinatal Neonatal Nurs. 2012;26(3):251–9. 15. Cioni G, D'Acunto G, Guzzetta A. Perinatal brain damage in children: neuroplasticity, early intervention, and molecular mechanisms of recovery. Prog Brain Res. 2011;189:139–54. 16. Vickers A, Ohlsson A, Lacy JB, Horsley A. Massage for promoting growth and development of preterm and/or low birth-weight infants. Cochrane Database Syst Rev. 2004 (Issue 2):CD000390. 17. Smith JR. Comforting touch in the very preterm hospitalized infant: an integrative review. Adv Neonatal Care. 2012;12(6):349–65. 18. Massaro AN, Hammad TA, Jazzo B, Aly H. Massage with kinesthetic stimulation improves weight gain in preterm infants. J Perinatol. 2009;29(5):352–7. 19. Authors’ contributions l d G l D Melissa M Lai and Giulia D'Acunto are equal first authors of this paper. MML, JF, RNB, PBC, KW and PL prepared the protocol. GD, AG, RNB, PBC, NN, SER, KP, SF, KW, PL and RSW contributed to protocol design and development. y 17. Smith JR. Comforting touch in the very preterm hospitalized infant: an integrative review. Adv Neonatal Care. 2012;12(6):349–65. 17. Smith JR. Comforting touch in the very preterm hospitalized infant: an integrative review. Adv Neonatal Care. 2012;12(6):349–65. 18. Massaro AN, Hammad TA, Jazzo B, Aly H. Massage with kinesthetic stimulation improves weight gain in preterm infants. J Perinatol. 2009;29(5):352–7. Competing interests Author details 1 1Perinatal Research Centre, School of Medicine, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia. 2University of Queensland Centre for Clinical Research, Level 4, Bldg 71/918, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia. 3Grantley Stable Neonatal Unit, Royal Brisbane & Women’s Hospital, Brisbane, Qld, Australia. 4Stella Maris Institute, The University of Pisa, Pisa, Italy. 5Queensland Cerebral Palsy and Rehabilitation Research Centre, School of Medicine, Faculty of Medicine and Biomedical Sciences, The University of Queensland, Brisbane, Qld, Australia. 6CSIRO, Australian e-Health Research Centre, Royal Brisbane and Women’s Hospital, Brisbane, Qld, Australia. 7UQ Child Health Research Centre, School of Medicine, The University of Queensland, Brisbane, Qld, Australia. 8School of Population Health, The University of Queensland, Brisbane, Qld, Australia. 24. Guzzetta A, D'Acunto MG, Carotenuto M, Berardi N, Bancale A, Biagioni E, et al. The effects of preterm infant massage on brain electrical activity. Dev Med Child Neurol. 2011;53 Suppl 4:46–51. 25. Procianoy RS, Mendes EW, Silveira RC. Massage therapy improves neurodevelopment outcome at two years corrected age for very low birth weight infants. Early Hum Dev. 2010;86(1):7–11. 26. Ho YB, Lee RS, Chow CB, Pang MY. Impact of massage therapy on motor outcomes in very low-birthweight infants: randomized controlled pilot study. Pediatr Int. 2010;52(3):378–85. 27. Diego MA, Field T, Hernandez-Reif M. Procedural pain heart rate responses in massaged preterm infants. Infant Behav Dev. 2009;32(2):226–9. 28. Jain S, Kumar P, McMillan DD. Prior leg massage decreases pain responses to heel stick in preterm babies. J Paediatr Child Health. 2006;42(9):505–8. Page 11 of 12 Page 11 of 12 Page 11 of 12 Lai et al. BMC Pediatrics (2016) 16:146 29. Hernandez-Reif M, Diego M, Field T. Preterm infants show reduced stress behaviors and activity after 5 days of massage therapy. Infant Behav Dev. 2007;30(4):557–61. 54. Huppi PS, Warfield S, Kikinis R, Barnes PD, Zientara GP, Jolesz FA, et al. Quantitative magnetic resonance imaging of brain development in premature and mature newborns. Ann Neurol. 1998;43(2):224–35. 30. Acolet D, Modi N, Giannakoulopoulos X, Bond C, Weg W, Clow A, et al. Changes in plasma cortisol and catecholamine concentrations in response to massage in preterm infants. Arch Dis Child. 1993;68(1 Spec No):29–31. 55. Ajayi-Obe M, Saeed N, Cowan FM, Rutherford MA, Edwards AD. Reduced development of cerebral cortex in extremely preterm infants. Lancet. 2000;356(9236):1162–3. 31. Mendes EW, Procianoy RS. Massage therapy reduces hospital stay and occurrence of late-onset sepsis in very preterm neonates. J Perinatol. 2008;28(12):815–20. 56. Author details 1 Inder TE, Huppi PS, Warfield S, Kikinis R, Zientara GP, Barnes PD, et al. Periventricular white matter injury in the premature infant is followed by reduced cerebral cortical gray matter volume at term. Ann Neurol. 1999;46(5):755–60. 32. Ang JY, Lua JL, Mathur A, Thomas R, Asmar BI, Savasan S, et al. A randomized placebo-controlled trial of massage therapy on the immune system of preterm infants. Pediatrics. 2012;130(6):e1549–58. 57. Boardman JP, Counsell SJ, Rueckert D, Kapellou O, Bhatia KK, Aljabar P, et al. Abnormal deep grey matter development following preterm birth detected using deformation-based morphometry. NeuroImage. 2006;32(1):70–8. 33. Basiri-Moghadam M, Basiri-Moghadam K, Kianmehr M, Jani S. The effect of massage on neonatal jaundice in stable preterm newborn infants: a randomized controlled trial. J Pak Med Assoc. 2015;65(6):602–6. 58. Inder TE, Warfield SK, Wang H, Huppi PS, Volpe JJ. Abnormal cerebral structure is present at term in premature infants. Pediatrics. 2005;115(2):286–94. 34. Smith SL, Lux R, Haley S, Slater H, Beachy J, Moyer-Mileur LJ. The effect of massage on heart rate variability in preterm infants. J Perinatol. 2013;33(1):59–64. 59. Malamateniou C, Adams ME, Srinivasan L, Allsop JM, Counsell SJ, Cowan FM, et al. The anatomic variations of the circle of Willis in preterm-at-term and term-born infants: an MR angiography study at 3T. AJNR Am J Neuroradiol. 2009;30(10):1955–62. 35. Feijo L, Hernandez-Reif M, Field T, Burns W, Valley-Gray S, Simco E. Mothers' depressed mood and anxiety levels are reduced after massaging their preterm infants. Infant Behav Dev. 2006;29(3):476–80. 60. Pannek K, Guzzetta A, Colditz PB, Rose SE. Diffusion MRI of the neonate brain: acquisition, processing and analysis techniques. Pediatr Radiol. 2012;42(10):1169–82. 36. Fallah R, Akhavan Karbasi S, Golestan M, Fromandi M. Sunflower oil versus no oil moderate pressure massage leads to greater increases in weight in preterm neonates who are low birth weight. Early Hum Dev. 2013;89(9):769–72. 61. Feldman HM, Yeatman JD, Lee ES, Barde LH, Gaman-Bean S. Diffusion tensor imaging: a review for pediatric researchers and clinicians. J Dev Behav Pediatr. 2010;31(4):346–56. 37. Saeadi R, Ghorbani Z, Shapouri MA. The effect of massage with medium-chain triglyceride oil on weight gain in premature neonates. Acta Medica Iran. 2015;53(2):134–8. 62. Pannek K, Scheck SM, Colditz PB, Boyd RN, Rose SE. Magnetic resonance diffusion tractography of the preterm infant brain: a systematic review. Dev Med Child Neurol. 2014;56(2):113–24. 38. Kumar J, Upadhyay A, Dwivedi AK, Gothwal S, Jaiswal V, Aggarwal S. Author details 1 2012;88 Suppl 1:S31–3. 70. Vigod SN, Villegas L, Dennis CL, Ross LE. Prevalence and risk factors for postpartum depression among women with preterm and low-birth-weight infants: a systematic review. BJOG. 2010;117(5):540–50. 47. Ricci D, Romeo DM, Serrao F, Cesarini L, Gallini F, Cota F, et al. Application of a neonatal assessment of visual function in a population of low risk full-term newborn. Early Hum Dev. 2008;84(4):277–80. 71. Korja R, Latva R, Lehtonen L. The effects of preterm birth on mother-infant interaction and attachment during the infant's first two years. Acta Obstet Gynecol Scand. 2012;91(2):164–73. 71. Korja R, Latva R, Lehtonen L. The effects of preterm birth on mother-infant interaction and attachment during the infant's first two years. Acta Obstet Gynecol Scand. 2012;91(2):164–73. 48. Bayley N. Bayley Scales of Infant and Toddler Development. 3rd ed. San Antonio: Harcourt Assessment; 2006. 72. Kraljevic M, Warnock FF. Early educational and behavioral RCT interventions to reduce maternal symptoms of psychological trauma following preterm birth: a systematic review. J Perinat Neonatal Nurs. 2013;27(4):311–27. 49. Luttikhuizen dos Santos ES, de Kieviet JF, Konigs M, van Elburg RM, Oosterlaan J. Predictive value of the Bayley scales of infant development on development of very preterm/very low birth weight children: a meta-analysis. Early Hum Dev. 2013;89(7):487–96. 73. Chertok IR, McCrone S, Parker D, Leslie N. Review of interventions to reduce stress among mothers of infants in the NICU. Adv Neonatal Care. 2014;14(1):30–7. 50. El-Dib M, Massaro AN, Bulas D, Aly H. Neuroimaging and neurodevelopmental outcome of premature infants. Am J Perinatol. 2010;27(10):803–18. 74. Davis L, Mohay H, Edwards H. Mothers' involvement in caring for their premature infants: an historical overview. J Adv Nurs. 2003;42(6):578–86. 51. Peterson BS, Vohr B, Staib LH, Cannistraci CJ, Dolberg A, Schneider KC, et al. Regional brain volume abnormalities and long-term cognitive outcome in preterm infants. J Am Med Assoc. 2000;284(15):1939–47. 75. Onozawa K, Glover V, Adams D, Modi N, Kumar RC. Infant massage improves mother-infant interaction for mothers with postnatal depression. J Affect Disord. 2001;63(1–3):201–7. 76. Feldman R, Eidelman AI. Neonatal state organization, neuromaturation, mother-infant interaction, and cognitive development in small-for- gestational-age premature infants. Pediatrics. 2006;118(3):e869–78. 52. Counsell SJ, Allsop JM, Harrison MC, Larkman DJ, Kennea NL, Kapellou O, et al. Diffusion-weighted imaging of the brain in preterm infants with focal and diffuse white matter abnormality. Pediatrics. 2003;112(1 Pt 1):1–7. 53. Rose SE, Hatzigeorgiou X, Strudwick MW, Durbridge G, Davies PS, Colditz PB. Author details 1 Effect of oil massage on growth in preterm neonates less than 1800 g: a randomized control trial. Indian J Pediatr. 2013;80(6):465–9. 63. Kwon SH, Vasung L, Ment LR, Huppi PS. The role of neuroimaging in predicting neurodevelopmental outcomes of preterm neonates. Clin Perinatol. 2014;41(1):257–83. 39. Badr LK, Abdallah B, Kahale L. A Meta-Analysis of Preterm Infant Massage: An Ancient Practice With Contemporary Applications. MCN Am J Matern Child Nurs. 2015;40(6):344–58. 64. Scher MS, Steppe DA, Sclabassi RJ, Banks DL. Regional differences in spectral EEG measures between healthy term and preterm infants. Pediatr Neurol. 1997;17(3):218–23. 40. Wang L, He JL, Zhang XH. The efficacy of massage on preterm infants: a meta-analysis. Am J Perinatol. 2013;30(9):731–8. 65. Scher MS, Sun M, Steppe DA, Banks DL, Guthrie RD, Sclabassi RJ. Comparisons of EEG sleep state-specific spectral values between healthy full-term and preterm infants at comparable postconceptional ages. Sleep. 1994;17(1):47–51. 41. Noble Y, Boyd R. Neonatal assessments for the preterm infant up to 4 months corrected age: a systematic review. Dev Med Child Neurol. 2012;54(2):129–39. 42. El-Dib M, Massaro AN, Glass P, Aly H. Neurodevelopmental assessment of the newborn: An opportunity for prediction of outcome. Brain Dev. 2011;33(2):95–105. 66. Scher MS, Sun M, Steppe DA, Guthrie RD, Sclabassi RJ. Comparisons of EEG spectral and correlation measures between healthy term and preterm infants. Pediatr Neurol. 1994;10(2):104–8. 67. Shany E, Berger I. Neonatal electroencephalography: review of a practical approach. J Child Neurol. 2011;26(3):341–55. 43. Spittle AJ, Doyle LW, Boyd RN. A systematic review of the clinimetric properties of neuromotor assessments for preterm infants during the first year of life. Dev Med Child Neurol. 2008;50(4):254–66. 68. Guzzetta A, Baldini S, Bancale A, Baroncelli L, Ciucci F, Ghirri P, et al. Massage accelerates brain development and the maturation of visual function. J Neurosci. 2009;29(18):6042–51. 44. Bosanquet M, Copeland L, Ware R, Boyd R. A systematic review of tests to predict cerebral palsy in young children. Dev Med Child Neurol. 2013;55(5):418–26. 69. Gulamani SS, Premji SS, Kanji Z, Azam SI. A review of postpartum depression, preterm birth, and culture. J Perinat Neonatal Nurs. 2013;27(1):52–9. quiz 60–1. 45. Einspieler C, Prechtl HF, Ferrari F, Cioni G, Bos AF. The qualitative assessment of general movements in preterm, term and young infants–review of the methodology. Early Hum Dev. 1997;50(1):47–60. 46. Mercuri E, Ricci D, Romeo DM. Neurological and visual assessments in very and late low-risk preterm infants. Early Hum Dev. Lai et al. BMC Pediatrics (2016) 16:146 Lai et al. BMC Pediatrics (2016) 16:146 Lai et al. BMC Pediatrics (2016) 16:146 78. Boyd RC, Le HN, Somberg R. Review of screening instruments for postpartum depression. Arch Womens Ment Health. 2005;8(3):141–53. 78. Boyd RC, Le HN, Somberg R. Review of screening instruments fo 79. Lovibond SL, PF. Manual for the Depression Anxiety Stress Scales. 2nd ed. Sydney: Psychology Foundation; 1995. 80. Parkitny L, McAuley J. The Depression Anxiety Stress Scale (DASS). J Physiother. 2010;56(3):204. 81. Brown TA, Chorpita BF, Korotitsch W, Barlow DH. Psychometric properties of the Depression Anxiety Stress Scales (DASS) in clinical samples. Behav Res Ther. 1997;35(1):79–89. 82. Amankwaa L, Pickler R. Measuring maternal responsiveness. ABNF J. 2007;18(1):4–15. 83. Nagata M, Nagai Y, Sobajima H, Ando T, Honjo S. Depression in the mother and maternal attachment–results from a follow-up study at 1 year postpartum. Psychopathology. 2003;36(3):142–51. 84. George CaS J. The caregiving behavioural system: A behavioural system approach to parenting. In: Shaver JCPR, editor. The Handbook of attachment: Theory, research and clinical applications. New York: Guilford Press; 2008. p. 833–56. 85. Barnard KE, Hammond MA, Booth CL, Bee HL, Mitchell SK, Spieker SJ. Measurement and meaning of parent–child interaction. New York: Academic; 1989. 86. Easterbrooks MA, Goldberg WA. Security of attachment to mother and to father during toddlerhood: Relation to children's sciopersonality functioning during kindergarten. In: Greenberg DC M, Cummings EM, editors. Attachment in the preschool years: Theory, research and intervention. Chicago: University of Chicago Press; 1990. p. 221–44. 87. Carter AS, Briggs-Gowan MJ, Jones SM, Little TD. The Infant-Toddler Social and Emotional Assessment (ITSEA): factor structure, reliability, and validity. J Abnorm Child Psychol. 2003;31(5):495–514. 88. Papile LA, Munsick-Bruno G, Schaefer A. Relationship of cerebral intraventricular hemorrhage and early childhood neurologic handicaps. J Pediatr. 1983;103(2):273–7. 89. Field TM, Schanberg SM, Scafidi F, Bauer CR, Vega-Lahr N, Garcia R, et al. Tactile/kinesthetic stimulation effects on preterm neonates. Pediatrics. 1986;77(5):654–8. 90. Definitions for audit. ANZNN Data Dictionary. 2013. p. 165. 91. Dubowitz LM, Dubowitz V, Palmer P, Verghote M. A new approach to the neurological assessment of the preterm and full-term newborn infant. Brain Dev. 1980;2(1):3–14. 92. Dubowitz L, Dubowitz, V., Mercuri E. Neurological Assessment of the Preterm and Fullterm Newborn Infant. Mac Keith Press, London; 1999 93. Dubowitz L, Mercuri E, Dubowitz V. An optimality score for the neurologic examination of the term newborn. J Pediatr. 1998;133(3):406–16. 94. Cunningham NK, Brown PM, Brooks J, Page AC. Lai et al. BMC Pediatrics (2016) 16:146 Lai et al. BMC Pediatrics (2016) 16:146 The structure of emotional symptoms in the postpartum period: is it unique? J Affect Disord. 2013;151(2):686–94. 95. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782–6. 95. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782–6. 96. Taylor A, Atkins R, Kumar R, Adams D, Glover V. A new Mother-to-Infant Bonding Scale: links with early maternal mood. Arch Womens Mental Health. 2005;8(1):45–51. 97. Kidokoro H, Neil JJ, Inder TE. New MR imaging assessment tool to define brain abnormalities in very preterm infants at term. AJNR Am J Neuroradiol. 2013;34(11):2208–14. 97. Kidokoro H, Neil JJ, Inder TE. New MR imaging assessment tool to define brain abnormalities in very preterm infants at term. AJNR Am J Neuroradiol. 2013;34(11):2208–14. 98. Tournier JD, Calamante F, Connelly A. Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution. NeuroImage. 2007;35(4):1459–72. Author details 1 Altered white matter diffusion anisotropy in normal and preterm infants at term-equivalent age. Magn Reson Med. 2008;60(4):761–7. 77. Feldman R, Eidelman AI, Sirota L, Weller A. Comparison of skin-to-skin (kangaroo) and traditional care: parenting outcomes and preterm infant development. Pediatrics. 2002;110(1 Pt 1):16–26. Page 12 of 12 Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit and we will help you at every step: 99. Ball G, Counsell SJ, Anjari M, Merchant N, Arichi T, Doria V, et al. An optimised tract-based spatial statistics protocol for neonates: applications to prematurity and chronic lung disease. NeuroImage. 2010;53(1):94–102. 100. Raffelt D, Tournier JD, Rose S, Ridgway GR, Henderson R, Crozier S, et al. Apparent Fibre Density: a novel measure for the analysis of diffusion- weighted magnetic resonance images. NeuroImage. 2012;59(4):3976–94. 100. Raffelt D, Tournier JD, Rose S, Ridgway GR, Henderson R, Crozier S, et al. Apparent Fibre Density: a novel measure for the analysis of diffusion- weighted magnetic resonance images. NeuroImage. 2012;59(4):3976–94. 101. Condon JTCC. The assessment of parent-to-infant attachment: Development of a self-report questionnaire instrument. J Reprod Infant Psychol. 1998;16(1):57–76. 101. Condon JTCC. The assessment of parent-to-infant attachment: Development of a self-report questionnaire instrument. J Reprod Infant Psychol. 1998;16(1):57–76. 102. Teti DM, Gelfand DM. Behavioral competence among mothers of infants in the first year: the mediational role of maternal self-efficacy. Child Dev. 1991;62(5):918–29. 102. Teti DM, Gelfand DM. Behavioral competence among mothers of infants in the first year: the mediational role of maternal self-efficacy. Child Dev. 1991;62(5):918–29.
https://openalex.org/W2905521051	https://journals.open.tudelft.nl/ejtir/article/download/4794/4970	English	null	Optimum Stop Spacing for Accessibility	Deleted Journal	2,021	cc-by	9,092	Optimum Stop Spacing for Accessibility Hao Wu1 School of Civil Engineering, University of Sydney, Australia. David Levinson2 School of Civil Engineering, University of Sydney, Australia. The cumulative opportunities measure accessibility is defined as the number of opportunities reachable under a given time threshold. The spacing between transit stations is fundamental for accessibility by transit, yet the stations cannot be easily relocated in built-up areas. This paper examines the relation between transit stop spacing and person-weighted accessibility for an urban train route through an analytical model, and identifies that for each type of transit (e.g., given some combination of vehicle acceleration, deceleration, top speed, dwell time, platform type), an optimal stop spacing exists that maximizes accessibility; neither short nor excessive stop spacing are efficient in providing accessibility. Rail is used as example, though the model and findings are applicable to bus services as well. This paper brings attention to the importance of stop spacing in accessibility, and provides guidelines for transit planning for the operational improvement of transit accessibility. Keywords: accessibility, stop spacing, transit. A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: h.wu@sydney.edu.au 2 A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: david.levinson@sydney.edu.au 1 A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: h.wu@sydney.edu.au 2 A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: david.levinson@sydney.edu.au Issue 21(2), 2021, pp. 1-18 https://doi.org/10.18757/ejtir.2021.21.2.4794 Issue 21(2), 2021, pp. 1-18 https://doi.org/10.18757/ejtir.2021.21.2.4794 1 A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: h.wu@sydney.edu.au ISSN: 1567-7141 http://ejtir.tudelft.nl/ 2 A: University of Sydney, Civil Engineering Building J05, NSW 2006, Australia E: david.levinson@ 1. Introduction Operational differences between various types of urban rail technologies (Metro, light rail, tram, etc.) including vehicle acceleration and deceleration characteristics, cruising speed, station dwelling potentially affect the level of accessibility achievable; these factors are investigated in this study. Though transit systems often improve accessibility, and choices of investment by private firms and approved by government have been shown to favor those which expand access (Levinson et al., 2016), maximizing accessibility is usually not the explicit objective of public transport agencies, nor is it often a quantified consideration when routes and stations are planned (Boisjoly and El- Geneidy, 2017). There is a lack of motivation in improving transit accessibility by transit operators, who often cannot directly capture the value access generates, and the role of stop spacing in transit accessibility is often not recognized. Transit agencies operate under tight budget constraints, with two main objectives to improve coverage, and to increase ridership (Walker, 2012). The transit coverage issue stems from the limited area covered by each transit stop, and requires more stops with fewer coverage over-laps to increase overall coverage along the route (and to contain operating cost) (Ibeas et al.,2010, Saka, 2001); the ridership issue requires more densely spaced stations and coverage overlaps to capture more passengers along the planned route. Transit systems with similar capital investments can produce different levels of accessibility depending on their modal and spatial configuration (Ermagun et al., 2015), which affects transit mode share (Owenand Levinson, 2015) and patronage. Coverage and ridership goals affect stop spacing; how transit accessibility can be affected will be investigated. While this paper focuses on accessibility and its relation to stop spacing, the importance of transit planning has been acknowledged by other articles. Tirachini et al. (2010) implicitly examines accessibility by comparing the operator and user costs from different transit planning scenarios. The trade-off between more transit stops and increased time penalty is acknowledged by Ceder et al. (2015). Van Nes and Stolk (2012) examines the accessibility of railway stations with space syntax. The stop location minimizing total cost is discussed by Wirasinghe and Ghoneim (1981) and Alonso et al. (2011), in which Alonso et al. (2011) examine stop location under congestion and elastic demand. Ibeas et al. (2010) proposed a model to minimize total social cost of public transport by varying stop spacing. 1. Introduction Transit systems should be designed to increase access, enabling customers to easily reach destinations. The level of transit accessibility depends on both the transit infrastructure (e.g. stations) and service provision. While scheduled transit services can be modified regularly, the configuration of transit stations are more permanent and stations cannot be easily relocated. Hence, because of the long term fixity of the location of transit stops, getting that location right should be a conscious strategic planning decision by transit operators to improve accessibility. Transit accessibility to jobs affects mode choice (Owen and Levinson, 2015, Wu et al.,2019a), transit patronage (Cervero et al., 2010, Wu et al., 2019b), property value (Debrezion et al., 2007, Mayor et al., 2012). and has economical impacts (Fernandez, 1994). While many other transit performance EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 2 Optimum Stop Spacing for Accessibility metrics, including the areal coverage, ridership, time headway, etc. are based on the transit operators’ perspective, the transit access to jobs directly measures the experience of transit passengers, and it provides an easily quantifiable measure of how transit service can be improved. metrics, including the areal coverage, ridership, time headway, etc. are based on the transit operators’ perspective, the transit access to jobs directly measures the experience of transit passengers, and it provides an easily quantifiable measure of how transit service can be improved. In this paper, accessibility is defined using the cumulative opportunities measure, reflecting the number of opportunities reachable within a given travel time (Hansen, 1959, Ingram,1971, Wachs and Kumagai, 1973, Wickstrom, 1971). The travel time, and thus the level of access is affected by the convenience of transfers between different routes, and transit systems in different cities have varying levels of complexity. We use the case of a single transit line to examine the discuss the implications of stop spacing on access, since the transfer locations can be viewed as destinations; this simplification also ensures that findings from the single-line setting to be easily understandable, and applies under any circumstances. Foremost, faster systems require greater distance between stops. Greater distance between stops increases the costs of access to and egress from those stops. Among other factors, distance between transit stations affects the level of accessibility provided by urban trains, and has long lasting effect due to the semi-permanent nature of rail infrastructure. 1. Introduction dell’Olio et al.(2006) proposed a mathematical programming model to study frequency and stop locations to minimize operating cost of the transit system. Saka (2001) proposed an analytical model based on passenger demand, bus acceleration, deceleration characteristics, cruise speeds, etc.to study the optimum bus stop spacing for reducing required fleet size. Transit route design and stop spacing often face a lack of objectives. Based on the transit users’ perspective, travel time is often used as a performance measure for the quality of route design (Kikuchi and Vuchic, 1982, van Nes and Bovy, 2000). However, such travel time objectives rely EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility 3 Optimum Stop Spacing for Accessibility heavily on complex behavioral assumptions on trip destinations. On the other hand, accessibility measure provides an intuitive and practical performance objective for transit stop spacing, using one simple assumption on the traveler’s travel time budget. This paper argues for the existence of an optimum stop spacing for accessibility, and illustrates through an analytical model how this optimum spacing associates with operational characteristics of transit services. The paper provides general accessibility guidelines for the planning of transit systems, and discusses the effectiveness of operational improvement measures for better accessibility. 2. The analytical model 2.1 Accessibility measure 2.1 Accessibility measure The accessibility to jobs measured by the analytical model is the number of job opportunities encapsulated by all possible paths of a one-way transit trip within a travel time threshold of 30minutes; the accessibility model is formulated as in Equation 1 and Equation 2. Opportunities covered within that 30-minute one-way travel time threshold are considered reachable. We use job locations to represent opportunities, since access to jobs is often used as a surrogate for work opportunities, services, and amenities (Giuliano et al., 2010, Owen and Levinson,2015). The 30- minute threshold reflects the common one-way commute travel time in large cities which has held stable over time (El-Geneidy and Levinson, 2006, Marchetti, 1994). The accessibility measures tend to be correlated across other time-thresholds and between cumulative opportunities and gravity measures, so we don’t think other metrics would affect the insights of the analysis. 𝐴𝑖= ∑ 𝑂𝑗 𝑛 𝑗=1 . 𝑓(𝐶𝑖𝑗) (1) 𝑓(𝐶𝑖𝑗) = {1 𝑖𝑓 𝐶𝑖𝑗< 𝐵 0 𝑖𝑓 𝐶𝑖𝑗≥𝐵 (2) (1) (2) 𝐴𝑖= ∑ 𝑂𝑗 𝑛 𝑗=1 . 𝑓(𝐶𝑖𝑗) 𝑓(𝐶𝑖𝑗) = {1 𝑖𝑓 𝐶𝑖𝑗< 𝐵 0 𝑖𝑓 𝐶𝑖𝑗≥𝐵 (1) (2) : cumulative accessibility for an average individual within service area of station i 𝐴𝑖: cumulative accessibility for an average individual within service area of station i 𝐴𝑖: cumulative accessibility for an average individual within service area of station i 𝑂𝑗: number of opportunities at location j 𝑂𝑗: number of opportunities at location j 𝑓(𝐶𝑖𝑗): travel impedance as a function of travel time from i to j, 𝐶𝑖𝑗 𝐴𝑖: travel time budget (30 minutes assumed by the model) 𝐴𝑖: travel time budget (30 minutes assumed by the model) 2.2 Spatial distribution of jobs and transit riders 2.2 Spatial distribution of jobs and transit riders p f j The model assumes the number of transit riders and job opportunities decrease with distance from the station. The assumption on the spatial distribution of jobs sets the amount of opportunities that exist within the 30-minute catchment area; the distribution of transit riders affects the person- weighted station access time. Spatial distribution of jobs and transit riders are represented by densities (𝑗𝑜𝑏𝑠/𝑚2 and 𝑟𝑖𝑑𝑒𝑟𝑠/𝑚2); as an illustration, the model assumes 30,000/𝑘𝑚2 for both jobs and riders. The density of jobs decrease linearly away from transit stations; the population density is assumed to be uniform, but the percentage of population using transit (transit riders) decreases linearly with distance to transit stations. For simplicity, the model assumes identical density of jobs and transit riders with zero distance to transit stations; both job and rider densities decrease at the same rate (k), as a function of distance from transit stations (d). This type of distribution assumption is elegant in that it does not require extra parameters Daganzo (2010). The percentage of transit users from outside the service area is low (McNeil et al., 2017). By limiting station service area with a walking distance of L, the model assumes when the distance to the closest station (d) reaches the limit of service area (L), the density of both job opportunities and the density of transit riders (ρ𝑑) drop to zero (Equation 3). ρd = ρ(1 − d L) (3) EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility Optimum Stop Spacing for Accessibility ρ𝑑: density of transit riders or jobs with a distance of d to the nearest transit stop ρ: density of riders or jobs with zero distance to transit stop 𝑑: distance to the nearest transit stop (Manhattan distance) 𝐿: walking distance defining service area of a transit station ρ𝑑: density of transit riders or jobs with a distance of d to the nearest transit stop ρ: density of riders or jobs with zero distance to transit stop 𝑑: distance to the nearest transit stop (Manhattan distance) 𝐿: walking distance defining service area of a transit station It is worth noting that the survival function calibrated for walking distances (Iacono et al.,2008) is very close to linear for most of its range. 2.3 Pedestrian network 2.3 Pedestrian network The model assumes a grid type pedestrian network, so walking distance for station access and egress uses the Manhattan distance. Figure 1 shows the rectangular shape of coverage area provided by each of the transit stop (stops in red dots). Size of the coverage area depends on the willingness to walk. An 800-meter (half-mile) walking distance threshold is used for the coverage areas (McNeil et al., 2017). Individuals vary in physical stamina, and willingness to walk long distances; a 400-meter walking distance threshold is usually considered acceptable for buses (Walker, 2012), although people are willing to walk longer to access express transit services such as the urban rail (Guerra et al., 2012) Stop spacing affects the coverage area by each stop, as well as station access time, and opportunities reachable during egressing from the destination stop. When spacing between stops decreases, overlapping coverage areas begin to form in between stations (blue areas in Figure 1), reducing station access distance; at the destination end, the number of opportunities covered per transit stop decreases with denser stop spacing. Figure 1. Station service areas using Manhattan Distance; red dots represent transit station; overlapping coverage areas shown in blue. Figure 1. Station service areas using Manhattan Distance; red dots represent transit station overlapping coverage areas shown in blue. 2.4 Components of travel time 2.4 Components of travel time p f The model uses a 30-minute travel time threshold for clarity. The analytical model defines a one- way transit trip as first accessing the closest transit stop (which may involve out-of-direction travel), on-board travel, and egressing from the terminal transit stop to the actual destination. The travel time for each one-way transit trip includes station access, egress, on-board travel time and waiting time on the platform. Transfers are ignored, as the concept presented here is the same for transfers except that the ‘actual destination’ is instead replaced by the transfer location. Theoretically, time spent waiting on the platform can be eliminated if passengers can time their departures according to the transit timetable, and if trains strictly run on schedule. The waiting time is further complicated by individual variations and transit timetable reliability issues; such behavioral aspects of waiting time introduces unnecessary complications, thus the waiting time on platform is set to null, assuming the travelers have knowledge of the timetable and would react rationally. The waiting time com-ponent is therefore excluded from this analysis. A one-way travel time 𝐶𝑖𝑗 can be expressed as Equation 4. 𝐶𝑖𝑗= 𝑇𝑎+ 𝑇𝑜+ 𝑇𝑒 (4) 𝐶𝑖𝑗= 𝑇𝑎+ 𝑇𝑜+ 𝑇𝑒 EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility 5 𝐶𝑖𝑗: one-way travel time 𝑇𝑎: station access time 𝑇𝑜: on-board travel time 𝑇𝑒: station egress time For a morning commute, the station access time (𝑇𝑎) is the time spent walking from the residence (trip origin) to the closest transit station. For the access time to be representative, it is calculated as the population-weighted average walking time to station for transit riders within the coverage area of the origin transit station (Figure 1). Station access time is calculated with Equation 5, its derivation is provided in Appendix; Table 1 provides a list of variables used in this study. Station access time is affected by the spacing between stations. When stations are spread wide apart to the extent there are no overlapping coverage areas, station access time is at its maximum. As the spacing between stations shrink and overlapping coverage areas begin to occur (shown in Figure 1), the person-weighted walking distance is reduced. The extent of access time reduction relates to how closely the stops are placed. 2.4 Components of travel time 𝑇𝑎= { 𝐿 2·𝑉𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 3+16𝐿3 (4𝐷𝑆 3−24𝐿𝐷𝑠+48𝐿2)·𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 𝑇𝑎= { 𝐿 2·𝑉𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 3+16𝐿3 (4𝐷𝑆 3−24𝐿𝐷𝑠+48𝐿2)·𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 𝑇𝑎= { 𝐿 2·𝑉𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 3+16𝐿3 (4𝐷𝑆 3−24𝐿𝐷𝑠+48𝐿2)·𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 (5) (5) Table 1. Variables and Abbreviations 𝐶𝑖𝑗 One-way travel time from station i and job j 𝐿 Walking distance defining station service area (m) 𝐴𝑖 Cumulative accessibility for an average individual within service are of station i 𝑘 Rate of density decay, 𝑘= ρ/L D Travel distance on rail track 𝐷𝑠 Distance between stops (m) 𝑇𝑎 Station access time ρ Population/Job density with zero distance to station (per 𝑚2) 𝑇𝑒 Station egress time B One-way trip travel time budget 𝑇𝑜 On-board travel time (Total) ρ𝑑 Population/Job density with a distance of d to station 𝑡𝑜1 On-board acceleration time v𝑤 Walking speed (m/s) 𝑡𝑜2 On-board cruising time (at top speed) 𝑣 Train top running (cruising) speed 𝑡𝑜3 On-board deceleration time v𝑜 Train operating speed, including station stops 𝑡𝑜4 On-board station dwell time 𝑎 Rate of train acceleration 𝑁𝑠 Number of stops on rail track with distance D and stop spacing 𝐷𝑠 𝑏 Rate of train deceleration 𝑃𝑐 Total population covered by a single transit station d𝑐 Total station access distance by transit riders within a single transit station Table 1. Variables and Abbreviations The on-board travel time between two stations (𝑇𝑜) depends on the distance between the origin and destination stop, vehicle operating speed, acceleration and deceleration characteristics of the train, the number of stops in between, and the dwell time the vehicle will spend boarding and alighting passengers for every stop it makes (National Academies of Sciences, Engineering, and Medicine, 2013). As a train travels between stations, it inevitably goes through the cycle of decelerating from its top running speed, come to a complete stop for passengers to board and alight, then accelerate back to the cruising speed again (National Academies of Sciences, Engineering, and Medicine, 2013). This speed cycle is necessary for passengers to utilize the transit system, however, all passengers already on-board incur extra time with each stop added to the route. 2.4 Components of travel time Time remaining from the travel time budget after arriving at the terminal stop (station egress time 𝑇𝑒) is used for reaching opportunities through walking. Opportunities within this walking catchment area are considered accessible. When stops are densely spaced, the number of opportunities assigned to each stop decreases. It can be observed that without the explicit intention to optimize for accessibility, different types of transit have different stop spacing configurations: metro and heavy rails tend to have longer stop spacing, streetcars and trams with lower speeds have stops located much closer together. Such arrangement in stop spacing improves accessibility, which will be discussed in the following sections. 2.4 Components of travel time Take two extreme cases for example: if a train does not make any stop during its one-way run to its EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 6 Optimum Stop Spacing for Accessibility destination, then the aggregated travel speed will almost be the same as the top operating speed; and if the train is making stops all the time, it will be excessively slow as it has to start decelerating as soon as it clears the previous stop. destination, then the aggregated travel speed will almost be the same as the top operating speed; and if the train is making stops all the time, it will be excessively slow as it has to start decelerating as soon as it clears the previous stop. The overall travel speed of the train will always be below the top running speed, and the more stops the train makes, the lower its operation speed becomes. The on-board travel time has four components: acceleration (𝑡𝑜1), deceleration time (𝑡𝑜2), time the train runs at top running (cruising) speed (𝑡𝑜3), and dwell time serving passengers on platforms (𝑡04) (shown by Equation 6). Stations with larger demand will usually have longer dwell time to serve passengers; an additional operating margin (buffer time) is usually added in addition to the dwell time to absorb service disruptions and improve timetable robustness (National Academies of Sciences, Engineering, and Medicine, 2013). The length of buffer time largely depends on the transit operator; we assume a default 45 seconds dwell time for all stops, including buffer time. (6) 𝑇𝑜= ∑𝑡𝑜1 + ∑𝑡𝑜2 + ∑𝑡𝑜3 + ∑𝑡𝑜4 Time remaining from the travel time budget after arriving at the terminal stop (station egress time 𝑇𝑒) is used for reaching opportunities through walking. Opportunities within this walking catchment area are considered accessible. When stops are densely spaced, the number of opportunities assigned to each stop decreases. It can be observed that without the explicit intention to optimize for accessibility, different types of transit have different stop spacing configurations: metro and heavy rails tend to have longer stop spacing, streetcars and trams with lower speeds have stops located much closer together. Such arrangement in stop spacing improves accessibility, which will be discussed in the following sections. 2.5 Operating speed 2.5 Operating speed Assuming there exists an optimum stop spacing for accessibility (i.e. non-monotonic relation between stop spacing and accessibility), it follows that the same level of accessibility (other than the peak accessibility) can be borne by two different stop spacing distances. Stop spacing affects both the number of opportunities reachable (accessibility), and the distance traveled on rail track in the same time frame. Operating speed measures how fast a train travels on its route, including station stops (National Academies of Sciences, Engineering, and Medicine, 2013). Under the same travel time budget, higher operating speeds enable passengers to travel further down the route in the same travel time. Although this higher operating speed doesn’t necessarily add to the person- weighted accessibility, it affects the spatial distribution of accessible opportunities. The operating speed is sensitive to the same set of parameters as the level of accessibility, including the rate of acceleration and deceleration (e.g. power-to-weight ratio and braking system), terrain gradient, dwell time, and stop spacing (National Academies of Sciences, Engineering, and Medicine, 2013). Improvements in those parameters increases operating speed, and thus distance reachable by transit users. Stop spacing is the most important consideration among the parameters, since stations are semi-permanent structures that can- not be easily relocated, yet their spacing have major ramifications for both accessibility and operating speed. We explore analytically how the operating speed is affected by stop spacing. For a typical configuration, a default top running (cruising) speed of 60 km/h is set for the train; the rate of acceleration (a) and deceleration (b) are both set at 1.3 𝑚/𝑠2, and each dwell time cost 45 seconds, per the Transit Capacity and Quality of Service Manual (National Academies of Sciences, Engineering, and Medicine, 2013). Equation 7 describes operating speed as a function of stop spacing; the derivation is shown in appendix. vo = D to = v·Ds v2( 1 2a+ 1 2b)+v·to4+Ds (7) 7 7 EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility 3. Results and discussion Stop spacing has a fundamental effect on accessibility, both short and excessive stop spacing are inefficient in delivering accessibility. Excessive stop spacing allocates long stretches of travel time en route without making stops, this point-to-point transit service generates very limited access to the land along the rail track. Reducing stop spacing causes chunks of stop coverage areas to overlap, it follows that diminishing marginal gains in accessibility per stop combines with accumulated time penalties by each additional stop will gradually reduce the value of increasing stop density. A critical point will be reached where the value of additional stop density is about to turn from positive to negative. The optimum stop spacing is reached at that point. The degree of optimum accessibility reachable, as well as the stop spacing by which this optimum is reached depends on the system hardware (vehicles’ rate of acceleration, deceleration, top running speed), and operational details (door type, bike carry policies, etc.). Improvements in system hardware and in operations has varying impact on the subsequent accessibility, and carry different costs to transit operators. This section discusses the sensitivity of accessibility to these system parameters, and feasibility of implementing these changes. 3.1 Top running speed 3.1 Top running speed 3.2 Train Acceleration and Deceleration Train acceleration and deceleration are necessary for reaching top running speed (cruising), and for slowing down before reaching the next station to serve passengers. Greater train acceleration and deceleration improves accessibility by increasing the proportion of on-board time at top- running speed, raising the overall speed. There is limited room for raising the rates of speed change, mostly due to physical acceptance of standing and seated passengers onboard. Standing passengers take up less space, but are more vulnerable to speed changes; most urban rails have standing areas for passengers to increase the capacity of the carriages. Fast acceleration and deceleration raise the risk of passengers losing balance and falling when speed changes exceeds the capacity of the human body to react (Powell and Palacín, 2015). Rate of speed change at around 1.3 𝑚/𝑠2is most often used (National Academies of Sciences, Engineering, and Medicine, 2013, Powell and Palacín, 2015) by urban railway vehicles, and is the default rate for our model. Research on urban rail has shown rates of acceleration close to 2 𝑚/𝑠2as not acceptable for most of the participants, and a rate of 1 𝑚/𝑠2 is mostly agreeable (Hiroaki, 1995). This narrow spectrum of physically acceptable rates of speed change defines the range to be studied in this analysis. 0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 30km/h 40km/h 50km/h 60km/h 70km/h 80km/h 8 8 Figure 2. Accessibility and top running speeds. 0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 30km/h 40km/h 50km/h 60km/h 70km/h 80km/h 0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 30km/h 40km/h 50km/h 60km/h 70km/h 80km/h Figure 2. Accessibility and top running speeds. 3.2 Train Acceleration and Deceleration 3.2 Train Acceleration and Deceleration Train acceleration and deceleration are necessary for reaching top running speed (cruising), and for slowing down before reaching the next station to serve passengers. Greater train acceleration and deceleration improves accessibility by increasing the proportion of on-board time at top- running speed, raising the overall speed. There is limited room for raising the rates of speed change, mostly due to physical acceptance of standing and seated passengers onboard. 3.1 Top running speed p g p The top running speed is the terminal, or cruising speed a train reaches after serving passengers and accelerating at a constant rate from the previous station. With very close stop spacing, the top running speed may not be reached before reaching the next station. Figure 2 plots the relation between top running speed, stop spacing and the person-weighted accessibility, using 45 seconds default dwell time. Accessibility increases with higher running speed at each stop spacing distance. Initial improvement in speed produces proportionally greater increase in accessibility; raising speed for services that already have considerable running speed has less effect on accessibility, so there is a diminishing marginal return to increasing speed. The transit system incurs infrastructure and maintenance cost for increasing speed; passenger throughput can be negatively impacted as higher speed requires longer time-headway separation between services (National Academies of Sciences, Engineering, and Medicine, 2013). Thus a balanced speed for accessibility can be considered in practice. The level of accessibility provided by a streetcar-type transit with 30 km/h top speed falls far behind that of a 50 km/h service at every stop spacing. Higher speed services generally require longer stop spacing to maximize accessibility. This has been operationalized, perhaps unintentionally, by many transit services; high speed rail, metro and BRT generally have separate right-of-way and have stations placed further apart than tram and streetcars in mixed traffic. For streetcars and trams (mostly in downtown or CBD), it become desirable to allow higher station density to improve accessibility. For rails with moderate speeds (60-70 km/h top speed), optimum stop spacing is slightly short of the coverage radius, so there is some overlapping coverage area both for station access and egress; slower services require more service overlaps to optimize for accessibility. Municipal decision makers need to be conscious of stop spacing issue and the resulting accessibility provided by services with different speeds for the choice of transit investment. Low speed transit services in mixed traffic are inherently disadvantaged in providing accessibility. Civil speed limits are imposed upon trains where there are curves and steep downgrades (National Academies of Sciences, Engineering, and Medicine, 2013), which implies that cities with unfavorable terrains and circuitous routes might be inhibited in their ability to provide accessibility through public transport. EJTIR 21(2), 2021, pp.1-18 8 Wu and Levinson Optimum Stop Spacing for Accessibility Figure 2. Accessibility and top running speeds. 3.1 Top running speed Standing passengers take up less space, but are more vulnerable to speed changes; most urban rails have standing areas for passengers to increase the capacity of the carriages. Fast acceleration and deceleration raise the risk of passengers losing balance and falling when speed changes exceeds the capacity of the human body to react (Powell and Palacín, 2015). Rate of speed change at around 1.3 𝑚/𝑠2is most often used (National Academies of Sciences, Engineering, and Medicine, 2013, Powell and Palacín, 2015) by urban railway vehicles, and is the default rate for our model. Research on urban rail has shown rates of acceleration close to 2 𝑚/𝑠2as not acceptable for most of the participants, and a rate of 1 𝑚/𝑠2 is mostly agreeable (Hiroaki, 1995). This narrow spectrum of physically acceptable rates of speed change defines the range to be studied in this analysis. Figure 3 shows changes in accessibility by varying the rate of acceleration and deceleration within human acceptable range, using 60 km/h as default running speed. The effect of different rates of speed change on accessibility proves limited, given its narrow range of variation. The improvement in accessibility by raising the acceleration from the current state of practice at 1.3 𝑚/𝑠2 to the near unacceptable rate at 1.9 𝑚/𝑠2proves minimal. 9 Wu and Levinson Optimum Stop Spacing for Accessibility Figure 3. Accessibility with different rates of speed change 0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 1.0 m/s^2 1.3 m/s^2 1.6 m/s^2 1.9 m/s^2 Figure 3. Accessibility with different rates of speed change 0 10000 20000 30000 40000 50000 60000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 1.0 m/s^2 1.3 m/s^2 1.6 m/s^2 1.9 m/s^2 Figure 3. Accessibility with different rates of speed change 3.3 Station dwell time Dwell time is the time a train stopped at station serving passengers. Each additional stop increases interaction between the transit line and the adjacent land, the dwell time that comes with additional stops is a time penalty that affects overall travel speed, line capacity and accessibility. Required dwell time depends on the flow of passengers boarding and alighting the trains, the effectiveness of the flow, and the built-in buffer time. 3.1 Top running speed Major components of dwell time include: door open and close time; passenger flow time; time the doors remain open after passenger flow ceases (National Academies of Sciences, Engineering, and Medicine, 2013). There is ample room for improvement with the dwell time by better design of the carriages and by operational improvements; such measures include more train doors and increasing the speed the doors operates, train doors on both sides of carriages, and installation of physical barriers on platform to smooth passenger flow. These measures combined have the potential to significantly reduce station dwell time. Figure 4 plots accessibility change with potential improvement in station dwell time. Lower dwell time reduces the stop spacing required for optimum accessibility. With closer spaced stops, the reduction in station access distance potentially captures more riders along the transit line, increasing its ridership. 10 Wu and Levinson Wu and Levinson Optimum Stop Spacing for Accessibility Figure 4. Accessibility under different dwell time 0 10000 20000 30000 40000 50000 60000 70000 80000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 45 s 30 s 22.5 s Figure 4. Accessibility under different dwell time 3.4 Platform configuration Stations are not points, the connection time between station entrance and train boarding and alighting depends on station layout, and platform design. While the station layout varies, how passengers board and alight through platforms remains a fundamental problem, and has accessibility implications. A station with ramps on both ends of the platform (e.g. Central Station, Sydney) has the same average walking distance as stations with access to the center, but reduces detour between the train and the actual destination; stations with access on only one end (Redfern, Erskineville Station, Sydney) provide the longest on-platform walking distance (Lahoorpoor and Levinson, 2020). We use Sydney trains as an example to explore how platform design affects accessibility. Platform design has a noticeable impact on accessibility. For example, a typical Sydney train carriage measures 19.5 meters in length, and a usual 8 car train configuration spans 160 meters (Transport NSW, 2018). The average walking distance on platform depends on the combination of origin and destination station. 3.1 Top running speed Passengers traveling through both origin and destination stations that have access to the center of the platform incur an average platform walking distance of 1/2 train length for the whole trip, and 1 train length for stations with ramps on different ends of the train between origin and destination stops. The extra walking distance is longer if their origin and destination is on the end of the platform without an entrance, requiring passengers to backtrack. Figure 5 shows the accessibility implication of platform design. Tram and street car type urban rails with no measurable on-platform connection time have a modest advantage in accessibility. Though stations with ramps on either end of the platform or a single center ramp have identical on-platform connection time, extra ramps reduce the connection time especially during rush hours when passenger flow exceeds the capacity of exit ramps. In light of the accessibility benefits and the technical practicality, it is recommended that the worst case scenario of having a single ramp on one end of the platform be avoided. 11 Wu and Levinson Figure 5. Accessibility with different platform configurations 3.5 Walking speed, bike and ride The speed passengers access and egress from transit stations affects accessibility. Difference in age, gender and physical stamina cause difference in walking speeds; difficulties in getting to and from transit stations as a result of being physically disadvantaged can potentially cause equity issues. Docked and dock-less bike sharing provides significant speed improvement. In this study, the use of bicycles is treated as having a higher walking speed, and a 16 km/h (10 mph) is used (El-Geneidy et al., 2007). Effects of varying walking speeds are shown in Figure 6. Slower walking speeds among younger and older travelers, and the physically disabled, may reduce accessibility compared to more physically capable groups. This disadvantage, however, appears small in proportion and should not be a major cause of concern for equity issues. The use of bikes greatly improves accessibility. 0 10000 20000 30000 40000 50000 60000 70000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) Center Different ends No connection time Figure 5. Accessibility with different platform configurations 0 10000 20000 30000 40000 50000 60000 70000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) Center Different ends No connection time Figure 5. 3.1 Top running speed Accessibility with different platform configurations 3.5 Walking speed, bike and ride 3.5 Walking speed, bike and ride g p The speed passengers access and egress from transit stations affects accessibility. Difference in age, gender and physical stamina cause difference in walking speeds; difficulties in getting to and from transit stations as a result of being physically disadvantaged can potentially cause equity issues. Docked and dock-less bike sharing provides significant speed improvement. In this study, the use of bicycles is treated as having a higher walking speed, and a 16 km/h (10 mph) is used (El-Geneidy et al., 2007). Effects of varying walking speeds are shown in Figure 6. Slower walking speeds among younger and older travelers, and the physically disabled, may reduce accessibility compared to more physically capable groups. This disadvantage, however, appears small in proportion and should not be a major cause of concern for equity issues. The use of bikes greatly improves accessibility. 12 Wu and Levinson Wu and Levinson Optimum Stop Spacing for Accessibility Figure 6. Accessibility and Speed of Station Access/Egress 0 10000 20000 30000 40000 50000 60000 70000 0 500 1000 1500 2000 2500 3000 Accessibility Stop Spacing (Meter) 1.2 m/s 1.4m/s 1.6m/s 4.5m/s Figure 6. Accessibility and Speed of Station Access/Egress Integration of bikes and public transport expands accessibility. There are three pronged accessibility benefits of the bike-and-ride mode, including reduced station access and egress time, potentially expanded station service areas and a higher patronage of transit service by people within service areas. As a feeder service to transit, biking not just increases the speed of station access, the distance people willing to bike might be longer than they are willing to walk (Iacono et al., 2008), since biking reduces the physical burden. Bike-and-ride is a common mode of access in some European countries like Denmark and Netherlands where public transport is more developed, and it accounts for over a quarter of all access trips (Cervero et al., 2013); in the San Francisco Bay Area, the percentage of access trips by bikes has reached over ten percent (Cervero et al., 2013). 3.6 How operating speed depends on stop spacing 3.6 How operating speed depends on stop spacing p g p p p p g Operating speed measures how fast a train can travel on its route, including station stops (National Academies of Sciences, Engineering, and Medicine, 2013). Shown in Figure 7 is the relationship between operating speed and stop spacing, based on Equation 7. The initial increase in stop spacing notably raises operating speed, but there are diminishing returns to operating speed; further extending stop spacing results in proportionally less speed improvement. The operating speed approaches the top running speed (60 km/h) with infinitely long stop spacing; with stop spacing optimized for accessibility, actual operating speed reaches less than half of the top speed. Unlike accessibility which has a convex curve and peaks with a certain stop spacing, the operating speed increases monotonically at a slower rate with longer stop spacing. This implies that for two stop spacing configurations that produce the same level of accessibility, the stop spacing with a lower operating speed (shorter stop spacing) focuses more on the local accessibility, where jobs reachable are clustered within a short linear distance along the transit route; hence it is a more suitable choice for reinforcing downtown or CBD. The same level of accessibility reached with a higher operating speed has a wider spatial distribution of job opportunities, which would be ideal for extending the envelope of CBD. Selection of the stop spacing dictates not just the level of EJTIR 21(2), 2021, pp.1-18 13 Wu and Levinson 13 13 EJTIR 21(2), 2021, pp.1-18 EJTIR 21(2), 2021, pp.1-18 Optimum Stop Spacing for Accessibility Optimum Stop Spacing for Accessibility accessibility, but also the spatial distribution of accessible opportunities, so it requires a conscious decision on the objective of the transit system, and its intended area and population. accessibility, but also the spatial distribution of accessible opportunities, so it requires a conscious decision on the objective of the transit system, and its intended area and population. Figure 7. How Operating Speed Depends on Stop Spacing Figure 7. Figure 7. How Operating Speed Depends on Stop Spacing Figure 7. 4. Conclusions This study finds there exists a stop spacing that maximizes person-weighted accessibility for fixed route transit services. For a transit route, the optimal stop spacing depends on the characteristics of the transit vehicle, mostly its speed in operation, and gate configurations. We show that in transit planning, optimizing for accessibility is not a situation where building more stops (thus increasing infrastructure and operational costs) will automatically improves accessibility; on the contrary, building more stops beyond the optimal number will lower accessibility. Basically, we show that there is an optimum balance between adding more access point through more stations, with how much transit service is slowed down by these additional stops. We suggest that the accessibility goal be considered during the transit planning stage. For transit systems that have already past the planning and construction phases, some transit operational changes improve accessibility more effectively than others. We list the general guidelines for operationally improving transit accessibility. For transit systems that have already past the planning and construction phases, some transit operational changes improve accessibility more effectively than others. We list the general guidelines for operationally improving transit accessibility.  Reduce station dwell time by managing passenger flows, facilitating boarding and alighting;  Maintain a moderate cruising speed for transit vehicles through managing transit right- of-way;  Implement stop-skipping;  Integrate bike-and-ride for transit. Reducing station dwell time through managing passenger flows, or introducing entrances and exists that open on both ends of the station platform has the most significant improvement on accessibility, and may be the most technically feasible measure. Lower dwell time also allows for closer stop spacing, which benefits ridership. Raising and maintaining a moderate cruising speed for trains (e.g. 60-70 km/h) produces satisfactory levels of accessibility, further increase in speed has diminishing returns to accessibility improvement. Higher rates of acceleration and deceleration only marginally improves accessibility while discomforting passengers, thus is not recommended for implementation. EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility 14 Well-studied and accepted design objectives often emphasize minimizing total costs. References Alonso, B., Moura, J. L., dell’Olio, L., and Ibeas, Á. (2011). Bus stop location under different levels of network congestion and elastic demand. Transport, 26(2):141–148. Boisjoly, G. and El-Geneidy, A. M. (2017). The insider: A planners’ perspective on accessibility. Journal of Transport Geography, 64:33–43. Ceder, A. A., Butcher, M., and Wang, L. (2015). Optimization of bus stop placement for routes on uneven topography. Transportation Research Part B: Methodological, 74:40–61. Cervero, R., Caldwell, B., and Cuellar, J. (2013). Bike-and-ride: build it and they will come. Journal of Public Transportation, 16(4):5. Cervero, R., Murakami, J., and Miller, M. (2010). Direct ridership model of bus rapid transit in Los Angeles county, California. Transportation Research Record, (2145):1–7. Daganzo, C. F. (2010). Structure of competitive transit networks. Transportation Research Part B: Methodological, 44(4):434–446. Debrezion, G., Pels, E., and Rietveld, P. (2007). The impact of railway stations on residential and commercial property value: a meta-analysis. The Journal of Real Estate Finance and Economics, 35(2):161– 180. dell’Olio, L., Moura, J., and Ibeas, A. (2006). Bi-level mathematical programming model for locating bus stops and optimizing frequencies. Transportation Research Record (1971):23–31. El-Geneidy, A. M., Krizek, K. J., and Iacono, M. (2007). Predicting bicycle travel speeds along different facilities using GPS data: A proof of concept model. In Proceedings of the 86th annual meeting of the Transportation Research Board, compendium of papers. TRB Washington, DC, USA. El-Geneidy, A. M. and Levinson, D. M. (2006). Access to destinations: Development of accessibility measures. Ermagun, A., Levinson, D., et al. (2015). Accessibility and transit performance. Ermagun, A., Levinson, D., et al. (2015). Accessibility and transit performance. Fernandez, R. M. (1994). Race, space, and job accessibility: Evidence from a plant relocation. Economic Geography, 70(4):390–416. Giuliano, G., Gordon, P., Pan, Q., and Park, J. (2010). Accessibility and residential land values: some tests with new measures. Urban Studies, 47(14):3103–3130. Guerra, E., Cervero, R., and Tischler, D. (2012). Half-mile circle: Does it best represent transit station catchments? Transportation Research Record (2276):101–109. Hansen, W. G. (1959). How accessibility shapes land use. Journal of the American Institute of Planners, 25(2):73–76. Hiroaki, S. (1995). Ability to withstand sudden braking. Railway Res Rev, 52(5):18–21. Iacono, M., Krizek, K., and El-Geneidy, A. M. (2008). Access to destinations: How close is close enough? estimating accurate distance decay functions for multiple modes and different purposes. Ibeas, Á., dell’Olio, L., Alonso, B., and Sainz, O. (2010). Optimizing bus stop spacing in urban areas. 4. Conclusions It has been noted that bus services (low speed, close stop spacing) often have the lowest total cost, but the more expensive and faster light-rail can outperform bus through its speed advantage (Tirachini et al., 2010); this is consistent with the optimal stop spacing, in that faster services generally have better accessibility, and fewer stops. More expensive transit options that provide better accessibility might be worth the additional investment; for instance, when one option provides much better accessibility, but is only marginally more expensive, especially when it increases patronage and fare box recovery. Different types of transit have different requirements for stop spacing in order to maximize access. Faster services with exclusive right-of-way generally need longer stop spacing than slower moving tram-type services in mixed traffic. Transit services with low cruising speeds are inherently disadvantaged in providing accessibility, and has limited room for operational improvements; although they may have lower access costs as they avoid some travel time associated with larger, grade-separated train stations. Integration of bike and transit should be encouraged, for it provides substantial improvement in accessibility. Substitution of biking for walking allows the optimum accessibility to be achieved under longer stop spacing, which allows for higher operating speed of transit vehicles. Hence transit stations need to be spaced strategically with accessibility as an important consideration. Stop spacing has significant implications for accessibility, and the effect would be long lasting due to the near-permanent nature of the transport infrastructure. While operating cost, and other considerations may change over time with higher levels of automation, the accessibility impact of stop locations are more fundamental, and longer lasting. This study contributes by raising the importance and accessibility implications of stop spacing, and demonstrates that the observed pattern of longer spacing for faster services is generally consistent with accessibility maximization. 15 EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility Optimum Stop Spacing for Accessibility Optimum Stop Spacing for Accessibility References Transportation Research part E: Logistics and Transportation Review,46(3):446–458. Ingram, D. R. (1971). The concept of accessibility: A search for an operational form. Regional Studies, 5(2):101–107. Kikuchi, S. and Vuchic, V. R. (1982). Transit vehicle stopping regimes and spacings. Transportation Science, 16(3):311–331. Lahoorpoor, B. and Levinson, D. M. (2020). Catchment if you can: The effect of station entrance and exit locations on accessibility. Journal of Transport Geography, 82:102556. 16 16 EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility Levinson, D. M., Giacomin, D., and Badsey-Ellis, A. (2016). Accessibility and the choice of network investments in the London Underground. Journal of Transport and Land Use,9(1):131–150. Marchetti, C. (1994). Anthropological invariants in travel behavior. Technological Forecasting and Social Change, 47(1):75–88. Mayor, K., Lyons, S., Duffy, D., and Tol, R. S. (2012). A hedonic analysis of the value of rail transport in the greater Dublin area. Journal of Transport Economics and Policy (JTEP),46(2):239–261. McNeil, N., Dill, J., DeVitis, D., Doubleday, R., Duncan, A., and Weigand, L. (2017). Manual on pedestrian and bicycle connections to transit. Technical report. National Academies of Sciences, Engineering, and Medicine (2013). Transit capacity andquality of service manual,third edition. Washington, DC: The National Academies Press. Owen, A. and Levinson, D. M. (2015). Modeling the commute mode share of transit using continuous accessibility to jobs. Transportation Research Part A: Policy and Practice,74:110–122. Powell, J. and Palacín, R. (2015). Passenger stability within moving railway vehicles: limits on maximum longitudinal acceleration. Urban Rail Transit, 1(2):95–103. Saka, A. A. (2001). Model for determining optimum bus-stop spacing in urban areas. Journal of Transportation Engineering, 127(3):195–199. Tirachini, A., Hensher, D. A., and Jara-Díaz, S. R. (2010). Comparing operator and users costs of light rail, heavy rail and bus rapid transit over a radial public transport network. Research in Transportation Economics, 29(1):231–242. Transport NSW (2018). Technical drawings and specifications. Transport NSW. Van Nes, A. and Stolk, E. (2012). Degrees of sustainable location of railway stations: Integrating space syntax and node place value model on railway stations in the province of north holland’s strategic plan for 2010-2040. In Proceedings of the 8th International Space Syntax Symposium, January 3-6, 2012, Santiago de Chile, Chile. Van Nes, R. and Bovy, P. H. (2000). Importance of objectives in urban transit-network design. Transportation Research Record, 1735(1):25–34. Wachs, M. and Kumagai, T. G. (1973). Physical accessibility as a social indicator. Socio-Economic Planning Sciences, 7(5):437–456. Wachs, M. and Kumagai, T. G. References (1973). Physical accessibility as a social indicator. Socio-Economic Planning Sciences, 7(5):437–456. Walker, J. (2012). Human transit. Island Press Wickstrom, G. V. (1971). Defining balanced transportation-a question of opportunity. Traffic Quarterly, 25(3). Wickstrom, G. V. (1971). Defining balanced transportation-a question of opportunity. Traffic Quarterly, 25(3). Wirasinghe, S. C. and Ghoneim, N. S. (1981). Spacing of bus-stops for many to many travel demand. Transportation Science, 15(3):210–221. Wirasinghe, S. C. and Ghoneim, N. S. (1981). Spacing of bus-stops for many to many travel demand. Transportation Science, 15(3):210–221. Wu, H., Levinson, D., and Owen, A. (2019a). Commute mode share and access to jobs acrossus metropolitan areas. Environment and Planning B: Urban Analytics and City Science, page 2399808319887394. Wu, H., Levinson, D., and Owen, A. (2019a). Commute mode share and access to jobs acrossus metropolitan areas. Environment and Planning B: Urban Analytics and City Science, page 2399808319887394. Wu, H., Levinson, D., and Sarkar, S. (2019b). How transit scaling shapes cities. Nature Sustainability, 2(12):1142–1148. Wu, H., Levinson, D., and Sarkar, S. (2019b). How transit scaling shapes cities. Nature Sustainability, 2(12):1142–1148. EJTIR 21(2), 2021, pp.1-18 Wu and Levinson O i S S i f A 17 EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility EJTIR 21(2), 2021, pp.1-18 Wu and Levinson Optimum Stop Spacing for Accessibility Optimum Stop Spacing for Accessibility Appendix A - Derivation of station access time 𝑇𝑎= { 𝐿 2𝑣𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 2+16𝐿3 (4𝐷𝑆 2−24𝐿𝐷𝑠+48𝐿2)𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 𝑇𝑎= { 𝐿 2𝑣𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 2+16𝐿3 (4𝐷𝑆 2−24𝐿𝐷𝑠+48𝐿2)𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 (10) 𝑇𝑎= { 𝐿 2𝑣𝑤 𝑖𝑓 𝐷𝑠≥2𝐿 𝐷𝑆 3−4𝐿𝐷𝑆 2+16𝐿3 (4𝐷𝑆 2−24𝐿𝐷𝑠+48𝐿2)𝑣𝑤 𝑖𝑓 𝐷𝑠< 2𝐿 (10) EJTIR 21(2), 2021, pp.1-18 Wu and Levinson 18 Wu and Levinson Optimum Stop Spacing for Accessibility Appendix A - Derivation of station access time The station access time used in measuring accessibility is the person weighted average travel time for transit riders whose residence is within the transit service area (defined as 800 meters Manhattan distance) to the closest train station. The model assumes a linear decay function with 100% trip likelihood at zero distance to station, and 0% at 800 meters at the edge of the service area. Population served by each of the transit station (Pc) is calculated as the original residential density (ρ) adjusted by a linear distance decay to reflect trip likelihood, as shown in Equation 8. Overlapping service areas reduces population covered by each of the station. A residential density of 5600 persons/km2 is assumed, although the relation between the density of transit riders and distance is scalable, so the assumption on density does not affect the person-weighted station access time. 𝑃𝑐= { 4 ∫ [(𝜌−𝑘√2𝑥)2𝑥] √2 2 𝐿 0 𝑑𝑥 𝑖𝑓 𝐷𝑠≥2𝐿 4 ∫ [(𝜌−𝑘√2𝑥)2𝑥] √2 2 𝐿 0 𝑑𝑥 − 4 ∫ [(𝜌−𝑘 𝐷𝑠 2 −𝑘√2𝑥)2𝑥] √2 2 (𝐿−𝐷𝑠 2 ) 0 𝑑𝑥 𝑖𝑓 𝐷𝑠< 2𝐿 (8) (8) 𝑃𝑐: population covered by each of the transit station Cumulative walking distance to transit station for all riders within a service area (𝐷𝑐) isgiven by Equation 9. 𝑑𝑐= { 4 ∫ [(𝜌−𝑘√2𝑥)2𝑥]√2 √2 2 𝐿 0 𝑥𝑑𝑥 𝑖𝑓 𝐷𝑠≥2𝐿 4 ∫ [(𝜌−𝑘√2𝑥)2𝑥]√2 √2 2 𝐿 0 𝑥𝑑𝑥−4 ∫ [(𝜌−𝑘 𝐷𝑠 2 −𝑘√2𝑥)2𝑥] √2 2 (𝐿−𝐷𝑠 2 ) 0 (√2𝑥+ 𝐷𝑠 2 )𝑑𝑥 𝑖𝑓 𝐷𝑠< 2𝐿 𝑖𝑓 𝐷𝑠≥2𝐿 𝑑𝑐= { 4 ∫ [(𝜌−𝑘√2𝑥)2𝑥]√2 √2 2 𝐿 0 𝑥𝑑𝑥−4 ∫ [(𝜌−𝑘 𝐷𝑠 2 −𝑘√2𝑥)2𝑥] √2 2 (𝐿−𝐷𝑠 2 ) 0 (√2𝑥+ 𝐷𝑠 2 )𝑑𝑥 𝑖𝑓 𝐷𝑠< 2𝐿 (9) ∫ [(𝜌 ) ] 0 ∫ [(𝜌 2 ) ] 0 ( 2 ) 𝑓 𝑠 (9) (9) The station access time for coverage area of a transit station is derived through dividing the total walking distance (derived in Equation 8) by the total service population from the service area of the origin transit stop (obtained in Equation 9), and the average walk or bike speed 𝒗𝒘. The resulting population weighted average station access time is Equation 10. Appendix B - Derivation of operating speed The number of stops a train makes on a one-way trip is intrinsically a discrete number. Here we derive the number of stops as a continuous variable of the total route distance divided by stop spacing between every two stops, i.e. 𝑁𝑠= 𝐷 𝐷𝑠. The distance traveled on the rail track (D) during a one-way trip comprises the distance run at top speed, and distance used for speed changes. This rail travel distance is given by Equation 11. 2 𝑎+ 𝑣2 2𝑎) · 𝐷 𝐷𝑠+ 𝑣· 𝑇𝑡𝑜𝑝 (11) 𝐷 = ( 𝑣2 2𝑎+ 𝑣2 2𝑎) · 𝐷 𝐷𝑠+ 𝑣· 𝑇𝑡𝑜𝑝 (11) The amount of time the train runs at top (cruising) speed is given by Equation 12. The amount of time the train runs at top (cruising) speed is given by Equation 12. 𝑡𝑜2 = 𝐷 𝑣(1 − 1 𝐷𝑠( 𝑣2 2𝑎+ 𝑣2 2𝑎)) 𝑡𝑜2 = 𝐷 𝑣(1 − 1 𝐷𝑠( 𝑣2 2𝑎+ 𝑣2 2𝑎)) (12) The total on-board travel time consists of the time spent for acceleration, deceleration, stopped time), and the time running at top speed, shown in Equation 13. The total on-board travel time consists of the time spent for acceleration, deceleration, stopped time), and the time running at top speed, shown in Equation 13. 𝑇𝑜= 𝐷 𝐷𝑠( 𝑣 𝑎+ 𝑣 𝑑+ 𝑡𝑜4) + 𝐷 𝑣(1 − 1 𝐷𝑠( 𝑣2 2𝑎+ 𝑣2 2𝑎)) (13) 𝑇𝑜= 𝐷 𝐷𝑠( 𝑣 𝑎+ 𝑣 𝑑+ 𝑡𝑜4) + 𝐷 𝑣(1 − 1 𝐷𝑠( 𝑣2 2𝑎+ 𝑣2 2𝑎)) 𝑇𝑜= 𝐷 𝐷𝑠( 𝑣 𝑎+ 𝑣 𝑑+ 𝑡𝑜4) + 𝐷 𝑣(1 − 1 𝐷𝑠( 𝑣2 2𝑎+ 𝑣2 2𝑎)) (13) (13) The operating speed (including station stops) as a function of stop spacing Ds is obtained by dividing rail travel distance (Equation 11) by on-board travel time (Equation 13), shown in Equation 14 below. The operating speed (including station stops) as a function of stop spacing Ds is obtained by dividing rail travel distance (Equation 11) by on-board travel time (Equation 13), shown in Equation 14 below. The operating speed (including station stops) as a function of stop spacing Ds is obtained by dividing rail travel distance (Equation 11) by on-board travel time (Equation 13), shown in Equation 14 below. 𝑣𝑜= 𝐷 𝑇𝑜= 𝑣·𝐷𝑠 𝑣2( 1 2𝑎+ 1 2𝑏)+𝑣·𝑡𝑜4+𝐷𝑠 (14) 𝑣𝑜= 𝐷 𝑇𝑜= 𝑣·𝐷𝑠 𝑣2( 1 2𝑎+ 1 2𝑏)+𝑣·𝑡𝑜4+𝐷𝑠 (14) 𝑣𝑜= 𝐷 𝑇𝑜= 𝑣·𝐷𝑠 𝑣2( 1 2𝑎+ 1 2𝑏)+𝑣·𝑡𝑜4+𝐷𝑠 (14)
https://openalex.org/W2339536475	http://wrap.warwick.ac.uk/79295/1/WRAP-SilE-intrinsically-disordered-periplasmic-%E2%80%98molecular-sponge%27-involved-bacterial-silver-resistance-Jenner-2016.pdf	English	null	SilE is an intrinsically disordered periplasmic “molecular sponge” involved in bacterial silver resistance	Molecular microbiology	2,016	cc-by	11,525	Original citation: Original citation: Asiani, Karishma R., Williams, Huw, Bird, Louise, Jenner, Matthew, Searle, Mark S., Jon L., Scott, David J. and Soultanas, Panos. (2016) SilE is an intrinsically disordered periplasmic ‘molecular sponge' involved in bacterial silver resistance. Molecular Microbiology, 101 (5). pp. 731-742. g Asiani, Karishma R., Williams, Huw, Bird, Louise, Jenner, Matthew, Searle, Mark S., Hobman, Jon L., Scott, David J. and Soultanas, Panos. (2016) SilE is an intrinsically disordered periplasmic ‘molecular sponge' involved in bacterial silver resistance. Molecular Microbiology, 101 (5). pp. 731-742. A note on versions: For more information, please contact the WRAP Team at: wrap@warwick.ac.uk Copyright and reuse: The Warwick Research Archive Portal (WRAP) makes this work of researchers of the University of Warwick available open access under the following conditions. This article is made available under the Creative Commons Attribution 4.0 International license (CC BY 4.0) and may be reused according to the conditions of the license. For more details see: http://creativecommons.org/licenses/by/4.0/ A note on versions: The version presented in WRAP is the published version, or, version of record, and may be cited as it appears here. Summary y Ag1 resistance was initially found on the Salmonella enetrica serovar Typhimurium multi-resistance plas- mid pMG101 from burns patients in 1975. The puta- tive model of Ag1 resistance, encoded by the sil operon from pMG101, involves export of Ag1 via an ATPase (SilP), an effluxer complex (SilCFBA) and a periplasmic chaperon of Ag1 (SilE). SilE is predicted to be intrinsically disordered. We tested this hypothe- sis using structural and biophysical studies and show that SilE is an intrinsically disordered protein in its free apo-form but folds to a compact structure upon optimal binding to six Ag1 ions in its holo- form. Sequence analyses and site-directed mutagen- esis established the importance of histidine and methionine containing motifs for Ag1-binding, and identified a nucleation core that initiates Ag1-medi- ated folding of SilE. We conclude that SilE is a molec- ular sponge for absorbing metal ions. Accepted 15 April, 2016. For correspondences. E-mail Panos.soul- tanas@nottingham.ac.uk; David.Scott@nottingham.ac.uk; Jon.Hob- man@nottingham.ac.uk. The copyright line for this article was changed on 23 August 2016 after original online publication. V C 2016 The Authors. Molecular Microbiology Published by John Wile This is an open access article under the terms of the Creative Comm reproduction in any medium, provided the original work is properly cit Ag1 resistance was initially found on the Salmonella enetrica serovar Typhimurium multi-resistance plas- mid pMG101 from burns patients in 1975. The puta- tive model of Ag1 resistance, encoded by the sil operon from pMG101, involves export of Ag1 via an ATPase (SilP), an effluxer complex (SilCFBA) and a periplasmic chaperon of Ag1 (SilE). SilE is predicted to be intrinsically disordered. We tested this hypothe- sis using structural and biophysical studies and show that SilE is an intrinsically disordered protein in its free apo-form but folds to a compact structure upon optimal binding to six Ag1 ions in its holo- form. Sequence analyses and site-directed mutagen- esis established the importance of histidine and methionine containing motifs for Ag1-binding, and identified a nucleation core that initiates Ag1-medi- ated folding of SilE. We conclude that SilE is a molec- ular sponge for absorbing metal ions. However, the sil operon comprises nine genes sil- PGABFCRSE (Silver, 2003; Mijnendonckx et al., 2013; Hobman and Crossman, 2015; Randall et al., 2015), organised into three transcriptional units, silCFBAGP, silRS and silE, each controlled by a different promoter (Silver, 2003). warwick.ac.uk/lib-publications warwick.ac.uk/lib-publications Molecular Microbiology (2016) 101(5), 731–742 Molecular Microbiology (2016) 101(5), 731–742 doi:10.1111/mmi.13399 First published online 7 May 2016 doi:10.1111/mmi.13399 First published online 7 May 2016 V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Introduction Karishma R. Asiani,1 Huw Williams,2 Louise Bird,3 Matthew Jenner,4 Mark S. Searle,2 Jon L. Hobman,1 David J. Scott1,5* and Panos Soultanas2* Karishma R. Asiani,1 Huw Williams,2 Louise Bird,3 Matthew Jenner,4 Mark S. Searle,2 Jon L. Hobman,1* David J. Scott1,5* and Panos Soultanas2* Silver is a soft, shiny, lustrous and precious metal (Lans- down, 2010) with high value as a human commodity in jewellery and as an investment, and with wide applica- tions in the electronics industry; approximately 24,000 tons of silver was mined and produced in 2012 (Mijnen- donckx et al., 2013). Silver has also been highly valued for its broad-spectrum antimicrobial properties and has been one of the most important antimicrobial agents prior to the discovery and introduction of antibiotics. The rapid emergence of antibiotic resistance among many bacteria has rejuvenated the interest in silver as a viable alternative antimicrobial agent (Holt and Bard, 2005; Atiyeh et al., 2007; Mijnendonckx et al., 2013). 1School of Biosciences, University of Nottingham, Sutton Bonington LE12 5RD, United Kingdom. 2Centre for Biomolecular Sciences, School of Chemistry, University Park, University of Nottingham, Nottingham NG7 2RD, United Kingdom. 3Oxford Protein Production Factory, Research Complex at Harwell, Rutherford Appleton Laboratory, Oxfordshire OX11 0FA, United Kingdom. 4Department of Chemistry, University of Warwick, Gibbet Hill, Coventry CV4 7AL, United Kingdom. 5ISIS Neutron and Muon Source and Research Complex at Harwell, Rutherford Appleton Laboratory, Oxfordshire OX11 0FA, United Kingdom. The widespread use of silver in medical and non- medical settings has resulted in the emergence of silver resistant bacteria. Initial reports of silver resistance date back to 1966 (Gupta et al., 1999; Mallard et al., 2012) and the first silver-resistant plasmid pMG101, a large 180 kb plasmid assigned to the IncHI incompatibility group and reportedly carrying the sil operon (conferring silver resistance), was isolated from Salmonella enterica serovar Typhimurium following the death from septicae- mia of several patients treated with silver nitrate, leading to the closure of the burn ward of the Massachusetts General Hospital (McHugh et al., 1975). The pMG101 sil resistance allowed growth of an Escherichia coli K-12 (E. coli) strain carrying pMG101, in standard Luria- Bertani (LB) broth containing 600 lM of Ag1, a concen- tration over six times of that known to be tolerated by E. coli strains K-12 strains lacking the plasmid (Gupta et al., 1999). Summary The corresponding proteins have been assigned putative roles based upon homology modelling compared with other known heavy metal resistant deter- minants of the Pco or Cus systems (Hobman and Crossman, 2015 and Fig. 1). silE is under the control of its own promoter and its expression is strongly induced Accepted 15 April, 2016. *For correspondences. E-mail Panos.soul- tanas@nottingham.ac.uk; David.Scott@nottingham.ac.uk; Jon.Hob- man@nottingham.ac.uk. The copyright line for this article was changed on 23 August 2016 after original online publication. Fig. 1. Silver resistance operon and functions of its corresponding proteins. Silver resistance proteins and their suggested active roles, deduced from homology, thought to partake in bacterial silver resistance. SilE—periplasmic metal-binding protein, SilR and SilS—responder and membrane sensor performing two-component transcription regulation, SilC—outer membrane protein, SilB—membrane fusion protein, SilA—chemi-osmotic antiporter, SilP—P-type cation ATPase and SilG (protein not depicted) and SilF—metal-binding chaperone protein. Dashed arrows highlight the hypothetical role of SilE. The bottom line shows the mRNAs, indicating the genes and open reading frames (including number of amino acids) with the orientation of their transcription (Silver, 2003). 732 K. R. Asiani et al. 732 K. R. Asiani et al. 732 K. R. Asiani et al. Fig. 1. Silver resistance operon and functions of its corresponding proteins. 732 K. R. Asiani et al. ability to bind copper ions (Gupta et al., 1999; Silver et al., 1999; Silver, 2003; Zimmerman et al., 2012) although there is no experimental data available to verify its precise function. Both PcoE and SilE have ten histi- dine residues that are spatially conserved (Mirolo et al., 2012) and have been proposed to be primary candi- dates for metal binding (Silver, 2003; Mirolo et al., 2012). Following a change in environmental pH, these residues could also partake in the release of Ag1 into the periplasmic space with the SilCBA efflux pump ejecting the toxic monovalent metal ion, out of the cell (Mirolo et al., 2012). This, however, contradicts other data showing increased binding of Ag1 ions to SilE under acidic conditions (Silver et al., 1999). In this paper, we provide experimental evidence that apo-SilE is an intrinsically disordered protein (IDP) that folds to a highly a-helical holo-SilE structure upon bind- ing to Ag1 ions. Summary SilE is an indispensable key component for the exogenous silver resistance phenotype (Randall et al., 2015), has been reported to bind between 5 and 38 Ag1, depend- ing on experimental conditions (Silver et al., 1999; Mij- nendonckx et al., 2013), and is often used as a marker when confirming the presence of silver resistance genes in microbes (Mirolo et al., 2012). It exhibits 48% sequence identity to the periplasmic copper-binding pro- tein PcoE which binds up to nine Cu1 or up to seven Ag1 ions (Zimmerman et al., 2012). The pcoE gene is within a cluster of seven genes (pcoABCDRSE) adja- cent to the sil operon on the large E.coli copper resist- ance plasmid pRJ1004. Expression of pcoE is controlled by the chromosomally located copper resistance cusRS system (Munson et al., 2000). The pco and sil operons have been found together in a single locus of identical arrangement in plasmids and on the chromosomes of many Gram-negative bacteria (Hobman and Crossman, 2015, Hao et al., 2015, Randall et al., 2015). PcoE, is believed to be unstructured in its apo-form but folds and dimerizes upon Cu1 binding, with some a-helical con- tent in its secondary structure (Zimmerman et al., 2012). Because of its similarity to PcoE, SilE is pre- sumed to have similar attributes as well as possess the Summary SilE can bind up to eight Ag1 ions or fewer of the harder divalent metal ions Cu21 (up to six), Zn21 (up to five) and Ni21 (up to two), indicating a higher capacity for complexing Ag1 compared to other metals. We show that metal-induced folding leads to a higher helical content with Ag1 followed by Cu21, Zn21 and Ni21, consistent with its higher selectivity for Ag1, and confirm from mutagenesis studies that conserved histidine and methionine residues within specific sequence motifs are involved in Ag1 binding. We pro- pose an Ag1-mediated nucleation folding mechanism for SilE and suggest that SilE acts as a “molecular sponge” and as a first line of defence against relatively low levels of Ag1 ions that enter the periplasm. Potential conse- quences of Ag1-mediated SilE folding relative to the combined bacterial silver resistance mechanism are also discussed. Fig. 1. Silver resistance operon and functions of its corresponding proteins. Silver resistance proteins and their suggested active roles, deduced from homology, thought to partake in bacterial silver resistance. SilE—periplasmic metal-binding protein, SilR and SilS—responder and membrane sensor performing two-component transcription regulation, SilC—outer membrane protein, SilB—membrane fusion protein, SilA—chemi-osmotic antiporter, SilP—P-type cation ATPase and SilG (protein not depicted) and SilF—metal-binding chaperone protein. Dashed arrows highlight the hypothetical role of SilE. The bottom line shows the mRNAs, indicating the genes and open reading frames (including number of amino acids) with the orientation of their transcription (Silver, 2003). Silver resistance proteins and their suggested active roles, deduced from homology, thought to partake in bacterial silver resistance. SilE—periplasmic metal-binding protein, SilR and SilS—responder and membrane sensor performing two-component transcription regulation, SilC—outer membrane protein, SilB—membrane fusion protein, SilA—chemi-osmotic antiporter, SilP—P-type cation ATPase and SilG (protein not depicted) and SilF—metal-binding chaperone protein. Dashed arrows highlight the hypothetical role of SilE. The bottom line shows the mRNAs, indicating the genes and open reading frames (including number of amino acids) with the orientation of their transcription (Silver, 2003). in the presence of Ag1. It codes for the 143-amino acid long periplasmic protein SilE whose precise role in silver resistance has not been experimentally confirmed. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Results The sequences displaying 48% identity, the “Percentage Identity” colour-scheme has a threshold of 80% or more being conserved residues (purple) and anything below 40% as non-conserved (white), with the colour gradient between clarifying other less conserved residues. Conserved, aligned histidine and methionine residues have been highlighted above with either an inverted triangle or a star, respectively. B. Repeat sequence identification and secondary structure fold, predicted on Jpred can be seen above the identification of repeat sequence patterns with a-helical predominance (six a-helices) in the folded conformation of SilE. C Two repeat sequence motifs (motif 1 and 2) are evident which include the conserved histidine and methionine residues g A. Sequence alignment of E. coli PcoE with SilE from Salmonella (lacking the leader sequence, residues 1–20). The sequences, aligned using Clustal Omega (McWilliam et al., 2013) have been coloured using the “Percentage Identity” colour-scheme in Jalview (Waterhouse et al., 2009). The sequences displaying 48% identity, the “Percentage Identity” colour-scheme has a threshold of 80% or more being conserved residues (purple) and anything below 40% as non-conserved (white), with the colour gradient between clarifying other less conserved residues. Conserved aligned histidine and methionine residues have been highlighted above with either an inverted triangle or a star, respectively. B. Repeat sequence identification and secondary structure fold, predicted on Jpred can be seen above the identification of repeat sequence patterns with a-helical predominance (six a-helices) in the folded conformation of SilE. C T t tif ( tif 1 d 2) id t hi h i l d th d hi tidi d thi i id p p ( ) C. Two repeat sequence motifs (motif 1 and 2) are evident, which include the conserved histidine and methionine residu SilE has been predicted to be an IDP. The SilE sequence was analysed through DisEMBL’s and the output indicated 56% loop/coil (Supporting Information Fig. S1), suggesting that SilE has a high proportion of intrinsic disorder. In paral- lel, screening for possible regions of secondary structure (a- helical, b-strand or random coil) within the SilE (using Jpred, Cole et al., 2008) predicted six a-helices across 54% of the protein sequence (Fig. 2B). Interestingly, the secondary structure predictions through Jpred are consistent with the far-UV CD data obtained for the holo-form of SilE (Support- ing Information Table S1). Results Furthermore, two characteristic motifs, MxxHxxxxxHxxMxx (motif 1) and HxxMxxxHxxMxx (motif 2), each repeated twice within the sequence have conserved histidines and methionines that constitute poten- tial metal-binding motifs (Fig. 2C). at 207 and 221 nm, are consistent with a minor fraction (<20%) of a-helical secondary structure for apo-SilE (Sreerama et al., 1999; Greenfield, 2007; Dodero et al., 2011). However, these bands become considerably more prominent (54%) when bound to Ag1 in the holo-SilE, with a strong negative band at 207 nm, a weaker negative ellipticity at 221 nm and a strong posi- tive band at 190 nm (Fig. 3A), consistent with the stabili- zation of helical structure. NMR spectroscopy was used to examine the tertiary structure with and without Ag1 (in 10 mM HEPES, 20 mM NaF, pH 7.5). The two-dimensional 1H-15N HSQC of apo- SilE (Fig. 3B) exhibited poorly dispersed peaks, character- istic of an unfolded and unstructured protein (Dyson et al., 2005), and provided further experimental evidence for the largely disordered and flexible nature of SilE under native conditions. Likewise, one-dimensional proton NMR experi- ments (in 10 mM sodium phosphate buffer) at several pH intervals between pH 9 and pH 5, showed no change in the local environment of apo-SilE (data not shown). How- ever, the dispersion within the NMR backbone amide 1H chemical shifts in the 1H-15N HSQC spectrum of holo-SilE (Fig. 3B) increased substantially in the presence of bound Ag1 ; showing a clear signature for the induction of hydro- gen bonded secondary structure. Results Significant primary sequence and structural features of SilE Significant primary sequence and structural features of SilE Sequence alignment of native SilE and its Cu-binding homologue PcoE was carried out using a general multi- purpose primary sequence alignment program for pro- teins Omega (McWilliam et al., 2013), and amino acid sequences were coloured according to the “Percentage Identity” between the two proteins in Jalview (Water- house et al., 2009), (Fig. 2A). The calculated sequence identity between the two homologue proteins was 48% (Zimmerman et al., 2012). SilE and PcoE are rich in potential metal ligand-binding histidines (ten each) and methionines (ten and fifteen, respectively). The primary sequence identity shows that the position of the ten his- tidine residues of the two proteins is completely con- served (Fig. 2A), suggesting a key role in metal binding. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Fig. 2. Protein sequence analysis and evaluations. A. Sequence alignment of E. coli PcoE with SilE from Salmonella (lacking the leader sequence, residues 1–20). The sequences, aligned using Clustal Omega (McWilliam et al., 2013) have been coloured using the “Percentage Identity” colour-scheme in Jalview (Waterhouse et al., 2009). The sequences displaying 48% identity, the “Percentage Identity” colour-scheme has a threshold of 80% or more being conserved residues (purple) and anything below 40% as non-conserved (white), with the colour gradient between clarifying other less conserved residues. Conserved, aligned histidine and methionine residues have been highlighted above with either an inverted triangle or a star, respectively. B. Repeat sequence identification and secondary structure fold, predicted on Jpred can be seen above the identification of repeat sequence patterns with a-helical predominance (six a-helices) in the folded conformation of SilE. C. Two repeat sequence motifs (motif 1 and 2) are evident, which include the conserved histidine and methionine residues. Intrinsically Disordered Periplasmic “Molecular Sponge” 733 Intrinsically Disordered Periplasmic “Molecular Sponge” 733 Intrinsically Disordered Periplasmic “Molecular Sponge” 733 Fig. 2. Protein sequence analysis and evaluations. Fig. 2. Protein sequence analysis and evaluations. Fig. 2. Protein sequence analysis and evaluations. A. Sequence alignment of E. coli PcoE with SilE from Salmonella (lacking the leader sequence, residues 1–20). The sequences, aligned using Clustal Omega (McWilliam et al., 2013) have been coloured using the “Percentage Identity” colour-scheme in Jalview (Waterhouse et al., 2009). V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Structural analysis of SilE with and without Ag1 by CD and NMR The secondary structure of SilE was determined by far- UV circular dichroism (CD) spectroscopy (Fig. 3A). With strong negative signals around 200 nm, the spectrum obtained for apo-SilE in 10 mM HEPES, 20mM NaF, pH7.5, is typical of an unstructured, random coil poly- peptide. Slight negative shoulders on the CD spectrum, V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 734 K. R. Asiani et al. 734 K. R. Asiani et al. We investigated the thermal stability of the Ag-bound SilE complex using CD by generating a melting curve and monitoring the change in ellipticity at 207 nm. (Fig. 3C). The unfolding showed a sigmoidal transition with a mid-point of 428C. Moreover, the relatively sharp transi- tion from the folded to unfolded form, (Honig et al., 2003), as the temperature increased suggested that only two conformational states are significantly populated. It appears that SilE folding upon binding to Ag1 occurs cooperatively, as indicated by the sigmoidal ther- mal stability curve deduced via far-UV CD (Fig. 3C). The monophasic, cooperative unfolding of the protein confirmed that with Ag1 present, the protein exists as a compact well-folded, stable structure up to a tempera- ture of 428C. Moreover, the relatively sharp transition from the folded to unfolded form, characteristic of two- state proteins (Horng et al., 2003), as the temperature increased suggested that only these two conformational states were present to any significant extent. SilE binds up to eight silver ions Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Histidine and methionine residues are involved in Ag1-binding In order to investigate the roles of conserved histidine and methionine residues in Ag1-binding and protein folding, we mutated nine histidines to alanines (H38A, H62A, H69A, H80A, H87A, H111A, H118A, H129A and H136A) and four methionines to leucines (M72L, M83L, M90L and M108L). The secondary structure contents of all the mutant SilE proteins were measured in the pres- ence of six Ag1 molar equivalents (Fig. 5A–C). H38A, slight decrease in a-helical structure (Fig. 5A and Sup porting Information Table S1). In contrast, H80A, H87 and H111A exhibited large reductions in a-helical con tent in comparison to wild-type holo-SilE (Fig. 5A an Supporting Information Table S1), implying that thes residues are important for Ag1-induced folding. Th largest change in the CD spectra was seen with H111 which exhibits only around 60% of the wild type ho SilE secondary structure content (Supporting Informa tion Table S1). Based on this data, we then made three doub Fig. 4. Determination of SilE metal ion-binding. After incubation with 2 mM: Ni21, Zn21, Cu21 and Ag1. A. native nano-ESI MS for SilE (25 lM in 25 mM ammonium acetate, pH 7.0) showing the 71 charge state. The digits together with the metal elemental symbol provide the numbe of atoms of metal ions bound to SilE (as deduced by the incremental mas increases) in the labeled peak. Apo-SilE has molecular mass of 13,271 Da. B. Far-UV circular dichroism data (5 mM in 10 mM HEPES, 20 mM NaF, pH 7.5) exhibiting more alpha-helica protein content when SilE is in the presence of Ag1 over the other divalent metal ions, especially Ni21. C. SilE titration at 75 mM in 10 mM HEPES, 20 mM NaF, pH 7.5, from 0 to 16 equivalents Ag1 using far-UV CD. No change in signal following Ag1 addition beyond 6 Ag1 equivalents to SilE. D. Plots of the CD signals as a function of the Ag1:SilE molar ratios at 190 and 221 nm produce sigmoidal curves indicative of co-operative uptake of Ag1 ions by SilE. Intrinsically Disordered Periplasmic “Molecular Sponge” 7 Intrinsically Disordered Periplasmic “Molecular Sponge” 735 Intrinsic Fig. 4. Determination of SilE metal ion-binding. Histidine and methionine residues are involved in Ag1-binding slight decrease in a-helical structure (Fig. 5A and Sup- porting Information Table S1). SilE binds up to eight silver ions Interactions of certain metal ions with SilE have been reported using nano-ESI-MS (nano-Electrospray Ionization-Mass Spectrometry) under non-denaturing conditions in volatile 25 mM ammonium acetate buffer at pH 7.0. Solutions of 2 mM of each metal ion: Ag1, Cu21, Zn21 and Ni21 were added to 25 lM of apo-SilE (molar ratio of 80:1). SilE-metal complexes were observed, in each case with the relative bound proportions dependent on the protein’s affinity and/or stoichiometry for the metal ion (Fig. 4A). SilE shows a distribution of Ag-bound com- plexes with species containing 5, 6 and 7 bound Ag1 ions particularly abundant, with a maximum number of 8 detected. Binding of other harder divalent metal ions Cu21 and Zn21 was also evident, but with different bind- ing stoichiometries (SilE:Cu21 of 1:6; SilE:Zn21 of 1:5, and a lower stoichiometry with Ni21of SilE:Ni21 of 1:2). Complementary titrations using far-UV CD (Fig. 4C) show no further change in secondary structure content after the addition of 6 molar equivalents of Ag1. Hence six Ag1 appears to be the optimum for full folding of the pro- tein, however, the MS data suggest a further two Ag1 ions are capable of being bound by the folded protein. The Ag1 CD titration shows a very clear isodichroic point around 206 nm, which is consistent with predominantly two spe- cies in solution, namely the apo-SilE and a predominant single ‘fully loaded’ Ag1-bound form, rather than a hetero- geneous mixture of different species with different binding stoichiometries. Furthermore, plots of the CD signals at 190 and 221 nm as a function of the Ag1/SilE molar ratio show some evidence for sigmoidal curves consistent with co-operative uptake of Ag1 by SilE (Fig. 4D). g ) p A. Far-UV CD spectrum obtained at a concentration of 57mM for apo-SilE (red curve) represents very little secondary structure (negative signal around 200 nm). Upon addition of Ag1 (blue curve) largely negative signals are present at 207 and 221 nm typically found in proteins with significant helical structure. B. Chemical shifts seen in peaks when comparing the 1H/15N HSQC spectra of apo-SilE (300 mM) (red peaks) and Ag1-bound SilE (350 mM) (blue peaks), in the presence of 2 mM AgNO3, confirm Ag1-induced folding. C. The sigmoidal temperature denaturation of holo-SilE (57 mM) suggests folding in a cooperative manner. V C 2016 The Authors. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Two ‘core-motifs’ central to the primary sequence of SilE likely form the nucleation site for Ag1-induced folding H111A. Far-UV CD spectra of these proteins (Fig. 5B) showed a decrease in secondary structure content greater than that found with the single histidine mutant proteins with a-helical content declining to half that of the wild type. Simi- lar measurements for the methionine mutants found that the M108L mutant had a comparable decrease in secondary structure content to the H111A protein (Fig. 5C). The other three methionine mutants (M90L, M83L and M72L) showed effects similar to those observed for the histidine mutants (compare Fig. 5A and C). H111A. Far-UV CD spectra of these proteins (Fig. 5B) showed a decrease in secondary structure content greater than that found with the single histidine mutant proteins with a-helical content declining to half that of the wild type. Simi- lar measurements for the methionine mutants found that the M108L mutant had a comparable decrease in secondary structure content to the H111A protein (Fig. 5C). The other three methionine mutants (M90L, M83L and M72L) showed effects similar to those observed for the histidine mutants (compare Fig. 5A and C). From our studies of mutant SilE proteins it appears that the H80A, H87A, H111A and M108L exhibited the largest folding defects upon Ag1 binding. These residues are located centrally within the sequence in an apparent core comprising a motif 1 (residues A77- M91) and a motif 2 (residues E110-F120) (Fig. 2B). Given that the secondary structure content has a high sensitivity to their mutation, we proposed that they provide initial nucleation sites for Ag1-induced folding. To test this hypothesis we engineered two truncated SilE polypeptides, one SilE46-128 that preserves the core region (P46 to P128) but lacks the peripheral sequences at the N- and C-terminal regions, and a second SilE21-98 that lacks the C-terminal region and motif 2 from the putative nucleation core (Fig. 6A). We then studied Ag1-mediated folding of these poly- peptides using far-UV CD (Fig. 6B). The SilE46-128 polypeptide exhibited Ag1-mediated folding similar to the wt SilE whereas the SilE21-98 polypeptide did not exhibit any signs of folding upon Ag1 binding. These data are consistent with a model where a nucleation core is formed by central core motifs 1 and 2 which then facilitates further folding as more Ag1 are bound to the rest of the polypeptide. We then measured the number of Ag1 bound to each of the SilE proteins using inductively coupled plasma mass spectrometry (ICP-MS). SilE binds up to eight silver ions In contrast, H80A, H87A and H111A exhibited large reductions in a-helical con- tent in comparison to wild-type holo-SilE (Fig. 5A and Supporting Information Table S1), implying that these residues are important for Ag1-induced folding. The largest change in the CD spectra was seen with H111A which exhibits only around 60% of the wild type holo SilE secondary structure content (Supporting Informa- tion Table S1). In order to investigate the roles of conserved histidine and methionine residues in Ag1-binding and protein folding, we mutated nine histidines to alanines (H38A, H62A, H69A, H80A, H87A, H111A, H118A, H129A and H136A) and four methionines to leucines (M72L, M83L, M90L and M108L). The secondary structure contents of all the mutant SilE proteins were measured in the pres- ence of six Ag1 molar equivalents (Fig. 5A–C). H38A, H62A, H69A, H118A, H129A and H136A gave only a In order to investigate the roles of conserved histidine and methionine residues in Ag1-binding and protein folding, we mutated nine histidines to alanines (H38A, H62A, H69A, H80A, H87A, H111A, H118A, H129A and H136A) and four methionines to leucines (M72L, M83L, M90L and M108L). The secondary structure contents of all the mutant SilE proteins were measured in the pres- ence of six Ag1 molar equivalents (Fig. 5A–C). H38A, H62A, H69A, H118A, H129A and H136A gave only a Based on this data, we then made three double mutant proteins, H80A/H87A, H80A/H111A and H87A/ V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 736 K. R. Asiani et al. Fig. 5. SilE histidine and methionine mutations affect protein Ag1 binding and folding. A–C. Far-UV CD spectra of wild-type holo-SilE, histidine single (A) and double (B) mutants, and methionine (C) mutants with the 6 Ag1 equivalents at 75 lM in 10 mM HEPES, 20 mM NaF, pH 7.5. D-F. ICP-MS analyses of wild-type holo-SilE, histidine single (D) and double (E) mutants, and methionine (F) mutants at 50 nM in 10 mM HEPES, 20 mM NaF, pH 7.5. 736 K. R. Asiani et al. Fig. 5. SilE histidine and methionine mutations affect protein Ag1 binding and folding. A–C. Far-UV CD spectra of wild-type holo-SilE, histidine single (A) and double (B) mutants, and methionine (C) mutants with the 6 Ag1 equivalents at 75 lM in 10 mM HEPES, 20 mM NaF, pH 7.5. D-F. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 SilE binds up to eight silver ions ICP-MS analyses of wild-type holo-SilE, histidine single (D) and double (E) mutants, and methionine (F) mutants at 50 nM in 10 mM HEPES, 20 mM NaF, pH 7.5. Fig. 5. SilE histidine and methionine mutations affect protein Ag1 binding and folding. A–C. Far-UV CD spectra of wild-type holo-SilE, histidine single (A) and double (B) mutants, and methionine (C) mutants with the 6 Ag1 equivalents at 75 lM in 10 mM HEPES, 20 mM NaF, pH 7.5. D-F. ICP-MS analyses of wild-type holo-SilE, histidine single (D) and double (E) mutants, and methionine (F) mutants at 50 nM in 10 mM HEPES, 20 mM NaF, pH 7.5. methionine mutations affect protein Ag binding and folding. of wild-type holo-SilE, histidine single (A) and double (B) mutants, and methionine (C) mutants with the 6 Ag1 0 mM HEPES, 20 mM NaF, pH 7.5. D-F. ICP-MS analyses of wild-type holo-SilE, histidine single (D) and double (E) F) mutants at 50 nM in 10 mM HEPES, 20 mM NaF, pH 7.5. Two ‘core-motifs’ central to the primary sequence of SilE likely form the nucleation site for Ag1-induced folding Two ‘core-motifs’ central to the primary sequence of SilE likely form the nucleation site for Ag1-induced folding Putative core important for nucleation and Ag1-induced folding of SilE. g A. Schematic diagram of wild-type (WT) SilE, after removal 20- amino acid periplasmic signal sequence, featuring proposed putative core motifs SilE76-124 (coloured black and labeled) alongside truncated polypeptides SilE46-128 and SilE21-98. B. Far-UV CD spectra of wild-type holo-SilE alongside truncated mutants SilE46-128 and SilE21-98 with the 6 Ag1 equivalents at 45 lM in 10 mM HEPES, 20 mM NaF, pH 7.5. C. Schematic representation of the speculative nucleation model for Ag1- mediated folding of SilE. As SilE binds Ag1 and following initial nucleation within the putative core motifs (highlighted) the rest of the polypeptide folds gradually to its complete holo-structure upon further Ag1-binding. Primary sequence alignment between SilE and its homologue PcoE revealed conserved histidine and methionine residues (Fig. 2) (Mirolo et al., 2012) which in a-helical structures are known to be a common fea- ture of many metal-binding proteins (Todd et al., 1991; Tanaka et al., 2004). We have further confirmed this to be the case in SilE by site-specific mutagenesis. Impor- tantly, our mutagenesis studies showed that some resi- dues—H80, H87, H111 and M108—play a much more significant role than others in the correct Ag1-mediated folding of SilE (Fig. 5). Secondary structure predictions were consistent with our experimental data, suggesting the presence of six a-helices, however, significant helical structure was only realised in the Ag-bound state (Fig. 2). In addition, two types of Ag1-binding motifs, motif 1 MxxHxxxxxxHxxMxx and motif 2 HxxMxxxHxxMxx, where identified from the sequence in which i to i13 or i14 spacing between residues ensures that they appear on the same face of a folded helix. This suggests that each single helical motif in itself is unable to bind Ag1, but several of these together are able to (co-operatively) co-ordinate multiple Ag1 ions leading to a compact protein fold. Two of these motifs are located at the centre of the primary SilE sequence between residues S76-G124 (Fig. 2) and we define them as the “core-motifs” (Fig. 6). Both H80 and H87 are located in the core-motif 2 whereas M108 and H111 are located in the core-motif 1. We suggest the possibil- ity of an Ag1-mediated folding mechanism whereby ini- tial Ag1 binding to these “core-motifs” induces a V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Two ‘core-motifs’ central to the primary sequence of SilE likely form the nucleation site for Ag1-induced folding The single histidine mutants H62A, H69A, H80A, H87A, H111A, H118A and H129A all bound on average one Ag1 less than the native holo-SilE (Fig. 5D). Two of the single mutants, H38A and H136A did not exhibit a clear reduction in the number of bound Ag1 (Fig. 5D). All three double histi- dine mutants showed a reduction of two bound Ag1 (Fig. 5E) whereas all the methionine mutants showed a reduction of one bound Ag1 (Fig. 5F). Collectively our data show that a number of con- served histidines and methionines are involved in Ag1 binding in SilE, and the ability to bind Ag1 has a direct effect on the holo-SilE structure. The inability to bind Ag1 to key residues results in a decrease in the amount of folded protein, as judged by the reduction in CD ellip- ticity in the 207-221 nm region. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Intrinsically Disordered Periplasmic “Molecular Sponge” 737 Fig. 6. Putative core important for nucleation and Ag1-induced folding of SilE. metries, and subsequently a lower degree of folding was seen (Fig. 4B), indicating that folding and binding are coupled and form part of the ion discrimination mechanism. This use of folding and binding to enable uptake of a specific ion gives SilE the functional role of a ‘molecular sponge’ in the mechanism of E. coli silver resistance. As such its relationship to the other components in the silver resistance mechanism must be one where the unfolded protein has a high affinity for silver, but a low affinity for other cellular components. In contrast the folded protein will have a high affinity for other cellular components and as such allow SilE to off-load its Ag1 to other components in the resistance mechanism. Accordingly, SilE may act either as a metal-ion chaper- one to the RND effluxer SilCBA or, as many IDPs do (Kosol et al., 2013) binds to or signals to the histidine kinase sensor SilS and thereby activates the remainder of the silver resistance machinery. Therefore, the the notion that SilE binds Ag1 and then initiates the entire mechanism of silver resistance (Sil proteins P, G, A, B, F, C, R and S, as well as positively autoregulating its own expression), by signalling or direct binding to SilS, in bacteria is a highly plausible hypothesis. Fig. 6. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Discussion Many proteins are intrinsically unstructured under physio- logical conditions, yet can fold when required to perform their biological functions (Dyson et al., 2005; Radivojac et al., 2007; Sibile and Bernado, 2012; Jensen et al., 2013; Kosol et al., 2013). Previous NMR studies of the SilE homologue protein PcoE have shown that it folds to a predominantly a-helical structure upon binding ligand- metals Cu1 and Ag1 (Zimmerman et al., 2012). Like- wise, SilE is thought to be a putative IDP that binds to Ag1 but no experimental evidence was previously avail- able to verify its structure and metal binding properties. The data presented in this paper show that that apo- SilEs is an IDP lacking significant secondary structure over a range of pH conditions (Fig. 3). However, we were then able to demonstrate strong coupling between folding and metal-uptake by showing that SilE binds Ag1 (Figs. 3 and 4) and folds into a moderately stable struc- ture of high a-helical content. Despite the involvement of multiple histidine residues in Ag1 binding, no pH induced folding of SilE was observed, verifying that SilE folding is specifically mediated by Ag1 binding. Each SilE molecule can bind up to a maximum of eight monovalent soft metal ions of Ag1 in solution (Fig. 4A) but is fully folded after binding six Ag1: the last two Ag1 ions must therefore bind only to the fully folded pro- tein. Cu21, Zn21 and Ni21 ions, exhibit lower stoichio- 738 K. R. Asiani et al. nucleation site around which the rest of the SilE struc- ture assembles facilitating further Ag1 binding (Fig. 6). This is consistent with the sigmoidal plots of CD signals at 190 and 221 nm as a function of the Ag1/SilE molar ratio (Fig. 4D) which are consistent co-operative Ag1 binding and a folding nucleus rather than a simple sequential Ag1 binding and folding model. We tested this model by engineering truncated SilE polypeptides. The SilE46-127 polypeptide contained both “core motifs” and exhibited Ag1-mediated folding in a similar manner to the native SilE whereas the SilE21-97 polypeptide lack- ing the core-motif 1 exhibited no detectable Ag1-medi- ated folding; a result consistent with our nucleation model. Our double histidine mutations H80AH87, H80H111A and H87AH111A revealed loss of two Ag1 ions and it is therefore likely that these pairs of histi- dines coordinate different Ag1 ions. PCR-based site-directed mutagenesis The SilE-pOPINF construct served as a template for PCR- based site-directed mutagenesis. Histidine to alanine (single and double) mutations as well as the methionine to leucine mutations were generated in 50 and 30 DNA fragments using the following primer combinations, where mutant introducing nucleotides are shown in bold, lowercase. The forward (Fwd) primer in each case was used to generate the mutated 50—cDNA fragment and the reverse (Rev) primer to generate the mutated 30-cDNA fragment, in pairs with the flanking primers SilEPpuMIF and SilEHindIIIR. SilEH38AFwd; 50—GGCACCTGCTgccCAGATGCAGT—30 SilEH38ARev; 50—ACTGCATCTGggcAGCAGGTGCC—30 SilEH62AFwd; 50—TATGGACCAGgccGAACAGGCCAT- TATTGCTCAT—30 SilEH62ARev; 50—CATGAGCAATAATGGCCTGTTCggc CTGGTCCAT—30 SilEH69AFwd; 50—CATTATTGCTgccGAAACCATGAC- GAACGG—30 SilEH69ARev; 50—CCGTTCGTCATGGTTTCggcAGCAA- TAATG—30 SilEH80AFwd; 50—GGCGGATGCGgccCAGAAAATGG—30 SilEH80ARev; 50—CCATTTTCTGggcCGCATCCGCC—30 SilEH87AFwd; 50—GGTGGAAAGTgccCAGAGGATGAT G—30 SilEH87ARev; 50—CATCATCCTCTGggcACTTTCCAC C—30 SilEH111AFwd; 50—AATGAATGAGgccGAAAGAGCTGC AGTTG—30 SilEH111ARev; 50—CAACTGCAGCTCTTTCggcCTCATT CATT—30 Discussion against the new buffer at 48C, using dialysis membranes with molecular weight cut-off at 3,500 Da (Spectrum Labo- ratories Inc.). The sources of the metal ions were always soluble and excess ions were removed via extensive dialy- sis or using a Vivaspin device. The metal ion salts used are as follows:- Ag1 from silver nitrate (AgNO3), Cu21 from cupric sulfate (CuSO4), (Zn21) from zinc chloride (ZnCl2) and Ni21 from nickel chloride (NiCl2). Construction of expression plasmid The DNA sequence of the gene encoding SilE, minus the 20-amino acid peptide leader sequence, was amplified by PCR from the E. coli plasmid pMG101 (Gupta et al., 1999). The two primer sequences are: for the forward primer 50- ACTGAAACCGTGAATATCCATG-30 and for the reverse primer 50-GCCTGCACTGAGCATGCG-30. To facilitate DNA cloning and protein expression, an Opti3CInffwd site was incorporated in the forward primer and in the reverse primer, an Infusion 3’ site including a stop codon were inte- grated. The PCR product was then built-into the expression vector pOPINF (OPPF), comprising the coding sequence of a His-tag, by In-Fusion Reactions (Bird, 2011). The expres- sion construct was validated to contain the correct gene sequence insert by PCR screening, using the same cloning primers. Complexes of Ag1 with histidine and imidazole can form with Ag1:L and Ag1:L2 stoichiometries (where L is the ligand; histidine or imidazole) with the latter being more stable, as indicated by higher stability constants and enthalpies of formation (Czoik et al., 2008). Ag1:L2 complexes adopt a linear geometry (Muller et al., 2005; Petrovac et al., 2012; Kumbhar et al., 2013) and have been proposed to coordinate Ag1 ions in SilE (Silver, 2003). Methionine residues are reported to co-ordinate Ag1 ions in non-linear geometries with Ag1:M2 and Ag1:M3 stoichiometries in the multicopper oxidase CueO via methionine-rich helices (Singh et al., 2011) while CusF co-ordinates Ag1 ions in a manner that employs histidine, methionine and tryptophan residues (Loftin et al., 2007). The human copper transporter 1 (hCtr1) protein in turn co-ordinates Cu1 and Ag1 ions via histidine, methionine and cysteine residues (Du et al., 2013; Zhu et al., 2014; Rubino et al., 2015). It is clear that coordination of Ag1 ions by metal binding pro- teins can employ a diverse collection of amino acids and variable co-ordination geometry. Our data collectively show that the previous theoretical model of SilE binding to five Ag1 via ten His residues (Silver, 2003) is not cor- rect but only a high resolution structure of the holo-SilE will reveal the molecular details of Ag1 coordination in this protein. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Materials All SilE over-expression plasmids (wild-type and mutants) were maintained in E. coli DH5a cells and transformed into E. coli BL21 Star (DE3) cells (Invitrogen) for SilE wild-type and mutant over-expressions. Unlabelled samples of SilE were produced from cells grown on LB medium. 15N labelled SilE was prepared in standard minimal media sup- plemented with 15NH4Cl. Each litre of medium, supple- mented with 50 mg carbenicillin or 100 mg ampicillin, was inoculated with an overnight culture (5-10 mL) of the trans- formed E. coli cells. The cells were grown aerobically with vigorous shaking at 200 rpm, 378C, to OD600 0.6-0.8 and isopropyl b-D-1-thiogalactopyranoside (IPTG) was added at a final concentration of 1mM to induce protein expression. After overnight induction at 378C (200 rpm) the cells were harvested by centrifugation prior to lysis by sonication (Soniprep 150) at an output frequency of 23 kHz 12 cycles of 20 seconds at an amplitude of 10 microns, followed by 30 seconds of recovery were carried out in buffer contain- ing 50 mM Tris- (tris(hydroxymethyl)aminoethane) HCl, 500 mM NaCl, pH 7.5, supplemented with 100 lg/mL lyso- zyme and 1 mL per 20 g cells of protease inhibitor cocktail (Sigma; for use in purification of His-tagged proteins, DMSO solution). The cell lysate was clarified by centrifuga- tion and the supernatant contained the soluble SilE protein. All SilE proteins were expressed with an N-terminal His- tag to allow purification by affinity chromatography on nickel-chelating Sepharose (GE Healthcare) (in column binding buffer 50 mM Tris-HCl, 500 mM NaCl, pH 7.5 with elution in 500 mM imidazole), after which they were further purified by size-exclusion chromatography (SEC) on a pre- equilibrated (20 mM Tris-HCl, 200 mM NaCl, pH 7.5) Superdex 75 column (10 mm x 300 mm) (GE Healthcare). The His-tag was then cleaved off by incubation at 48C over- night, with 2.5 lg HRV 3C protease per 10 mL of protein sample and all forms of SilE finally were further purified by a second round of affinity chromatography on nickel- chelating Sepharose (GE Healthcare), whereby the protein was eluted in 50 mM Tris-HCl, 500 mM NaCl, 30 mM-50 mM imidazole (with the histidine mutants and truncated polypeptides requiring a lower concentration imidazole), pH 7.5. The purity and identity of the SilE proteins was con- firmed by; SDS-PAGE and nano-ESI MS, which yielded a molecular mass of 13,268 (1/-1.6) Da for wild-type (Suppl. Fig. Engineering of SilE truncated polypeptides The SilE-pOPINF construct served as a template for PCR- based production of the truncated polypeptides SilE46-128 and SilE21-98, using the following primers – SilE46-128 F; 50—ATTCCCCGGAGTTAATCCGGGACCTT- TAATTC – 30 SilE46-128 R; 50 – GCTAATGAAAGCTTCGGTTATTAAGG GGAAACGG – 30 SilE21-98 F; 50—CGATCGGGGCCCGCCTGTCGGGATC- CAGGGG – 30 SilE21-98 R; 50 – GCGCTTCAAGCTTGGCTTATTATGGG CCAG – 30 SilE46-128 F; 50—ATTCCCCGGAGTTAATCCGGGACCTT- TAATTC – 30 SilE46-128 R; 50 – GCTAATGAAAGCTTCGGTTATTAAGG GGAAACGG – 30 SilE21-98 F; 50—CGATCGGGGCCCGCCTGTCGGGATC- CAGGGG – 30 SilE21-98 R; 50 – GCGCTTCAAGCTTGGCTTATTATGGG CCAG – 30 Materials S2), corresponding to the values calculated from the sequences of SilE. All the purified SilE proteins contained no detectable metals. The 50 and 30 SilE mutated fragments were used as mega primers in a PCR including the flanking primers to generate the cDNA containing the entire translated region of the mutated His-tagged SilE. The final cDNA constructs were cloned into the PpuMI/HindIII site, of the prokaryotic expression vector pOPINF. The introduction of the muta- tions as well as the absence of undesired spontaneous mutations was confirmed by sequencing. Materials Chemicals and reagents (analytical grade) were obtained from Sigma Chemical, unless otherwise stated. All buffer exchanges were completed either by centrifugal ultrafiltra- tion through a high flux polyethersulphone membrane with molecular weight cut-off at 5,000 Da on a Vivaspin 20 or 2 devices or, by three cycles of four-eight hours of dialysis V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 739 Intrinsically Disordered Periplasmic “Molecular Sponge” SilEH118AFwd; 50—TGCAGTTGCCgccGAATTTATGAA- TAACG—30 SilEH118ARev; 50—CGTTATTCATAAATTCggcGGCAAC TGCA—30 SilEH129AFwd; 50—GTCTGGCCCAgccCAGGCCATG G—30 SilEH129ARev; 50—CCATGGCCTGggcTGGGCCAGA C—30 SilEH136AFwd; 50—GGCCGAAGCGgccCGTCGCATG C—30 SilEH136ARev—50—GCATGCGACGggcCGCTTCGGC C—30 SilEM72LFwd; 50—TCATGAAACCctgACGAACGGGTC—30 SilEM72LRev; 50 – GACCCGTTCGTcagGGTTTCATGA—30 SilEM83LFwd; 50—GCACCAGAAActgGTGGAAAGTCAT- CAG 230 SilEM83LRev; 50—CTGATGACTTTCCACcagTTTCTGGT GC – 30 SilEM90LFwd; 50—TCATCAGAGGctgATGGGAAGTCA- GAC 230 SilEM90LRev; 50 – GTCTGACTTCCCATcagCCTCTGAT GA – 30 SilEM108LFwd; 50—ATTAGCGGCActgAATGAGCATGAA AG 230 SilEM108LRev; 50 – CTTTCATGCTCATTcagTGCCGCT AAT – 30 SilEPpuMIF; 50—ATTCCCCGGAGTTAATCCgggacctTTAA TTC 230 SilEHindIIIR; 50—ATCACAAACTGGTCTAGAaagcttTAGC CTGC 230 Protein overexpression and purification All SilE over-expression plasmids (wild-type and mutants) were maintained in E. coli DH5a cells and transformed into E. coli BL21 Star (DE3) cells (Invitrogen) for SilE wild-type and mutant over-expressions. Unlabelled samples of SilE were produced from cells grown on LB medium. 15N labelled SilE was prepared in standard minimal media sup- plemented with 15NH4Cl. Each litre of medium, supple- mented with 50 mg carbenicillin or 100 mg ampicillin, was inoculated with an overnight culture (5-10 mL) of the trans- formed E. coli cells. The cells were grown aerobically with vigorous shaking at 200 rpm, 378C, to OD600 0.6-0.8 and isopropyl b-D-1-thiogalactopyranoside (IPTG) was added at a final concentration of 1mM to induce protein expression. After overnight induction at 378C (200 rpm) the cells were harvested by centrifugation prior to lysis by sonication (Soniprep 150) at an output frequency of 23 kHz 12 cycles of 20 seconds at an amplitude of 10 microns, followed by 30 seconds of recovery were carried out in buffer contain- ing 50 mM Tris- (tris(hydroxymethyl)aminoethane) HCl, 500 mM NaCl, pH 7.5, supplemented with 100 lg/mL lyso- zyme and 1 mL per 20 g cells of protease inhibitor cocktail (Sigma; for use in purification of His-tagged proteins, DMSO solution). The cell lysate was clarified by centrifuga- tion and the supernatant contained the soluble SilE protein. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 ICP-MS The instrument was run employing three operational modes, including (i) a collision-cell (Q cell) using He with kinetic energy discrimination (He-cell) to remove polyatomic interferences, (ii) standard mode (STD) in which the collision cell is evacuated and (iii) hydrogen mode (H2-cell) in which H2 gas is used as the cell gas. Samples were introduced from an autosampler (Cetac ASX-520) incorporating an ASXpressTM rapid uptake mod- ule through a PEEK nebulizer (Burgener Mira Mist). An internal standard Rh (10 mg L21) in 2% trace analysis grade (Fisher Scientific, UK) HNO3 was introduced to the sample stream on a separate line via the ASXpress unit. External multi-element calibration standards (Claritas-PPT grade CLMS-2 from SPEX Certiprep Inc., Metuchen, NJ) included Ag, Al, As, Ba, Be, Cd, Ca, Co, Cr, Cs, Cu, Fe, K, Li, Mg, Mn, Mo, Na, Ni, P, Pb, Rb, S, Se, Sr, Tl, U, V and Zn, in the range 0 – 100 mg L21 (0, 20, 40, 100 mg L21). A bespoke external multi-element calibration solution (PlasmaCAL, SCP Science, France) was used to create Ca, Mg, Na and K standards in the range 0–30 mg L21. Phosphorus, boron and sulphur calibration utilized in-house standard solutions (KH2PO4, K2SO4 and H3BO3). In-sample switching was used to measure B and P in STD mode, Se in H2-cell mode and all other elements in He-cell mode. Peak dwell times were 10 ms for the element with 150 scans per sam- ple. Sample processing was undertaken using QtegraTM software (Thermo-Fisher Scientific) utilizing external cross- calibration between pulse-counting and analogue detector modes when required with data being acquired in mg L21. Protein concentrations were measured before and after the experiment. All glassware and plasticware used for these experiments were washed with 10% nitric acid to remove contaminating metal. The stability of apo and holo-SilE (2 mM Ag1) to temper- ature denaturation was tested and determined by following changes in the CD spectra. The changes in the intensity of the maximal negative signal at 200 nm for apo and positive signal at 190 nm for holo- (2 mM Ag1) were recorded as a function of increasing temperature from 18 to 808C. The temperature was gradually increased at increments of 18C per minute and protein samples were allowed to equilibrate at each temperature, prior to recordings at intervals of 58C. In each case, spectra were acquired from 100 ml protein samples at 57 lM. ICP-MS The CD data was converted to a per- centage change of the maximum CD (mdeg). 740 K. R. Asiani et al. and transfer collision voltage, 6 and 5 V, respectively; back- ing pressure, 1.6–1.8 mbar; trap pressure, 2.1 3 1022 mbar; TOF region pressure, 1.5 3 1026 mbar. Instrument control as well as data processing was carried out with the Waters MassLynx 4.1 data system. All spectra were acquired in ion positive mode and the TOF analyser oper- ated on V-mode. Minimum smoothing and background sub- traction was applied to the obtained spectra prior to analysis. Scientific NanoDrop 2000c spectrophotometer (Scopes, 1974). Both calculations gave similar concentration values in the mg/mL range. Scientific NanoDrop 2000c spectrophotometer (Scopes, 1974). Both calculations gave similar concentration values in the mg/mL range. Nano-electrospray ionisation mass spectrometry Experiments were carried out and spectra were recorded on a SYNAPT High Definition Mass Spectrometry (HDMS) (Waters) a hybrid quadrupole ion mobility time-of-flight MS instrument, with travelling-wave ion mobility (TWIM) capabil- ity, equipped with the standard z-spray source. The instru- ment conditions were optimized to provide the highest relative signals for apo- and 2 mM metal ligand ion (Ag1, Cu21, Zn21, Ni21) bound-SilE complexes at a protein con- centration of 25 lM, sprayed from 25 mM ammonium ace- tate (C2H3O2NH4), pH 7.0. The nano-ESI capillary voltage was 1.5 kV; cone voltage, 20 V; extraction voltage, 5 V; transfer voltage, 5 V. Other settings were as follow: trap Protein Concentrations SilE lacks light absorbing tryptophan tyrosine residues and its concentration was estimated via two alternative meth- ods. Firstly, acquiring the Brix coefficient using an Atago DD-7 Digital Differential Refractometer allowed us to calcu- late the protein concentration using the following formula as explained elsewhere (Theisen et al., 2000): The final cDNA constructs were cloned into the PpuMI/HindIII site for SilE21-98 and into the ApaI/HindIII site for SilE46-128, of the prokaryotic expression vector pOPINF, ensuring the presence of the N-terminal His-tag remained. Confirmation of complete cloning of truncated polypeptides as well as the absence of undesired spon- taneous mutations was confirmed by sequencing. (dndc sucrose/dndc protein sample) x 7.8883 x Brix coef- ficient 5 (0.15/0.18) x 7.8883 x Brix coefficient Secondly, by measuring the absorbance of the peptide bond at the ultraviolet wavelength of 205 nm on a Thermo V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 10 Circular dichroism spectroscopy Far-UV CD was used to determine the secondary structure of apo- and holo-SilE in solution. CD experiments were conducted at CD1 beam line at the ASTRID2 storage ring facility at Aarhus University, Aarhus, Denmark (Hertela and Hoffman, 2011). Data was acquired at 258C from 50 ll apo- and metal ligand ion (Ag1, Cu21, Zn21, Ni21) bound-SilE protein samples at 57 or 75 lM, in a quartz suprasil cylin- drical cell (Hellma type 121.000) in a 10 mM HEPES (4-(2- hydroxyethyl)21-piperazineethanesulfonic acid), 20 mM NaF buffer at pH 7.5, with either 2 mM or titrated quantities (1-16 equiv. Ag1) of metal ions added to the holo samples. Spectra were recorded from 170 to 280 nm, with the protein sample being in a 0.05 cm path length cell; scan speed of 20 nm min-1 and a response time of 1 s, with each spec- trum representing an average of three accumulations, with an average of 15 scans per point. A scan of buffer alone was subtracted from the protein curve. Data were con- verted to molar CD per residue and spectra analysis was carried out by comparing the profile of the obtained curve to those illustrated and quantified in literature (Sreerama et al., 1999; Greenfield et al., 2007; Dodero et al., 2011). Secondary structure percentages were calculated using the DichroWeb (Lobley et al., 2002) interfaces analysis pro- gramme CONTINLL, which implements the locally linear- ised algorithm in selecting protein sets from the reference database (Provencher and Glockner, 1981; Van Stokkum et al., 1990; Sreerama and Woody, 2000). Far-UV CD was used to determine the secondary structure of apo- and holo-SilE in solution. CD experiments were conducted at CD1 beam line at the ASTRID2 storage ring facility at Aarhus University, Aarhus, Denmark (Hertela and Hoffman, 2011). Data was acquired at 258C from 50 ll apo- and metal ligand ion (Ag1, Cu21, Zn21, Ni21) bound-SilE protein samples at 57 or 75 lM, in a quartz suprasil cylin- drical cell (Hellma type 121.000) in a 10 mM HEPES (4-(2- hydroxyethyl)21-piperazineethanesulfonic acid), 20 mM NaF buffer at pH 7.5, with either 2 mM or titrated quantities (1-16 equiv. Ag1) of metal ions added to the holo samples. ICP-MS Ag1 elemental analysis of 50 nM protein diluted, with and without Ag1 in 5 mL solutions of 10 mM HEPES, 20 mM NaF, pH 7.5, in 1% HNO3, was undertaken by ICP-MS (Thermo-Fisher Scientific iCAP-Q; Thermo Fisher Scientific, Bremen, Germany). The instrument was run employing three operational modes, including (i) a collision-cell (Q cell) using He with kinetic energy discrimination (He-cell) to remove polyatomic interferences, (ii) standard mode (STD) in which the collision cell is evacuated and (iii) hydrogen mode (H2-cell) in which H2 gas is used as the cell gas. Samples were introduced from an autosampler (Cetac ASX-520) incorporating an ASXpressTM rapid uptake mod- ule through a PEEK nebulizer (Burgener Mira Mist). An internal standard Rh (10 mg L21) in 2% trace analysis grade (Fisher Scientific, UK) HNO3 was introduced to the sample stream on a separate line via the ASXpress unit. External multi-element calibration standards (Claritas-PPT grade CLMS-2 from SPEX Certiprep Inc., Metuchen, NJ) included Ag, Al, As, Ba, Be, Cd, Ca, Co, Cr, Cs, Cu, Fe, K, Li, Mg, Mn, Mo, Na, Ni, P, Pb, Rb, S, Se, Sr, Tl, U, V and Zn, in the range 0 – 100 mg L21 (0, 20, 40, 100 mg L21). A bespoke external multi-element calibration solution (PlasmaCAL, SCP Science, France) was used to create Ca, Mg, Na and K standards in the range 0–30 mg L21. Phosphorus, boron and sulphur calibration utilized in-house standard solutions (KH2PO4, K2SO4 and H3BO3). In-sample switching was used to measure B and P in STD mode, Se in H2-cell mode and all other elements in He-cell mode. Peak dwell times were 10 ms for the element with 150 scans per sam- ple. Sample processing was undertaken using QtegraTM software (Thermo-Fisher Scientific) utilizing external cross- calibration between pulse-counting and analogue detector modes when required with data being acquired in mg L21. Protein concentrations were measured before and after the experiment. All glassware and plasticware used for these experiments were washed with 10% nitric acid to remove contaminating metal. Ag1 elemental analysis of 50 nM protein diluted, with and without Ag1 in 5 mL solutions of 10 mM HEPES, 20 mM NaF, pH 7.5, in 1% HNO3, was undertaken by ICP-MS (Thermo-Fisher Scientific iCAP-Q; Thermo Fisher Scientific, Bremen, Germany). V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Bioinformatics Dyson, H.J. and Wright P.E. (2005) Intrinsically unstruc- tured proteins and their functions. Nature Rev Mol Cell Biol 6: 197–208. Both native protein sequences (excluding their periplasm exporting leader sequence, residues 1-20) of PcoE and SilE were aligned, using a general multipurpose alignment program for protein primary sequences – Clustal Omega, which finds the best alignment over the entire length of each sequence submitted (McWilliam et al., 2013). Their percentage identity was then calculated in Jalview (Water- house et al., 2009). The sequences were coloured using the “Percentage Identity” colour-scheme in Jalview too, to clarify the sequence similarities at a more obvious intensity. Greenfield N.J. (2007) Using circular dichroism spectra to estimate protein secondary structure. Nat Protoc 1: 2876–2890. Gupta, A., Matsui, K., Lo, J.F., and Silver, S. (1999) Molec- ular basis for resistance to silver cations in Salmonella. Nat Med 5: 183–188. Hao, X., L€uthje, F. L., Qin, Y., McDevitt, S. F., Lutay, N., Hobman, J. L., Asiani, K., Soncini, F. C., German, N., Zhang, S., Zhu, Y. G., and Rensing, C. (2015) Survival in amoeba–a major selection pressure on the presence of bacterial copper and zinc resistance determinants? Iden- tification of a “copper pathogenicity island”. Appl Microbiol Biotechnol. 99:5817–5824. Based on the SilE sequence, Jpred (a secondary struc- ture prediction server that incorporates the Jnet algorithm to make more accurate predictions) was used to predict a- helices, b-strands and random coil (Cole et al., 2008). Addi- tionally, the intrinsic protein disorder predictor DisEMBL, which utilises the PDB (Protein Data Bank), was used to predict disordered loops in SilE (Linding et al., 2003). Hertela, N. and Hoffman S.V. (2011) ASTRID2: A new Dan- ish low-emittance SR source. Synchrotron Radiat News 24:19–23. Hobman, J.L. and Crossman, L.C. (2015) Bacterial antimi- crobial metal ion resistance. J Med Microbiol 64: 471– 497. Acknowledgements The authors thank Nykola Jones and Søren Vrønning Hoff- mann at the CD Beamline, ADTRID 2, ISA Storage Ring Facilities, Department of Physics at Aarhus University, Scott Young, ICP-MS Analytical Services, School of Biosciences, Sutton Bonington Campus, University of Nottingham. KA was supported by the Biotechnological and Biological Science Research Council (UK) through a studentship award to DJS, JH and PS, and by a BBSRC grant to PS (BB/K021540/1). The OPPF-UK is funded by the Medical Research Council (Grant MR/K018779/1). D.J.S is a Senior Molecular Biology and Neutron Fellow supported by the Science and Technology Facilities Council (UK). Holt, K.B. and Bard, A.J. (2005) Interaction of Silver(I) Ions with the respiratory chain of Escherichia coli: An electro- chemical and scanning electrochemical microscopy study of the antimicrobial mechanism of micromolar Ag1. Bio- chemistry 44: 13214–13223. Horng, J.C., Moroz, V., and Raleigh, D.P. (2003) Rapid cooperative two-state folding of a miniature alpha-beta protein and design of a thermostable variant. J Mol Biol 326: 1261–1270. Jensen, M.R., Ruigrok, R.W.H., and Blackledge, M. (2013) Describing intrinsically disordered proteins at atomic reso- lution by NMR. Current Opin Struct Biol 23: 426–435. Kosol, S., Contreras-Martos, S., Cede~no, C., and Tompa, P. (2013) Structural characterization of intrinsically disor- dered proteins by NMR spectroscopy. Molecules 18: 10802–10828. Nuclear magnetic resonance spectroscopy One-dimensional 1H nuclear magnetic resonance (NMR) spectra were taken at 258C on an 800 MHz Bruker Avance NMR machine, using 500 lM protein in 10 mM sodium phosphate (Na2HPO4/NaH2PO4) buffer, at various pHs ranging from 5-9. The solvent water peak was attenuated using pulsed field gradients or by pre-saturation. Two-dimensional NMR based 15N/1H Heteronuclear Sin- gle Quantum Coherence (HSQC) spectra were acquired at V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 Intrinsically Disordered Periplasmic “Molecular Sponge” 741 Cole, C., Barber, J.D., and Barton G. J., (2008) The Jpred 3 secondary structure prediction server. Nucleic Acid Res 36 (Suppl 2): W197–W201. Cole, C., Barber, J.D., and Barton G. J., (2008) The Jpred 3 secondary structure prediction server. Nucleic Acid Res 36 (Suppl 2): W197–W201. 258C, on an 800 MHz Bruker Avance spectrometer. Data were collected from 0.6 mL samples of 300–350 lM 15N labelled sample of apo- and holo-Ag1-bound SilE in 10 mM HEPES buffer, 20 mM NaF, 10% D2O buffer at pH 7.5. Due to the higher protein concentration required for NMR experiments, Ag1 was dialysed into the holo sample a concentration a final concentration of 500 mM and excess ions were removed through further dialysis using. NMR data was processed and analyzed using Topspin package (Bruker). Czoik, R., Heintz, E. J., and Marczak, W. (2008) complexes of silver with histidine and imidazole and potentiometric methods. Acta Phys Polonica A 114: A51–A56. Dodero, V.I., Quirolo, Z.B., and Sequeira, M.A., (2011) Bio- molecular studies by circular dichroism. Front Biosci 16: 61–73. Du, X. B., Li, H. Y., Wang, X. H., Liu, Q., Ni, J. Z., and Sun, H. Z. (2013) Kinetics and thermodynamics of metal binding to the N-terminus of a human copper transporter, hCTR1. Chem Commun 49: 9134–9136. NOTE ADDED IN PROOF: Since submission of this manuscript a paper on evolution of pco/sil has been published (Staehlin et al., 2016) Kumbhar, S., Johannsen, S., Sigel, R. K. O., Waller, M. P., and Muller, J. (2013) A QM/MM refinement of an experi- mental DNA structure with metal-mediated base pairs. J Inorg Biochem 127: 203–210. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 742 K. R. Asiani et al. Silver, S., Gupta, A., Matsui, K., and Lo, J.F. (1999) Resist- ance to Ag1 cations in bacteria: environments, genes and proteins. Metal-Based Drugs 6: 315–320. Lobley, A., Whitmore, L., and Wallace B.A. (2002). DICHROWEB: an interactive website for the analysis of protein secondary structure from circular dichroism spec- tra. Bioinformatics 18: 211–212. Singh, S.K., Roberts, S.A., McDevitt, S.F., Weichsel, A., Wildner, G.F., Grass, G.B., Rensing, C., and Montfort, W.R. (2011) Crystal structures of multicopper oxidase CueO bound to copper(i) and silver(I): functional role a methionine-rich sequence. J Biol Chem 286: 37849– 37857. Loftin, I.R., Franke, S., Blackburn, N.J., and McEnvoy, M.M. (2007) Unusual Cu(I)/Ag(I) coordination of Secherichia coli CusF as revealed by atomic resolution crystallogra- phy and X-ray absorption spectroscopy. Protein Sci 16: 2287–2293. Mallard, J.Y. and Hartemann, P. (2012) Silver as an antimi- crobial: Facts and gaps in knowledge. Crit Rev Microbiol 39: 373–383. Sreerama, N., Venyaminov, S.Y., and Wood,y R. W. (1999) Estimation of the number of a-helical and b-strand seg- ments in proteins using circular dichroism spectroscopy. Protein Sci 8: 370–380. McHugh, G.L., Moellering, R.C., Hopkins, C.C., and Swartz, M.N. (1975) Salmonella typhimurium resistant to silver nitrate, chloramphenicol, and ampicillin. Lancet 1: 235–240. Sreerama, N. and Woody R.W. (2000) Estimation of protein secondary structure from circular dichroism spectra: Comparison of CONTIN, SELCON, and CDSSTR meth- ods with an expanded reference set. Anal Biochem 287: 252–560. McWilliam, H., Li, W., Uludag, M., Squizzato, S., Park, Y.M., Buso, N., et al. (2013) Analysis Tool Web Services from the EMBL-EBI. Nucleic Acids Res 41(Web Server issue): W597–W600. Staehlin, B. M., Gibbons, J. G., Rokas, A., O’Halloran, T.V., Slot, J. C. (2016) Evolution of a Heavy Metal Homeosta- sis/Resistance Island Reflects Increasing Copper Stress in Enterobacteria. Genome Biol Evol 8: 811–826. Mijnendonckx, K,. Leys, N., Mahillon, J., Silver, S., and Houdt, R.V. (2013) Antimicrobial silver: Uses, toxicity and potential for resistance. BioMetals 26: 1–13. Mirolo, L., Schmidt, T., Eckhardt, S., Meuwly, M., and Fromm, K. M., (2012) pH-Dependent Coordination of AgI Ions by Histidine: Experiment, Theory, and a Model for SilE Chemistry—A Eur J, 19: 1754–1761. Tanaka, T., Mizuno, T., Fukui, S., Hiroaki, H., Oku, J., Kanaori, K., et al. (2004). Two-metal ion, Ni(II) and Cu(II), binding alpha-helical coiled coil peptide. J Am Chem Soc 126; 14023–14028. Muller, J., Bohme, D., Lax, P., Morell Cerdia, M., and Roitzsch, M. (2005) Metal ion coordination to azole nucleosides. Chem Eur J 11: 6246–6253. 742 K. R. Asiani et al. (2012) PcoE-a metal sponge expressed to the periplasm of copper resistance Escherichia coli. Implication of its function role in copper resistance. J Inorg Biochem 115: 186–197. Zimmerman, M., Udagedara, S.R., Sze, C.M., Ryan, T.M., Howlett, G.J., Xiao, Z., et al. (2012) PcoE-a metal sponge expressed to the periplasm of copper resistance Escherichia coli. Implication of its function role in copper resistance. J Inorg Biochem 115: 186–197. Zimmerman, M., Udagedara, S.R., Sze, C.M., Ryan, T.M., Howlett, G.J., Xiao, Z., et al. (2012) PcoE-a metal sponge expressed to the periplasm of copper resistance Escherichia coli. Implication of its function role in copper resistance. J Inorg Biochem 115: 186–197. Scopes, R.K. (1974). Measurement of protein by spectro- photometry at 205 nm. Anal Biochem 59: 277–282. Sibile, N. and Bernado P. (2012) Structural characterization of intrinsically disordered proteins by the combined use of NMR and SAXS. Biochem Soc Trans 40: 955–962. References Atiyeh, B.S., Costagliola, M., Hayek, S.N., and Dibo, S.A. (2007) Effect of silver on burn wound infection control and healing: Review of the literature. Burns 33: 139–148. Lansdown, A.B.G. (2010) A Pharmacological and toxicological profile of silver as an antimicrobial agent in medical devices. Adv Pharmacol Sci 2010 (Article ID 910686): 1–16. Bird, L.E. (2011) High throughput construction and small- scale expression screening of multi-tag vectors in Esche- richia coli. Methods 55: 29–37. Bird, L.E. (2011) High throughput construction and small- scale expression screening of multi-tag vectors in Esche- richia coli. Methods 55: 29–37. Linding R., Jensen L.J., Diella F., Bork P., Gibson T.J., and Russell R.B. (2003) Protein disorder prediction: implica- tions for structural proteomics. Structure 11: 1453–1459. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742 742 K. R. Asiani et al. 742 K. R. Asiani et al. Theisen, A., Johann, C., deacon, A.P., and Harding, S.E. (2000) Refractive increment data-book for polymer and biomolecular scientists. 1st ed. Nottingham University Press, 5–45. Munson, G.P., Lam, D.L., Outten, F.W., O’Halloran, T.V., (2000) Identification of a copper-responsive two-compo- nent system on the chromosome of Escherichia coli K- 12. J Bacteriol 182: 5864–5871. Todd, R.J., Van Dam, M.E., Casimir, o D., Haymor, e B.L., and Arnold, F.H. (1991) Cu(II)-binding properties of a cytochrome c with a synthetic metal-binding site: His-X3- His in an alpha-helix. Proteins, 10; 156–161. Petrovec, K., Ravoo, B. J., and Muller, J. (2012) Coopera- tive formation of silver(I)-mediated base pairs. Chem Commun 48: 11844–11846. Van Stokkum, I.H.M., Spoelder, H.J.W., Bloemendal, M., Van Grondelle, R., and Groen F.C.A. (1990) Estimation of protein secondary structure and error analysis from CD spectra. Anal Biochem 191: 110–118. Provencher, S.W., and Glockner J. (1981) Estimation of globular protein secondary structure from circular dichro- ism Biochemistry 20: 33–37. Waterhouse, A.M., Procter, J.B., Martin, D.M.A., Clamp, M., and Barton G.J. (2009) Jalview Version 2—A multiple sequence alignment editor and analysis workbench. Bio- informatics 25: 1189–1191. Radivojac, P., Iakoucheva L.M., Oldfield C.J., Obradovic, Z., Uversky, V.N., and Dunker A. K., (2007) Intrinsic disorder and functional proteomics. Biophys J, 92: 1439–1456. Randall, C.P., Gupta, A., Jackson, N., Busse, D., and O’Neill, A.J. (2015) Silver resistance in Gram-negative bacteria: a dissection of endogenous and exogenous mechanisms. J Antimicrob Chemother 70: 1037–1046. Zhu, Y. C., Wang, E. Q., Ma, G. L., Kang, Y. B., Zhao, L. H., and Liu, Y. Z. (2014) Interaction of C-terminal metal- binding domain of copper transport protein with Ag1 and Hg21. Acta Phys Chim Sin 30: 1–7. Zimmerman, M., Udagedara, S.R., Sze, C.M., Ryan, T.M., Howlett, G.J., Xiao, Z., et al. (2012) PcoE-a metal sponge expressed to the periplasm of copper resistance Escherichia coli. Implication of its function role in copper resistance. J Inorg Biochem 115: 186–197. Zhu, Y. C., Wang, E. Q., Ma, G. L., Kang, Y. B., Zhao, L. H., and Liu, Y. Z. (2014) Interaction of C-terminal metal- binding domain of copper transport protein with Ag1 and Hg21. Acta Phys Chim Sin 30: 1–7. Rubino, J. T., Riggs-Gelasco, P., and Franz, K. J. (2010) Methionine motifs of copper transport proteins provide general and flexible thioether-onl binding sites for Cu(I) and Ag(I). J Biol Inorg Chem 15: 1033–1049. Zimmerman, M., Udagedara, S.R., Sze, C.M., Ryan, T.M., Howlett, G.J., Xiao, Z., et al. Supporting information Silver, S. (2003) Bacterial silver resistance: molecular biol- ogy and uses and misuses of silver compounds. FEMS Microbiol Rev 27: 341–353. Additional supporting information may be found in the online version of this article at the publisher’s web-site. V C 2016 The Authors. Molecular Microbiology Published by John Wiley & Sons Ltd., Molecular Microbiology, 101, 731–742
https://openalex.org/W3099000198	https://authors.library.caltech.edu/94335/2/aa35259-19.pdf	English	null	The role of dissipative evolution for three-planet, near-resonant extrasolar systems	Astronomy & astrophysics	2,019	cc-by	17,303	1. Introduction require convergent migration for systems such as Trappist-1, a star famously hosting seven planets with period ratios very close to small integer ratios (Gillon et al. 2016, 2017, Luger et al. 2017). Indeed, Gillon et al. (2017) performed N-body integra- tions with the orbital ﬁts as initial conditions and these went unstable over timescales 10 000 times shorter than the estimated age of the system; in contrast, Tamayo et al. (2017) remarked that if an initial condition which results from capture into reso- nance through migration is chosen, then the system is stable over timescales two orders of magnitude longer then the ones found in Gillon et al. (2017). They also note that the addition of tidal eccentricity damping should help maintain the evolution stable over the system’s age. Other good examples of systems neces- sarily sculpted by migration are the four sub-Neptune planets of Kepler-223 (Mills et al. 2016) and the now-classic example of Laplace-like resonance in GJ-876 (Rivera et al. 2010; Batygin et al. 2015). The search for exoplanets in recent years has uncovered a mul- titude of planetary systems, the study of which is the key to an understanding of planetary formation and evolution. Currently, the exoplanet population is dominated by Kepler’s transit detec- tions, making the planetary physical radii and orbital periods the better constrained parameters of the sample. Concerning the ﬁrst aspect, much work has been done recently to understand how photoevaporation sculpts the physical radii of planets (Fulton et al. 2017, and references therein). In this work we address the second, complementary problem of the orbital period distribu- tion. One of the most notable aspects of the Kepler data is that the distribution of the period ratios of neighbouring planets in multi-planets systems shows two seemingly conﬂicting features: on the one hand, it appears relatively broad and smooth, without any single, unmistakably emerging feature; on the other hand, a slight preference for near-resonant conﬁgurations is evident upon close examination. In fact, it is often pointed out that there is a lack of planet pairs in correspondence with period ratios very close to low-integer ratios, and a deﬁnite excess just wide of these values, especially the 2:1 and 3:2, see Fig. 1. ABSTRACT Early dynamical evolution of close-in planetary systems is shaped by an intricate combination of planetary gravitational interactions, orbital migration, and dissipative effects. While the process of convergent orbital migration is expected to routinely yield resonant plan- etary systems, previous analyses have shown that the semi-major axes of initially resonant pairs of planets will gradually diverge under the inﬂuence of long-term energy damping, producing an overabundance of planetary period ratios in slight excess of exact commen- surability. While this feature is clearly evident in the orbital distribution of close-in extrasolar planets, the existing theoretical picture is limited to the speciﬁc case of the planetary three-body problem. In this study, we generalise the framework of dissipative divergence of resonant orbits to multi-resonant chains, and apply our results to the current observational census of well-characterised three-planet systems. Focusing on the 2:1 and 3:2 commensurabilities, we identify three three-planet systems, whose current orbital architecture is consistent with an evolutionary history wherein convergent migration ﬁrst locks the planets into a multi-resonant conﬁguration and subsequent dissipation repels the orbits away from exact commensurability. Nevertheless, we ﬁnd that the architecture of the overall sample of multi-planetary systems is incompatible with this simple scenario, suggesting that additional physical mechanisms must play a dominant role during the early stages of planetary systems’ dynamical evolution. Key words. planets and satellites: dynamical evolution and stability – planets and satellites: formation Open Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4. which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. g p 2 Division of Geological and Planetary Sciences, California Institute of Technology, 1200 E. California Blvd., Pasadena, CA 91125 USA g p Division of Geological and Planetary Sciences, California Institute of Technology, 1200 E. California Blvd., Pasadena CA 91125, USA Received 12 February 2019 / Accepted 13 March 2019 Received 12 February 2019 / Accepted 13 March 2019 Astronomy & Astrophysics Astronomy & Astrophysics Astronomy & Astrophysics A&A 625, A7 (2019) https://doi.org/10.1051/0004-6361/201935259 © G. Pichierri et al. 2019 A&A 625, A7 (2019) https://doi.org/10.1051/0004-6361/201935259 © G. Pichierri et al. 2019 Gabriele Pichierri1, Konstantin Batygin2, and Alessandro Morbidelli1 Gabriele Pichierri1, Konstantin Batygin2, and Alessandro Morbidelli1 Observatoire de la Côte d’Azur, CNRS, Laboratoire Lagrange, Université Côte d’Azur, Nice, France e-mail: gabriele.pichierri@oca.eu e mail: gabriele.pichierri@oca.eu 2 Division of Geological and Planetary Sciences, California Institute of Technology, 1200 E. California Blvd., Pasadena, CA 91125, USA 1. Introduction In any case, the fact that slow convergent orbital trans- port strongly favours resonant captured states is well supported both analytically and numerically, as well as by the speciﬁc observations of multi-planets systems mentioned above. Papaloizou & Terquem (2010) considered the speciﬁc case of the K-dwarf HD 40307, which hosts1 3 hot super-Earths/mini- Neptunes with both pairs wide of the 2:1 mean motion reso- nance, and planetary masses obtained with Radial Velocity. They showed that as tidal interaction between the planets and the star reduces the eccentricities, the system maintains the libration of the resonant angles even when the period ratios are considerably far away from exact commensurability. Subsequently, Batygin & Morbidelli (2013a), Lithwick & Wu (2012) and Delisle & Laskar (2014) showed that two planets in mean motion resonance repel each other as energy is lost during tidal evolution. They thus proposed this as a viable mechanism to explain the observed distribution of period ratios in exoplanetary systems. We note that, for two planets, this repulsion can be easily understood if one considers that any process that dissipates the energy, E ∝−m1/a1 −m2/a2, and at the same time conserves angular momentum, L ∝m1 √a1 + m2 √a2, should give rise to such an evolution. Indeed, this study applies to any dissipative evolu- tion that maintains constant the angular momentum, not just tidal dissipation, and not just resonant coupling (Delisle et al. 2012). In this paper we focus on slow convergent Type-I migration in a disc of gas, and adopt simple synthetic analytical formulæ for the work and the torque generated by the disc on the plan- ets (Cresswell & Nelson 2006, 2008); the requirement that exact prescriptions for the interaction between the planets and the disc be implemented will be relaxed, invoking the aforementioned arguments favouring the plausibility of mean motion resonant capture. A similar reasoning can be applied for the post-disc phase. In order to simulate the dissipative forces that can act on the planetary system, we will implement tidal dissipation. Of course, the tidal parameters for these planets are not known (as we do not yet have a precise understanding of the interior struc- ture of these bodies or the speciﬁc physical mechanisms that dominate the dissipation), which would pose additional ques- tions concerning for example the timescales over which this type of dissipation takes place. 1. Introduction ) In light of the fact that convergent migration should lock planets into mean motion commensurability, it is pertinent to ask how we can explain the lack of planets with exactly reso- nant period ratios and the excess just wide of them. Analytical models of resonance do predict that a pair of planets in a ﬁrst order mean motion resonance need not satisfy the exact reso- nance condition a1/a2 = ((k −1)/k)2/3 (where a1 and a2 are the semi-major axes of the inner and outer planet, respectively, and k is a positive integer), but they can reside wide of resonance while the resonant angles are still librating. This divergence of the resonant equilibrium conﬁgurations happens at vanish- ingly low eccentricities and is linked to a fast precession of the Numerical simulations show that compact chains of mean motion resonances are a common outcome of slow, conver- gent orbital transport of planets within protoplanetary discs. Although details of disc-driven migration remain an active topic of research, it is clear that such a process should play some role in the dynamical history of planetary systems. For example, it is not easy to envision a formation narrative which does not A7, page 1 of 14 A&A 625, A7 (2019) A&A 625, A7 (2019) 2 3 4 5 6 7 8 # planets 0 100 200 300 400 # systems (a) 2:1 3:2 4:3 1.5 2.0 2.5 3.0 3.5 4.0 period ratios 0 10 20 30 40 50 60 # planet pairs (b) Fig. 1. Observations of planet-hosting stars reveal that multi-planetary systems are not rare, hosting over 1600 conﬁrmed planets (panel a). The period ratio distribution of neighbouring planets is shown in panel b. One can observe an overall broad distribution as well as a number of peaks slightly wide of resonant ratios, especially the 2:1 and 3:2 commensurabilities. Data was obtained from the Nasa Exoplanet Archive (https://exoplanetarchive.ipac.caltech.edu/). Panel a: histogram of multi-planetary system by number of planets in each system. Panel b: distribution of period ratios of neighbouring planets in exoplanetary systems. 2 3 4 5 6 7 8 # planets 0 100 200 300 400 # systems (a) 2:1 3:2 4:3 1.5 2.0 2.5 3.0 3.5 4.0 period ratios 0 10 20 30 40 50 60 # planet pairs (b) (b) (a) Fig. 1. 1. Introduction Observations of planet-hosting stars reveal that multi-planetary systems are not rare, hosting over 1600 conﬁrmed planets (panel a). The period ratio distribution of neighbouring planets is shown in panel b. One can observe an overall broad distribution as well as a number of peaks slightly wide of resonant ratios, especially the 2:1 and 3:2 commensurabilities. Data was obtained from the Nasa Exoplanet Archive (https://exoplanetarchive.ipac.caltech.edu/). Panel a: histogram of multi-planetary system by number of planets in each system. Panel b: distribution of period ratios of neighbouring planets in exoplanetary systems. perihelia, which is well understood analytically. However some Kepler systems are so wide of resonance that, after the resonant conﬁguration is attained and the disc of gas is dissipated, an aux- iliary mechanism might need to be invoked which actively drives these planets farther away from the exact resonance. As we see in Sect. 3, observations show that a signiﬁcant fraction of nearly resonant systems lie up to 50 times wider from the resonance than the typical resonant width, and at lower eccentricities than are expected for such planets captured in resonance via migra- tion in protoplanetary discs. These observations can potentially be interpreted as evidence for dissipative processes acting on the planetary systems after the disc phase. three planets are embedded in the protoplanetary disc in which they formed; they interact with the disc, which ultimately results in a resonant capture. Then, after the disc is slowly depleted, the dissipative effects mentioned above are introduced, leading to orbital divergence. Naturally this is a simpliﬁed and idealised scenario. In reality, we still do not know with enough accuracy the ﬁnal con- ﬁguration obtained by multi-systems migrating in a disc of gas. One approach towards a better approximation would be to per- form full hydrodynamic simulations of planets immersed in their protoplanetary disc accounting for various disc parameters (such as disc surface density, turbulence, opacity, etc.). This approach would however be very expensive computationally. Moreover, to date we have virtually no direct observations of the speciﬁc phys- ical processes acting during planet formation and evolution in the early epochs of the disc phase, so these simulations, no matter how exhaustive in terms of the implementation of the plausi- ble physics, cannot yet be directly constrained by the available data. 1 We note that since the publication of the aforementioned paper, more planets have been observed in the same system, including two conﬁrmed planets HD 40307 f and HD 40307 g. For this reason, we will not con- sider this system in the current work, although we draw inspiration from the analysis of Papaloizou & Terquem (2010). 1. Introduction In order to answer this question, we examined the NASA Exoplanet Archive2 and selected conﬁrmed three-planet systems for which both planet pairs lie close to a ﬁrst order mean motion resonance, in particular the 2:1 and 3:2 resonances, as these seem to be the most common in the Kepler data. Our aim is to analyse these systems’ orbital parameters and to evaluate quantitatively how close they are to a multi-resonant chain, which would be indicative of a dynamical history characterised by the physical processes described above. ( y ) For the purposes of this study, we can limit ourselves to an analysis to ﬁrst order in the planetary eccentricities. Indeed, the eccentricities that are expected for planets that have been cap- tured into mean motion resonance by slow convergent migration in a disc are of order √τe/τa ∼h, where τa and τe are the timescales of migration and eccentricity damping respectively, and h = H/r ∼0.05 is the aspect ratio of the disc (Goldreich & Schlichting 2014, Pichierri et al. 2018). Since discs with high aspect ratios are not expected, the limit of small eccentricity is justiﬁed, and even more so in the phase of dissipative tidal evo- lution, which further damps the eccentricities. Moreover, given that these are transiting planets, and that during the disc phase any mutual inclination of the planets would be damped out, we assume coplanar orbits for simplicity. Another useful piece of information which is available to us is the radii of the planets. This could in principle be used to infer the planetary masses (e.g. Wu & Lithwick 2013). However the radius–mass relationship in Kepler planets is marked by extreme scatter (Weiss et al. 2013), and we therefore choose to keep the planetary masses as a free parameter. More speciﬁcally, we are only interested in the mass ratios m1/m2 and m2/m3, since, as we will show, they are the only dynamically signiﬁcant quantities that can affect the values of the semi-major axis ratios (see also Appendix A). p Evidence suggesting that planets around Trappist-1 and Kepler-223 truly reside in a resonant conﬁguration has recently been marshalled from the observed libration of the three-body Laplace angles. 1. Introduction To this end, recall that if two neighbouring pairs of planets in a multi-planet systems are in the k(in):(k(in) −1) and k(out):(k(out) −1) resonances respectively (so that the resonant angles k(in)λ2 −(k(in) −1)λ1 −ϖ2 and k(out)λ3 −(k(out) −1)λ2 −ϖ2 are librating), then the Laplance angle ϕL = (k(in) −1)λ1 −(k(in) + k(out) −1)λ2 + k(out)λ3 will be automatically librating as well. The advantage of examining this three-body angle over the two-body resonant angles is that the latter contain the longitudes of the pericenters ϖ, whose precession rates are poorly constrained by the data, while the former only includes the mean longitudes λ whose derivatives in time are directly deduced by the transit observations. However, solutions for which the resonant angles were originally in libration around a resonant equilibrium point can become circulating when the eccentricity of the equilibrium point becomes small enough under the effect of tidal damp- ing (Delisle et al. 2015), and, similarly, even a small distance from the equilibrium point could be responsible for breaking the libration of the three-body Laplace angle when the equi- librium eccentricity becomes small enough. Therefore, even if such circulations of the angles were observed, this would not be in disagreement with the envisioned scenario of resonant cap- ture and subsequent dissipative evolution. In other words, the libration of the Laplace angle is a sufﬁcient, but not neces- sary condition for past resonant capture in a chain of ﬁrst-order resonances. j pp The remainder of this paper is organised as follows. In Sect. 2 we obtain an analytical model for three planets in a chain of ﬁrst order mean motion resonances, valid in the limit of small eccen- tricities. With this analytical model, we ﬁnd the stable resonant conﬁgurations and map them in terms of the orbital elements. Finally we obtain an analytical conﬁrmation of resonant repul- sion for three-planets systems undergoing dissipation. In Sect. 3 we detail our study, employing both analytical and numerical methods. We select systems of three planets near mean motion resonances, focusing on the 2:1 and 3:2 resonances, and we anal- yse their orbital conﬁguration using the available data in order to evaluate if they are consistent with the process of resonant capture and subsequent dissipative evolution. We present our results in Sect. 4 and we ﬁnally conclude by discussing their signiﬁcance in Sect. 5. 1. Introduction However, the speciﬁc choice of tidal dissipation is only one possible example of a process such that ˙E < 0 and ˙L = 0. We conclude that our speciﬁc implementation of Type-I migration and tidal dissipation after the disc removal is therefore not restrictive, which makes our results generalisable to any other equivalent processes. In the light of these consider- ations, we ask if it is possible to reproduce the observed orbital Thanks to these studies, the case of two planet system is well understood. However the data also contains numerous systems of more than two planets (Fig. 1). Accordingly, in this paper we aim to expand the study to detected extrasolar systems of three planets. More speciﬁcally, we envision the following scenario for the formation and evolution of these planetary systems. First, A7, page 2 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems not constrained observationally) but two observables, that is the two pairs’ semi-major axis ratios. conﬁguration of real exoplanetary systems which reside close to resonance, assuming that planets are captured into resonance and undergo dissipative evolution after the disc phase. In other words we ask if the aforementioned physical processes are compatible with the distribution of near-resonant period ratios that emerges from available data. As detailed below, we carried out our study of ﬁnding orbital conﬁgurations that match the observed data using both an ana- lytical and a numerical approach. The semi-major axes of the planets may be inferred from the orbital periods once the stellar mass is known, however this quantity is not yet well constrained in all cases. Nonetheless, all we will be interested in will be the semi-major axes ratios a2/a1 and a3/a2, which can be obtained without any knowledge of the mass of the star directly from the period ratios and using Kepler’s third law. This is tantamount to renormalising all separations by some arbitrary length, which does not affect the underlying physics since the dynamics only depends on the ratios of the semi-major axes and not on their individual values (only the timescale of the evolution does). 1. Introduction We therefore perform here a different analysis of the observed data, where we do not attempt to verify that a given sys- tem resides formally in resonance at the present day, but instead we evaluate the distance of a system from the considered res- onance chain and the probability that this proximity is due to mere chance. In order to do this, we look for resonant solu- tions that provide the closest match to the observed planetary orbital conﬁgurations, that is the semi-major axis ratios. It is worth anticipating here the following important point. As it will be clear later (see Sect. 3.2), in the case of only two resonant planets residing wide of resonance it is always possible to ﬁnd a resonant conﬁguration which matches the observed data. This is because the eccentricities of these planets are at the present day not well constrained observationally, making the total orbital momentum of the system L a free parameter: it is therefore always possible to ﬁnd a value of L that reproduces the observed a2/a1 with resonance-locked orbits. However, this is not the case for systems of three planets, since we still have only one free parameter L (whose value is linked to the initial captured state, 2 https://exoplanetarchive.ipac.caltech.edu 2. Planetary Hamiltonian The Hamiltonian of two resonant planets in the limit of low eccentricities has been studied extensively in the literature (e.g. Batygin & Morbidelli 2013b, and references therein). Collec- tively these studies have pointed out that even if both planets are massive and to ﬁrst order in eccentricity it is possible to reduce the problem to an Hamiltonian that is analogous to the well-known Hamiltonian of the restricted, circular three-body problem of a massless particle in resonance with a massive A7, page 3 of 14 A7, page 3 of 14 A&A 625, A7 (2019) unperturbed body. In particular, such a Hamiltonian is integrable and is equivalent to the so-called second fundamental model for resonance (Henrard & Lamaitre 1983). This is, however, not the case for three planets. Nonetheless, it is useful to extend an ana- lytical description of the resonant dynamics at low amplitude of libration of the resonant angles in the case of three planets orbit- ing a star. In this section, we introduce the Hamiltonian of the system, derive curves representing the loci of its stable equilib- rium points, and show how these can provide a description of a system along the dissipative evolution. We will apply this model to real Kepler system in Sect. 3.2. planet are not included in Hres as this is a higher order effect. Note also that by dropping the higher order terms the problem is reduced to a planar one. In order to maintain the canonical nature of the equations of motion, we introduce for each planet the modiﬁed Delaunay action-angle variables (Λ, Γ, λ, γ) (omit- ting the subscripts 1,2,3 for simplicity), which are given in terms of the orbital elements by Λ = µ p G(M∗+ m)a ≃m p GM∗a, λ = ℓ+ ϖ, Γ = Λ(1 − √ 1 −e2) ≃Λe2/2, γ = −ϖ, (5) (5) Consider three planets of masses m1, m2 and m3, orbiting around a star of mass M∗in a canonical heliocentric reference frame (Poincare 1892). Indices 1, 2 and 3 will refer to the inner, middle and outer planet, respectively. As usual, we consider the planetary Hamiltonian, which we write as where µ = M∗m/(M∗+ m) is the reduced mass, and ℓ= E − e sin E is the mean anomaly (E being the eccentric anomaly). 2. Planetary Hamiltonian In these variables, the Keplerian part Hkepl of the Hamiltonian (1) takes the form Hkepl = − 3 X i=1 G2(M∗+ mi)2µ3 i 2Λ2 i ≃− 3 X i=1 m3 i 2 GM∗ Λi !2 , (6) (6) H = Hkepl + Hpert, (1) Hkep (1) H = Hkepl + Hpert, where the keplerian part is given by while the resonant Hamiltonian writes while the resonant Hamiltonian writes Hkepl = −GM∗m1 2a1 −GM∗m2 2a2 −GM∗m3 2a3 , (2) (2) Hres = −G2M∗m1m3 2 Λ2 2 (2) Λ2 × f (1,in) res r 2Γ1 Λ1 cos kinλ2 −(kin −1)λ1 + γ1 + f (2,in) res r 2Γ2 Λ2 cos kinλ2 −(kin −1)λ1 + γ2  − G2M∗m2m3 3 Λ2 3 and describes the (integrable) motion of the three planets due to their interaction with the star, to which the small perturba- tion Hpert is added, which includes all the mutual interactions between the planets. We now assume that the inner pair of planets is close to a kin:(kin −1) mean motion resonance, and that the outer pair of planets is close to a kout:(kout −1) mean motion resonance, where kin, kout > 1 are two positive integers. In other words, we assume the resonance conditions n1/n2 ≃ kin/(kin −1), n2/n3 ≃kout/(kout −1), where for each planet n = and describes the (integrable) motion of the three planets due to their interaction with the star, to which the small perturba- tion Hpert is added, which includes all the mutual interactions between the planets. We now assume that the inner pair of planets is close to a kin:(kin −1) mean motion resonance, and that the outer pair of planets is close to a kout:(kout −1) mean motion resonance, where kin, kout > 1 are two positive integers. In other words, we assume the resonance conditions n1/n2 ≃ kin/(kin −1), n2/n3 ≃kout/(kout −1), where for each planet n = p G(M∗+ m)a−3 is the mean motion. Since we are interested in the resonant interaction between the planets only, we will aver- age the Hamiltonian over the fast evolving angles so that only combinations of the resonant angles kinλ2 −(kin −1)λ1 −ϖ1, kinλ2 −(kin −1)λ1 −ϖ2, koutλ3 −(kout −1)λ2 −ϖ2, and koutλ3 − (kout −1)λ2 −ϖ3 remain in the Hamiltonian, where λ is the mean longitude of a planet, and ϖ is its longitude of the periastron. 2. Planetary Hamiltonian × f (1,out) res r 2Γ2 Λ2 cos koutλ3 −(kout −1)λ2 + γ2 + f (2,out) res r 2Γ3 Λ3 cos koutλ3 −(kout −1)λ2 + γ3 , (7) (7) we note that in going from Eqs. (3) to (7) we have made use of the approximation e ≃√2Γ/Λ, which holds at ﬁrst order in e. The resonant perturbing Hamiltonian expanded to ﬁrst order in the eccentricities reads This Hamiltonian is clearly not integrable. However, one can perform a series of changes of variables that allow us to reduce by two the number of degrees of freedom. The ﬁrst canonical transformation is Hres = −Gm1m2 a2 f (1,in) res e1 cos kinλ2 −(kin −1)λ1 −ϖ1 + f (2,in) res e2 cos kinλ2 −(kin −1)λ1 −ϖ2 −Gm2m3 a3 f (1,out) res e2 cos koutλ3 −(kout −1)λ2 −ϖ2 + f (2,out) res e3 cos koutλ3 −(kout −1)λ2 −ϖ3 , (3) K = Λ1 + kin −1 kin Λ2 + (kin −1)(kout −1) kinkout Λ3, κ = λ1, Θ(1) = 1 kin Λ2 + kout −1 kinkout Λ3, θ(1) = kinλ2 −(kin −1)λ1, Θ(2) = 1 kout Λ3, θ(2) = koutλ3 −(kout −1)λ2; (8 (3) (8) it is straightforward to check using the Poisson bracket criterion (Morbidelli 2002) that it is indeed canonical. Now, the new angle κ does not appear explicitly in the Hamiltonian, which makes K a constant of motion. The signiﬁcance of K has already been discussed for two planets (e.g. Michtchenko et al. 2008, Batygin & Morbidelli 2013b), and it has to do with the location of exact resonance. As we have already mentioned, neighbouring planets can still be in resonance while their semi-major axis ratios do not satisfy exactly the resonant condition ai/ai+1 = ((k −1)/k)2/3, therefore the observed ai,obs do not alone reveal how close the planets are to resonance, nor do they represent the nominal ¯ai that do satisfy it. However by calculating from ai,obs the value of the constant of motion K, it is straightforward to check using the Poisson bracket criterion (Morbidelli 2002) that it is indeed canonical. Now, the new angle κ does not appear explicitly in the Hamiltonian, which makes K a constant of motion. The signiﬁcance of K has already been discussed for two planets (e.g. Michtchenko et al. 2.1. Resonant equilibrium points (5) and (12)); the Hamiltonian in these variables reads ¯H = Hkepl + Hres, Hkepl = Hkepl Ψ(1) 1 , Ψ(2) 1 ; K, Ω , Hres = Hres Ψ(1) 1 , Ψ(2) 1 , Ψ(1) 2 , Ψ(2) 2 , ψ(1) 1 , ψ(2) 1 , δγ(1), δγ(2); K, Ω , (13) ¯H = Hkepl + Hres, Hkepl = Hkepl Ψ(1) 1 , Ψ(2) 1 ; K, Ω , Hkepl = Hkepl Ψ(1) 1 , Ψ(2) 1 ; K, Ω , Hres = Hres Ψ(1) 1 , Ψ(2) 1 , Ψ(1) 2 , Ψ(2) 2 , ψ(1) 1 , ψ(2) 1 , δγ(1), δγ(2); K, Ω , Hres = Hres Ψ(1) 1 , Ψ(2) 1 , Ψ(1) 2 , Ψ(2) 2 , ψ(1) 1 , ψ(2) 1 , δγ(1), δγ(2); K, Ω , (13) 1 1 2 2 1 1 (13) (13) where the explicit dependence of each term can be obtained by direct substitution. We now consider the stable equilibria of this system. We look for equilibrium points of this Hamiltonian by simultaneously solving the set of equations ∂¯H ∂Ψ(1) 1 = 0, ∂¯H ∂Ψ(2) 1 = 0, ∂¯H ∂Ψ(1) 2 = 0, ∂¯H ∂Ψ(2) 2 = 0, ∂¯H ∂ψ(1) 1 = 0, ∂¯H ∂ψ(2) 1 = 0, ∂¯H δγ(1) = 0, ∂¯H δγ(2) = 0. (14) Hkepl = − 3 X i=1 G2(M∗+ mi)2µ3 i 2 ×  1 ¯Λ2 i −2 1 ¯Λ3 i δΛi + 3 1 ¯Λ4 i δΛ2 i + O(δΛ3 i ) , (10) (14) (10) We note that by the functional form of the Hamiltonian, any combination of values in {0, π} for the angles immediately sat- isﬁes the last line. These are known as the symmetric equilibria. Asymmetric equilibria are possible (e.g. Beaugé et al. 2006), but they do not play a role at the low eccentricities at which we are limiting ourselves here. which, inserting the deﬁnition of δΛi and dropping the unim- portant constant term and the higher order terms, reduces to: Hkepl = 3 X i=1 4¯niΛi −3 2 ¯h1Λ2 i ! , (11) (11) Plugging in speciﬁc values for the angles in {0, π} reduces the problem of solving the four equations that appear in the ﬁrst line to ﬁnd the stable equilibria of the system. 2.1. Resonant equilibrium points We note that although the Hamiltonian depends on both Ωand K, the latter assumes a natural value for any speciﬁc problem at hand (that is, any values of m1, m2, m3 and of kin, kout) by rescaling the units so that for example ¯a1 = 1. To trace out the loci of the resonant equilibria, we then simply change the value of Ω(which corre- sponds to changing the angular momentum L, at constant K) and solve Eq. (14) to ﬁnd Ψ(1) 1,eq, Ψ(2) 1,eq, Ψ(1) 2,eq, Ψ(2) 2,eq , which are then translated into orbital elements working backwards through the canonical transformations. where ¯ni = p G(M∗+ mi)/¯a3 is the nominal mean motion and ¯hi = ¯ni/ ¯Λi = 1/(µi¯a2 i ) can be interpreted as the inverse of the moment of inertia of a circular orbit around the star. As we will see below, for the purposes of our study, considering the expanded Keplerian Hamiltonian up to order O(δΛ2) does not introduce any signiﬁcant inaccuracy in our calculations. Con- cerning the resonant Hamiltonian (Eq. (7)), we can evaluate it at the nominal values Λ = ¯Λ as it is already of order O(e). where ¯ni = p G(M∗+ mi)/¯a3 is the nominal mean motion and ¯hi = ¯ni/ ¯Λi = 1/(µi¯a2 i ) can be interpreted as the inverse of the moment of inertia of a circular orbit around the star. As we will see below, for the purposes of our study, considering the expanded Keplerian Hamiltonian up to order O(δΛ2) does not introduce any signiﬁcant inaccuracy in our calculations. Con- cerning the resonant Hamiltonian (Eq. (7)), we can evaluate it at the nominal values Λ = ¯Λ as it is already of order O(e). y Finally, one last canonical change of variable is made: Ω= Θ(1) + Θ(2) −(Γ1 + Γ2 + Γ3), θ′ = θ(1), Ψ(1) 1 = Γ1 + Γ2 + Γ3 −Θ(2), ψ(1) 1 = θ(1) + γ1, Ψ(2) 1 = Θ(2), ψ(2) 1 = θ(2) + γ2, Ψ(1) 2 = −Γ2 −Γ3 + Θ(2), δγ(1) = γ1 −γ2, Ψ(2) 2 = −Γ3, δγ(2) = γ2 −γ3. 2. Planetary Hamiltonian 2008, Batygin & Morbidelli 2013b), and it has to do with the l ti f t A h l d ti d where the orbital elements are constructed from heliocentric positions and barycentric velocities (Poincare 1892). The coef- ﬁcients fres are typically of order unity, and it is straightforward to determine the strength of each resonant harmonic, and incor- porate direct and indirect terms. They depend (weakly) on the semi-major axis ratios, and their expressions may be found in Murray & Dermott (2000). We therefore write the Hamiltonian after the averaging procedure as ¯H = Hkepl + Hres + O(e2, I2), (4) ¯H = Hkepl + Hres + O(e2, I2), (4) and then drop the higher order terms. We note that terms that describe the mutual inﬂuence of the innermost and outermost and then drop the higher order terms. We note that terms that describe the mutual inﬂuence of the innermost and outermost A7, page 4 of 14 A7, page 4 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems and imposing in the formula Again, we see that the new angle θ′ does not appear in the Hamiltonian, making Ωanother constant of motion of the system (we note that here Ωdoes not denote the longitude of the node which does not appear in our model, since the problem is planar). We are therefore left with a four-degree-of-freedom Hamiltonian which depends parametrically on the constants of motion K, Ω. We already mentioned the meaning of K; for Ω, one can easily show that K + Ω= (Λ1 −Γ1) + (Λ2 −Γ2) + (Λ3 −Γ3) ≡L, the total angular momentum of the system, which is to be expected knowing that it is a conserved quantity. K Λ3 = µ1 µ3 r M∗+ m1 M∗+ m3 a1 a3 + kin −1 kin µ2 µ3 r M∗+ m2 M∗+ m3 a2 a3 + (kin −1)(kout −1) kinkout , (9) (9) the condition of exact resonance, ai/ai+1 = ((k −1)/k)2/3, for all pairs i = 1, 2, we derive the nominal value of ¯Λ3. From this, one easily obtains ¯a3 from ¯a3 = ( ¯Λ3/m3)2/(GM∗), and then recursively ¯a2 = ((kout −1)/kout)2/3¯a3, and ﬁnally ¯a1 = ((kin −1)/kin)2/3¯a2. 2.1. Resonant equilibrium points (( ) ) It is worth brieﬂy recalling here why even in resonance the planets’ semi-major axes do not coincide exactly with their nom- inal values. As an example, for the inner planet, the condition for resonance is that the resonant angle kinλ2 −(kin −1)λ1 −ϖ1 is librating, 0 ∼kin˙λ2 −(kin −1)˙λ1 −˙ϖ1 = kinn2 −(kin −1)n1 −˙ϖ1, which together with the condition of exact nominal resonance kinn2 −(kin −1)n1 = 0 would imply ˙ϖ1 ∼0; however from the Hamiltonian (Eq. (7)) we have ˙ϖ1 = −˙γ1 ∝Γ−1/2 1 ∼1/e1 which grows as e1 ↘0, meaning that at low eccentricities ˙ϖ1 / 0, which in turn forces kin˙λ2 −(kin −1)˙λ1 = kinn2 −(kin −1)n1 / 0 in order to maintain the libration of the resonant angle. The reso- nant equilibrium points will therefore correspond to semi-major axes ai which may well deviate farther and farther from ¯ai as ei approaches 0. We can however already greatly simplify the calculations given that we will only consider deviations of the semi-major axis ratios from the nominal ones of no more than 5% (moreover, very small values of the eccentricities are obser- vationally disfavoured for Kepler systems, Hadden & Lithwick 2017). In this limit, we can expand the Keplerian part to sec- ond order in δΛi = Λi −¯Λi, where ¯Λi = µ1 √G(M∗+ mi)¯a1 is the nominal resonant value of Λi, and write Let us brieﬂy summarise our work so far. We have obtained a 4-degrees-of-freedom Hamiltonian ¯H which is a func- tion of the actions Ψ(1) 1 , Ψ(2) 1 , Ψ(1) 2 , Ψ(2) 2 and the angles ψ(1) 1 , ψ(2) 1 , δγ(1), δγ(2), and depends parametrically on the values of K and Ω(which are linked to the orbital elements as expressed in Eqs. 2.1. Resonant equilibrium points Assuming that the dissipative evolution is slow com- pared to that of the resonant variables, which has a characteristic timescale given by the libration period at vanishing amplitude of libration, the system will remain bound to the equilibrium curves. Since the eccentricities are damped by the dissipation, we conclude that the semi-major axes are expected to diverge. We also conclude that systems of three planets that are close to a given ﬁrst-order mean motion resonance but for which one or both pairs is narrow of the resonance can only be explained by a resonant conﬁguration if the amplitude of libration around the resonant equilibrium point is large, and temporarily takes the planets to period ratios that are lower than the exact resonant period ratios. Fig. 2. Equilibrium curves showing the loci of the stable resonant equi- librium points calculated as explained in the text, in the case of a 3:2 – 3:2 mean motion resonance chain, with m1 = m2 = m3 = 10−4M∗. The full curves are calculated using the expanded Keplerian Hamiltonian (Eq. (11)), while the dashed curves are calculated using the unexpanded Keplerian Hamiltonian (Eq. (6)), showing very little difference down to very small eccentricities and for reasonable values of the nearly exactly resonant semi-major axis ratio. The location of the nominal resonant semi-major axis ratio (3/2)2/3 is shown by a vertical orange line. We also superimpose the numerically computed evolution of a three-planet system deep in the 3:2 mean motion resonance (for both pairs) and undergoing dissipative evolution depicted with transparent lines: the system follows the locations of the equilibrium points, which are close to the curves calculated analytically. p We ﬁnally note here a property of these curves that will be used later. The Hamiltonian (Eq. (13)) can be rescaled by a parameter which encapsulates all of the information regarding how the dynamics scales with mass ratios and physical sizes of the orbits. This is analogous to the rescaling found in Batygin & Morbidelli (2013b) for the 2-planets case, and only works when using an expanded Hamiltonian and for semi-major axes close to the nominal resonant ones, which are our working assump- tions anyway. Therefore, after rescaling all planetary masses by a certain factor ˜m, the corresponding loci of the resonant equilib- ria are also simply rescaled, and can be immediately calculated. 4 At higher eccentricities the main term which might shift the equi- libria in a2/a1, a3/a2 to the left of exact resonance is the second order (secular) term which describes the interaction between the inner planet and outer planet; however, we checked that adding this term to the Hamiltonian, even at high eccentricities and for a very massive outer planet the picture does not change. 2.1. Resonant equilibrium points Ω= Θ(1) + Θ(2) −(Γ1 + Γ2 + Γ3), θ′ = θ(1), Ψ(1) 1 = Γ1 + Γ2 + Γ3 −Θ(2), ψ(1) 1 = θ(1) + γ1, Ψ(2) 1 = Θ(2), ψ(2) 1 = θ(2) + γ2, Ψ(1) 2 = −Γ2 −Γ3 + Θ(2), δγ(1) = γ1 −γ2, Ψ(2) 2 = −Γ3, δγ(2) = γ2 −γ3. We show in Fig. 2 one example of equilibrium curves for three equal-mass planets down to eccentricities of order 10−3, where we also show that the expanded Keplerian Hamiltonian provides an accurate description of the system. These curves are then matched against the result of full N-body numerical simula- tions of a system with the same physical parameters which starts deep in resonance and evolves dissipatively so to slowly follow (12) A7, page 5 of 14 A7, page 5 of 14 A7, page 5 of 14 A7, page 5 of 14 A&A 625, A7 (2019) a2/a1 vs. e1, (δΛ2) a3/a2 vs. e2, (δΛ2) a2/a1 vs. e1, no exp. a3/a2 vs. e2, no exp. 1.32 1.34 1.36 1.38 ai+1/ai 0.001 0.010 0.100 1 ei a2/a1 vs. e1, (δΛ2) a3/a2 vs. e2, (δΛ2) a2/a1 vs. e1, no exp. a3/a2 vs. e2, no exp. 1.32 1.34 1.36 1.38 ai+1/ai 0.001 0.010 0.100 1 ei Fig. 2. Equilibrium curves showing the loci of the stable resonant equi- librium points calculated as explained in the text, in the case of a 3:2 – 3:2 mean motion resonance chain, with m1 = m2 = m3 = 10−4M∗. The full curves are calculated using the expanded Keplerian Hamiltonian (Eq. (11)), while the dashed curves are calculated using the unexpanded Keplerian Hamiltonian (Eq. (6)), showing very little difference down to very small eccentricities and for reasonable values of the nearly exactly resonant semi-major axis ratio. The location of the nominal resonant semi-major axis ratio (3/2)2/3 is shown by a vertical orange line. We also superimpose the numerically computed evolution of a three-planet system deep in the 3:2 mean motion resonance (for both pairs) and undergoing dissipative evolution depicted with transparent lines: the system follows the locations of the equilibrium points, which are close to the curves calculated analytically. where the analytically computed resonant equilibria agree very well with our numerical simulations4. Consider now a resonant three-planet system that is close to some resonant equilibrium point and is subjected to (tidal) dissipation. 3 We note that even away from nominal resonance all four resonant angles can continue to librate when the system is sufﬁciently close to the resonant equilibrium point, unlike what has been erroneously stated in Sect. 4 of Batygin & Morbidelli (2013a). 2.2. Resonant repulsion for three-planets systems The equilibrium curves in the ai+1/ai vs. ei plane show that the resonant repulsion during energy dissipation is expected also for three-planets systems. For systems of two planets, it is well known that for ﬁrst order resonances the resonant equilibria always reside wide of the exact resonant ratio of the semi- major axes. That is, because the resonant condition requires k˙λ2 −(k −1)˙λ1 −˙ϖ1,2 ≃0 and since the perihelion precession is always retrograde, ˙ϖ1,2 < 0, one necessarily has k˙λ2 −(k − 1)˙λ1 < 0, that is a2/a1 > (k/(k −1))2/3. More concretely, at low enough eccentricities and at semi-major axis ratios close to the nominal ones, one ﬁnds directly using the resonant Hamilto- nian expanded to ﬁrst order in e that ˙ϖ1 = αf (1) res n1(m2/M∗)e−1 1 , ˙ϖ2 = −f (2) res n2(m1/M∗)e−1 2 with f (1) res < 0, f (2) res > 0: this means that the lower are the eccentricities, the wider are the equilibria from the exact commensurability. At higher eccentricities the secular terms, of O(e2), become more important, and they contribute a positive contribution (that is constant in e) in ˙ϖ; however, one still ﬁnds ˙ϖ1,2 < 0 at higher eccentricities as well (e.g. Pichierri et al. 2018). 2.1. Resonant equilibrium points More speciﬁcally, one can easily see that for given semi-major axis ratios, the values of the eccentricities that correspond to the resonant equilibrium point are just rescaled by ˜m, since the eccentricities and the planetary masses appear as a prod- uct in the perturbing Hamiltonian. This can be easily understood using the previous formula ˙ϖ ∝(mpl/M∗)e−1, and noticing that ﬁxing the semi-major axis ratio ultimately ﬁxes ˙ϖ by the res- onance condition; therefore, rescaling the planetary masses, at ﬁxed ˙ϖ1 = ˙ϖ2 = ˙ϖ3 (i.e. at constant semi-major axis ratios), the eccentricities are simply rescaled by the same factor. This implies that for a given equilibrium conﬁguration of the semi- major axis ratio a2,eq/a1,eq, the corresponding equilibrium of the ratio a3,eq/a2,eq will be independent of ˜m, that is independent of the absolute value of the planetary masses. Only the ratios m1/m2 and m2/m3 are signiﬁcant, meaning that if one changes one of these ratios, the equilibrium in a3,eq/a2,eq corresponding to the same a2,eq/a1,eq will have changed (see Appendix A for an explicit presentation of this rescaling procedure). the resonant equilibrium points (transparent lines)3. These N-body integrations with the addition of dissipative effects will be detailed below in Sect. 3.3. the resonant equilibrium points (transparent lines)3. These N-body integrations with the addition of dissipative effects will be detailed below in Sect. 3.3. 3.1. Choice of systems Recall that the only orbital parameters that are well constrained by transit data are the orbital periods, which allow us to obtain the semi-major axis ratios even without knowing the mass of the star. The orbital periods listed in the NASA Exoplanet Archive catalogue are, for the cases considered below, obtained by fase folding the observed signal. Since we will be considering short- period planets this is equivalent to obtaining the proper value of the periods, so that we can directly compare the corresponding observed semi-major axis ratios with the ones coming from our averaged model of resonance5. The masses of the planets could be obtained starting from the estimates for the planetary radii, and making use of a scaling relation such as the one found in Wu & Lithwick (2013), mpl = 3m⊕(Rpl/R⊕), where mpl, Rpl are the mass and radius of the exoplanet, and R⊕, m⊕those of the Earth. However this is only a statistical law and the uncertainties on the mean densities usually preclude accurate estimates for the masses, which are indeed not yet known. We will therefore use the masses as free parameters of our study. In fact, as we already saw, the only signiﬁcant quantities for our study are the mass ratios between the planet; this follows from the discussion at the ∆k/(k−1) := k −1 k P2 P1 −1, (15) (15) called the normalised distance from (exact) resonance. When ∆> 0 the planets reside wide of the k : k −1 resonance, while when ∆< 0 the planets are narrow of the resonance. ﬁxes the eccentricities of all three planets since they are all linked by the p We will be selecting planetary systems of three planets with both pairs close to some ﬁrst order mean motion resonance such that \|∆\| ≲0.05 holds for both pairs, with k = 2, 3 as they appear to be the most common resonances. We recall that the normalised width of a resonance is of order ∼(m/M∗)2/3 (Deck et al. 2013; Batygin 2015), where the average planetary mass for Kepler systems is of order m ∼3 × 10−5M∗, and moreover most planets in each system appear to be quite homogeneous in mass (Weiss et al. 2018; Millholland et al. 2017). 3. A scenario for dissipative evolution of three-planet systems We note a clear peak to the right of the value ∆= 0 (corresponding to the exact nominal resonance, indicated by a vertical orange line), which is most prominent for 0 < ∆≲0.05. Out of the 358 pairs plotted here, there are 123 total pairs in this conﬁguration. For the three-planets systems only, 48 pairs have 0 < ∆≲0.05, out of the 121 shown in the purple histogram. 0.00 Fig. 3. Distribution of the (signed) normalised distance from ﬁrst order mean motion resonances ∆= k−1 k P2 P1 −1 with k = 2, 3 in all exoplanetary systems (yellow) and for three-planet systems (purple). We note a clear peak to the right of the value ∆= 0 (corresponding to the exact nominal resonance, indicated by a vertical orange line), which is most prominent for 0 < ∆≲0.05. Out of the 358 pairs plotted here, there are 123 total pairs in this conﬁguration. For the three-planets systems only, 48 pairs have 0 < ∆≲0.05, out of the 121 shown in the purple histogram. end of Sect. 2.2. In practice, we will choose a total planetary mass mtot = m1 + m2 + m3 by using for each system the mean planetary radius (R1 + R2 + R3)/3, the aforementioned scaling relationship from Wu & Lithwick (2013) to obtain an average planetary mass mpl,avg, and setting mtot = 3mpl,avg. Again, this is simply a choice that we are forced to make in order to run N-body simulations, but it does not in any way change our result, which is therefore not sensitive to the uncertainties on the radii (or to the lack of their knowledge, as will be the case for YZ Ceti). Note however that since individual Kepler systems seem to show a homogeneity in planetary radii and masses (Weiss et al. 2018; Millholland et al. 2017), this choice likely constitutes a good approximation to the real architecture of these systems. Following the depletion of the gas, dissipation is introduced which removes orbital energy at constant angular momentum; this is done here implementing tidal dissipation but again the method is general. During this phase of dissipative evolution, the planets will follow again the equilibrium curves of the resonant Hamiltonian for changing Ω, this time decreasing their eccen- tricities and hence increasing the semi major axis ratios ai+1/ai for each planet–planet pair. 3. A scenario for dissipative evolution of three-planet systems The planets keep approaching each other and their eccentricities keep increasing due to the curvature of the locus of the equilibrium points in the (ai+1/ai, ei) diagram, until an equilibrium is reached when the damping of e balances such resonant eccentricity pumping. This is why planets are expected to be (close to) a resonant equilibrium point in the ﬁrst place, and a chain of resonances can be formed. The disc is then slowly depleted and the planets maintain their conﬁguration. inclinations), and an exchange in angular momentum between the planets and the disc which causes the planets’ semi-major axes to change (planetary migration). Planets captured in mean motion resonance are usually attributed to inward migration (that is, the planet looses orbital angular momentum to the disc), and we will consider this case in this paper, but these results are general to any form of convergent evolution. In our implemen- tation, when the inner planet reaches the inner edge of the disc, migration is stopped, and the second incoming planet will cross a commensurability with the ﬁrst; ﬁnally, the third planet will cross a commensurability with the second. We note that the timescale over which the eccentricity is damped is usually of order ∼100 times shorter than that over which the semi major axis changes. Planets are therefore expected to approach these commensurabilities with vanishingly low eccentricity. As we can see from Fig. 2, however, this conﬁguration with semi-major axis ratio wide of resonance and low eccentricity is very close to the resonant equilibrium points. The planets keep approaching each other and their eccentricities keep increasing due to the curvature of the locus of the equilibrium points in the (ai+1/ai, ei) diagram, until an equilibrium is reached when the damping of e balances such resonant eccentricity pumping. This is why planets are expected to be (close to) a resonant equilibrium point in the ﬁrst place, and a chain of resonances can be formed. The disc is then slowly depleted and the planets maintain their conﬁguration. -0.05 0.00 0.05 0.10Δ 0 10 20 30 40 50 # of pairs -0.05 0.00 0.05 0.10Δ 0 10 20 30 40 50 # of pairs Fig. 3. Distribution of the (signed) normalised distance from ﬁrst order mean motion resonances ∆= k−1 k P2 P1 −1 with k = 2, 3 in all exoplanetary systems (yellow) and for three-planet systems (purple). 5 We should remark however that even the periods are not known with arbitrary precision, meaning that there might be small discrepancies in the values that are used in different works. In this paper, we will use the ones listed in the NASA Exoplanet Archive catalogue without considering error bars. This is enough for the scope of our analysis. 3. A scenario for dissipative evolution of three-planet systems We note that Ωchanges because K changes (since do the semi-major axes as a consequence of the dissipation of energy) and L has to stay constant for this kind of dissipation. Given a pair of neighbouring planets with periods P1 and P2 respectively, which are close to a given ﬁrst order resonance, P1/P2 ≃(k −1)/k, one can deﬁne (see for example Lithwick et al. 2012; Hadden & Lithwick 2014) 3. A scenario for dissipative evolution of three-planet systems In this section, we select near-resonant systems of 3 plan- ets from the NASA Exoplanet Archive catalogue, and discuss whether or not their observed orbital conﬁguration is compatible with the dynamical evolution driven by the following physical processes. As we already mentioned, planets are expected to dynamically interact with the protoplanetary disc in which they formed. This has two effects: a damping of the eccentricities (and For systems with three planets, since we used a ﬁrst order expansion in e in the analytical model and therefore we are not including the mutual interaction between the inner planet and the outer planet, the perihelion precession will still be retrograde and it will remain true that for each pair of planets the resonant equi- libria lie wide of the exact nominal resonance, and that, in the limit of small enough eccentricities, the separations grow with diminishing eccentricities. This is indeed what we see in Fig. 2, A7, page 6 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar system inclinations), and an exchange in angular momentum between the planets and the disc which causes the planets’ semi-major axes to change (planetary migration). Planets captured in mean motion resonance are usually attributed to inward migration (that is, the planet looses orbital angular momentum to the disc), and we will consider this case in this paper, but these results are general to any form of convergent evolution. In our implemen- tation, when the inner planet reaches the inner edge of the disc, migration is stopped, and the second incoming planet will cross a commensurability with the ﬁrst; ﬁnally, the third planet will cross a commensurability with the second. We note that the timescale over which the eccentricity is damped is usually of order ∼100 times shorter than that over which the semi major axis changes. Planets are therefore expected to approach these commensurabilities with vanishingly low eccentricity. As we can see from Fig. 2, however, this conﬁguration with semi-major axis ratio wide of resonance and low eccentricity is very close to the resonant equilibrium points. 3.1. Choice of systems This gives a typical resonance width of order ∆∼10−3 in normalised units, meaning that in selecting systems with 0 < ∆≲0.05 we are generously including planetary pairs with separation 50 times larger than the typical resonant width. Moreover, the available data shows that for systems close to mean motion resonance, the distribution of ∆favours values between 0 and ≲0.05 (Hadden & Lithwick (2017), see also Fig. 3). Additionally, we will require that ∆> 0, which is justiﬁed by our results in the previous section. eccentricities of all three planets, since they are all linked by the equilibrium curves. Of all three-planet systems, only 8 satisfy \|∆\| < 0.05 for both pairs, that is, appear to be close to a multi-resonant chain (they A7, page 7 of 14 A&A 625, A7 (2019) that (a3/a1)\|eq ≡(a3/a1)\|obs automatically ﬁxes the equilibrium values of both semi-major axis ratios a2/a1 and a3/a2. Consid- ering for example the corresponding equilibrium (a3/a2)\|eq, we obtain the weighted difference 101 102 Kepler-289 Kepler-31 Kepler-53 Kepler-305 Kepler-114 Kepler-326 YZ Cet Kepler-207 R⊕ 2:1 2:1 −4.4 × 10−2 −4.7 × 10−2 2:1 2:1 2.2 × 10−2 2.8 × 10−2 2:1 2:1 −4.4 × 10−2 3.4 × 10−2 2:1 3:2 9.4 × 10−3 7.4 × 10−3 3:2 3:2 3.3 × 10−2−2.4 × 10−2 2:1 3:2 1.9 × 10−2 −1.5 × 10−2 3:2 3:2 3.6 × 10−2 1.4 × 10−2 2:1 2:1 −4.7 × 10−2 −4.5 × 10−2 period (days) ¯δ(a3/a2) := ((a3/a2)\|eq −(a3/a2)\|obs) (a3/a2)\|obs (16) (16) between (a3/a2)\|eq and the observed value (a3/a2)\|obs. The same can be done for a2/a1, which gives a similar (absolute) result, \|¯δ(a2/a1)\| ≃\|¯δ(a3/a2)\|. Maintaining this procedure, we loop over different planetary mass ratios. p y We applied this procedure to the three systems selected above, starting with Kepler-305, which resides close to a 3:2– 2:1 mean motion resonance chain. First of all, to better represent what these analytical maps intend to show, we draw in Fig. 5 equilibrium curves (equivalent to those shown in Fig. 2) which describe the locations of the resonant equilibria for this reso- nant chain and for one speciﬁc choice of mass ratios m1/m2 = m2/m3 = 1. The observed values of the semi-major axis ratios are indicated by dashed vertical lines; then, we indicate with a red dot the location of the speciﬁc equilibrium point that is selected when we impose (a3/a1)\|eq ≡(a3/a1)\|obs; ﬁnally, we obtain ¯δ(a3/a2) as deﬁned in Eq. 3.1. Choice of systems For each system, we place a circle in correspondence of the period of each planet, and indi- cate between pairs of planets the ﬁrst order mean motion resonance in which they are envisioned to reside (below) and the normalised distance to that resonance ∆(above); the sign of ∆indicates if the pair of planets are narrow (∆< 0) or wide (∆> 0) of the resonance. For our analy- sis, we will only consider systems for which ∆> 0 for both pairs (the systems enclosed by a box). The size of the circle is an indication of the estimated radius of the planet (the small dot in the top right cor- ner demonstrates the size of Earth). For YZ Ceti this information is not available, but this does not pose a problem for our study. are Kepler-31, Kepler-53, Kepler-114, Kepler-207, Kepler-289, Kepler-305, Kepler-326, YZ Ceti). The architecture of these sys- tems is shown in Fig. 4; of these, only 3 satisfy ∆> 0 for both pairs. These are Kepler-31, Kepler-305 and YZ Ceti. For these systems, we consider whether or not their observed orbital conﬁguration can be consistent with the scenario envisioned above. 3.2. Analytical maps With our analytical model of resonance at hand, we can construct analytical maps of resonant equilibrium points for different reso- nant chains and different planetary mass ratios. For the purposes of the current study, we proceed as follows. For an arbitrary sys- tem, we assume to have observations for the orbital period ratios and obtain the values of k(in) and k(out). We then pick both mass ratios m2/m1 and m3/m2 and construct the equilibrium curves as explained in Sect. 2.1 (in practice, we work with the afore- mentioned Hamiltonian rescaled by the common planetary mass factor ˜m, see Appendix A). Then, we ﬁnd the resonant equilib- rium point (i.e. the value of L) that corresponds to a value of a3/a1 equal to the observed semi-major axis ratio, and therefore put (a3/a1)\|eq ≡(a3/a1)\|obs. The determined value of L ﬁxes the eccentricities of all three planets, since they are all linked by the equilibrium curves. 3.1. Choice of systems (16) (and similarly for ¯δ(a2/a1)). As explained above, imposing (a3/a1)\|eq ≡(a3/a1)\|obs automatically ﬁxes all equilibrium eccentricities e1,eq, e2,eq, e3,eq, and we can store their maximum max{e1,eq, e2,eq, e3,eq} in to bet- ter describe the orbital conﬁguration at the selected equilibrium point. We choose to consider the quantity max eeq rescaled by a common planetary mass factor ˜m in order to obtain results that are independent of the planet-to-star mass ratio, since again the latter quantity does not effect equilibrium semi-major axis ratio conﬁgurations, and therefore does not affect ¯δ(a3/a2). Panels a and b in Fig. 6 are the result of this procedure spanning different planetary mass ratios, showing maps of ¯δ(a2/a1) and max e/ ˜m using the observed semi-major axis ratios of Kepler-305 (the bottom panels c and d show the results of numerical simulations which will be detailed in Sect. 3.3, and are only intended to val- idate the analytical results). We show analogous results for the system YZ Cet, residing close to a 3:2 – 3:2 chain, in Fig. 7, and for For Kepler-31, a chain (close to the) 2:1 – 2:1 mean motion resonances, in Fig. 8. period (days) Fig. 4. Orrery of the three-planets systems sufﬁciently close to ﬁrst order mean motion resonances k:k −1 with k = 2, 3, with a normalised distance to resonance \|∆\| < 0.05 for both pairs. For each system, we place a circle in correspondence of the period of each planet, and indi- cate between pairs of planets the ﬁrst order mean motion resonance in which they are envisioned to reside (below) and the normalised distance to that resonance ∆(above); the sign of ∆indicates if the pair of planets are narrow (∆< 0) or wide (∆> 0) of the resonance. For our analy- sis, we will only consider systems for which ∆> 0 for both pairs (the systems enclosed by a box). The size of the circle is an indication of the estimated radius of the planet (the small dot in the top right cor- ner demonstrates the size of Earth). For YZ Ceti this information is not available, but this does not pose a problem for our study. Fig. 4. Orrery of the three-planets systems sufﬁciently close to ﬁrst order mean motion resonances k:k −1 with k = 2, 3, with a normalised distance to resonance \|∆\| < 0.05 for both pairs. G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems (a2/a1)obs 1.315 1.320 1.325 1.330a2/a1 50 100 500 1000 5000 e1/m (a3/a2)obs 1.590 1.595 1.600 1.605 1.610 1.615 a3/a2 50 100 500 1000 5000 e2/m Fig. 5. Locations of the resonant equilibrium points in the ai+1/ai vs. ei/ ˜m planes, i = 1, 2, for three equal-mass planets in a 3:2 – 2:1 mean motion resonance chain, close to which Kepler-305 resides. Orange vertical lines show the exact nominal commensurability, while dashed vertical lines show the observed a2/a1 and a3/a2 in the case of Kepler-305. As explained in the text, we select one equilibrium conﬁguration (indicated by the red dot in both panels) by requiring that (a3/a1)\|eq ≡(a3/a1)\|obs, which automatically ﬁxes all orbital elements a2/a1, a3/a2, e1, e2 and e3 along the equilibrium curves. (a2/a1)obs 1.315 1.320 1.325 1.330a2/a1 50 100 500 1000 5000 e1/m (a3/a2)obs 1.590 1.595 1.600 1.605 1.610 1.615 a3/a2 50 100 500 1000 5000 e2/m Fig. 5. Locations of the resonant equilibrium points in the ai+1/ai vs. ei/ ˜m planes, i = 1, 2, for three equal-mass planets in a 3:2 – 2:1 mean motion resonance chain, close to which Kepler-305 resides. Orange vertical lines show the exact nominal commensurability, while dashed vertical lines show the observed a2/a1 and a3/a2 in the case of Kepler-305. As explained in the text, we select one equilibrium conﬁguration (indicated by the red dot in both panels) by requiring that (a3/a1)\|eq ≡(a3/a1)\|obs, which automatically ﬁxes all orbital elements a2/a1, a3/a2, e1, e2 and e3 along the equilibrium curves 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.00208725 0.00208794 0.00208863 0.00208932 0.00209001 0.00209070 0.00209139 0.00209208 0.00209277 0.00209346 (a) 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 164 205 246 287 328 369 410 451 492 533 (b) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 (a3/a2eq-a3/a2obs) a3/a2obs 0.001768 0.001820 0.001872 0.001924 0.001976 0.002028 0.002080 0.002132 0.002184 0.002236 (c) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 Max (e) m 164 205 246 287 328 369 410 451 492 533 (d) ig. 6. Top row: analytical maps constructed for Kepler-305 as explained in Sect. 3.2. G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems Panel a: we plot the quantity ¯δ(a3/a2), which represents ow close the system is now to some resonant equilibrium point, for different mass ratios m1/m2 and m2/m3 (each point in this plot is constructed y repeating the procedure described in Fig. 5). We notice that ¯δ(a3/a2) changes very little with the mass ratios, and is of the order of ∼0.002. omparing with Fig. 9, we see that this can be the case by pure chance only in ∼15% of randomly generated systems close to the 3:2–2:1 mean otion resonance chain. Panel b: we show a map of the quantity max eeq/ ˜m representing the equilibrium orbital conﬁguration that is selected at ach ﬁxed value of m1/m2 and m2/m3 by imposing (a3/a1)\|eq ≡(a3/a1)\|obs. Bottom row: numerical maps constructed for Kepler-305 as explained in ect. 3.3. We show numerical maps of ¯δ(a3/a2) in panel c and max eeq/ ˜m in panel d, analogous to the analytical plots above (over a slightly smaller nge of mass ratios for simplicity). These are intended to validate the analytical maps, and show very good agreement between corresponding anels. 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 164 205 246 287 328 369 410 451 492 533 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m /m m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.00208725 0.00208794 0.00208863 0.00208932 0.00209001 0.00209070 0.00209139 0.00209208 0.00209277 0.00209346 m1/m2 (a) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 (c) m1/m2 (b) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 Max (e) m 164 205 246 287 328 369 410 451 492 533 (d) (a) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 (a3/a2eq-a3/a2obs) a3/a2obs 0.001768 0.001820 0.001872 0.001924 0.001976 0.002028 0.002080 0.002132 0.002184 0.002236 (c) (d) (c) Fig. 6. Top row: analytical maps constructed for Kepler-305 as explained in Sect. 3.2. Panel a: we plot the quantity ¯δ(a3/a2), which represents how close the system is now to some resonant equilibrium point, for different mass ratios m1/m2 and m2/m3 (each point in this plot is constructed by repeating the procedure described in Fig. 5). We notice that ¯δ(a3/a2) changes very little with the mass ratios, and is of the order of ∼0.002. Comparing with Fig. 3.3. Numerical simulations In order to check the validity of our analytical calculations, we turned to numerical simulations by performing the following study. We simulated the process of capture into a chain of ﬁrst- order mean motion resonances by placing the planets relatively wide of the desired resonances, according to the speciﬁc val- ues of kin and kout of each case, and simulating the effects of the protoplanetary discs by adding ﬁctitious forces which mimic the interaction with the disc (Cresswell & Nelson 2006, 2008) to the N-body integrator swift_symba. To ensure convergent migration for all planetary mass ratios, we stopped the migration of the inner planet by adding at the desired location a so-called planetary trap, which reproduces the effect of the inner edge of the disc and describes a disc cavity around a star (Masset et al. 2006, Pichierri et al. 2018). As we mentioned above, the mass ratios were kept as free parameters. Since again we are interested in mass ratios of order unity, in our simulations we limited ourselves to m1/m2 and m2/m3 between ∼0.2 and ∼5, and repeated the same set of simulations. Recall that we only have one free parameter to select the cho- sen equilibrium conﬁguration: the angular momentum; however, we have two observables that we want to match, which are both semi-major axis ratios a2/a1 and a3/a2. This is unlike the case of only two resonant planets, where one has one free parame- ter (again the angular momentum L) and only one observable (the single a2/a1 ratio): in this case, it would always be possible to ﬁnd a suiting value of L which gives a resonant equilibrium conﬁguration such that the semi-major axis ratio is equal to the observed one (provided that the latter is wide of the nominal value (k/(k−1))2/3). In the case of three planets, choosing L such Recall that the timescale for planetary migration depends on the mass of the planet, meaning that changing the mass ratios A7, page 8 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar system G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems 9, we see that this can be the case by pure chance only in ∼15% of randomly generated systems close to the 3:2–2:1 mean motion resonance chain. Panel b: we show a map of the quantity max eeq/ ˜m representing the equilibrium orbital conﬁguration that is selected at each ﬁxed value of m1/m2 and m2/m3 by imposing (a3/a1)\|eq ≡(a3/a1)\|obs. Bottom row: numerical maps constructed for Kepler-305 as explained in Sect. 3.3. We show numerical maps of ¯δ(a3/a2) in panel c and max eeq/ ˜m in panel d, analogous to the analytical plots above (over a slightly smaller range of mass ratios for simplicity). These are intended to validate the analytical maps, and show very good agreement between corresponding panels. A7, page 9 of 14 A7, page 9 of 14 A&A 625, A7 (2019) 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.0103920 0.0103935 0.0103950 0.0103965 0.0103980 0.0103995 0.0104010 0.0104025 0.0104040 0.0104055 (a) 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 49.7 56.8 63.9 71.0 78.1 85.2 92.3 99.4 106.5 (b) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 (a3/a2eq-a3/a2obs) a3/a2obs 0.010032 0.010120 0.010208 0.010296 0.010384 0.010472 0.010560 0.010648 0.010736 0.010824 (c) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 Max (e) m 56.8 63.9 71.0 78.1 85.2 92.3 99.4 106.5 (d) Fig. 7. Same as in Fig. 6, but for the system YZ Cet, residing close to a 3:2–3:2 chain. The value of ¯δ(a3/a2) ∼0.01 across all planetary mass ratios an be matched against the corresponding curve in Fig. 9, where we ﬁnd that ∼80% of randomly generated systems lie this close to the 3:2–3:2 hain. 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 49.7 56.8 63.9 71.0 78.1 85.2 92.3 99.4 106.5 (b) 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.0103920 0.0103935 0.0103950 0.0103965 0.0103980 0.0103995 0.0104010 0.0104025 0.0104040 0.0104055 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 (a3/a2eq-a3/a2obs) a3/a2obs 0.010032 0.010120 0.010208 0.010296 0.010384 0.010472 0.010560 0.010648 0.010736 0.010824 (c) 0.2 0.5 1 2 5 0.2 0.5 1 2 5 m2/m1 m3/m2 Max (e) m 56.8 63.9 71.0 78.1 85.2 92.3 99.4 106.5 (d) (d) (c) Fig. 7. Same as in Fig. 4.1. Probabilistic measure of a resonant conﬁguration in Kepler-305, YZ Cet and Kepler-31 Using the maps of ¯δ(a3/a2) shown in Figs. 6–8 for the three selected systems Kepler-305, YZ Cet and Kepler-31, we draw the following conclusions. First of all, one might expect that the quantity ¯δ(a3/a2) should change with the different choices of mass ratios, thus allowing one to make a prediction on their (so far unknown) values of m1/m2 and m2/m3 under the assump- tion that these systems are indeed in resonance and evolving dissipatively. Follow-up monitoring of these systems could then produce new observations from which to obtain the real masses of the planets, and so validate or disprove the hypothesis. How- ever, in practice we ﬁnd that these analytical maps show very little dependence on m1/m2 and m2/m3 spanning reasonable values. Note also that for all three systems ¯δ(a3/a2) is small, but never vanishing, which would represent an analytically com- puted equilibrium conﬁguration such that ¯δ(a3/a2) = 0, that is, a resonant equilibrium point which satisﬁes (a3/a2)\|eq = (a3/a2)\|obs and (a2/a1)\|eq = (a2/a1)\|obs. But even if this hap- pened to be the case, the level curve ¯δ(a3/a2) = 0 would still span a broad range of mass ratios: given moreover the uncertainty in the observed period ratios of exoplanetary systems, this would make any determination of m2/m1 or m3/m2 using observed data, in general, inconclusive. Consider a choice of the mass ratios, and a simulation of the dissipative evolution of two pairs of resonant planets. The semi- major axis ratios a3/a1 and a3/2 will increase in time. When, for two consecutive outputs of the simulation, the ratio a3/a1 crosses the observed one (a3/a1)\|obs, we store the correspond- ing value of a3/a2 from the simulation (an average of it at the two consecutive outputs). Since the system might be librating around the equilibrium points with some amplitude, and there are additional short period terms, this will happen many times for a single simulation, and we obtain a list of a3/a2 values. Then, we report the average of this list, and again compare this quan- tity with the observed (a3/a2)\|obs (since they are obtained from the mean period extracted from the data) by computing the rela- tive difference as in Eq. (16). We then loop over different choices of planetary mass ratios and obtain a map that can be compared with the analytical maps of the previous section. G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems Therefore, we used a ﬁctitious τa which is kept equal for all planets and constant along the different simulations of resonant capture. This has the sole effect of making it easier to automate the simulations, and does not affect our results. We need only to make sure that at the end of the disc-migration phase the semi-major axis ratios are smaller than the observed ones, since the subsequent evolution dominated by tidal dissipation will only cause the semi-major axis ratios to expand. We note that this is always possible, since one can obtain different ﬁnal eccentricities at the captured state by changing the ratio of the eccentricity damping τe and the migration rate τa, and thus obtain different corresponding equi- librium values of the semi-major axis ratios; the latter approach the exact commensurabilities as the eccentricities grow (Fig. 2), and since for the systems that we are studying both pairs reside wide enough of the nominal resonance, the ﬁnal eccentricities need not be too high, of order a few 10−2 for a typical planetary mass of order 10−5M∗. undergoing fast oscillation while they cross the observed value (a3/a1)\|obs, but the typical value of ¯δ(a3/a2) is similar to the one found analytically. The case of Kepler-31, which resides close to a 2:1–2:1 mean motion resonant chain, is a bit different, since the 2:1 reso- nance capture might be only temporary if the librations around equilibrium are overstable (Goldreich & Schlichting 2014). This behaviour has been already investigated thoroughly in the case of two planets (Delisle et al. 2015, Deck & Batygin 2015), however it has been shown to be dependent on the speciﬁc implemen- tation of the disc-planet forces, and to disappear in some cases (Xu et al. 2018). In this work we do not intend to expand on these matters, since the formulas that mimic the planet-disc inter- actions represent only approximate implementations of the real forces that are felt by the planets from the disc, which themselves remain observationally unconstrained. We therefore take a prac- tical approach, and note that in the numerical simulations where resonant capture was successful (typically for m1/m2, m2/m3 ≳ 1) the numerical results agree very well with the analytical ones; moreover we still observe that the theoretical value of ¯δ(a3/a2) varies extremely little with m1/m2 and m2/m3 (Fig. 8), so the latter simulations can be considered as enough support for the analytical calculations. G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems 6, but for the system YZ Cet, residing close to a 3:2–3:2 chain. The value of ¯δ(a3/a2) ∼0.01 across all planetary mass ratios can be matched against the corresponding curve in Fig. 9, where we ﬁnd that ∼80% of randomly generated systems lie this close to the 3:2–3:2 chain. 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.0034011600 0.0034020000 0.0034028400 0.0034036800 0.0034045200 0.0034053600 0.0034062000 0.0034070400 0.0034078800 0.0034087200 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 (a3/a2eq-a3/a2obs) a3/a2obs 0.0034011600 0.0034020000 0.0034028400 0.0034036800 0.0034045200 0.0034053600 0.0034062000 0.0034070400 0.0034078800 0.0034087200 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 35.2 44.0 52.8 61.6 70.4 79.2 88.0 96.8 105.6 114.4 Fig. 8. Same as in Fig. 6, but for the system Kepler-31. Only analytical maps are shown in this case since in some simulations capture into resonance was unsuccessful due to overstability of the captured state for different planetary mass ratios. As explained in the text, this issue is model-dependent and is not within the scope of our analysis. Moreover, in the simulations where capture was successful, the results agree well with the analytical calculations, showing ¯δ(a3/a2) ∼0.003. Comparing with Fig. 9 we see that there is a ∼20% probability that Kepler-31 lies this close to the 2:1 – 2:1 chain by pure chance. 0.1 0.5 1 5 10 0.1 0.5 1 5 10 m1/m2 m2/m3 Max (e) m 35.2 44.0 52.8 61.6 70.4 79.2 88.0 96.8 105.6 114.4 Fig. 8. Same as in Fig. 6, but for the system Kepler-31. Only analytical maps are shown in this case since in some simulations capture into resonance was unsuccessful due to overstability of the captured state for different planetary mass ratios. As explained in the text, this issue is model-dependent and is not within the scope of our analysis. Moreover, in the simulations where capture was successful, the results agree well with the analytical calculations, showing ¯δ(a3/a2) ∼0.003. Comparing with Fig. 9 we see that there is a ∼20% probability that Kepler-31 lies this close to the 2:1 – 2:1 chain by pure chance. A7, page 10 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems will change the relative speeds at which each planet’s semi-major axis decreases, which is practically inconvenient. G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems After the desired resonant state is obtained, we slowly depleted the gas. Finally, we added the effects of tidal dissipa- tion (following Mardling & Lin 2002), using arbitrary quality factors for the planets but large enough to ensure that the dissi- pative evolution be slow compared to the resonant evolution of the two planets’ pairs. This allows us to perform efﬁcient integra- tions without breaking the adiabatic approximation which keeps the system close to the resonant equilibrium points. We note that we have little to no information on the internal structures of exo- planets, so we would not be able to conﬁdently assign realistic eccentricity damping timescales anyway. Moreover, as we have already mentioned, tides are only one example of dissipation (that is, loss of orbital energy E) at constant angular momentum L, so that these results are in fact generalisable to any dissipative process such that ˙E < 0 and ˙L = 0. Therefore, a resonant system undergoing any such process is expected to follow the loci of the resonant equilibria, and the divergence of the semi-major axes is obtained as a general result. We now explain how we obtain maps similar to those drawn in the previous section from these numerical simulations. 4. Results 4.1. Probabilistic measure of a resonant conﬁguration in Kepler-305, YZ Cet and Kepler-31 4.1. Probabilistic measure of a resonant conﬁguration in Kepler-305, YZ Cet and Kepler-31 A similar pro- cedure can be applied to a2/a1 (which gives again similar values to that of a3/a2 as we mentioned in Sect. 3.2), as well as the quantity max e/ ˜m. Secondly, we note that we do obtain in all three cases small values for ¯δ(a3/a2), meaning that these systems are indeed close to some equilibrium point of the Hamiltonian (Eq. (13)) and therefore could potentially reside in a multi-resonant chain. However, these values by themselves do not contain any mean- ingful information. The quantity ¯δ(a3/a2) should indeed be calibrated if we intend to use it as a measure of the probabil- ity that the actual system (with its real unknown eccentricities) is in resonance, which in turn would yield a measure of how consistent the orbital architecture of such a system is with the envisioned scenario described above. To this end, for various resonant chains we randomly generate period ratios of ﬁctional systems such that 0 < ∆< 0.05 for each pair, and extract the This analysis has been performed for the three selected sys- tems. For Kepler-305 and YZ Cet, we show the resulting plots on the bottom panels c and d of Figs. 6 and 7 respectively, and notice very good agreement with the analytical results. The noise that is observed in the panels c relative to the quantity ¯δ(a3/a2) is due to the fact that the numerically simulated systems are A7, page 11 of 14 A&A 625, A7 (2019) 3:2 - 2:1 (Kepler-305) 2:1 - 2:1 (Kepler-31) 3:2 - 3:2 (YZ Cet) 2:1 - 3:2 0.005 0.010 0.015 0.020 δ(a3/a2) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Fig. 9. Cumulative distribution functions for \|¯δ(a3/a2)\| for randomly generated systems close to chains of any possible combinations of the 2:1 and 3:2 mean motion resonances. We indicate the chains that repre- sent selected systems from Fig. 4; for each of them, a point indicates the observed ¯δ(a3/a2). From this, we obtain on the vertical axis the probability that these systems could have this value of ¯δ(a3/a2) by pure chance. 4.2. The 5:4 – 4:3 resonant chain on Kepler-60 and other near-resonant systems with k > 3 While in this work we have concentrated on the 2:1 and 3:2 mean motion commensurabilities, it is worthwhile to point out that more compact resonant chains are possible, and Kepler-60 represents a notable example. This system hosts three planets with mean observed periods of ≃5.49, ≃8.29 and ≃16.74 days respectively. Their masses have been constrained via TTV to be ≃4 M⊕for all planets (Jontof-Hutter et al. 2016). The mean motions of the planets satisfy 4˙λ1 −8˙λ2 + 4˙λ3 ≃−0.02◦days−1, hinting at a resonant conﬁguration. Indeed, Go´zdziewski et al. (2016) found that the TTV signal for these planets is consis- tent with a true three-body Laplace-like resonance as well as a chain of 5:4–4:3 two-body mean motion resonances. Using the system’s parameters we can ﬁnd a resonant equilibrium con- ﬁguration as in Sect. 3.2, by imposing a3/a1 to be equal to the observed (a3/a1)\|obs. This gives ¯δ(a3/a2) of order 4 × 10−5. Using an analogous argument to that of Fig. 9, we ﬁnd that there is only a 0.25% probability that Kepler-60 lies this close to a 5:4–4:3 resonant chain by pure chance. The eccentricities that we ﬁnd at the selected resonant equilibrium point are of order e1 ≃0.02, e2 ≃0.03, e3 ≃0.01 for the observed planetary masses. These numbers are quite close to the ones consistent for the two-body mean motion resonance chain solution found in Go´zdziewski et al. (2016). Note in passing that their solution is for non-vanishing libration amplitude of the four resonant angles (however their mean values are the same found here for a stable resonant equilibrium). corresponding ¯δ(a3/a2) (calculated for the choice of mass ratios m1/m2 = m2/m3 = 1 for simplicity, since as we saw above the ¯δ(a3/a2) value depends extremely weakly on the mass ratios). From the cumulative distribution of \|¯δ(a3/a2)\| that arises from this procedure we can obtain the probability that any given system has a given (small) ¯δ(a3/a2) purely by chance. y g p y y Since we are interested mainly in the 2:1 and 3:2 mean motion resonances, we show in Fig. 9 these cumulative distri- butions for systems close to any possible combinations of these resonances. The results show that the proximity of YZ Cet to the 3:2–3:2 resonance is not statistically signiﬁcant, since in ∼80% of randomly generate systems close to the 3:2–3:2 chain we obtain an equivalent or smaller value of ¯δ(a3/a2). 4.2. The 5:4 – 4:3 resonant chain on Kepler-60 and other near-resonant systems with k > 3 Instead, Kepler-305 and Kepler-31 are likely to be in resonance at the 1σ level (i.e. the probability that their value of ¯δ(a3/a2) is smaller than the determined value by chance is less than 32%): for Kepler-305 there is a ∼15% chance that this particular system lies this close to resonance by chance, while for Kepler-31 the probability is ∼20%. We should remark that these speciﬁc values for the probabil- ities that each system is this close to exact resonance by chance are calibrated by the choice 0 < ∆< 0.05 for both pairs of plan- ets, which is used in generating the ﬁctional systems. This value is however not arbitrary. For, it must be consistent with the choice made in Sect. 3.1, which produced only these three systems with both the inner and outer planet pair this close to ﬁrst order mean motion resonance: there, the choice \|∆\| < max ∆= 0.05 was dic- tated by the observation of the location of the peak wide of nominal resonance (Hadden & Lithwick 2017 and our Fig. 3), so restricting the interval in ∆values with smaller max ∆might have resulted in excluding potential systems. On the other hand, increasing max ∆in the generation of the ﬁctional systems would have the only effect to decrease the calculated probabilities. Therefore, we conclude that our choice of max ∆= 0.05 is not arbitrary, and gives a reasonable upper bound to the probabilities that each system ﬁnds itself so close to exact resonance by pure chance. For completeness, we cite other near-resonant systems of three planets with k > 3 that are found in the catalogue. The only ones which satisfy our criterion \|∆\| < 0.05 for both pairs are K2-239 (close to a 3:2–4:3 chain), Kepler-289 (close to a 2:1– 2:1 chain), Kepler-226 (close to a 4:3–3:2 chain) and Kepler-431 (close to a 5:4–4:3). Of these, only the latter two satisfy ∆> 0 for both pairs. 4.1. Probabilistic measure of a resonant conﬁguration in Kepler-305, YZ Cet and Kepler-31 3:2 - 2:1 (Kepler-305) 2:1 - 2:1 (Kepler-31) 3:2 - 3:2 (YZ Cet) 2:1 - 3:2 0.005 0.010 0.015 0.020 δ(a3/a2) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 point (and hence libration of the Laplace angle) the average of the (a, e) oscillation falls on the equilibrium point, while in case of circulation, the average falls off the equilibrium point curve. Consequently, the circulation of the Laplace angle implies that the libration amplitude is larger than the distance of the equi- librium point from the axis e = 0, and that ¯δ(a3/a2) cannot be zero. However, this does not mean that the system did not reach that point via divergent migration: being the (a, e) equi- librium so close to to e = 0, even a minute perturbation can induce circulation of the Laplace angle. Hence the libration of the Laplace angle is a sufﬁcient but not necessary condition to conclude that a system’s dynamical history has been shaped by resonant capture and subsequent resonant repulsion driven by dissipation. Fig. 9. Cumulative distribution functions for \|¯δ(a3/a2)\| for randomly generated systems close to chains of any possible combinations of the 2:1 and 3:2 mean motion resonances. We indicate the chains that repre- sent selected systems from Fig. 4; for each of them, a point indicates the observed ¯δ(a3/a2). From this, we obtain on the vertical axis the probability that these systems could have this value of ¯δ(a3/a2) by pure chance. 4.2. The 5:4 – 4:3 resonant chain on Kepler-60 and other near-resonant systems with k > 3 5. Conclusions In this work, we have generalised the formalism of dissipa- tive divergence of resonant orbits to multi-resonant chains. The analytical study performed in Sect. 2 allows us to predict the orbital conﬁgurations of systems of planets deep in a chain of ﬁrst order mean motion resonances, and therefore, even though at a lesser degree of precision, of systems that are in resonance with a ﬁnite amplitude of libration. Then, we showed in Sect. 2.2 that under the effect of slow dissipation a nearly-resonant sys- tem is expected to follow the loci of the equilibrium points of the resonant Hamiltonian (Eq. (13)) maintaining the amplitude of libration in an adiabatic manner. Therefore, if the orbital For completeness, we report the observed variation of the three-body Laplace angle ˙ϕL in these systems, since its libration can be in principle a sufﬁcient condition to conclude that they are indeed resonant. For Kepler-305 we checked that the Laplace angle ϕL = 2λ1 −4λ2 + 2λ3 satisﬁes ˙ϕL ≃0.5◦days−1 given the observed transits periods; for Kepler-31, ϕL = λ1 −3λ2 + 2λ3 sat- isﬁes ˙ϕL ≃0.1◦days−1; ﬁnally for YZ Cet, ϕL = 2λ1 −5λ2 + 3λ3 satisﬁes ˙ϕL ≃9.4◦days−1. As we argued in the Introduction, in case of libration of the resonant angles around the equilibrium A7, page 12 of 14 G. Pichierri et al.: The role of dissipative evolution for three-planet, near-resonant extrasolar systems probabilities of resonant association that we have found indicate that between 1/3 and 2/3 of the systems with 0 < ∆< 0.05 show memory of the processes of resonant capture; this is consistent with the histogram of Fig. 3, where the peak wide of the reso- nance is about 2 times higher than the underlying random-like ﬂat distribution. architecture of a system is found near one of these equilibrium points, it is strongly suggestive that the envisioned scenario of slow convergent orbital migration leading to capture into res- onance and subsequent orbital divergence due to dissipative evolution really occurred for the system. In the light of the results presented above, we can draw the following conclusions. p g On the one hand, we must face the fact that the orbital archi- tecture of a signiﬁcant fraction of the systems of three planets is actually not consistent with these physical mechanisms. References Batygin, K. 2015, MNRAS, 451, 2589 Batygin, K., & Adams, F. C. 2017, AJ, 153, 120 Batygin, K., & Morbidelli, A. 2013a, AJ, 145, 1 Batygin, K., & Morbidelli, A. 2013b, A&A, 556, A28 Batygin, K., Deck, K. M., & Holman, M. J. 2015, AJ, 149, 167 Beaugé, C., Michtchenko, T. A., & Ferraz-Mello, S. 2006, MNRAS, 365, 1160 Cresswell, P., & Nelson, R. P. 2006, A&A, 450, 833 Cresswell, P., & Nelson, R. P. 2008, A&A, 482, 677 Deck, K. M., & Batygin, K. 2015, ApJ, 810, 119 Deck, K. M., Payne, M., & Holman, M. J. 2013, ApJ, 774, 129 Delisle, J.-B., & Laskar, J. 2014, A&A, 570, L7 Delisle, J.-B., Laskar, J., Correia, A. C. M., & Boué, G. 2012, A&A, 546, A71 Delisle, J.-B., Correia, A. C. M., & Laskar, J. 2015, A&A, 579, A128 Fulton, B. J., Petigura, E. A., Howard, A. W., et al. 2017, AJ, 154, 109 Gillon, M., Jehin, E., Lederer, S. M., et al. 2016, Nature, 533, 221 Gillon M Triaud A H M J Demory B O et al 2017 Nature 542 456 Deck, K. M., & Batygin, K. 2015, ApJ, 810, 119 Deck, K. M., Payne, M., & Holman, M. J. 2013, ApJ, 774, 129 Delisle, J.-B., & Laskar, J. 2014, A&A, 570, L7 Delisle, J.-B., Laskar, J., Correia, A. C. M., & Boué, G. 2012, A&A ulton, B. J., Petigura, E. A., Howard, A. W., et al. 2017, AJ, 154, 10 Gillon, M., Jehin, E., Lederer, S. M., et al. 2016, Nature, 533, 221 Gillon, M., Triaud, A. H. M. J., Demory, B.-O., et al. Gillon, M., Triaud, A. H. M. J., Demory, B.-O., et al. 2017, Nature, 542, 456 Goldreich, P., & Schlichting, H. E. 2014, AJ, 147, 32 Go´zdziewski, K., Migaszewski, C., Panichi, F., & Szuszkiewicz, E. 2016, MNRAS 455 L104 g MNRAS, 455, L104 Hadden, S., & Lithwick, Y. 2014, ApJ, 787, 80 , p On the other hand, systems with orbital properties that are compatible with a (near) resonant state do exist in the exoplanet census. These include the already known examples mentioned above of Trappist-1, Kepler-223, GJ867 and Kepler-60, and, potentially, some of the systems analysed in this paper. 5. Conclusions More precisely, the majority of the systems are not close to resonance, implying that either they never captured in resonance in the ﬁrst place, or they escaped from resonance due to a violent insta- bility (Izidoro et al. 2017) losing any memory of their resonant dynamical past. To ponder these two possibilities, consider ﬁrst of all that some form of orbital transport is expected to take place: for example, the majority of planets with Rpl ≳1.6 R⊕ have H/He gaseous atmospheres that cannot be explained by production of volatiles after the formation of the planet (Rogers 2015), implying that these planets formed while the protoplane- tary disc was still present. The associated planet–disc interaction would then force the planets’ orbital elements to change, in other words, force the planets to migrate. However, orbital migration may not be convergent (Migaszewski 2015, 2016), that is, not leading to resonant capture. Moreover, some mechanisms have been proposed to inhibit the capture even in the case of conver- gent migration, such as turbulence in the disc or e-dependent migration rates. Nevertheless, these processes alone do not ade- quately explain the lack of resonance in the exoplanet sample (e.g. Batygin & Adams 2017; Deck & Batygin 2015; Xu et al. 2018). For these reasons, it is more likely that capture into mean motion resonance is a common outcome of the early epochs of disc-planet interaction, but the subsequent evolution after the disc removal is subject to instabilities which break the compact conﬁguration. This approach seems to be able to reproduce the observed distribution of period ratios if these instabilities are extremely common (Izidoro et al. 2017). The primary mecha- nism through which planets are ejected from resonance, however, remains elusive, and a topic of active research. The architecture of many planetary systems observed by transit is not well constrained by observations. Opportunities for more extensive characterisation will come from missions such as the Transiting Exoplanet Survey Satellite (TESS) or the PLAnetary Transits and Oscillations of stars (PLATO), which are designed to target bright stars to allow for follow-up via fur- ther ground-based and space-based observations (with methods such as radial velocity). This will allow for a better quantiﬁcation of planetary masses, radii, ages of the systems and eccentricities. 5. Conclusions In the light of this augmented perception that we can expect to acquire, our study outlines the groundwork for further dynam- ical characterisation of the physical processes that shaped the present-day architectures of extrasolar planetary systems. References S., Morbidelli, A., Crida, A., & Ferreira, J. 2 Michtchenko, T. A., Beaugé, C., & Ferraz-Mello, S. 2008, MNR Millholland, S., Wang, S., & Laughlin, G. 2017, ApJ, 849, L33 , , g, , g , , p , , Mills, S. M., Fabrycky, D. C., Migaszewski, C., et al. 2016, Nature, 533, 509 Migaszewski C 2015 MNRAS 453 1632 g g p Mills, S. M., Fabrycky, D. C., Migaszewski, C., et al. 2016, Nature, 533, 509 y y g Migaszewski, C. 2015, MNRAS, 453, 1632 Migaszewski, C. 2016, MNRAS, 458, 2051 g Morbidelli, A. 2002, Modern Celestial Mechanics: Aspects of Solar System D i (L d T l & F i ) Morbidelli, A. 2002, Modern Celestial Mechanics: Asp Morbidelli, A. 2002, Modern Celestial Mechanics: Aspects of Solar System Dynamics (London: Taylor & Francis) Dynamics (London: Taylor & Francis) y y Murray, C. D., & Dermott, S. F. 2000, Solar System Dynamics (Cambridge, UK: Murray, C. D., & Dermott, S. F. 2000, Solar System Dynamics (Cambridge, UK: Cambridge University Press) Murray, C. D., & Dermott, S. F. 2000, Solar y y y g Cambridge University Press) Cambridge University Press) g y ) Papaloizou, J. C. B., & Terquem, C. 2010, MNRAS, 405, 573 p , , q , , , , Pichierri, G., Morbidelli, A., & Crida, A. 2018, Celest. Mech. Dyn. Astron., 130, 54 Poincare, H. 1892, Les Méthodes nouvelles de la mécanique céleste (Paris Gauthier-Villars et ﬁls) Rivera, E. J., Laughlin, G., Butler, R. P., et al. 2010, ApJ, 719, 890 Rogers, L. A. 2015, ApJ, 801, 41 Tamayo, D., Rein, H., Petrovich, C., & Murray, N. 2017, ApJ, 840, L19 Weiss, L. M., Marcy, G. W., Rowe, J. F., et al. 2013, ApJ, 768, 14 Weiss, L. M., Marcy, G. W., Petigura, E. A., et al. 2018, AJ, 155, 48 Wisdom, J. 1980, AJ, 85, 1122 Wu, Y., & Lithwick, Y. 2013, ApJ, 772, 74 Xu, W., Lai, D., & Morbidelli, A. 2018, MNRAS, 481, 1538 Poincare, H. 1892, Les Méthodes nouvelles de la mécanique céleste (Paris: Gauthier-Villars et ﬁls) Rivera, E. J., Laughlin, G., Butler, R. P., et al. 2010, ApJ, 71 Tamayo, D., Rein, H., Petrovich, C., & Murray, N. 2017, ApJ, 840, L19 Weiss, L. M., Marcy, G. W., Rowe, J. F., et al. 2013, ApJ, 768, 14 y p Weiss, L. M., Marcy, G. W., Petigura, E. A., et al. References That is, while it is difﬁcult to prove deﬁnitively that a given sys- tem is now in resonance in a formal sense (the resonant angles are in libration), in this work we have developed a method to quantitatively test the consistency of a given orbital architecture with a dynamical history characterised by resonant capture and subsequent dissipative evolution. This is achieved through the calculation of the quantity ¯δ(a3/a2) deﬁned in Eq. (16), which is obtained directly from the observed semi-major axis ratios and, as we have shown, depends very weakly on the mass ratios between the planets, making the observational uncertainties on the latter quantities irrelevant. Then, using the approach illus- trated in Fig. 9, this “indicator” can be turned into a quantitative probability that the system is related to the considered resonant chain. In this sense, we have found that there is a ∼15, ∼20 and ∼80% probability that Kepler-305, Kepler-31 and YZ Cet respec- tively ﬁnd themselves this close to resonance merely by chance. Multiplying these probabilities we ﬁnd that there is only a ∼2% probability that all three of these systems lie close to resonance just by chance. This suggests strongly that at least some of them should have had a resonant dynamical history. Although the sample is clearly too small to make any meaningful inference, the Hadden, S., & Lithwick, Y. 2017, AJ, 154, 5 Henrard, J., & Lamaitre, A. 1983, Celest. Mech., 30, 197 Izidoro, A., Ogihara, M., Raymond, S. N., et al. 2017, MNRAS, 470, 1750 J f H D F d E B R J F l 2016 A J 820 39 Izidoro, A., Ogihara, M., Raymond, S. N., et al. 2017, MNRAS, 470, 1750 Jontof-Hutter, D., Ford, E. B., Rowe, J. F., et al. 2016, ApJ, 820, 39 ontof-Hutter, D., Ford, E. B., Rowe, J. F., et al. 2016, ApJ, 820, 39 Lithwick, Y., & Wu, Y. 2012, ApJ, 756, L11 Lithwick, Y., Xie, J., & Wu, Y. 2012, ApJ, 761, 122 uger, R., Sestovic, M., Kruse, E., et al. 2017, Nat. Astron., 1, 0129 Luger, R., Sestovic, M., Kruse, E., et al. 2017, Nat. Astro Mardling, R. A., & Lin, D. N. C. 2002, ApJ, 573, 829 Mardling, R. A., & Lin, D. N. C. 2002, ApJ, 573, 829 Masset, F. S., Morbidelli, A., Crida, A., & Ferreira, J. 2006, ApJ, 642, 478 Masset, F. Appendix A: Reduced Hamiltonian to a common planetary mass factor (A.1) Inverting this expression we easily get all planetary masses in terms of ˜m, m = c ˜m := 3β1β2M∗ ˜m Hres = ˜m −G2M∗c1c3 2 ¯Λ2 2 × f (1,in) res s 2Γ1 ¯Λ1 cos kinλ2 −(kin −1)λ1 + γ1 (2 in) s 2Γ2 in in  1) es Hres = ˜m −G2M∗c1c3 2 ¯Λ2 2 × f (1,in) res s 2Γ1 ¯Λ1 cos kinλ2 −(kin −1)λ1 + γ1 + f (2,in) res s 2Γ2 ¯Λ2 cos kinλ2 −(kin −1)λ1 + γ2 + − G2M∗c2c3 3 ¯Λ2 3 ˜m = m1 + m2 + m3 3 M∗ = m1(1 + β−1 1 + β−1 1 β−1 2 ) 3 M∗ . (A.1) Hres = ˜m −G2M∗c1c3 2 ¯Λ2 2 s Inverting this expression we easily get all planetary masses in terms of ˜m, m1 = c1 ˜m := 3β1β2M∗ 1 + β2 + β1β2 ˜m, m2 = c2 ˜m := 3β1M∗ 1 + β2 + β1β2 ˜m, m3 = c3 ˜m := 3M∗ 1 + β2 + β1β2 ˜m, (A.2) m1 = c1 ˜m := 3β1β2M∗ 1 + β2 + β1β2 ˜m, m2 = c2 ˜m := 3β1M∗ 1 + β2 + β1β2 ˜m, m3 = c3 ˜m := 3M∗ 1 + β2 + β1β2 ˜m, (A.2) × f (1,out) res s 2Γ2 ¯Λ2 cos koutλ3 −(kout −1)λ2 + γ2 + f (2,out) res s 2Γ3 ¯Λ3 cos koutλ3 −(kout −1)λ2 + γ3  . (A.5) (A.2) with coefﬁcients c depending on M∗, β1 and β2 only. with coefﬁcients c depending on M∗, β1 and β2 only. We introduce the modiﬁed Delaunay action-angle variables (Λ, Γ, λ, γ) as in Eq. (5), but we rescale the actions by the com- mon mass factor ˜m: this gives the following deﬁnition for the Λ’s (we maintain the same notation as the non-rescaled actions for simplicity) From here, the sequence of changes of variables detailed in Sect. 2 can be performed using the same formal deﬁnitions for the new rescaled variables. Appendix A: Reduced Hamiltonian to a common planetary mass factor and the same formal deﬁnition of Γ = Λe2/2 at lowest order in e. We now introduce the Hamiltonian for the problem, that is the sum of the Keplerian Hamiltonian (Eq. (6)) and the resonant interaction Hamiltonian (Eq. (7)) to ﬁrst order in e. However, since we have rescaled the actions by ˜m, in order for the Hamilton equations to be conserved we must also rescale the Hamiltonian itself by ˜m. As in Sect. 2, since we are not considering large deviations in the semi-major axes from their nominal values, and since the resonant Hamiltonian is already of order O(e), we eval- uate the resonant Hamiltonian on the nominal values ¯Λ deﬁned from ¯a using Eq. (A.3). It is then easy to see that the rescaled Keplerian Hamiltonian takes the form In the course of the paper we make implicit use of a reduced Hamiltonian which incorporates the planetary masses through a common planet-to-star mass factor ˜m. In this appendix, we detail the construction of this Hamiltonian and its use in the paper. Consider three planets, whose physical parameters are labelled 1, 2, and 3 for the inner, middle and outer planet respec- tively, orbiting around a star of mass M∗on the same plane. Suppose that the planets are (close to) a chain of mean-motion resonance, with nominal semi-major axes ¯a and that the devia- tions of the semi-major axes from the nominal values are small, and assume that the eccentricities are small enough, so that an analysis to ﬁrst order in e is valid. These are the working assump- tions throughout Sect. 2. Having ﬁxed the planet–planet mass ratios m1/m2 = β1 and m2/m3 = β2, we introduce the average planet–star mass ratio Hkepl = − 3 X i=1 c3 i 2 GM∗ Λi !2 , (A.4) (A.4) and is therefore independent of ˜m, while the rescaled resonant part will have a multiplicative coefﬁcient ˜m: ˜m = m1 + m2 + m3 3 M∗ = m1(1 + β−1 1 + β−1 1 β−1 2 ) 3 M∗ . References 2018, AJ, 155, A7, page 13 of 14 A7, page 13 of 14 A&A 625, A7 (2019) Appendix A: Reduced Hamiltonian to a common planetary mass factor Λ1 = 3β1β2M∗ 1 + β2 + β1β2 p GM∗a1 = c1 p GM∗a1, Λ2 = 3β1M∗ 1 + β2 + β1β2 p GM∗a2 = c2 p GM∗a2, Λ3 = 3M∗ 1 + β2 + β1β2 p GM∗a3 = c3 p GM∗a3, (A.3) Λ1 = 3β1β2M∗ 1 + β2 + β1β2 p GM∗a1 = c1 p GM∗a1, Λ2 = 3β1M∗ 1 + β2 + β1β2 p GM∗a2 = c2 p GM∗a2, Λ3 = 3M∗ 1 + β2 + β1β2 p GM∗a3 = c3 p GM∗a3, (A.3) Already from the Hamiltonian written in terms of the rescaled variables Λ and Γ ∝e2 one can see the following. Assuming a ﬁxed equilibrium value of the semi-major axes (that is, of the Λ’s), a change in the planet-star mass factor ˜m will have the only effect to rescale the equilibrium values of all √ Γ ∝e by the same quantity. In this conﬁguration, the equilib- ria of the semi-major axis ratios (a2/a1)eq and (a3/a2)eq remain independent of ˜m. Already from the Hamiltonian written in terms of the rescaled variables Λ and Γ ∝e2 one can see the following. (A.3) A7, page 14 of 14
https://openalex.org/W2097816357	https://ccforum.biomedcentral.com/counter/pdf/10.1186/cc11657	English	null	Glycemic variability and glucose complexity in critically ill patients: a retrospective analysis of continuous glucose monitoring data	Critical care	2,012	cc-by	7,387	Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Open Access RESEARCH * Correspondence: ulrike.holzinger@meduniwien.ac.at 1Department of Medicine III, Division of Gastroenterology and Hepatology, ICU 13H1, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria Full list of author information is available at the end of the article © 2012 Brunner et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Introduction: Glycemic variability as a marker of endogenous and exogenous factors, and glucose complexity as a marker of endogenous glucose regulation are independent predictors of mortality in critically ill patients. We evaluated the impact of real time continuous glucose monitoring (CGM) on glycemic variability in critically ill patients on intensive insulin therapy (IIT), and investigated glucose complexity - calculated using detrended fluctuation analysis (DFA) - in ICU survivors and non-survivors. Methods: Retrospective analysis were conducted of two prospective, randomized, controlled trials in which 174 critically ill patients either received IIT according to a real-time CGM system (n = 63) or according to an algorithm (n = 111) guided by selective arterial blood glucose measurements with simultaneously blinded CGM for 72 hours. Standard deviation, glucose lability index and mean daily delta glucose as markers of glycemic variability, as well as glucose complexity and mean glucose were calculated. Results: Glycemic variability measures were comparable between the real time CGM group (n = 63) and the controls (n = 111). Glucose complexity was significantly lower (higher DFA) in ICU non-survivors (n = 36) compared to survivors (n = 138) (DFA: 1.61 (1.46 to 1.68) versus 1.52 (1.44 to 1.58); P = 0.003). Diabetes mellitus was significantly associated with a loss of complexity (diabetic (n = 33) versus non-diabetic patients (n = 141) (DFA: 1.58 (1.48 to 1.65) versus 1.53 (1.44 to 1.59); P = 0.01). Conclusions: IIT guided by real time CGM did not result in significantly reduced glycemic variability. Loss of glucose complexity was significantly associated with mortality and with the presence of diabetes mellitus. Glycemic variability and glucose complexity in critically ill patients: a retrospective analysis of continuous glucose monitoring data Richard Brunner1, Gabriel Adelsmayr1, Harald Herkner2, Christian Madl1 and Ulrike Holzinger1* Introduction covered by mean glucose, glycemic variability has been suggested as an additional measure for glucose control. Glycemic variability is represented by standard deviation (SD), mean daily δ blood glucose or glucose lability index (GLI). SD is the most commonly used parameter and is calculated as the square-root of the average of the squared differences between individual glucose values and the mean. Mean daily δ blood glucose describes the mean of the daily difference between minimum and maximum blood glucose. These two measures do not take order and timing of measurements into account. GLI is the squared difference between consecutive blood glucose levels per unit of actual time between the samples. GLI considers the time between and the order of measurements. Although no gold standard of measuring glycemic variabil- ity has been established yet, SD seems to be the best pre- dictor of mortality [10]. Glucose control in critically ill patients has been a highly disputed topic since 2001, when van den Berghe et al. showed that intensive insulin therapy (IIT) (mean glucose levels ≤6.11 mmol/L) could reduce the morbidity and mortality of patients in surgical ICUs by 42% [1]. However, subsequent studies came to inconclusive findings [2]. Recently, several retrospective trials found glycemic varia- bility per se to be associated with mortality in critically ill patients, independent of mean glucose concentration [3-9]. The measure glycemic variability describes fluctuations of blood glucose over time. As glucose fluctuations are not Page 2 of 9 Page 2 of 9 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 In a prospective study of septic critically ill patients a significant association between high glycemic variability and mortality was found [11]. These results are consis- tent with in vitro data showing that short-time fluctua- tions of glucose levels induce endothelial cell damage and apoptosis [12]. Moreover, a significant association between glycemic variability and 8-iso prostaglandin F2a, a marker of oxidative stress and potential mediator of organ dysfunction, has been shown in diabetic type 2 patients [13]. Minimal glycemic variability has been pro- posed to become the gold standard of glycemic control in diabetic patients [14]. proposed as a description of the endogenous glucose reg- ulation system that is independent from exogenous fac- tors. Lundelin et al. Introduction hypothesized that in a healthy regulatory system, glucose levels are corrected frequently and result in a ‘complex’ glucose profile [19]. However, in critically ill patients a decomplexification of glucose regulation was suggested [20]. Low complexity represents the inability of the patient to correct glucose fluctuations frequently and quickly. In a small number of patients glycemic profile was shown to be more complex in ICU survivors than in ICU non-survivors [19]. However, this has not been confirmed in a large group of critically ill patients. Therefore, we investigated the role of glucose complexity in a large group of critically ill patients. Glycemic variability depends on both endogenous patient-specific factors such as severity of disease and diabetes status [15], as well as exogenous factors such as type and quality of glucose monitoring, the glucose algo- rithm used for the calculation of the insulin rate, compli- ance of the nursing staff with the recommendations of the protocol and application of medication including ent- eral and parenteral nutrition. As the endogenous glucose regulation system can hardly be influenced, glycemic variability needs to be improved by acting on the exogen- ous factors. Besides well-trained nursing staff and contin- uous application of medication, appropriate glucose monitoring is suggested to minimize blood glycemic variability. Therefore, the need for real time continuous glucose level reporting has been emphasized numerous times [16,17]. Subcutaneous continuous glucose monitor- ing (CGM) systems provide both real time capability and adequate accuracy in medical critically ill patients includ- ing those requiring vasopressors [18]. The primary hypothesis is that real-time CGM gui- dance of IIT is associated with decreased glycemic varia- bility in critically ill patients. The secondary hypothesis of the study is that glucose complexity is independently associated with increased mortality. Patients and setting This is a post-hoc analysis of two prospective, randomized controlled trials conducted in an eight-bed closed medi- cal ICU at the University Hospital of Vienna, Austria [21,22]. During the period from April 2005 to August 2008 a total of 983 critically ill patients were admitted, 728 of whom were ventilated. A total of 174 consecutive, mechanically ventilated and sedated patients fulfilling the inclusion criteria (age ≥18, expected to stay ≥48 hours in the ICU after initiation of IIT) were enrolled in the study within 48 hours after ICU admission. Patients were not enrolled in the study if any of the following criteria were present: ICU stay expected to be <48 hours, mechanical ventilation not expected for >48 hours, inclusion in another study, no CGM device available during the screening phase or glucose values in the normal range without insulin therapy. Although IIT may be associated with increased glyce- mic variability [7] we hypothesized that IIT guided by real time glucose monitoring would decrease glycemic variability. Thus, we aimed to evaluate the impact of real time CGM on glycemic variability in critically ill patients. More importantly, by using CGM devices valuable insights into glucose regulation are possible. While glu- cose variability can be calculated using conventional blood glucose measurements every four to six hours, glucose complexity calculation requires the availability of continuous glucose data. Recently, in critically ill patients glucose complexity has been proposed as a marker of endogenous glucose regulation [19]. The original objectives of the two studies were to evaluate the impact of circulatory shock requiring norepinephrine therapy on the accuracy and reliability of a subcutaneous CGM sensor in critically ill patients [21] and to evaluate the impact of real-time CGM on glycemic control and risk of hypoglycemia in critically ill patients [22]. Glucose complexity is a dynamic measure of glucose time series and, therefore, seems to provide more power- ful information on endogenous glucose regulation than does conventional glycemic variability analysis. In con- trast to glycemic variability that describes the magnitude of glucose fluctuations over several hours, glucose com- plexity is proposed as a measure of short-term glucose oscillations. Glycemic variability depends on endogenous and exogenous factors, whereas glucose complexity is Materials and methods Data analysis was approved by the ethics committee of the Medical University of Vienna. Because of the retro- spective character of the analysis the need for informed consent was waived by the institutional review board. Statistical analysis In the control group insulin infusion rates were guided by selective arterial blood glucose measurements, obtained using an automated blood gas analyzer (Radio- meter ABL 700®, Copenhagen, Denmark). IIT was per- formed by the nursing staff according to a previously described dynamic paper-based insulin titration algo- rithm [23] based on the algorithm used in the Leuven studies [1,24]. This algorithm prescribes the insulin infu- sion rate, time of next glucose measurement (between one and six hours), and, in the case of hypoglycemia, dex- trose administration depending on glucose levels and glu- cose trends. Consequently, it defines nine different states requiring different actions, although leaving space for interpretation by the responsible nurse [22]. In the con- trol group glucose levels were additionally recorded con- tinuously using the Continuous Glucose Monitoring System® (CGMS, Medtronic MiniMed, Northridge, CA, USA), but were blinded and available only in retrospect. To evaluate the impact of real time CGM on glucose variability as a marker of exogenous glucose regulation, we used linear regression analyses with glycemic variabil- ity, represented as SD (GluSD), glucose lability index ({∑(Glun - Glun+1 (mmol/L))2(hn+1 - hn)-1)(number of readings)-1 or mean daily delta (difference between mini- mum and maximum) glucose as primary outcome and real time versus concealed CGMS (controls) as predictor. Secondary endpoints were coefficient of variation (CV) of glucose (GluSD/Glucose mean (%)), variability or mean of glucose during the first 24 hours, maximum glucose during ICU stay. To investigate glycemic dynamics and its relation with mortality in critically ill patients we used glucose complex- ity as the main risk factor and ICU mortality as outcome. Glucose complexity is proposed as the representation of the endogenous glucoregulatory process. Although it is similar to glucose variability it is able to detect minor sys- temic alterations in endogenous glucose regulation. In a healthy regulatory system, glucose levels are corrected fre- quently and result in a ‘complex’ glucose profile. However, in critically ill patients a decomplexification of glucose reg- ulation was suggested [20]. In the real time CGM group, IIT was performed by the nursing staff and insulin infusion rates were guided by continuously available glucose levels using the Guar- dian® real time CGMS (Medtronic MiniMed) according to the algorithm used in the control group. Statistical analysis In contrast to the control group, nurses were requested to take real time glucose readings in close intervals according to clinical necessity at personal discretion, however at least every two hours [22]. Glucose complexity was calculated using DFA. DFA is a unitless metric that estimates the degree of long-range correlations within a signal, analyzing how the time series and its linear regression diverge as the ‘time window’ considered increases. As a rule of thumb, higher com- plexities are displayed as lower DFA (until a minimum of 0.5). Details on DFA can be found elsewhere [19]. To investigate glucose complexity detrended fluctua- tion analysis (DFA) [19] was calculated for ICU survivors (n = 138) and ICU non-survivors (n = 36). Furthermore, we evaluated glucose complexity in diabetic (n = 33) and non-diabetic (n = 141) patients. Glucose complexity was normally distributed as assessed by visual inspection of a histogram but not line- arly related with mortality as assessed with a test for deviation from linearity in a logistic regression model. We found a good fit with a quadratic function and entered a quadratic term into the model as a conse- quence. We assessed whether this association was modi- fied by diabetes by testing the significance of the interaction term in such a model using a Wald test. To allow for the influence of other variables on the effect of glucose complexity we entered SAPS II (Simplified Acute Physiology Score) score, age (years), sex, and diabetes (yes versus no) as covariables into our model. These cov- ariables were selected a priori. As we merged patients from two studies into this database we used a multivari- ate mixed effects logistic regression model to additionally allow for potential clustering within each of the two stu- dies. As sensitivity analyses we used robust estimations instead which yielded virtually the same results. Research design To evaluate the impact of real time CGM on glucose variability, data on patients allocated to two groups (real Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Page 3 of 9 time CGM, n = 63 and controls, n = 111) were analyzed. recorded continuously in an internal monitor blinded to the study team and obtained after the trial for further analysis. All included patients were treated with IIT to main- tain glucose levels between 4.44 and 6.11 mmol/L. Real time continuous glucose monitoring system (real time CGMS) The Guardian® real time CGMS has been described in detail previously [22]. Briefly, it displays a mean of 30 glucose measurements over the last five minutes on a monitor, allowing glucose monitoring in real time. The real time CGMS was calibrated against blood glucose measurements, obtained using an automated blood gas analyzer, at least four times per day (every five to six hours). Sensors were planned to stay in place for 72 hours. Baseline characteristics Baseline characteristics of 174 critically ill patients receiving IIT either guided by a real time CGM system or by selective arterial blood glucose measurements with simultaneously blinded CGM can be found in Table 1. Mean CGM time was 59.2 ± 14.7 hours (RT CGM group); 7.0 ± 1.6 BGA readings per patient in 24 hours were taken in the control group. The difference of glucose complexity between survivors and non-survivors was confirmed in a binary logistic regression analysis with ICU mortality as outcome and glucose complexity, as well as age, BMI, gender, diabetes status and SAPS II as co-factors. In this model only DFA was significantly associated with mortality. Continuous glucose monitoring system (CGMS) The CGMS has been described in detail previously [22]. Briefly, it is equivalent to the Guardian® real time CGMS, except it is lacking the capability to display glucose con- centration in real time. Glucose concentrations were Page 4 of 9 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Differences in glycemic metrics between ICU survivors and non-survivors We assessed the impact of the method of glucose determination (CGMS versus blood gas analyzer (BGA)) on the glucose variability measures SD, CV, GLI and mean daily delta glucose by calculating them from CGMS and BGA values in all patients. Measures of glycemic variability, mean glucose and hypo- glycemia were similar between ICU survivors and ICU non-survivors, whereas glucose complexity was signifi- cantly lower in non-survivors (Table 3). Data are presented as mean ± standard deviation, median (25th to 75th percentile) or absolute count and relative frequency. For bivariate comparisons we tabu- lated data and used simple one-way analysis of variance (ANOVA), Mann-Whitney rank sum test or a chi2- test as appropriate to test the null hypothesis of no difference. These measures were similar in hospital survivors and non-survivors (data not shown). Multivariate analysis of glucose complexity and mortality Multivariate analysis of glucose complexity and mortality Although glucose complexity was significantly lower in non-survivors, relation between glucose complexity and mortality was not linear but can be described best with a quadratic function: logodds (ICU survival) = (-0.09 * DFA decile)2 + 0.64 * DFA decile + 1.07, where DFA dec- ile (from 0 to 9) represents one tenth of DFA values in increasing sequence (Figure 1). Data are means ± SD, unless otherwise stated. ALF, acute liver failure; BMI, body mass index; CPR, Cardiopulmonary resuscitation; SAPSII, Simplified Acute Physiology Score; SD, standard deviation; SOFA, Sequential Organ Failure Assessment. Diabetic status and glucose complexity The presence of diabetes was significantly associated with a loss of complexity (higher DFA) (diabetic (n = 33) versus non-diabetic patients (n = 141): DFA 1.58 (1.48 to 1.65) versus 1.53 (1.44 to 1.59); P = 0.01). This difference per- sisted even after correcting for survival (P = 0.027). For data management and analyses we used Excel for Mac 2011 and STATA 11.0 for Mac (Stata Corp., Col- lege Station, TX, USA). Generally a two-sided P-value <0.05 was considered statistically significant. Differences in glycemic metrics between patients using real time CGM and controls The use of real time CGM did not have any impact on the measures of glycemic variability, glucose complexity and maximum glucose (Table 2). The glycemic variability measures SD, CV and GLI were significantly higher while mean daily delta glucose Table 1 Admission reason and baseline characteristics. Real time CGM [21] Controls [21,22] Total [21,22] Included Patients 63 111 174 Admission reason Number of patients (% of patients in the category) Respiratory failure 15 (24) 23 (21) 38 (22) CPR 12 (19) 27 (24) 39 (22) Sepsis/Septic shock 13 (20) 24 (22) 37 (21) Heart failure 8 (13) 21 (19) 29 (17) Neurologic disease/Coma 9 (14) 10 (9) 19 (11) Pulmonary embolism 3 (5) 3 (3) 6 (3) GI-bleeding/ALF 3 (5) 2 (2) 5 (3) Necrotising pancreatitis 0 (0) 1 (1) 1 (1) History of diabetes 12 (19) 21 (19) 33 (19) Age (years) 59 ± 15 62 ± 16 61 ± 16 Gender (male/female) 41/22 64/47 105/69 BMI (kg/m²) 27.1 ± 5.1 26.4 ± 3.9 26.7 ± 4.4 SAPS II 60 ± 16 58 ± 16 59 ± 16 SOFA on admission day 11.5 ± 3.8 10.9 ± 3.5 11 ± 4 Data are means ± SD, unless otherwise stated. ALF, acute liver failure; BMI, body mass index; CPR, Cardiopulmonary resuscitation; SAPSII, Simplified Acute Physiology Score; SD, standard deviation; SOFA, Sequential Organ Failure Assessment. Table 1 Admission reason and baseline characteristics. Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Page 5 of 9 Page 5 of 9 Table 2 Glycemic metrics in the real time CGM and control group. Table 2 Glycemic metrics in the real time CGM and control group. Table 2 Glycemic metrics in the real time CGM and control group. Differences in glycemic metrics between patients using real time CGM and controls Real time CGM (number = 63) [21] Controls (number = 111) [21,22] P-value Measures of glycemic variability Variability of glucose (SD) (mmol/L) 1.19 ± 0.49 1.27 ± 0.54 0.330 Variability of glucose (GLI) 81 (43 to 197) 126 (64 to 222) 0.247 Variability of glucose (δ) (mmol/L) 4.47 ± 2.02 4.76 ± 2.07 0.336 Coefficient of variation (%) 20 ± 7 21 ± 8 0.547 Variability of glucose during first 24 hours (SD) (mmol/L) 0.84 (0.65 to 1.33) 1.04 (0.66 to 1.40) 0.395 Variability of glucose during first 24 hours (GLI) 85 (38 to 190) 118 (60 to 207) 0.348 Variability of glucose during first 24 hours (δ) (mmol/L) 5.72 ± 2.42 5.73 ± 2.40 0.966 Measures of glucose Mean of glucose during first 24 hours (mmol/L) 5.70 (5.19 to 6.47) 5.96 (5.48 to 6.36) 0.099 Maximum glucose (mmol/L) 9.43 ± 2.12 9.77 ± 2.26 0.329 Glucose complexity 1.54 ± 0.11 1.52 ± 0.11 0.210 Measures of IIT Cumulative daily dose of insulin (I.U.) 45 ± 27 40 ± 23 0.239 Number of changes of the insulin infusion/24 hours 3.8 ± 1.5 3.6 ± 1.3 0.573 Extent of insulin change (%) 57 ± 19 62 ± 23 0.150 Number of BGA measurements/24 hours 7.4 ± 2.1 7.0 ± 1.6 0.217 Number of BGA not required by the algorithm/24 ha 0 (0 to 0) 0 (0 to 1) 0.238 asafety/double check BGA measurements. Data are shown as mean ± SD or median (25th to 75th percentile). BGA, blood gas analyzer; CGM, continuous glucose monitoring; GLI, glucose lability index; IIT, intensive insulin therapy. asafety/double check BGA measurements. Data are shown as mean ± SD or median (25th to 75th percentile). BGA, blood gas analyzer; CGM, continuous glucose monitoring; GLI, glucose lability index; IIT, intensive insulin therapy. glucose. The loss of glucose complexity was found to be independently associated with mortality and with the presence of diabetes mellitus. was significantly lower when calculated from BGA com- pared to CGMS values (Table 4). Data are shown as mean ± SD or median (25th to 75th percentile). CGM, continuous glucose monitoring; GLI, glucose lability index; SD, standard deviation. Glucose variability Figure 1 Relation between glucose complexity and mortality. Relation between glucose complexity and mortality can be described best with a quadratic function showing a pronounced increase in mortality with higher DFA and a moderate increase in mortality with very low DFA. exogenous factors. It is associated with mortality and strategies are being sought for reducing glycemic variabil- ity. Improved glucose variability with real time CGM was reported in diabetic patients [25]. A possible explanation of real time CGM not reducing glycemic variability in the present study may be the use of an already well-estab- lished insulin titration algorithm in the control group. This algorithm has, in combination with the experienced nurses and frequent BGA measurements, already shown excellent results regarding glucose control. Hence, the use of real time CGM may have a larger benefit in envir- onments with less experienced and established ICU staff. Unlike numerous reports in the literature [3,4,8-10,26], we did not find a significant association between mortality and glycemic variability or between hypoglycemia and mortality, because our analysis was not powered for these purposes. The method and frequency of glucose determination has a significant impact on variability measures as a h c g w S b q r H fr fo G G o in a T Table 4 Impact of the method of glucose determination on va Variability measures derived from: CGMS value (174 patients Number of glucose measurements 140 209 GLI 178 ± 188 Mean daily delta (mmol/L) 4.65 ± 2.06 SD (mmol/L) 1.24 ± 0.52 BGA, blood gas analyzer; CV, coefficient of variation; GLI, glucose lability index: SD, st exogenous factors. It is associated with mortality and strategies are being sought for reducing glycemic variabil- ity. Improved glucose variability with real time CGM was reported in diabetic patients [25]. A possible explanation of real time CGM not reducing glycemic variability in the present study may be the use of an already well-estab- lished insulin titration algorithm in the control group. This algorithm has, in combination with the experienced nurses and frequent BGA measurements, already shown excellent results regarding glucose control. Hence, the use of real time CGM may have a larger benefit in envir- onments with less experienced and established ICU staff. already shown with mean absolute glucose change per hour [27]. Glucose variability Mean daily delta was naturally higher when calculated from CGMS compared to BGA values as the gap between minimum and maximum glucose increases with the number of measured values. The increase in SD, CV and GLI calculated from BGA values may be based on the fact that blood gases are taken more fre- quently when a patient’s glucose levels are unstable, resulting in virtually higher glucose variability values. However, glucose variability measures were calculated from CGM values in both groups in our study. There- fore, measures between these groups are comparable. Unlike numerous reports in the literature [3,4,8-10,26], we did not find a significant association between mortality and glycemic variability or between hypoglycemia and mortality, because our analysis was not powered for these purposes. Glucose variability In this post-hoc analysis of CGM data the use of a real time CGM did not have an impact on measures of gly- cemic variability, glucose complexity and maximum Glycemic variability describes fluctuations of blood glu- cose over time and is influenced by endogenous and Glycemic variability describes fluctuations of blood glu- cose over time and is influenced by endogenous and Table 3 Glycemic metrics in ICU survivors and non-survivors. Real Time CGM and controls [21,22] Survivors (number = 138) Non-survivors (number = 36) P-value Measures of glycemic variability Variability of glucose (SD) (mmol/L) 1.21 ± 0.49 1.39 ± 0.59 0.067 Variability of glucose (GLI) 112 (62 to 214) 126 (56 to 223) 0.468 Variability of glucose (δ) (mmol/L) 4.54 ± 2.00 5.08 ± 2.20 0.158 Coefficient of variation (%) 20 ± 7 22 ± 8 0.169 Variability of glucose during first 24 hours (SD) (mmol/L) 0.92 (0.64 to 1.32) 1.03 (0.73 to 1.46) 0.083 Variability of glucose during first 24 hours (GLI) 101 (54 to 200) 123 (54 to 198) 0.232 Variability of glucose during first 24 hours (δ) (mmol/L) 5.55 ± 2.23 6.42 ± 2.91 0.051 Measures of glucose Mean of glucose (mmol/L) 6.03 ± 0.57 6.23 ± 0.09 0.097 Mean of glucose during first 24 hours (mmol/L) 5.89 (5.40 to 6.36) 5.74 (5.27 to 6.39) 0.414 Maximum glucose (mmol/L) 9.52 ± 2.01 10.16 ± 2.82 0.121 Glucose complexity 1.51 ± 0.10 1.58 ± 0.14 0.003 Measures of hypoglycemia Time below 4.44 mmol/L (min/24hours) 109 (27 to 262) 120 (29 to240) 0.874 Time below 3.33 mmol/L (min/24hours) 0 (0 to 29) 0 (0 to 50) 0.864 Time below 2.22 mmol/L (min/24 hours) 0 (0 to 0) 0 (0 to 0) 0.116 Data are shown as mean ± SD or median (25th to 75th percentile). CGM, continuous glucose monitoring; GLI, glucose lability index; SD, standard deviation. Table 3 Glycemic metrics in ICU survivors and non-survivors. Data are shown as mean ± SD or median (25th to 75th percentile). CGM, continuous glucose monitoring; GLI, glucose lability index; SD, standard deviation. Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Page 6 of 9 Figure 1 Relation between glucose complexity and mortality. Relation between glucose complexity and mortality can be described best with a quadratic function showing a pronounced increase in mortality with higher DFA and a moderate increase in mortality with very low DFA. Glucose complexity Glucose complexity has been hypothesized as descriptive of the endogenous glucoregulatory process and is an independent predictor of mortality in critically ill patients as reported by Lundelin et al. for the first time [19]. These findings have now been confirmed in a larger The method and frequency of glucose determination has a significant impact on variability measures as Table 4 Impact of the method of glucose determination on variability measures. Variability measures derived from: CGMS values (174 patients) BGA values (174 patients) P value Number of glucose measurements 140 209 3497 GLI 178 ± 188 301 ± 380 <0.01 Mean daily delta (mmol/L) 4.65 ± 2.06 3.10 ± 1.50 <0.01 SD (mmol/L) 1.24 ± 0.52 1.35 ± 0.57 <0.01 BGA, blood gas analyzer; CV, coefficient of variation; GLI, glucose lability index: SD, standard deviation. Table 4 Impact of the method of glucose determination on variability measures. Variability measures derived from: CGMS values (174 patients) BGA values (174 patients) P value Number of glucose measurements 140 209 3497 GLI 178 ± 188 301 ± 380 <0.01 Mean daily delta (mmol/L) 4.65 ± 2.06 3.10 ± 1.50 <0.01 SD (mmol/L) 1.24 ± 0.52 1.35 ± 0.57 <0.01 BGA, blood gas analyzer; CV, coefficient of variation; GLI, glucose lability index: SD, standard deviation. Table 4 Impact of the method of glucose determination on variability measures. Variability measures derived from: CGMS values (174 patients) BGA values (174 patients) P value Number of glucose measurements 140 209 3497 GLI 178 ± 188 301 ± 380 <0.01 Mean daily delta (mmol/L) 4.65 ± 2.06 3.10 ± 1.50 <0.01 SD (mmol/L) 1.24 ± 0.52 1.35 ± 0.57 <0.01 BGA, blood gas analyzer; CV, coefficient of variation; GLI, glucose lability index: SD, standard deviation. Table 4 Impact of the method of glucose determination on variability measures. Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Page 7 of 9 Page 7 of 9 Thus, a systematic error based on the CGMS would influence both groups in the same way. patient group in a medical ICU. Loss of complexity in glucose time series was significantly associated with higher mortality. However, the relation between glucose complexity and mortality was not linear but can be described best with a quadratic function with a pro- nounced increase in mortality with higher DFA and a moderate increase in mortality with very low DFA. Glucose complexity In our opinion, the advantages of the real time CGM could not be fully utilized based on the following fac- tors. In our algorithm decisions are primarily based on the value of blood glucose, but not on the actual glucose trend [21]. However, glucose trend data is, in our opi- nion, one of the essential strengths of CGM. Therefore, we hypothesize that a (computer-based) algorithm using trend data for its decision process and capable of pro- cessing the great number of glucose values of CGM would be superior to the conventional algorithm used in our trials. The underlying hypothesis is that the ability of a healthy organism to detect even minor changes in glu- cose concentration and then to follow promptly with counter regulatory measures leads to a complex glucose profile. In contrast, an impaired regulatory system responds slowly and imprecisely to varying glucose con- centrations and, therefore, displays low glucose complex- ity [19]. The unexpected mortality increase with very complex profiles, which has not been discussed by Lun- delin et al. [19], cannot be explained by the present data and needs further investigation. Therefore, the biological explanation of the association between glucose complex- ity and mortality in critically ill patients should still be seen as a hypothesis. Moreover, the CGM devices in the study were used off- label and were originally designed for outpatients. There- fore, the display was very small and trend data could not be visualized. Due to the study design and because of the impossibility of making alarms adequately audible in an ICU environment alarm functions were not used. Devices compensating for these shortcomings are currently under development by several manufacturers, but were not available when we conducted our trial. Glucose complexity, but not SAPS II score, was signifi- cantly associated with mortality in a binary logistic regression analysis. However, this study was not powered to address this association. Consequently, despite the availability of real time data, the frequency of changes of the insulin infusion did not increase in the CGM group (Table 2). Furthermore, the number of BGA measurements was equal in both groups. However, BGA are not only used for glucose measurement in our ICU. Based on these data we con- clude that the use of our CGMS device did not have a significant impact on the behavior of the nursing staff in the real time CGM group compared to the control group. Strengths and weaknesses The present findings may be influenced by the accuracy and method of glucose monitoring. Glucose variability and inaccuracy of glucose monitoring may be positively correlated [30] and glycemic variability may be underes- timated with a higher time span between glucose mea- surements [31]. Glucose complexity Moreover, the complexity of the glycemic profile was significantly lower in diabetic, compared to non-diabetic, critically ill patients. This is consistent with several stu- dies assessing glucose complexity in non-critically ill dia- betic patients [28,29] and in critically ill patients after controlling for mortality [19]. Glucose complexity was similar between the real time CGM and the control group. We expected this finding, as glucose complexity reflects the endogenous fundamental glucose regulation, which seems autonomous from exogenous stimuli such as insulin treatment. Unlike in the study of Lundelin et al. [19], glucose complexity metrics were convincing in our analysis based on a relatively large patient group, standardized beginning of CGM and calibration of CGM devices with glucose values determined by very accurate blood gas analyzers [32]. Conclusions IIT guided by real time CGM did not result in signifi- cantly reduced glycemic variability. The loss of glucose complexity was significantly associated with mortality and with the presence of diabetes mellitus. Thus, glu- cose complexity is an excellent measure of the endogen- ous glucose regulation and a robust parameter of the severity of disease in critically ill patients. In the future - when continuous glucose monitoring may become stan- dard in the ICU - glucose complexity may add to clini- cal scores in this regard. We regard the estimation of glycemic variability in our study acceptable based on the following factors: the CGMS is relatively accurate, the accuracy of the CGMS is constant at all glucose levels [18] and the time span between glucose measurements is very small (5 minutes). Calculation of glucose complexity is possible only with CGM. The CGMS we used in the trial was the most accurate system available at that time. References 1. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R: Intensive insulin therapy in critically ill patients. N Engl J Med 2001, 345:1359-1367 1. van den Berghe G, Wouters P, Weekers F, Verwaest C, Bruyninckx F, Schetz M, Vlasselaers D, Ferdinande P, Lauwers P, Bouillon R: Intensive insulin therapy in critically ill patients. N Engl J Med 2001, 345:1359-1367. 24. Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, Van Wijngaerden E, Bobbaers H, Bouillon R: Intensive insulin therapy in the medical ICU. N Engl J Med 2006, 354:449-461. py y p g 2. Sacks DB: Therapy: intensive glucose control in the ICU: is sugar nice? Nat Rev Endocrinol 2009, 5:473-474. 2. Sacks DB: Therapy: intensive glucose control in the ICU: is sugar nice? Nat Rev Endocrinol 2009, 5:473-474. 25. Danne T, de Valk HW, Kracht T, Walte K, Geldmacher R, Solter L, von dem Berge W, Welsh ZK, Bugler JR, Lange K, Kordonouri O: Reducing glycaemic variability in type 1 diabetes self-management with a continuous glucose monitoring system based on wired enzyme technology. Diabetologia 2009, 52:1496-1503. 3. Egi M, Bellomo R, Stachowski E, French CJ, Hart G: Variability of blood glucose concentration and short-term mortality in critically ill patients. Anesthesiology 2006, 105:244-252. 3. Egi M, Bellomo R, Stachowski E, French CJ, Hart G: Variability of blood glucose concentration and short-term mortality in critically ill patients. Anesthesiology 2006, 105:244-252. gy 4. Mackenzie I, Whitehouse T, Nightingale P: The metrics of glycaemic control in critical care. Intensive Care Med 2011, 37:435-443. 26. Krinsley JS, Grover A: Severe hypoglycemia in critically ill patients: risk factors and outcomes. Crit Care Med 2007, 35:2262-2267. 26. Krinsley JS, Grover A: Severe hypoglycemia in critically ill patients: risk factors and outcomes. Crit Care Med 2007, 35:2262-2267. control in critical care. Intensive Care Med 2011, 37:435-443. 5. Ali NA, O’Brien JM, Dungan K, Phillips G, Marsh CB, Lemeshow 5. Ali NA, O’Brien JM, Dungan K, Phillips G, Marsh CB, Lemeshow S, Connors AF, Preiser J-C: Glucose variability and mortality in patients with sepsis. Crit Care Med 2008, 36:2316-2321. 27. Harmsen R, Van Braam Houckgeest F, Spronk P, Schultz M, Abu-Hanna A: Blood glucose variability, measured as mean absolute glucose, strongly depends on the frequency of blood glucose level measurements. Crit Care 2011, 15:P392. 6. Author details 1 f 1Department of Medicine III, Division of Gastroenterology and Hepatology, ICU 13H1, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria. 2Department of Emergency Medicine, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria. 18. Brunner R, Kitzberger R, Miehsler W, Herkner H, Madl C, Holzinger U: Accuracy and reliability of a subcutaneous continuous glucose- monitoring system in critically ill patients. Crit Care Med 2011, 39:659-664. 19. Lundelin K, Vigil L, Bua S, Gomez-Mestre I, Honrubia T, Varela M: Differences in complexity of glycemic profile in survivors and nonsurvivors in an intensive care unit: a pilot study. Crit Care Med 2010, 38:849-854. Acknowledgements We thank all the health professionals of the ICU 13H1 for their commitment to this study, as well as Dr. Manuel Varela for providing us with the program to calculate glucose complexity. 15. Krinsley JS: Glycemic variability and mortality in critically ill patients: the impact of diabetes. J Diabetes Sci Technol 2009, 3:1292-1301. This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector. Study equipment (two CGMS 16. Bochicchio GV, Scalea TM: Glycemic control in the ICU. Adv Surg 2008, 42:261-275. 16. Bochicchio GV, Scalea TM: Glycemic control in the ICU. Adv Surg 2008, 42:261-275. This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector. Study equipment (two CGMS Gold monitors, two Guardian monitors and 180 sensors) was donated by Medtronic Austria. the public, commercial or not-for-profit sector. Study equipment (two CGMS Gold monitors, two Guardian monitors and 180 sensors) was donated by Medtronic Austria. the public, commercial or not for profit sector. Study equipment (two CGMS Gold monitors, two Guardian monitors and 180 sensors) was donated by Medtronic Austria. 17. Corstjens AM, Ligtenberg JJM, van der Horst ICC, Spanjersberg R, Lind JSW, Tulleken JE, Meertens JHJM, Zijlstra JG: Accuracy and feasibility of point- of-care and continuous blood glucose analysis in critically ill ICU patients. Crit Care 2006, 10:R135. • The loss of glucose complexity is associated with mortality and diabetes mellitus in critically ill patients • The loss of glucose complexity is associated with mortality and diabetes mellitus in critically ill patients 11. Waeschle RM, Moerer O, Hilgers R, Herrmann P, Neumann P, Quintel M: The impact of the severity of sepsis on the risk of hypoglycaemia and glycaemic variability. Crit Care 2008, 12:R129. 12. Risso A, Mercuri F, Quagliaro L, Damante G, Ceriello A: Intermittent high glucose enhances apoptosis in human umbilical vein endothelial cells in culture. Am J Physiol Endocrinol Metab 2001, 281:924-930. Abbreviations bl d BGA: blood gas analyzer; CGM: continuous glucose monitoring; CGMS: continuous glucose monitoring system; CV: coefficient of variation; DFA: detrended fluctuation analysis; GLI: glucose lability index; GluSD: overall glucose variability measured by SD; IIT: Intensive insulin therapy. 13. Monnier L, Mas E, Ginet C, Michel F, Villon L, Cristol J-P, Colette C: Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia in patients with type 2 diabetes. JAMA 2006, 295:1681-1687. 14. Hirsch IB, Brownlee M: Should minimal blood glucose variability become the gold standard of glycemic control? J Diabetes Complications 2005, 19:178-181. Authors’ contributions RB collected data, carried out the statistical analyses and interpretation, and drafted the manuscript. GA collected data and revised the manuscript critically. HH performed statistical analyses. CM and UH designed and coordinated the study, collected data and helped to draft the manuscript. All authors read and approved the final manuscript. 20. Goldstein B, Fiser DH, Kelly MM, Mickelsen D, Ruttimann U, Pollack MM: Decomplexification in critical illness and injury: relationship between heart rate variability, severity of illness, and outcome. Crit Care Med 1998, 26:352-357. 21. Holzinger U, Warszawska J, Kitzberger R, Herkner H, Metnitz PGH, Madl C: Impact of shock requiring norepinephrine on the accuracy and reliability of subcutaneous continuous glucose monitoring. Intensive Care Med 2009, 35:1383-1389. Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Competing interests The authors declare that they have no competing interests. All authors were paid by the Medical University of Vienna. The authors declare that they have no competing interests. All authors were paid by the Medical University of Vienna. 22. Holzinger U, Warszawska J, Kitzberger R, Wewalka M, Miehsler W, Herkner H, Madl C: Real time continuous glucose monitoring in critically ill patients - a prospective, randomized trial. Diabetes Care 2010, 33:467-72. Received: 30 April 2012 Revised: 10 September 2012 A d O b P bli h d O b Received: 30 April 2012 Revised: 10 September 2012 Accepted: 2 October 2012 Published: 2 October 2012 Accepted: 2 October 2012 Published: 2 October 2012 23. Holzinger U, Feldbacher M, Bachlechner A, Kitzberger R, Fuhrmann V, Madl C: Improvement of glucose control in the intensive care unit: an interdisciplinary collaboration study. Am J Crit Care 2008, 17:150-156. References Dossett LA, Cao H, Mowery NT, Dortch MJ, Morris JM, May AK: Blood glucose variability is associated with mortality in the surgical intensive care unit. Am Surg 2008, 74:679-685. 28. Churruca J, Vigil L, Luna E, Ruiz-Galiana J, Varela M: The route to diabetes: Loss of complexity in the glycemic profile from health through the metabolic syndrome to type 2 diabetes. Diabetes Metab Syndr Obes 2008, 1:3-11. g 7. Meyfroidt G, Keenan DM, Wang X, Wouters PJ, Veldhuis JD, Van den Berghe G: Dynamic characteristics of blood glucose time series during the course of critical illness: effects of intensive insulin therapy and relative association with mortality. Crit Care Med 2010, 38:1021-1029. 8. Krinsley JS: Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med 2008, 36:3008-3013. 9. Hermanides J, Vriesendorp TM, Bosman RJ, Zandstra DF, Hoekstra JB, g 7. Meyfroidt G, Keenan DM, Wang X, Wouters PJ, Veldhuis JD, Van den Berghe G: Dynamic characteristics of blood glucose time series during the course of critical illness: effects of intensive insulin therapy and relative association with mortality. Crit Care Med 2010, 38:1021-1029. 29. Ogata H, Tokuyama K, Nagasaka S, Ando A, Kusaka I, Sato N, Goto A, Ishibashi S, Kiyono K, Struzik ZR, Yamamoto Y: Long-range negative correlation of glucose dynamics in humans and its breakdown in diabetes mellitus. Am J Physiol Regul Integr Comp Physiol 2006, 291:1638-1643. y 8. Krinsley JS: Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med 2008, 36:3008-3013. 8. Krinsley JS: Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med 2008, 36:3008-3013. 9. Hermanides J, Vriesendorp TM, Bosman RJ, Zandstra DF, Hoekstra JB, DeVries JH: Glucose variability is associated with intensive care unit mortality. Crit Care Med 2010, 38:838-842. 30. Breton MD, Kovatchev BP: Impact of blood glucose self-monitoring errors on glucose variability, risk for hypoglycemia, and average glucose control in type 1 diabetes: an in silico study. J Diabetes Sci Technol 2010, 4:562-570. 10. Meynaar IA, Eslami S, Abu-Hanna A, van der Voort P, de Lange DW, de Keizer N: Blood glucose amplitude variability as predictor for mortality in surgical and medical intensive care unit patients: a multicenter cohort study. J Crit Care 2012, 27:119-124. 31. Kalfon P, Chilles M: Impact of the type of glucose monitoring on the assessment of glycemic variability in critical care patients. Key messages • IIT guided by real time CGM did not lead to reduced glycemic variability Furthermore, glucose variability and complexity mea- sures are calculated from CGMS values in all patients. Page 8 of 9 Page 8 of 9 References Crit Care 2012, 16:P169. 31. Kalfon P, Chilles M: Impact of the type of glucose monitoring on the assessment of glycemic variability in critical care patients. Crit Care 2012, 16:P169. Page 9 of 9 Page 9 of 9 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 Brunner et al. Critical Care 2012, 16:R175 http://ccforum.com/content/16/5/R175 32. Corstjens AM, Ligtenberg JJ, van der Horst IC, Spanjersberg R, Lind JS, Tulleken JE, Meertens JH, Zijlstra JG: Accuracy and feasibility of point-of- care and continuous blood glucose analysis in critically ill ICU patients. Crit Care 2006, 10:R135. doi:10.1186/cc11657 Cite this article as: Brunner et al.: Glycemic variability and glucose complexity in critically ill patients: a retrospective analysis of continuous glucose monitoring data. Critical Care 2012 16:R175. 32. Corstjens AM, Ligtenberg JJ, van der Horst IC, Spanjersberg R, Lind JS, Tulleken JE, Meertens JH, Zijlstra JG: Accuracy and feasibility of point-of- care and continuous blood glucose analysis in critically ill ICU patients. Crit Care 2006, 10:R135. doi:10.1186/cc11657 Cite this article as: Brunner et al.: Glycemic variability and glucose complexity in critically ill patients: a retrospective analysis of continuous glucose monitoring data. Critical Care 2012 16:R175. Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit
https://openalex.org/W4206898544	https://peerj.com/articles/12726v0.3/submission	English	null	Peer Review #3 of "Prediction of active ingredients in Salvia miltiorrhiza Bunge. based on soil elements and artificial neural network (v0.1)"	null	2,022	cc-by	11,104	Prediction of active ingredients in Salvia miltiorrhiza Bunge. based on soil elements and artiﬁcial neural network Yu Liu 1 , Ke Wang 1, 2 , Zhuyun Yan Corresp., 1 , Xiaofeng Shen 3 , Xinjie Yang 4 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, School of Pharmacy, Che Medicine, Chengdu, Sichuan, China 2 School of Big Data and Artiﬁcial Intelligence, Chengdu Technological University, Chengdu, Sichuan, China 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, China 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China Corresponding Author: Zhuyun Yan Email address: yanzhuyun@cdutcm.edu.cn 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, China 2 School of Big Data and Artiﬁcial Intelligence, Chengdu Technological University, Chengdu, Sichuan, China 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, China 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China Corresponding Author: Zhuyun Yan Email address: yanzhuyun@cdutcm.edu.cn 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, China 2 School of Big Data and Artiﬁcial Intelligence, Chengdu Technological University, Chengdu, Sichuan, China 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, China 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China Corresponding Author: Zhuyun Yan Email address: yanzhuyun@cdutcm.edu.cn The roots of Salvia miltiorrhiza Bunge. are commonly used in the treatment of cardiovascular diseases, and tanshinones and salvianolic acids are its main active ingredients. However, the composition and content of active ingredients of S. miltiorrhiza planted in diﬀerent regions of the soil environment are also quite diﬀerent, which adds new diﬃculties to the large-scale and standardization of artiﬁcial cultivation. Therefore, in this study, we measured the active ingredients in the roots of S. miltiorrhiza and the contents of rhizosphere soil elements from 25 production areas in 8 provinces in China, and used the data to develop a prediction model based on BP (back propagation) neural network. The results showed that the active ingredients had diﬀerent degrees of correlation with soil macronutrients and trace elements, the prediction model had the best performance (MSE=0.0203, 0.0164; R 2 =0.93, 0.94). The artiﬁcial neural network model was shown to be a method that can be used to screen the suitable cultivation sites and proper fertilization. It can also be used to optimize the fertilizer application at speciﬁc sites. It also suggested that soil testing formula fertilization should be carried out for medicinal plants like S. Manuscript to be reviewed 1 Prediction of active ingredients in Salvia miltiorrhiza 2 Bunge. based on soil elements and artificial neural 3 network It PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 4 Yu Liu1, Ke Wang1,2, Zhuyun Yan1, Xiaofeng Shen3, Xinjie Yang4 5 6 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, 7 School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 8 China 9 2 School of Big Data and Artificial Intelligence, Chengdu Technological University, Chengdu, 10 Sichuan, China 11 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, 12 China 13 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China 14 15 Corresponding Author: 16 Zhuyun Yan1 17 Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, P.R. China 18 Email address: yanzhuyun@cdutcm.edu.cn 19 20 Abstract 21 The roots of Salvia miltiorrhiza Bunge. are commonly used in the treatment of cardiovascular 22 diseases, and tanshinones and salvianolic acids are its main active ingredients. However, the 23 composition and content of active ingredients of S. miltiorrhiza planted in different regions of 24 the soil environment are also quite different, which adds new difficulties to the large-scale and 25 standardization of artificial cultivation. Therefore, in this study, we measured the active 26 ingredients in the roots of S. miltiorrhiza and the contents of rhizosphere soil elements from 25 27 production areas in 8 provinces in China, and used the data to develop a prediction model based 28 on BP (back propagation) neural network. The results showed that the active ingredients had 29 different degrees of correlation with soil macronutrients and trace elements, the prediction mode 2 17 Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, P.R. China 18 Email address: yanzhuyun@cdutcm.edu.cn Prediction of active ingredients in Salvia miltiorrhiza Bunge. based on soil elements and artiﬁcial neural network Yu Liu 1 , Ke Wang 1, 2 , Zhuyun Yan Corresp., 1 , Xiaofeng Shen 3 , Xinjie Yang 4 miltiorrhiza, which is grown in multiple origins, rather than promoting the use of "special fertilizer" on a large scale. Therefore, the model is helpful for eﬃcient, rational, and scientiﬁc guidance of fertilization management in the cultivation of S. miltiorrhiza. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed 1 Prediction of active ingredients in Salvia miltiorrhiza 2 Bunge. based on soil elements and artificial neural 3 network 4 Yu Liu1, Ke Wang1,2, Zhuyun Yan1, Xiaofeng Shen3, Xinjie Yang4 5 6 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, 7 School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 8 China 9 2 School of Big Data and Artificial Intelligence, Chengdu Technological University, Chengdu, 10 Sichuan, China 11 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, 12 China 13 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China 14 15 Corresponding Author: 16 Zhuyun Yan1 17 Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, P.R. China 18 Email address: yanzhuyun@cdutcm.edu.cn 19 1 Prediction of active ingredients in Salvia miltiorrhiza 2 Bunge. based on soil elements and artificial neural 3 network 2 Bunge. based on soil elements and artificial neural 3 network 4 Yu Liu1, Ke Wang1,2, Zhuyun Yan1, Xiaofeng Shen3, Xinjie Yang4 5 6 1 State Key Laboratory of Characteristic Chinese Medicine Resources in Southwest China, 7 School of Pharmacy, Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 8 China 9 2 School of Big Data and Artificial Intelligence, Chengdu Technological University, Chengdu, 10 Sichuan, China 11 3 Institute of Medicinal Plant Development, Chinese Academy of Medical Sciences, Beijing, 12 China 13 4 College of Pharmacy, Shaanxi University of Chinese Medicine, Xianyang, Shaanxi, China 14 15 Corresponding Author: 16 Zhuyun Yan1 17 Chengdu University of Traditional Chinese Medicine, Chengdu, Sichuan, 611137, P.R. China 18 Email address: yanzhuyun@cdutcm.edu.cn 19 20 Abstract 21 The roots of Salvia miltiorrhiza Bunge. are commonly used in the treatment of cardiovascular 22 diseases, and tanshinones and salvianolic acids are its main active ingredients. However, the 23 composition and content of active ingredients of S. miltiorrhiza planted in different regions of 24 the soil environment are also quite different, which adds new difficulties to the large-scale and 25 standardization of artificial cultivation. Therefore, in this study, we measured the active 26 ingredients in the roots of S. miltiorrhiza and the contents of rhizosphere soil elements from 25 27 production areas in 8 provinces in China, and used the data to develop a prediction model based 28 on BP (back propagation) neural network. The results showed that the active ingredients had 29 different degrees of correlation with soil macronutrients and trace elements, the prediction mod 30 had the best performance (MSE=0.0203, 0.0164; R2 =0.93, 0.94). The artificial neural network 31 model was shown to be a method that can be used to screen the suitable cultivation sites and 32 proper fertilization. It can also be used to optimize the fertilizer application at specific sites. Manuscript to be reviewed 33 also suggested that soil testing formula fertilization should be carried out for medicinal plants 34 like S. miltiorrhiza, which is grown in multiple origins, rather than promoting the use of "special 35 fertilizer" on a large scale. Therefore, the model is helpful for efficient, rational, and scientific 36 guidance of fertilization management in the cultivation of S. miltiorrhiza. 33 also suggested that soil testing formula fertilization should be carried out for medicinal plants 34 like S. miltiorrhiza, which is grown in multiple origins, rather than promoting the use of "special 35 fertilizer" on a large scale. Therefore, the model is helpful for efficient, rational, and scientific 36 guidance of fertilization management in the cultivation of S. miltiorrhiza. 33 also suggested that soil testing formula fertilization should be carried out for medicinal plants 34 like S. miltiorrhiza, which is grown in multiple origins, rather than promoting the use of "special 35 fertilizer" on a large scale. Therefore, the model is helpful for efficient, rational, and scientific 36 guidance of fertilization management in the cultivation of S. miltiorrhiza. 20 Abstract 21 The roots of Salvia miltiorrhiza Bunge. are commonly used in the treatment of cardiovascular 22 diseases, and tanshinones and salvianolic acids are its main active ingredients. However, the 23 composition and content of active ingredients of S. miltiorrhiza planted in different regions of 24 the soil environment are also quite different, which adds new difficulties to the large-scale and 25 standardization of artificial cultivation. Therefore, in this study, we measured the active 26 ingredients in the roots of S. miltiorrhiza and the contents of rhizosphere soil elements from 25 27 production areas in 8 provinces in China, and used the data to develop a prediction model based 28 on BP (back propagation) neural network. The results showed that the active ingredients had 29 different degrees of correlation with soil macronutrients and trace elements, the prediction model 30 had the best performance (MSE=0.0203, 0.0164; R2 =0.93, 0.94). The artificial neural network 31 model was shown to be a method that can be used to screen the suitable cultivation sites and 32 proper fertilization. It can also be used to optimize the fertilizer application at specific sites. It PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 37 Introduction 38 Salvia miltiorrhiza Bunge. is a persistent herb of the genus Salvia of the mint family, 39 Lamiaceae (Chinese Flora Commission of the Chinese Academy of Sciences 1974). In China, 40 Japan, and the United States (Su et al. 2015), its dried roots are widely used as one of the most 41 commonly used traditional medicinal herbs to treat cardiovascular diseases, especially angina 42 pectoris and myocardial infarction (Wang et al. 2017). Researches have shown that diterpenoid 43 quinones and hydrophilic phenolic acids are its principal bioactive components (Mei et al. 2019). 44 As cardiovascular disease is a common and frequently-occurring disease (Jagannathan et al. 45 2019), the wild resources are unable to meet the ever-increasing market demand. Currently, 19 46 provinces in China, including Shandong, Henan, Sichuan, Hebei, Hubei, Jiangsu, Shanxi, and 47 Shaanxi (Lu 2019), have been introduced and cultivated S. miltiorrhiza, but there are significant 48 differences in the content of active ingredients in the same species planted in different regions 49 (Huang et al. 2019; Yang et al. 2011b). Soil is the material basis for the survival of plants, and 50 the abundance and deficiency of major and trace elements in soil affect plant growth and 51 development and physiological and biochemical metabolism, which also affect the yield and the 52 composition and content of active ingredients in medicinal plants (Zhang et al. 2018). 53 Fertilization is the primary way often used in agricultural production to improve soil nutrition, 54 and proper fertilization is an important measure to ensure the yield and quality of medicinal 55 plants. Therefore, analyzing the relationship between soil elements and active components of 56 medicinal plants and establishing related models can provide data support and technical guidance 57 for selecting suitable sites and fertilization management in the cultivation of S. miltiorrhiza. 58 Although the use of models based on the relationship between soil environment and crop 59 quality or the quality of medicinal plants to predict suitable land, crop yield, and quality has 60 received widespread attention, most studies have used regression analysis models to predict the 61 relationship between the two (Li et al. 2020). Regression analysis models are built on the 62 assumption of some idealized linear relationship between predictor and response variables. Manuscript to be reviewed 68 different disciplines (e.g., in vitro culture (Hesami & Jones 2020), remote sensing studies (Wang 69 et al. 2019)). The artificial neural network methods can learn and create nonlinear and complex 70 relationships and can flexibly solve the complex problems of multiple interacting variables 71 (Bayat et al. 2019). It is one of the best techniques for extracting information from inaccurate and 72 nonlinear data (Caselli et al. 2009). Among them, BP neural network model is the most widely 73 used neural network model in the fields of economics, engineering, and botany (Armaghani et al. 74 2018; Chen et al. 2019). The BP neural network has been used to monitor crop growth and crop 75 yield prediction (Akbar et al. 2018; Wang et al. 2019). However, few reports on the application 76 of the BP neural network predict active ingredients of medicinal plants based on soil elements. 77 Therefore, in this study, we took S. miltiorrhiza as the research object, measured the active 78 ingredients in the roots of S. miltiorrhiza. and the contents of rhizosphere soil elements from 25 79 production areas in 8 provinces in China, analyzed the relationship between them, and 80 established a BP neural network model for the prediction of active ingredients using soil 81 elements as input values, which expands the application of artificial neural network methods in 82 the field of medicinal botany and also provides references for its application in other directions 83 in the field of medicinal plant cultivation, and verifies the feasibility of using artificial neural 84 network model to effectively improve the accuracy of the prediction of the content of active 85 ingredients of plants. 37 Introduction 63 However, environmental factors show very high intra- and inter-individual variability, which 64 means that the biological responses of plants in response to the environment are uncertain and 65 nonlinear in nature (Gago et al. 2010). Therefore, many biological interactions cannot be 66 explained by simple stepwise algorithms or exact formulations, especially in complex or noisy 67 data. Artificial neural network methods have proven effective in solving such problems in 38 Salvia miltiorrhiza Bunge. is a persistent herb of the genus Salvia of the mint family, 39 Lamiaceae (Chinese Flora Commission of the Chinese Academy of Sciences 1974). In China, 40 Japan, and the United States (Su et al. 2015), its dried roots are widely used as one of the most 41 commonly used traditional medicinal herbs to treat cardiovascular diseases, especially angina 42 pectoris and myocardial infarction (Wang et al. 2017). Researches have shown that diterpenoid 43 quinones and hydrophilic phenolic acids are its principal bioactive components (Mei et al. 2019). 44 As cardiovascular disease is a common and frequently-occurring disease (Jagannathan et al. 45 2019), the wild resources are unable to meet the ever-increasing market demand. Currently, 19 46 provinces in China, including Shandong, Henan, Sichuan, Hebei, Hubei, Jiangsu, Shanxi, and 47 Shaanxi (Lu 2019), have been introduced and cultivated S. miltiorrhiza, but there are significant 48 differences in the content of active ingredients in the same species planted in different regions 49 (Huang et al. 2019; Yang et al. 2011b). Soil is the material basis for the survival of plants, and 50 the abundance and deficiency of major and trace elements in soil affect plant growth and 51 development and physiological and biochemical metabolism, which also affect the yield and the 52 composition and content of active ingredients in medicinal plants (Zhang et al. 2018). PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 103 Determination of the content of active ingredients 103 Determination of the content of active ingredients 104 A high-performance liquid chromatographic method was used to determine the contents of 105 water-soluble components and lipid-soluble components (Yang et al. 2010; Yang et al. 2011a). 106 The working parameters for the determination of water-soluble components was as follows: the 107 column was a phenomenex Gemini C18 column (250×4.6mm, 5µm, Guangzhou Philomen 108 Scientific Instruments Co., Ltd.); the mobile phases were water-acetonitrile-formic acid 109 (90:10:0.4) (phase A) and acetonitrile (phase B), with gradient elution, phase A: 0~40min, 110 100%~70%; phase B: 0~40min, 0%~30%; the detection wavelength was 280 nm; the flow rate 111 was 1 mL/min; the column temperature was 25 °C; the injection volume was 20 µL. The 112 working parameters for the determination of liposoluble components were as follows: the 113 column was a Welchrom™ C18 column (Analytial 4.6×250 mm, 5 µm, Welch Corporation, 114 USA); the mobile phases were methanol (A phase) and water (B phase), with gradient elution, A 115 phase: 0~25 min, 67%~67%; 25~45 min, 67%~90%; B phase: 0-25 min, 33%-33%; 25-45 min, 116 33%-10%. The detection wavelength was 270 nm; the flow rate was 1 mL/min; the column 117 temperature was 25 °C; the injection volume was 20 µL. 87 Sample collection and processing 87 Sample collection and processing 88 Roots and rhizosphere soil of S. miltiorrhiza (cultivation or wild) were collected from 25 89 producing areas in 8 provinces of China from mid-November to early December 2007 (Fig. 1, 90 Table. S1). The samples were collected by the "S" parallel sampling and multi-point mixing 91 method, each sample was collected at 20 to 25 points, and the rhizosphere soil was collected by 92 the root shaking method, and finally, the samples for analysis were obtained by the quadratic 93 method, of which 2 kg/sample of rhizosphere soil and 5 kg/sample of medicinal parts were 94 retained. After the samples were collected and quickly transported back to the laboratory, the 95 herbs were processed routinely, and the soil samples were air-dried and prepared. 96 Determination of inorganic elements 97 The available potassium (K), copper (Cu), zinc (Zn), and manganese (Mn) were determined 98 by atomic absorption spectrophotometry after ammonium bicarbonate- 99 diethylenetriaminepentaacetic acid (AB-DTPA) extraction (Lu 2000); the available nitrogen (N) 100 was determined by alkaline diffusion method (Lin 2004); available phosphorus (P) was 101 determined by sodium bicarbonate extraction-molybdenum antimony anti-colorimetric method 102 (Zhao et al. 2020). 88 Roots and rhizosphere soil of S. miltiorrhiza (cultivation or wild) were collected from 25 89 producing areas in 8 provinces of China from mid-November to early December 2007 (Fig. 1, 90 Table. S1). The samples were collected by the "S" parallel sampling and multi-point mixing 91 method, each sample was collected at 20 to 25 points, and the rhizosphere soil was collected by 92 the root shaking method, and finally, the samples for analysis were obtained by the quadratic 93 method, of which 2 kg/sample of rhizosphere soil and 5 kg/sample of medicinal parts were 94 retained. After the samples were collected and quickly transported back to the laboratory, the 95 herbs were processed routinely, and the soil samples were air-dried and prepared. 97 The available potassium (K), copper (Cu), zinc (Zn), and manganese (Mn) were determined 98 by atomic absorption spectrophotometry after ammonium bicarbonate- 99 diethylenetriaminepentaacetic acid (AB-DTPA) extraction (Lu 2000); the available nitrogen (N) 100 was determined by alkaline diffusion method (Lin 2004); available phosphorus (P) was 101 determined by sodium bicarbonate extraction-molybdenum antimony anti-colorimetric method 102 (Zhao et al. 2020). PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 118 BP (back propagation) neural network 119 Bootstrap is a statistical inference method based on resampling and data simulation. Due to 120 the nonlinear nature of small sample sizes and the difficulty of characterizing the overall 121 distribution, the Bootstrap method can be applied to improve the estimation accuracy of the 122 model (Wang et al. 2018). Therefore, the Bootstrap method has the potential to be widely applied 123 to modeling estimations of small sample sizes. The steps are described as follows: (i) perform 124 resampling to select a certain number (given) of samples and to allow repeat sampling; (ii) 125 calculate the given statistics T based on the given samples; and (iii) repeat the above steps N 126 times to gain N number of statistics T (Wang et al. 2018). The back propagation neural network 127 proposed by Rumelhart in 1986 is a multi-layer feedforward network trained according to the 128 error back propagation algorithm and is one of the most widely used neural network models 129 (Rumelhart et al. 1986). The BP neural network model algorithm consists of two aspects: the 130 forward propagation of the signal and the back propagation of the error. In other words, the 131 actual output is calculated according to the direction from input to output. However, when the 132 actual output contradicts the expected output, the back propagation of error is performed 133 according to the direction from output to input, and the output error of each layer neuron is 134 calculated layer by layer. Then, the weights and thresholds of each layer are adjusted according 135 to the error gradient descent method to achieve the final output value that can be close to the 136 expected value. The model-building process was described in detail as follows. 137 In this study, a BP neural network model was constructed using the neural fitting tool 138 (nftool) in MATLAB 2019b mathematical software and trained with soil element parameters as 137 In this study, a BP neural network model was constructed using the neural fitting tool 138 (nftool) in MATLAB 2019b mathematical software and trained with soil element parameters as PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed 139 input and the content of plant active ingredients as output. The input variables were as follows: 140 Mn, Cu, Zn, N, P, and K; output variables: water-soluble components (danshensu, protocatechuic 141 aldehyde, caffeic acid, rosmarinic acid, salvianolic acid B, salvianolic acid A), and lipid-soluble 142 components (dihydrotanshinone I, cryptotanshinone, tanshinone I, and tanshinone IIA). In the 143 solution process, 70% of the samples (n = 18) were used to obtain samples using the bootstrap 144 method and build a BP neural network model. The remaining 30% of the samples (n = 8) were 145 used to verify the model. First, to prevent the negative impact of different ranges of input 146 variables on the model's efficiency, the input variables in the model were therefore normalized 147 for the specified range (0-1) (Equation 1): 𝐹𝑗= 𝑋𝑗‒ 𝑋𝑚𝑖𝑛 𝑋𝑚𝑎𝑥‒ 𝑋𝑚𝑖𝑛 𝐹𝑗= 𝑋𝑗‒ 𝑋𝑚𝑖𝑛 𝑋𝑚𝑎𝑥‒ 𝑋𝑚𝑖𝑛 (1) 𝐹𝑗= 𝑋𝑗‒ 𝑋𝑚𝑖𝑛 𝑋𝑚𝑎𝑥‒ 𝑋𝑚𝑖𝑛 (1) (1) 148 In the equation, Fj, Xj, Xmin and Xmax are the standardized value, original value, minimum 149 value and maximum value of the input variable respectively. 148 In the equation, Fj, Xj, Xmin and Xmax are the standardized value, original value, minimum 149 value and maximum value of the input variable respectively. 150 Second, the number of hidden layers and the number of neurons in each hidden layer in a 151 BP neural network impact the overall neural network structure. Currently, many empirical 152 formulas are applied to determine the number of neurons in the hidden layer, and one of these 153 formulas is as follows (Equation 2): 150 Second, the number of hidden layers and the number of neurons in each hidden layer in a 151 BP neural network impact the overall neural network structure. Currently, many empirical 152 formulas are applied to determine the number of neurons in the hidden layer, and one of these 153 formulas is as follows (Equation 2): ℎ= 𝑚+ 𝑛+ 𝑎 (2) ℎ= 𝑚+ 𝑛+ 𝑎 ℎ= 𝑚+ 𝑛+ 𝑎 (2) (2) 154 In the equation, m (m=6) is the number of nodes in the input layer, n (n=2) is the number of 155 nodes in the output layer, and a (1≤a≤10) is a constant. 154 In the equation, m (m=6) is the number of nodes in the input layer, n (n=2) is the number of 155 nodes in the output layer, and a (1≤a≤10) is a constant. Manuscript to be reviewed 156 According to equation (2), the number of nodes in the hidden layer was set as an integer 157 between 4 and 12. Eventually, the number of neurons in the hidden layer was determined to be 8 158 by iterative trials. A neural network model with 6-8-2 structure was finally established (Fig. 2), 159 in which the input layer consisted of 6 neurons corresponding to the 6 input variables, and the 160 output layer had 2 neurons representing the content of active components in the model. 161 Third, another problem in establishing neural network models is the choice of network 162 learning or training algorithms. Since the Levenberg-Marquardt algorithm minimizes the sum of 163 the error function of the form (Equation 3), thus the Levenberg-Marquardt algorithm has the best 164 performance (R-value) compared to other training algorithms (i.e., Bayesian regularization (BR) 165 and scaled conjugate gradient (SCG)) (Mahmoudi & Mahmoudi 2014). Therefore, the 166 Levenberg-Marquardt algorithm will be used to train the network. The training epochs, learning 167 rate, and minimum performance gradient were set as 1000, 0.01, and 1e-7. 156 According to equation (2), the number of nodes in the hidden layer was set as an integer 157 between 4 and 12. Eventually, the number of neurons in the hidden layer was determined to be 8 158 by iterative trials. A neural network model with 6-8-2 structure was finally established (Fig. 2), 159 in which the input layer consisted of 6 neurons corresponding to the 6 input variables, and the 160 output layer had 2 neurons representing the content of active components in the model. 161 Third, another problem in establishing neural network models is the choice of network 162 learning or training algorithms Since the Levenberg Marquardt algorithm minimizes the sum of 156 According to equation (2), the number of nodes in the hidden layer was set as an integer 157 between 4 and 12. Eventually, the number of neurons in the hidden layer was determined to be 8 158 by iterative trials. A neural network model with 6-8-2 structure was finally established (Fig. 2), 159 in which the input layer consisted of 6 neurons corresponding to the 6 input variables, and the 160 output layer had 2 neurons representing the content of active components in the model 160 output layer had 2 neurons representing the content of active components in the model. Manuscript to be reviewed Manuscript to be reviewed Where ek is the error in the kth exemplar or pattern and e is a vector with element 168 Where ek is the error in the kth exemplar or pattern and e is a vector with element ek. 169 During the BP training process, a sigmoid function (Equation 4) is used to describe the 170 nonlinear relationship between the input and output of each neuron in the hidden layer as follows 171 (Yi et al. 2007). 169 During the BP training process, a sigmoid function (Equation 4) is used to describe the 170 nonlinear relationship between the input and output of each neuron in the hidden layer as follows 171 (Yi et al. 2007). 𝑓(𝑥) = 1 1 + 𝑒𝑥 (4) (4) 172 The output yj of the hidden layer neuron j is calculated by Equation 5: 172 The output yj of the hidden layer neuron j is calculated by Equation 5: The output yj of the hidden layer neuron j is calculated by Equation 5: 172 𝑦𝑗= ∅( 𝑛 ∑ 𝑖= 1 𝑊𝑖𝑗𝑋𝑖+ 𝜃𝑗) 𝑦𝑗= ∅( 𝑛 ∑ 𝑖= 1 𝑊𝑖𝑗𝑋𝑖+ 𝜃𝑗) (5) (5) 173 173 In the equation, ø(x) represents the activation function of the hidden layer, n is the number 174 of neurons in the input layer, Xi is the input of the input layer neuron i, Wij is the weight from the 175 input layer neuron i to the hidden layer neuron j, and θj is the threshold value of the hidden layer 176 neuron j. 𝑧𝑘= 𝑓( 𝑚 ∑ 𝑗= 1 𝑉𝑗𝑘𝑦𝑗+ 𝛼𝑘) 𝑧𝑘= 𝑓( 𝑚 ∑ 𝑗= 1 𝑉𝑗𝑘𝑦𝑗+ 𝛼𝑘) (6) (6) 178 In the equation, f(x) represents the activation function of the output layer, and m is the 179 number of neurons in the hidden layer, Vjk is the weight of the hidden layer neuron j to the output 180 layer neuron k, and αk is the threshold value of the output layer neuron k. 181 The network weights adjustment is defined as follows (Equation 7) (Huang et al. Manuscript to be reviewed 𝐸= 1 2∑𝑘(𝑒𝑘)2 = 1 2‖𝑒‖2 (3) 𝐸= 1 2∑𝑘(𝑒𝑘)2 = 1 2‖𝑒‖2 (3) PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Manuscript to be reviewed 192 To evaluate the predictive power of the model, the MSE (mean square error) and the R2 193 (coefficient of determination) were used to evaluate the overall performance of the model 194 (Equation 8-9): 192 To evaluate the predictive power of the model, the MSE (mean square error) and the R2 193 (coefficient of determination) were used to evaluate the overall performance of the model 194 (Equation 8-9): 192 To evaluate the predictive power of the model, the MSE (mean square error) and the R2 193 (coefficient of determination) were used to evaluate the overall performance of the model 194 (Equation 8-9): ( ) p 194 (Equation 8-9): 𝑅2 = 1 ‒∑ n i = 1 (𝑌i ‒ 𝑌j)2 ∑ n i = 1 (𝑌𝑖- 𝑌𝑖)2 𝑀𝑆𝐸= 1 𝑛∑ 𝑛 𝑖= 1(𝑌𝑖‒ 𝑌𝑖)2 (8) 𝑀𝑆𝐸= 1 𝑛∑ 𝑛 𝑖= 1(𝑌𝑖‒ 𝑌𝑖)2 (9) 𝑀𝑆𝐸= 1 𝑛∑ 𝑛 𝑖= 1(𝑌𝑖‒ 𝑌𝑖)2 (9) (9) 195 In the equation, Yi is the experimental value of the evaluation model, Ŷ 195 In the equation, Yi is the experimental value of the evaluation model, Ŷi is the corresponding 196 predicted data, Ῡi is the average of the experimental data and n is the number of experimental 197 data. 195 In the equation, Yi is the experimental value of the evaluation model, Ŷi is the corresponding 196 predicted data, Ῡi is the average of the experimental data and n is the number of experimental 197 data. 195 In the equation, Yi is the experimental value of the evaluation model, Ŷi is the corresponding 196 predicted data, Ῡi is the average of the experimental data and n is the number of experimental 197 data. 198 Data analysis 199 In this study, correlation plots of the active ingredients of S. miltiorrhiza and rhizosphere 200 soil elements were created using the “corrplot” package (version 0.9; cran.r- 201 project.org/web/packages/corrplot/index.html) in R. The Pearson's correlation test was used to 202 demonstrate the correlation between the active ingredients of S. miltiorrhiza and rhizosphere soil 203 elements. P<0.05 was considered to indicate a statistically significant difference. BP neural 204 network model was created using the neural fitting tool (nftool) in MATLAB 2019b 205 mathematical software (MathWorks Corporation, America). Scatter and regression plots were 206 created using the “ggplot2” package (version 3.3.5; cran.r- 207 project.org/web/packages/ggplot2/index.html) in R. The study used the kappa value to evaluate 208 the multicollinearity of characteristic indicators by the “Kappa” package in R. The map plot was 209 created using the “plyr” package (version 1.8.6; cran.r- 210 project.org/web/packages/plyr/index.html), “maptools” package (version 1.1-2; cran.r- 211 project.org/web/packages/maptools/index.html), and “ggplot2” package (version 3.3.5; cran.r- 212 project.org/web/packages/ggplot2/index.html) in R. 213 Results 214 Correlation analysis of active ingredients and rhizosphere soil elements 207 project.org/web/packages/ggplot2/index.html) in R. The study used the kappa value to evaluate 208 the multicollinearity of characteristic indicators by the “Kappa” package in R. The map plot was 209 created using the “plyr” package (version 1.8.6; cran.r- 210 project.org/web/packages/plyr/index.html), “maptools” package (version 1.1-2; cran.r- 211 project.org/web/packages/maptools/index.html), and “ggplot2” package (version 3.3.5; cran.r- 210 project.org/web/packages/plyr/index.html), “maptools” package (version 1.1-2; cran.r- 211 project.org/web/packages/maptools/index.html), and “ggplot2” package (version 3.3.5; cran.r- 212 project.org/web/packages/ggplot2/index.html) in R. 211 project.org/web/packages/maptools/index.html), and “ggplot2” package (version 3.3.5; cran.r- 212 project.org/web/packages/ggplot2/index.html) in R. Manuscript to be reviewed 1996): ∆𝜔(𝑡) =‒ 𝛼 ∂𝐸 ∂𝜔(𝑡) + 𝛽∆𝜔(𝑡‒ 1) ∆𝜔(𝑡) =‒ 𝛼 ∂𝐸 ∂𝜔(𝑡) + 𝛽∆𝜔(𝑡‒ 1) (7) ∆𝜔(𝑡) =‒ 𝛼 ∂𝐸 ∂𝜔(𝑡) + 𝛽∆𝜔(𝑡‒ 1) (7) (7) 182 Where α and β are assumed constants, called the learning rate and momentum factor, 183 respectively, E is the error function (in which MSE was used), ω is the weight vector, and t is the 184 iteration number (epoch in MATLAB). 185 Finally, the steps of training can be summarized as follows: (i) applying the input vectors, 186 (ii) calculating the output of the network and comparing it with the corresponding target vectors, 187 (iii) feeding the difference (error) through the network, and (iv) changing the weights according 188 to the algorithm, which tends to minimize the error. The vectors of the training set are applied 189 sequentially. This process is repeated several times using the entire training set until the error is 190 within acceptable criteria or until the output does not change significantly. 185 Finally, the steps of training can be summarized as follows: (i) applying the input vectors, 186 (ii) calculating the output of the network and comparing it with the corresponding target vectors, 187 (iii) feeding the difference (error) through the network, and (iv) changing the weights according 188 to the algorithm, which tends to minimize the error. The vectors of the training set are applied 189 sequentially. This process is repeated several times using the entire training set until the error is 190 within acceptable criteria or until the output does not change significantly. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Manuscript to be reviewed 220 (D10)) (Fig. S2). From the perspective of linear relationship, Figure 3 shows the correlation 221 between the content of different active ingredients and soil elements. Most active ingredients are 222 related to major and trace elements (r>0). In addition, the relationships between soil elements 223 (Fig. 3) showed that Mn had a significant positive correlation with Zn (p=0.004< 0.01); Cu 224 showed a weak correlation with Zn, K (r<0.4); N showed a weak correlation with P (r<0.4); Zn 225 showed a weak correlation with K (r<0.4). It is suggested that there were synergistic properties 226 in the utilization of elements by S. miltiorrhiza. Then, the study used the kappa value to evaluate 227 the multicollinearity of typical indicators. A k value below 100 was interpreted as low 228 multicollinearity, and a k value exceeding 1000 indicates high multicollinearity (Ma et al. 2021). 229 The kappa value (k=480.153) was exceeded 100. These results indicated multicollinearity among 230 fundamental soil indicators, and linear regression cannot be performed directly. From the 231 perspective of a nonlinear relationship, Fig. 4 and Fig. 5 showed the scatter and regression plots 232 between the active ingredients of S. miltiorrhiza and the soil elements. The results showed that 233 there was not a simple increase or decrease between the active ingredients of S. miltiorrhiza and 234 soil elements, but an inevitable fluctuation would follow, which means that as the content of soil 235 elements increases, the content of active ingredients of S. miltiorrhiza will increase and decrease. 236 These results indicated that the active ingredients of S. miltiorrhiza had a nonlinear relationship 237 with multiple soil elements, and a particular element index cannot be used to express the 238 effective ingredients of S. miltiorrhiza. The artificial neural network methods can learn and 239 create nonlinear and complex relationships and can flexibly solve the complex problems of 240 multiple interacting variables (Bayat et al. 2019). It is one of the best techniques for extracting 241 information from inaccurate and nonlinear data (Caselli et al. 2009). Therefore, in this study, a 242 BP neural network model with soil elements as input values was established to explore the 243 relationship between soil elements and the active ingredients for predicting the active 244 i di t 214 Correlation analysis of active ingredients and rhizosphere soil elements 215 In this study, the active ingredients and rhizosphere soil elements of S. miltiorrhiza from 25 216 producing areas in 8 provinces of China were determined, and the contents of 6 elements (Mn, 217 Cu, Zn, N, P, K) (Fig. S1) and 10 the kinds of active ingredients (danshensu (D1), protocatechuic 218 aldehyde (D2), caffeic acid (D3), rosmarinic acid (D4), salvianolic acid B (D5), salvianolic acid 219 A (D6), dihydrotanshinone I (D7)), cryptotanshinone (D8), tanshinone I (D9), tanshinone IIA PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 245 Predictive performance of BP neural network model 246 In this study, a BP neural network model was constructed using the neural fitting tool 247 (nftool) in MATLAB 2019b mathematical software and trained with soil element parameters as 248 input and the content of plant active ingredients as output. The input variables were as follows: 249 Mn, Cu, Zn, N, P, and K; output variables: water-soluble components (danshensu, protocatechuic 250 aldehyde, caffeic acid, rosmarinic acid, salvianolic acid B, salvianolic acid A), and lipid-soluble 251 components (dihydrotanshinone I, cryptotanshinone, tanshinone I, and tanshinone IIA). Through 252 repeated training, a neural network model with 6-8-2 structure was finally established (Fig. 2). In 253 order to evaluate the predictive ability of the established BP neural network model, the MSE 254 (mean square error) and the coefficient of determination (R2) were used to evaluate the overall 255 performance of the model. Ideally. The closer the MSE value is to zero, the closer the R2 value is PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed The coefficient of determination R2 between 279 the predicted and real values of the inversion model was 0.94, the linear regression line between 280 the measured and predicted values were close to 1:1 (i.e., linear), and the MSE (MSE=0.0164) 281 was low. Therefore, the predictive ability of the model was relatively high. Our results indicate 282 that the BP neural network model based on the content of soil elemental can be a powerful tool 283 for predicting the content of active ingredients of S. miltiorrhiza. 274 To further evaluate the model's generalizability, the developed neural network model was 275 tested using 41 new datasets. The new datasets were derived from the results of Zhang's study 276 (Zhang et al. 2018b). The correlation coefficient, mean square error, was used to evaluate the 277 generalization ability of the model. A scatter diagram (Fig. 6B) was constructed between the 278 predicted and real values of the inversion model. The coefficient of determination R2 between 279 the predicted and real values of the inversion model was 0.94, the linear regression line between 280 the measured and predicted values were close to 1:1 (i.e., linear), and the MSE (MSE=0.0164) 281 was low. Therefore, the predictive ability of the model was relatively high. Our results indicate 282 that the BP neural network model based on the content of soil elemental can be a powerful tool 283 for predicting the content of active ingredients of S. miltiorrhiza. Manuscript to be reviewed 256 to 1, which indicates that the average training and testing performance is appropriate. The BP 257 model showed fast training and high simulation accuracy in model testing (Figure 6A). The 258 relationship between the predicted and measured active ingredients content was favorable, 259 thereby indicating that this model has a basic consistency and a high degree of simulation. A 260 scatter diagram was constructed between the predicted and real values of the inversion model. 261 The coefficient of determination R2 between the predicted and real values of the inversion model 262 was 0.93, the linear regression line between the measured and predicted values were close to 1 263 (i.e., linear), and the MSE (MSE=0.0203) was low. Therefore, the predictive ability of the model 264 was relatively high and showed strong nonlinear fitting ability, indicating that the soil elemental 265 content could be used for accurate inversion of the active ingredient content. 266 From the results obtained, the empirical equation based on the Levenberg–Marquardt 267 algorithm for predicting active ingredient content in normalized form is Equation 10: (𝐷1,𝐷2) = 3 ∑ 𝑗= 1[𝑝𝑢𝑟𝑒𝑙𝑖𝑛{ 8 ∑ 𝑖= 1 6 ∑ 𝑗= 1 𝑙𝑜𝑔𝑠𝑖𝑔[(𝑀𝑛𝑗1 + 𝐶𝑢𝑗2 + 𝑍𝑛𝑗3 + 𝑁𝑗4 + 𝑃𝑗5 + 𝐾𝐽6) + 𝑏𝑖]} × 𝐿𝑤𝑖,𝑗+ 𝑏𝑘𝑖] (10) (10) The transfer function “purelin” correlated the linear relationship between the i 268 The transfer function “purelin” correlated the linear relationship between the input and 269 output variables, while “logsig” calculated the layer’s output from the network input. The 270 variables j1, j2, j3, j4, j5, and j6 were input weights from the input layer to the hidden layer. Also, 271 Lw1, Lw2 were hidden layer weights from the hidden layer to the output layer. The variables bi 272 and bk were biases connected to the hidden layer neurons and output layer neurons, respectively. 273 The values for these weights and biases for the model were in Table 1. 274 To further evaluate the model's generalizability, the developed neural network model was 275 tested using 41 new datasets. The new datasets were derived from the results of Zhang's study 276 (Zhang et al. 2018b). The correlation coefficient, mean square error, was used to evaluate the 277 generalization ability of the model. A scatter diagram (Fig. 6B) was constructed between the 278 predicted and real values of the inversion model. Manuscript to be reviewed 288 production of secondary metabolites of the plant. Isopentenyl diphosphate (IPP) and its isomer 289 dimethylallyl diphosphate (DMAPP) are the precursors of all terpenoids (Ma et al. 2015). The 290 synthesis of both requires P. The phenylpropanoid pathway and the tyrosine-derived pathway are 291 the main pathways for the biosynthesis of salvianolic acids (Ma et al. 2015). The synthesis of 292 phenylalanine, the starting point of the phenylpropane pathway, requires the participation of N. 293 K not only regulates plant water metabolism but also acts as an activator of enzymes involved in 294 respiration and photosynthesis (e.g., glutathione synthase) (Liu 2009). Trace elements (Mn, Cu, 295 Zn) are involved in the biosynthesis of plant active ingredients as cofactors, constituent elements, 296 and activators of some enzymes (Guo et al. 2005). Therefore, the biosynthesis of the active 297 ingredients of S. miltiorrhiza cannot be achieved without the participation of soil elements. In 298 this study, the results of correlation analysis between soil elements and active ingredients based 299 on the harvesting period of the herb showed that the active ingredients of S. miltiorrhiza 300 interacted with several soil elements and showed a nonlinear relationship, which may be 301 different from the results of fertilization experiments, for example, Han and Liang (2005) 302 reported that phosphorus fertilization showed a positive effect on the accumulation of danshensu 303 and tanshinone IIA, but this effect was not observed for the effective phosphorus content in this 304 study. Combined with the Mao's research results (Mao et al. 2009), it is believed that it is a 305 matter of elemental availability or antagonism in the utilization between elements. Moreover, 306 there were different levels of correlation between the affective states of the elements, suggesting 307 a synergistic or antagonistic effect on elemental uptake by S. miltiorrhiza. The results of our 308 study further suggest that soil testing formula fertilization should be implemented for medicinal 309 plants like S. miltiorrhiza, which is grown in multiple origins, rather than promoting the use of 310 "special fertilizer" on a large scale. The relationship between fertilizer requirements of medicinal 311 plants and soil supply should be coordinated. The required nutrients should be supplemented in a 312 targeted manner to achieve a balanced supply of nutrients and guide the fertilization management 313 of medicinal plants in an efficient, rational, and rational scientific manner. 284 Discussion 285 Soil is the material basis for the survival of plants. The large and trace elements in the soil 286 can provide nutrients for plants, provide good growth and metabolic status, and enhance their 287 resistance to adversity (Alam & Naik 2009; Zhang et al. 2018). These are the basis for the PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Therefore, it is 314 imperative to establish a method that can be used for screening suitable cultivation sites and 315 rational fertilization of medicinal plants. 288 production of secondary metabolites of the plant. Isopentenyl diphosphate (IPP) and its isomer 289 dimethylallyl diphosphate (DMAPP) are the precursors of all terpenoids (Ma et al. 2015). The 290 synthesis of both requires P. The phenylpropanoid pathway and the tyrosine-derived pathway are 291 the main pathways for the biosynthesis of salvianolic acids (Ma et al. 2015). The synthesis of 292 phenylalanine, the starting point of the phenylpropane pathway, requires the participation of N. 293 K not only regulates plant water metabolism but also acts as an activator of enzymes involved in 294 respiration and photosynthesis (e.g., glutathione synthase) (Liu 2009). Trace elements (Mn, Cu, 295 Zn) are involved in the biosynthesis of plant active ingredients as cofactors, constituent elements, 296 and activators of some enzymes (Guo et al. 2005). Therefore, the biosynthesis of the active 297 ingredients of S. miltiorrhiza cannot be achieved without the participation of soil elements. In 298 this study, the results of correlation analysis between soil elements and active ingredients based 299 on the harvesting period of the herb showed that the active ingredients of S. miltiorrhiza 300 interacted with several soil elements and showed a nonlinear relationship, which may be 301 different from the results of fertilization experiments, for example, Han and Liang (2005) 302 reported that phosphorus fertilization showed a positive effect on the accumulation of danshensu 303 and tanshinone IIA, but this effect was not observed for the effective phosphorus content in this 304 study. Combined with the Mao's research results (Mao et al. 2009), it is believed that it is a 305 matter of elemental availability or antagonism in the utilization between elements. Moreover, 306 there were different levels of correlation between the affective states of the elements, suggesting 307 a synergistic or antagonistic effect on elemental uptake by S. miltiorrhiza. The results of our 308 study further suggest that soil testing formula fertilization should be implemented for medicinal 309 plants like S. miltiorrhiza, which is grown in multiple origins, rather than promoting the use of 310 "special fertilizer" on a large scale. The relationship between fertilizer requirements of medicinal 311 plants and soil supply should be coordinated. Manuscript to be reviewed 324 acids, tanshinones, and dihydrotanshinone I using the site's soil elemental characteristics 325 conditions. We recorded minimal differences between the predicted and observed data (MSE 326 =0.0203, 0.0164; R2 =0.93,0.94), and the excellent agreement between the results support the 327 high efficiency of artificial neural networks while demonstrating that the use of mathematical 328 models (e.g., artificial neural networks) to predict the active ingredient content in medicinal 329 plants can reduce the time and cost required for analytical methods. Artificial neural network 330 models offer significant advantages over traditional mathematical models by using nonlinear 331 network connections and allowing analysis that explores the efficacy of all input variables 332 simultaneously, thereby improving the accuracy of results (Alam & Naik 2009; Tušek et al. 333 2018). However, the capability of artificial neural network models is usually also limited by 334 drawbacks such as slow learning speed, overfitting, and local minima, which suggests that 335 building some hybrid neural network models to reduce their drawbacks and improve their 336 performance is an essential direction for future research. 337 Many medicinal plants have greater medicinal productivity in their original habitat than in 338 cultivated areas. Soil nutrient characteristics similar to the original habitat must be best suited for 339 the production of active metabolites (Alam & Naik 2009). Based on the results of this study, an 340 artificial neural network model can be used to screen suitable cultivation sites for medicinal 341 plants and simulate soil conditions similar to the original soil conditions through soil 342 management, thus increasing the yield of active ingredients. The specific measures are as 343 follows: firstly, measuring the key nutrient factors of the soil, such as massive elements and trace 344 elements; and finally balancing fertilization, such as targeted supplementation of required 345 nutrients during plant growth and harvest, to achieve a balanced supply of nutrients and to 346 efficiently, rationally and scientifically guide the fertilization management of medicinal plants. 347 However, the soil matrix is a complex organic ecosystem, while the plants' secondary 348 metabolites themselves are a complex physiological process, so the effect of soil on plant active 349 ingredients is far more complex than what is involved in the current study. So future work could 350 also extend this modeling process by using a more inclusive range of biotic and abiotic variables 351 to obtain more accurate estimates. Manuscript to be reviewed The required nutrients should be supplemented in a 312 targeted manner to achieve a balanced supply of nutrients and guide the fertilization management 313 of medicinal plants in an efficient, rational, and rational scientific manner. Therefore, it is 314 imperative to establish a method that can be used for screening suitable cultivation sites and 315 rational fertilization of medicinal plants. 316 Here, we presented and validated the application of artificial neural networks in predicting h i i di f di i l l A ifi i l l k h i d 316 Here, we presented and validated the application of artificial neural networks in predicting 317 the active ingredients of medicinal plants. Artificial neural networks have received more 318 attention in various applications due to their sensitivity, accuracy, non-destructive nature, and 319 rapidity (Wu et al. 2012). Many researchers have used artificial neural networks to monitor crop 320 growth and crop yield prediction (Akbar et al. 2018; Wang et al. 2019; Yang et al. 2018), but 321 there are few reports on the use of artificial neural networks to predict active ingredients of 322 medicinal plants based on soil elements. Our results showed that artificial neural network models 323 provided more appropriate predictions of the content of protocatechuic aldehydes, salvianolic PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) 366 References 367 Akbar A, Kuanar A, Patnaik J, Mishra A, and Nayak S. 2018. Application of artificial neural 368 network modeling for optimization and prediction of essential oil yield in turmeric 369 (Curcuma longa L.). Computers and Electronics in Agriculture 148:160-178. 370 https://doi.org/10.1016/j.compag.2018.03.002 371 Alam MA, and Naik PK. 2009. Impact of soil nutrients and environmental factors on 372 podophyllotoxin content among 28 Podophyllum hexandrum populations of northwestern 373 Himalayan region using linear and nonlinear approaches. Communications in soil 374 science and plant analysis 40:2485-2504. https://doi.org/10.1080/00103620903111368 371 Alam MA, and Naik PK. 2009. Impact of soil nutrients and environmental factors on 372 podophyllotoxin content among 28 Podophyllum hexandrum populations of northwestern 373 Himalayan region using linear and nonlinear approaches. Communications in soil 374 science and plant analysis 40:2485-2504. https://doi.org/10.1080/00103620903111368 375 Armaghani DJ, Hasanipanah M, Mahdiyar A, Abd Majid MZ, Amnieh HB, and Tahir MM. 2018. 376 Airblast prediction through a hybrid genetic algorithm-ANN model. Neural Computing 377 and Applications 29:619-629. https://doi.org/10.1007/s00521-016-2598-8 378 Bayat M, Ghorbanpour M, Zare R, Jaafari A, and Pham BT. 2019. Application of artificial neural 379 networks for predicting tree survival and mortality in the Hyrcanian forest of Iran. 380 Computers and Electronics in Agriculture 164:104929. 381 https://doi.org/10.1016/j.compag.2019.104929 382 Caselli M, Trizio L, De Gennaro G, and Ielpo P. 2009. A simple feedforward neural network for 383 the PM 10 forecasting: Comparison with a radial basis function network and a 384 multivariate linear regression model. Water, Air, and Soil Pollution 201:365-377. 385 https://doi.org/10.1007/s11270-008-9950-2 382 Caselli M, Trizio L, De Gennaro G, and Ielpo P. 2009. A simple feedforward neural network for 383 the PM 10 forecasting: Comparison with a radial basis function network and a 384 multivariate linear regression model. Water, Air, and Soil Pollution 201:365-377. 385 https://doi.org/10.1007/s11270-008-9950-2 386 Chen RP, Zhang P, Kang X, Zhong ZQ, Liu Y, and Wu HN. 2019. Prediction of maximum 387 surface settlement caused by earth pressure balance (EPB) shield tunneling with ANN 388 methods. Soils and Foundations 59:284-295. https://doi.org/10.1016/j.sandf.2018.11.005 386 Chen RP, Zhang P, Kang X, Zhong ZQ, Liu Y, and Wu HN. 2019. Prediction of maximum 387 surface settlement caused by earth pressure balance (EPB) shield tunneling with ANN 388 methods. Soils and Foundations 59:284-295. https://doi.org/10.1016/j.sandf.2018.11.005 389 Chi Fl C i i f h Chi A d f S i 19 4 Fl f Chi V l 389 Chinese Flora Commission of the Chinese Academy of Sciences. 1974. Manuscript to be reviewed 359 salvianolic acids, tanshinones, and dihydrotanshinone I. This further validates the feasibility tha 360 artificial neural network models can effectively improve the accuracy of the prediction of the 361 active ingredient content of medicinal plants. In addition, the model can provide broader 362 applicability for ranch managers, manufacturers, and producers of medicinal plants to screen 363 suitable sites for medicinal plant cultivation in a robust and reproducible manner. It can also 364 optimize the fertilizer application at specific sites and guide the fertilization management of 365 medicinal plants in an efficient, rational, and rational scientific manner. 359 salvianolic acids, tanshinones, and dihydrotanshinone I. This further validates the feasibility that 360 artificial neural network models can effectively improve the accuracy of the prediction of the 361 active ingredient content of medicinal plants. In addition, the model can provide broader 362 applicability for ranch managers, manufacturers, and producers of medicinal plants to screen 363 suitable sites for medicinal plant cultivation in a robust and reproducible manner. It can also 364 optimize the fertilizer application at specific sites and guide the fertilization management of 365 medicinal plants in an efficient, rational, and rational scientific manner. 352 Conclusions 353 In this study, an artificial neural network model for predicting the active ingredients of S. 354 miltiorrhiza using rhizosphere soil elements as input values were developed to localize the 355 effects of these factors on the active ingredients. This expands the application of artificial neural 356 network methods in medicinal botany and provides a reference for other directions in the field of 357 medicinal plant cultivation. The results show that, in combination with soil data, we can use 358 artificial neural network models to successfully predict the content of protocatechuic aldehyde, PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed 397 Han JP, and Liang ZS. 2005. Regulation of Salvia miltiorrhizae growth and danshensu and 398 tanshion Ⅱ A accumulation under nitrogen and phosphorus. Chinese Traditional and 399 Herbal Drugs:756-759. 400 Hesami M, and Jones AMP. 2020. Application of artificial intelligence models and optimization 401 algorithms in plant cell and tissue culture. Applied Microbiology and Biotechnology 402 104:9449-9485. https://doi.org/10.1007/s00253-020-10888-2 403 Huang XL, Xiong L, Wang YM, Xu F, Liang YQ, Wang XP, and Wu HM. 2019. Evaluation of 404 quality of Salvia miltiorrhiza Bunge from different provenances by HPLC-DAD fingerprint 405 combined with Chemometrics Method. IOP Conference Series: Earth and Environmental 406 Science: IOP Publishing. p 012163. 407 Huang ZH, Shimeld J, Williamson M, and Katsube J. 1996. Permeability prediction with artificial 408 neural network modeling in the Venture gas field, offshore eastern Canada. Geophysics 409 61:422-436. https://doi.org/10.1190/1.1443970 410 Jagannathan R, Patel SA, Ali MK, and Narayan KV. 2019. Global updates on cardiovascular 411 disease mortality trends and attribution of traditional risk factors. Current diabetes 412 reports 19:1-12. https://doi.org/10.1007/s11892-019-1161-2 413 Li ZT, Liu DX, Zhan LJ, and Li LH. 2020. Mineral elements and active ingredients in root of wild 414 Paeonia lactiflora growing at Duolun County, Inner Mongolia. Biological trace element 415 research 193:548-554. https://doi.org/10.1007/s12011-019-01725-3 416 Lin DY. 2004. Soil Science Experiment Guidance. Beijing: China Forestry Publishing House. 417 Liu YH. 2009. The role of nitrogen, phosphorus and potassium elements and the symptoms of 418 different levels. The Journal of Hebei Forestry Science and Technology 000:57. 419 Lu RK. 2000. Soil agricultural chemical analysis method. Beijing: China Agricultural Science and 420 Technology Press. 421 Lu SF. 2019. Salvia miltiorrhiza: An Economically and Academically Important Medicinal Plant. 422 The Salvia miltiorrhiza Genome: Springer, 1-15. https://doi.org/10.1007/978-3-030- 423 24716-4_1 424 Ma BQ, Hauer RJ, Xu CY, and Li WJ. 2021. Visualizing evaluation model of human perceptions 425 and characteristic indicators of landscape visual quality in urban green spaces by using 426 nomograms. Urban Forestry & Urban Greening 65:127314. 427 https://doi.org/10.1016/j.ufug.2021.127314 428 Ma XH, Ma Y, Tang JF, He YL, Liu YC, Ma XJ, Shen Y, Cui GH, Lin HX, and Rong QX. 2015. 429 The biosynthetic pathways of tanshinones and phenolic acids in Salvia miltiorrhiza. 430 Molecules 20:16235-16254. https://doi.org/10.3390/molecules200916235 431 Mahmoudi S, and Mahmoudi A. 2014. Water saturation and porosity prediction using back- 432 propagation artificial neural network (BPANN) from well log data. Journal of Engineering 433 and Technology (JET) 5:1-8. 366 References Flora of China: Volume 390 66. Beijing: Science Press. 389 Chinese Flora Commission of the Chinese Academy of Sciences. 1974. Flora of China: Volume 390 66. Beijing: Science Press. 391 Gago J, Martínez-Núñez L, Landín M, and Gallego P. 2010. Artificial neural networks as an 392 alternative to the traditional statistical methodology in plant research. Journal of plant 393 physiology 167:23-27. https://doi.org/10.1016/j.jplph.2009.07.007 391 Gago J, Martínez-Núñez L, Landín M, and Gallego P. 2010. Artificial neural networks as an 392 alternative to the traditional statistical methodology in plant research. Journal of plant 393 physiology 167:23-27. https://doi.org/10.1016/j.jplph.2009.07.007 394 Guo XH, Gao WY, Chen HX, and Huang LQ. 2005. Effects of mineral cations on the 395 accumulation of tanshinone ⅡA and protocatechuic aldehyde in the adventitious root 396 culture of Salvia niltiorrhiza. China Journal of Chinese Materia Medica 30:885-888. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed 437 Mei XD, Cao YF, Che YY, LI J, Shang ZP, Zhao WJ, Qiao YJ, and Zhang JY. 2019. Danshen: a 438 phytochemical and pharmacological overview. Chinese journal of natural medicines 439 17:59-80. https://doi.org/10.1016/S1875-5364(19)30010-X 440 Rumelhart DE, Hinton GE, and Williams RJ. 1986. Learning representations by back- 441 propagating errors. nature 323:533-536. https://doi.org/10.1038/323533a0 442 Su CY, Ming QL, Rahman K, Han T, and Qin LP. 2015. Salvia miltiorrhiza: Traditional medicinal 443 uses, chemistry, and pharmacology. Chinese journal of natural medicines 13:163-182. 444 https://doi.org/10.1016/S1875-5364(15)30002-9 442 Su CY, Ming QL, Rahman K, Han T, and Qin LP. 2015. Salvia miltiorrhiza: Traditional medicinal 443 uses, chemistry, and pharmacology. Chinese journal of natural medicines 13:163-182. 444 https://doi.org/10.1016/S1875-5364(15)30002-9 445 Tušek AJ, Benković M, Valinger D, Jurina T, Belščak-Cvitanović A, and Kljusurić JG. 2018. 446 Optimizing bioactive compounds extraction from different medicinal plants and prediction 447 through nonlinear and linear models. Industrial Crops and Products 126:449-458. 448 https://doi.org/10.1016/j.indcrop.2018.10.040 449 Tang ZY, and Fishwick PA. 1993. Feedforward neural nets as models for time series 450 forecasting. ORSA journal on computing 5:374-385. https://doi.org/10.1287/ijoc.5.4.374 451 Tušek AJ, Benković M, Valinger D, Jurina T, Belščak-Cvitanović A, and Kljusurić JG. 2018. 452 Optimizing bioactive compounds extraction from different medicinal plants and prediction 453 through nonlinear and linear models. Industrial Crops and Products 126:449-458. 454 https://doi.org/10.1016/j.indcrop.2018.10.040 455 Wang L, Wang PX, Liang SL, Qi X, Li L, and Xu LX. 2019. Monitoring maize growth conditions 456 by training a BP neural network with remotely sensed vegetation temperature condition 457 index and leaf area index. Computers and Electronics in Agriculture 160:82-90. 458 https://doi.org/10.1016/j.compag.2019.03.017 459 Wang LL, Ma RF, Liu CY, Liu HX, Zhu RY, Guo SZ, Tang MK, Li Y, Niu JZ, and Fu M. 2017. 460 Salvia miltiorrhiza: a potential red light to the development of cardiovascular diseases. 461 Current Pharmaceutical Design 23:1077-1097. 462 https://doi.org/10.2174/13816128226666161010105242 463 Wang XP, Zhang F, Ding JL, Latif A, and Johnson VC. 2018. Estimation of soil salt content 464 (SSC) in the Ebinur Lake Wetland National Nature Reserve (ELWNNR), Northwest 465 China, based on a Bootstrap-BP neural network model and optimal spectral indices. 466 Science of the Total Environment 615:918-930. 467 https://doi.org/10.1016/j.scitotenv.2017.10.025 468 Wu B, Liu Y, and Lu JQ. 2012. New results on global exponential stability for impulsive cellular 469 neural networks with any bounded time-varying delays. Mathematical and Computer 470 Modelling 55:837-843. Manuscript to be reviewed 434 Mao Y, Yuan Y, He XR, and Cui GH. 2009. The Effects of Different Element on the Growth of 435 Salvia Miltiorrhiza Hairy Roots and the Accumulate of Tanshinones in Salvia Miltiorrhiza. 436 Chinese Journal of Experimental Traditional Medical Formulae 15:6-8. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed https://doi.org/10.1016/j.mcm.2011.09.009 471 Yang SX, Feng QS, Liang TG, Liu BK, Zhang WJ, and Xie HJ. 2018. Modeling grassland 472 above-ground biomass based on artificial neural network and remote sensing in the 473 Three-River Headwaters Region. Remote Sensing of Environment 204:448-455. 474 https://doi.org/10.1016/j.rse.2017.10.011 475 Yang XJ, Wan DG, Lin GB, Zhang Q, Wang JY, and Yan ZY. 2010. Analysis on geographical 476 distribution of liposoluble con-stituents in Salvia miltiorrhiza. Chinese Traditional and 477 Herbal Drugs 41:809-812. 475 Yang XJ, Wan DG, Lin GB, Zhang Q, Wang JY, and Yan ZY. 2010. Analysis on geographical 476 distribution of liposoluble con-stituents in Salvia miltiorrhiza. Chinese Traditional and 477 Herbal Drugs 41:809-812. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Manuscript to be reviewed 478 Yang XJ, Wan DG, Liu M, Lin GB, Wang JY, and Yan ZY. 2011a. Analysis on Geographical 479 Distribution of Hydrophilic Components in Radix Salviae Miltiorrhizae. Natural Product 480 Research and Development 23:98-102. 478 Yang XJ, Wan DG, Liu M, Lin GB, Wang JY, and Yan ZY. 2011a. Analysis on Geographic 479 Distribution of Hydrophilic Components in Radix Salviae Miltiorrhizae. Natural Prod 480 Research and Development 23:98-102. 481 Yang XJ, Wan DJ, Lin GB, Guo XH, Liu T, and Yan ZY. 2011b. The quality analysis of 482 introduced Salvia miltiorrhiza from different provenances. Journal of Chengdu Med 483 College 006:291-295. 484 Zhang F, Yu X, Ma Y, Wang L, and Zhang YQ. 2018. Effects of Soil factors on the growth 485 Lonicera japonica Thunb. and the quality of honeysuckle flower. Lishizhen Medicin 486 Materia Medica Research 29:187-189. https://doi.org/10.3969/j.issn.1008- 487 0805.2018.07.063 488 Zhang XD, Yu YG, Yang DF, Qi ZC, Liu RZ, Deng FT, Cai ZX, Li Y, Sun YF, and Liang ZS 489 2018b. Chemotaxonomic variation in secondary metabolites contents and their 490 correlation between environmental factors in Salvia miltiorrhiza Bunge from natura 491 habitat of China. Industrial Crops and Products 113:335-347. 492 https://doi.org/10.1016/j.indcrop.2018.01.043 493 Zhao L, Li JN, Tie XW, and Li XY. 2020. Determination of soil available phosphorus based 494 sodium bicarbonate extrac-tion-molybdenum antimony spectrophotometry. Techno 495 and Economic Guide 28:86. 481 Yang XJ, Wan DJ, Lin GB, Guo XH, Liu T, and Yan ZY. 2011b. The quality analysis of 482 introduced Salvia miltiorrhiza from different provenances. Journal of Chengdu Medical 483 College 006:291-295. 484 Zhang F, Yu X, Ma Y, Wang L, and Zhang YQ. 2018. Effects of Soil factors on the growth of 485 Lonicera japonica Thunb. and the quality of honeysuckle flower. Lishizhen Medicine and 486 Materia Medica Research 29:187-189. https://doi.org/10.3969/j.issn.1008- 487 0805.2018.07.063 488 Zhang XD, Yu YG, Yang DF, Qi ZC, Liu RZ, Deng FT, Cai ZX, Li Y, Sun YF, and Liang ZS. 489 2018b. Chemotaxonomic variation in secondary metabolites contents and their 490 correlation between environmental factors in Salvia miltiorrhiza Bunge from natural 491 habitat of China. Industrial Crops and Products 113:335-347. 492 https://doi.org/10.1016/j.indcrop.2018.01.043 493 Zhao L, Li JN, Tie XW, and Li XY. 2020. Determination of soil available phosphorus based on 494 sodium bicarbonate extrac-tion-molybdenum antimony spectrophotometry. Technology 495 and Economic Guide 28:86. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Figure 1 Part of China map indicating the location of the sample sites. Part of China map indicating the location of the sample sites. Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Figure 2 Structure of the back propagation neural network. Structure of the back propagation neural network. D1, D2 represents water-soluble components and fat-soluble components. D1, D2 represents water soluble components and fat soluble components. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Figure 3 Correlation diagram between active ingredients and soil elements. D1-D10 represents danshensu, protocatechuic aldehyde, caﬀeic acid, rosmarinic acid, dihydrotanshinone I, cryptotanshinone, tanshinone I, tanshinone IIA, salvianolic acid B, salvianolic acid A. Blue indicates positive correlation, red indicates negative correlation. The numbers and the size of the circle represented the correlation coeﬃcients. The larger the area and number, the greater the correlation. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Figure 4 The scatter and regression plots between the water-soluble components of S. miltiorrhiza and the soil elements. D1-D6 represents danshensu, protocatechuic aldehyde, caﬀeic acid, rosmarinic acid, salvianolic acid B, salvianolic acid A. PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Figure 5 The scatter and regression plots between the lipid-soluble components of S. miltiorrhiza and the soil elements. The scatter and regression plots between the lipid-soluble components of S. miltiorrhiza and the soil elements. D7-D10 represents dihydrotanshinone I, cryptotanshinone, tanshinone I, tanshinone IIA. Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Figure 6 Scatter plots of the BP neural network outputs versus targets values. Scatter plots of the BP neural network outputs versus targets values. (A) Scatter plot of outputs versus targets values of dataset (n=26); (B) Scatter plot of outputs versus targets values of new dataset (n=41). (A) Scatter plot of outputs versus targets values of dataset (n=26); (B) Scatter plot of outputs versus targets values of new dataset (n=41). PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed Table 1(on next page) Table 1(on next page) Weights and biases of the model PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021) Manuscript to be reviewed 1 Table 1 Weights and biases of the model 1 Table 1 Weights and biases of the model 1 Table 1 Weights and biases of the model i Input weights Output layer biases Hidden layer weights Output layer biases j1(Mn) j2(Cu) j3(Zn) j4(N) j5(P) j6(K) b1 Lw1 Lw2 bk 1 0.996143 -0.02048 2.317871 -0.37965 -0.56815 0.54681 -2.080196453 1.472587 0.699387 -0.212174053 2 0.838481 0.545057 1.347549 1.166583 0.741616 0.673065 -1.217495956 -0.60922 0.526448 0.540334675 3 -0.88541 1.155593 -0.07306 -0.36007 0.91443 0.29891 -0.874658689 -0.28682 1.10661 4 -0.78139 1.617126 -0.09689 1.246001 0.545242 0.535554 1.953430407 0.985855 -1.55521 5 1.154236 1.305964 -0.34628 1.030944 -1.48709 -1.23566 -0.164765862 -0.18309 -0.66367 6 -1.56626 3.327984 1.205375 0.906212 -2.31703 1.897674 0.667870142 -0.55526 0.804189 7 0.275443 1.076501 -1.06431 1.218909 0.808466 -0.84648 1.880952801 -0.10458 1.113067 8 2.092393 0.073101 -0.1403 0.38235 -0.49332 0.539668 2.641857041 0.501492 0.781672 2 2 PeerJ reviewing PDF \| (2021:07:63562:2:0:NEW 6 Dec 2021)
https://openalex.org/W920177536	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0128593&type=printable	English	null	SLC2A9 Genotype Is Associated with SLC2A9 Gene Expression and Urinary Uric Acid Concentration	PloS one	2,015	cc-by	9,092	Methods The association between urinary UA concentrations and single nucleotide polymorphisms (SNPs) within the SLC2A9 gene region, expression levels of genes in the uricosuric path- way, and dietary protein intake were analyzed for a sample of non-Hispanic white partici- pants from the Genetic Epidemiology Network of Arteriopathy (GENOA) cohort. The SLC2A9 SNP most significantly associated with urinary UA concentration was then tested for associations with gene expression levels from uric acid absorption/secretion associated genes. Models including interactions between dietary protein (total, animal, and vegetable) and genetic factors were also assessed. Copyright: © 2015 Ware et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: Due to the sensitive nature of the individual genetic data, data are from the GENOA study whose authors may be contacted at skardia@umich.edu. The reason for this is that the consent form for the GENOA study did not specifically allow for public posting of gene expression or genotype data to publically available sites. Objectives SLC2A9 gene variants have been associated with urinary uric acid (UA) concentration, but little is known about the functional mechanism linking these gene variants with UA. SLC2A9 encodes a UA transporter present in the proximal tubule of the kidney, and gene expression levels of SLC2A9 and other genes in the uricosuric pathway (ABCG2, SLC17A1, SLC17A3, and SLC22A12) could potentially mediate the relationship between SLC2A9 gene variants and urinary UA excretion. Editor: Dana C Crawford, Case Western Reserve University, UNITED STATES Received: July 29, 2014 Accepted: April 28, 2015 Published: July 13, 2015 d to a a C C a o d, Case este ese e University, UNITED STATES Received: July 29, 2014 Accepted: April 28, 2015 Published: July 13, 2015 Copyright: © 2015 Ware et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. RESEARCH ARTICLE SLC2A9 Genotype Is Associated with SLC2A9 Gene Expression and Urinary Uric Acid Concentration Erin B. Ware1,2☯, Ellen Riehle1☯, Jennifer A. Smith1, Wei Zhao1, Stephen T. Turner3, Sharon L. R. Kardia1, John C. Lieske3,4 1 Department of Epidemiology, School of Public Health, University of Michigan, Ann Arbor, MI, United States of America, 2 Institute for Social Research, University of Michigan, Ann Arbor, MI, United States of America, 3 Division of Nephrology and Hypertension, Mayo Clinic, Rochester, MN, United States of America, 4 Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, United States of America ☯These authors contributed equally to this work. skardia@umich.edu ☯These authors contributed equally to this work. * skardia@umich.edu OPEN ACCESS Citation: Ware EB, Riehle E, Smith JA, Zhao W, Turner ST, Kardia SLR, et al. (2015) SLC2A9 Genotype Is Associated with SLC2A9 Gene Expression and Urinary Uric Acid Concentration. PLoS ONE 10(7): e0128593. doi:10.1371/journal. pone.0128593 Citation: Ware EB, Riehle E, Smith JA, Zhao W, Turner ST, Kardia SLR, et al. (2015) SLC2A9 Genotype Is Associated with SLC2A9 Gene Expression and Urinary Uric Acid Concentration. PLoS ONE 10(7): e0128593. doi:10.1371/journal. pone.0128593 Editor: Dana C Crawford, Case Western Reserve University, UNITED STATES Received: July 29, 2014 Accepted: April 28, 2015 Published: July 13, 2015 Copyright: © 2015 Ware et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. D A il bili S D h i i ☯These authors contributed equally to this work. * skardia@umich.edu Introduction Nephrolithiasis (NL), or kidney stone formation, represents a substantial public health burden affecting approximately 1 in 11 Americans [1–3]. Specific diet patterns–particularly protein consumption–and familial history are both known risk factors for NL [4–11]. Most stones in the general population (70–80%) contain a majority of calcium oxalate (CaOx), while approxi- mately 5–10% contain uric acid (UA) or a combination of UA and CaOx [3, 12, 13]. Hyperuri- cosuria is a risk factor for not only UA stone formation, but can also induce salting out of CaOx from solution, promoting CaOx stone formation [14–18]. Thus, factors influencing uri- nary UA excretion are likely important modifiers of kidney stone risk [11, 19, 20]. The concentration of UA in urine is determined by a combination of renal filtration fol- lowed by reabsorption and secretion in the nephron. SLC2A9 encodes a key UA transporter expressed as two isoforms in the apical and basolateral membranes of the renal proximal tubule. This transporter also mediates uptake of UA from the blood into the liver [21–25]. Genetic variation in this gene has been associated with both mild and severe UA phenotypes, as well as NL [26–29]. Previous genome-wide association studies have identified single nucleo- tide polymorphisms (SNPs) within SLC2A9 that are associated with serum and urinary UA concentration [30–36]. Several other genes have been found to be involved in excretion of UA in the urine as well as reabsorption within the kidney, though to a lesser extent than SLC2A9, and include SLC22A12, SLC17A1, SLC17A3, and ABCG2 [37–40]. In addition, diet also plays a key role in UA homeostasis, and several studies have shown that higher protein (and purine) intake is associated with higher urinary UA concentration [41–44]. To better understand the contributions of genetic and dietary factors in urinary UA excre- tion, we examined the influence of SLC2A9 gene variation on urinary UA, gene expression from uric acid absorption/secretion associated genes, and potential interactions with dietary protein. Since our gene expression measures are taken from transformed β lymphocytes, they are best thought of as a cumulative measure of the many genetic variations in each gene that are influencing its expression. Specifically, we identified the SNP within the SLC2A9 gene region (rs12509955) that was most significantly associated with urinary UA excretion using a linear mixed model (LMM), and then examined its association with SLC2A9, ABCG2, SLC17A1, SLC17A3, and SLC22A12 gene expression levels. Conclusion Our results illustrate that SNPs in the SLC2A9 gene influence SLC2A9 gene expression as well as urinary UA excretion. Evidence is also suggestive that gene-by-diet interactions may disproportionately increase urinary UA in genetically susceptible individuals that con- sume higher amounts of protein. Competing Interests: The authors have declared that no competing interests exist. SLC2A9 Genotype, Gene Expression, and Uric Acid Gene expression level of one SLC2A9 transcript had a significant interaction with dietary animal protein (SLC2A9-001 ENST00000506583, p = 0.01) and a marginally significant interaction with total dietary protein (p = 0.07) on urinary UA. and Grant Number UL1 TR000135 from the National Center for Advancing Translational Sciences (NCATS), all funded by the National Institutes of Health. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Results The most significant SLC2A9 SNP associated with urinary UA (rs12509955, corrected p = 0.001) was also associated with SLC2A9 gene expression levels (corrected p = 0.0084); however, SLC2A9 gene expression levels were not significantly associated with urinary UA concentrations (p = 0.509). The interactions between rs12509955 and total die- tary protein, and SLC2A9 gene-level gene expression and dietary vegetable protein on the outcome of urinary UA were marginally significant (p = 0.11 and p = 0.07, respectively). Funding: This work was supported by R01 DK077950, R01 DK073537, U01 HL054457, R01 HL087660, R01 HL119443, and the Mayo Clinic O’Brien Urology Research Center U54 DK100227, PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 1 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 Ethics Statement This study protocol was approved by the Institutional Review Board at the Mayo Clinic under “Genetic Determinants of Urine Lithogenicity–Working Protocol” 08–006238. Study Population This study included non-Hispanic whites from Rochester, Minnesota that were enrolled in the Genetic Epidemiology Network of Arteriopathy (GENOA) study (Phase I: 1995–2000, n = 1,583; Phase II: 2000–2005, n = 1,241). GENOA is a multicenter, community-based study of hypertensive sibships initially collected to identify genes that influence blood pressure and target organ damage due to hypertension. The Genetic Determinants of Urinary Lithogenicity (GDUL) study (2008–2012) is an ancillary study of the Phase III GENOA Genetics of Chronic Kidney Disease Study (R01 DK073537), undertaken to investigate predictors of urinary super- saturation and risk of kidney stone diseases among participants without end-stage renal failure (Stage 5 Chronic Kidney Disease). GENOA participants were invited to join the GDUL study which consisted of a study visit; one (or preferably two or three) 24-hour urine collections for determination of quantitative urinary lithogenic factors, and a food frequency questionnaire (FFQ) (Viocare Technologies, Princeton, NJ, USA). Age, sex, and body mass index (BMI) were ascertained at the time of the GDUL exam. Of the 811 participants enrolled in GDUL, two were excluded from the initial SNP discovery analysis because their urinary UA concentration measure was >4 standard deviations from the mean. A subset who had gene expression data (n = 541 participants in 318 sibships) were used for all other association studies. A total of 424 of these participants also had animal and total dietary protein information (393 had vegetable protein) from the food frequency questionnaire. While urinary uric acid is the primary out- come of interest, blood serum uric acid was available from Phase II of the GENOA study, con- ducted approximately five years prior to the GDUL study. Serum uric acid was investigated to evaluate whether the effects of SNP variation and gene expression on urinary uric acid were consistent for serum uric acid. No participants were excluded for missing serum uric acid, as this was a supplemental analysis. All serum uric acid analyses can be found in supplemental material. Urine Collection After the initial visit, subjects completed one, two, or three 24-hour urine collections. Urine was collected with toluene as a preservative. Measurements were averaged for subjects with more than one 24-hour urine collection sample to account for the day-to-day variability in uri- nary composition. Twenty four-hour urine uric acid (mg/day) and sodium (mmol) concentra- tions were measured in the Mayo Clinic Renal Testing Laboratory. For additional details, see Lieske, et al. 2014 [45]. Introduction We further investigated gene-by- dietary protein interaction effects on urinary UA to determine whether rs12509955 genotype or gene expression from uric acid absorption/secretion associated genes interacts with dietary protein intake, since purine metabolism is a known modifier of UA phenotypes. SLC17A1, SLC17A3, and SLC22A12 gene expression levels. We further investigated gene-by- dietary protein interaction effects on urinary UA to determine whether rs12509955 genotype or gene expression from uric acid absorption/secretion associated genes interacts with dietary protein intake, since purine metabolism is a known modifier of UA phenotypes. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 2 / 15 SLC2A9 Genotype, Gene Expression, and Uric Acid Food frequency Participants completed a food frequency questionnaire (Viocare Technologies, Princeton, NJ, USA) at a study visit. The Women's Health Initiative FFQ was adapted by Viocare in an elec- tronic computer-administered format, and is similar to the Willett or Harvard Food Frequency Questionnaire. It assesses food intake over the last 3 months and employs the Minnesota NDS nutritional analysis database for nutrient analyses. The FFQ was validated and compares well with nutrient intake against 4-day food records and 24-hr dietary recalls [46]. The FFQ has validity for variables of interest to this study, when compared to a 4-day diet diary, namely for 3 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid protein intake (r = 0.41) and sodium intake (r = 0.31). Dietary protein was assessed as total die- tary protein, and also subdivided into animal or vegetable protein sources. Genotyping and Imputation All GENOA participants were genotyped on the Affymetrix Genome-Wide Human SNP Array 6.0. A small portion of the GENOA stored blood samples contained poor quality DNA, and some of these were successfully genotyped using the Illumina Human 1M-Duo BeadChip. Since the two platforms used for genotyping contain only a small number of overlapping SNPs (~200,000), association analyses were performed using only imputed data. Prior to imputation, SNPs and samples with a call rate less than 95% were excluded. Imputation was performed using a single-step approach implemented in MACH v1.0.16 using the CEU reference panel of HapMap2 (release 22). The SLC2A9 gene region of interest (including all SNPs within the gene, plus 200kb upstream and downstream) included a total of 880 SNPs. For analyses including the rs12509955 SNP, participants were categorized into imputed genotype (i) calls: (i <0.5 = 0; 0.5< i <1.5 = 1; i>1.5 = 2). Gene Expression Assessment and Quality Control Blood samples for beta-lymphocyte extraction were collected from a subset of participants dur- ing GENOA Phase I and Phase II. RNA samples were extracted using standard protocols. RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Foster City, CA) and quantified by spectrophotometry using the Nanodrop ND-1000 (Nanodrop Inc., Wilmington, DE). Array quality control was performed at the transcript level with the Affymetrix Expression Console (v 1.1) using core-level probe sets. All array images passed visual inspection. Hybrid- ization controls were all detectable by signal increases that followed concentration. Signal intensity plots were examined for raw and processed data to identify outliers. Raw intensity data were processed using the Affymetrix Power Tool software. Probe summarization and set normalization were performed using Robust Multi-array Analysis (RMA), including back- ground correction, quantile normalization, log2-transformation, and probe set summarization. Probe sets which are known to cross-hybridize and those with undetectable expression were also excluded. Gene-level expression was assessed by averaging all core probe sets for that gene. Transformed beta-lymphocyte gene expression measurements were adjusted for age, sex, and batch prior to analysis. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 Statistical Methods Data management and statistical analyses were conducted in SAS version 9.3 [47]. Urinary phenotypes had approximately normal distributions; thus, no variable transformations were applied. The sample of 809 GENOA participants with genotype and urinary UA excretion data was used to identify the SNP with the most significant p-value for association between SNP dosage (as an additive effect) and urinary UA. Fixed effects linear mixed modeling (LMM) was used to account for familial correlations within sibships. Age, sex, body mass index (BMI), and urinary sodium were included in the model because they significantly predict urinary UA in bivariate analyses (Table 1) or have been included as a covariate in other genetic models of urinary UA [32]. In order to investigate whether there were multiple independent signals within the SLC2A9 region, we also tested for association between each SNP in the region and urinary UA, conditioned on the SNP mostly significantly associated with UA. 4 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 doi:10.1371/journal.pone.0128593.t001 SLC2A9 Genotype, Gene Expression, and Uric Acid nd bivariate associations with Urinary UA excretion and gene expression levels from uric acid absorp- Table 1. Characteristics of study sample and bivariate associations with Urinary UA excretion and gene expression levels from uric acid absorp- tion/secretion associated genes, n = 541. Table 1. Characteristics of study sample and bivariate associations with Urinary UA excretion and gene expression levels from uric acid absorp- tion/secretion associated genes, n = 541. Urinary UA Serum UA SLC2A9 Gene Expression Mean (sd) / n (%) B P B P B P Age (yrs) 65.5 (9.3) -1.27 0.09 0.04 < .0001* 0.08 0.06 Male 215 (39.7%) 132.65 < .0001* 1.18 < .0001* -0.01 0.78 BMI (kg/m2) 30.4 (5.6) 5.50 < .0001* 0.09 < .0001* 0.02 0.71 Urinary sodium (mmol) 138.6 (56.9) 1.76 < .0001* 0.01 < .0001* 0.00 0.93 Urinary UA (mg/day) 435.3 (161.7) 0.00 0.002* 0.03 0.51 Serum uric acid (mg/dL) 6.0 (1.6) 11.88 0.01* 0.05 0.28 Dietary protein intake Vegetable (g) (n = 393) 27.8 (13.3) 0.79 0.18 -0.002 0.74 0.03 0.59 Total (g) (n = 424) 80.5 (33.9) 0.82 <0.001* 0.002 0.27 -0.05 0.36 Animal (g) (n = 424) 52.6 (25.5) 1.25 <0.001* 0.004 0.13 -0.07 0.15 Gene expression ABCG2 4.6 (0.2) -13.42 0.68 -0.11 0.72 0.32 < .0001* SLC17A1 5.0 (0.2) -31.93 0.34 0.11 0.72 0.40 < .0001* SLC17A3 3.6 (0.3) 6.59 0.77 0.00 0.99 0.25 < .0001* SLC22A12 6.7 (0.3) -39.22 0.16 0.17 0.52 0.20 < .0001* SLC2A9 5.4 (0.3) 17.95 0.51 0.27 0.30 SLC2A9-001 (n = 515) 5.7 (1.0) 1.72 0.82 0.02 0.74 0.01 0.58 SLC2A9-002 6.1 (1.0) -3.85 0.59 -0.04 0.58 0.06 < .0001* Genotype rs12509955 33.32 0.01* -0.47 <0.001* 0.07 <0.001* 0 367 (67.8%) 1 153 (28.3%) 2 21 (3.9%) statistically signiﬁcant association at α = 0.05. UA = uric acid; BMI = body mass index; Bivariate association results were obtained from linear mixed models accounting for sibship (n = 541). ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. β coefﬁcients for association with gene expression were standardized. Table 1. Characteristics of study sample and bivariate associations with Urinary UA excretion and gene exp tion/secretion associated genes, n = 541. Table 1. Characteristics of study sample and bivariate associations with Urinary UA excretion and gene expression levels from uric acid absorp- tion/secretion associated genes, n = 541. Descriptive Statistics Table 1 provides descriptive statistics regarding demographic, anthropometric, dietary and genetic risk factors for the study sample and their regression relationships to urinary UA excretion and SLC2A9 gene expression. In our sample, as expected, males had significantly higher urinary UA excretion (β = 132.6, p-value < 0.0001) compared to females. BMI and uri- nary sodium excretion were positively associated with urinary UA excretion (β = 5.5, p- value < 0.0001; β = 1.8, p-value < 0.0001). Total dietary protein and dietary animal protein were positively associated with urinary UA excretion (β = 0.8, p-value = 0.0003; β = 0.7, p- value = 0.004, respectively). Participants who were in the larger SNP discovery analysis sample (n = 809) but lacked gene expression data (n = 268) were not significantly different with regard to urinary UA excretion than the 541 participants with gene expression data (p-value = 0.10). Table 1 also displays the distribution of rs12509955 genotypes in our study sample (copies of the coded, minor allele (T): 0 = 67.8%; 1 = 28.3%; 2 = 3.9%) and their corresponding mean uri- nary UA excretion levels. The rs12509955 genotype was significantly associated with urinary UA (β = 33.32, p-value = 0.01), serum UA (β = -0.47, p-value <0.001), and SLC2A9 gene expression (β = 0.07, p-value <0.001). Correlations for gene-level gene expression levels for the five investigated genes were moderate and significant (all p-values <0.0001) and ranged from r = -0.61 (SLC2A9-001 and SLC2A9-002) to r = 0.55 (SLC17A1 and SLC17A3). A correlation table is provided in the supplementary material (S1 Table). SLC2A9 SNP association with urinary UA excretion Of the 880 SNPs tested for an association with urinary UA (adjusted for sibships, age, sex, BMI, and urinary sodium in the LMM), the most significant SNP was rs12509955 (intron 2, coded allele = T, β = 40.0, corrected p-value = 0.001). P-values were corrected for the number of independent test in the SLC2A9 gene region using principal components to account for cor- related SNPs. After running a principal component analysis, 28 principal components were required to explain 95% of the variation in genotype and thus p-values were corrected by a fac- tor of 28 [49]. As seen in Fig 1, the SCL2A9 gene region has a strong linkage disequilibrium pattern that extends across the SLC2A9 gene, as well as the intergenic regions upstream. There are many SNPs that are strongly associated with urinary UA (S2 Table); however, after condi- tioning upon the most significant SNP, rs12509955, no additional SNPs in the region were sig- nificantly associated with urinary UA after multiple testing correction (S3 Table). We performed this analysis to detect any alternate independent predictive gene variants in the region in the context of a strong linkage disequilibrium pattern. Thus, for subsequent analyses, rs12509955 was selected to represent genetic variation in the SLC2A9 region associated with urinary UA excretion. Since a subsample of 541 participants with gene expression data was used for the remainder of the analysis, we repeated the associations between SNPs and UA in this subsample and again found that rs12509955 was the most statistically significant (β = 40.1, corrected p-value = 0.001). Equivalent analyses for serum uric acid can be found in supplemen- tal material (S4 Table and S5 Table, S1 Fig and S2 Fig). statistically signiﬁcant association at α 0.05. UA = uric acid; BMI = body mass index; Bivariate association results were obtained from linear mixed models accounting for sibship (n = 541). ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. β coefﬁcients for association with gene expression were standardized. s a s ca y s g ca assoc a o a 0 05 UA = uric acid; BMI = body mass index; Bivariate association results were obtained from linear mixed models accounting for sibship (n = 541). ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. β coefﬁcients for association with gene expression were standardized. y g UA = uric acid; BMI = body mass index; Bivariate association results were obtained from linear mixed models accounting for sibship (n = 541). ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. β coefﬁcients for association with gene expression were standardized. doi:10.1371/journal.pone.0128593.t001 In a subsample of 541 GDUL participants with gene expression data, LMM was used to assess whether there was an association between the rs12509955 genotype and gene expression levels from uric acid absorption/secretion associated genes. To further evaluate whether a genetic effect on gene expression levels had a corresponding influence on urinary UA, we used LMMs to test the association between gene expression and urinary UA excretion, adjusting for BMI and urinary sodium. LocusZoom plots were used to visualize results of the SNP associa- tions with UA and gene expression [48]. To assess gene-by-diet interactions on urinary UA on rs12509955, LMMs were adjusted for age, sex, BMI, and urinary sodium, and included terms for SNP, dietary protein, and the inter- action between dietary protein and SNP. Similarly, we assessed the potential interaction between each dietary protein measure and gene expression levels from uric acid absorption/ secretion associated genes on urinary UA excretion. Interaction term β coefficients were con- sidered to be statistically significant at an alpha level of 0.05. 5 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid SLC2A9 SNP association with gene expression Next, 880 SNPs within the SLC2A9 region were tested for association with SLC2A9 gene expression. Of the 880 SNPs tested, 429 SNPs within the region were significantly associated with SLC2A9 gene expression (corrected p-value < 0.05; Fig 2), with the most significant being rs2240724 (β = -0.08, corrected p-value = 6.10x10-16). As displayed in Table 1, a significant 6 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid Fig 1. Urinary UA association results for 880 SNPs in the SLC2A9 gene region. Left Y-axis:–log10(p-value) from association between SNPs and urinary UA, adjusted for age, sex, BMI, and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs12509955 (purple diamond). Fig 1. Urinary UA association results for 880 SNPs in the SLC2A9 gene region. Left Y-axis:–log10(p-value) from association between SNPs and urinary UA, adjusted for age, sex, BMI, and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs12509955 (purple diamond). doi:10.1371/journal.pone.0128593.g001 association between rs12509955 genotype and standardized SLC2A9 gene-level gene expres- sion was observed (β = 0.16, corrected p-value = 0.008). rs12509955 was not in linkage disequi- librium with rs2240724, and thus had an independent, more modest influence on gene expression than rs2240704 and the many other SNPs in the region that were associated with SLC2A9 expression (Fig 2). rs2240704 was not investigated further because this SNP was not significantly and independently associated with urinary UA excretion, our main outcome of interest, in a model conditioned upon rs12509955. Overall, the tight linkage disequilibrium pattern indicated that gene expression levels are representing variation in over 48% of SNPs. h f l d h f No SNPs in the SLC2A9 gene region were significantly associated with gene expression of the other four genes of interest after multiple testing correction (α = 0.001). The rs12509955 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 7 / 15 SLC2A9 Genotype, Gene Expression, and Uric Acid Fig 2. SLC2A9 gene expression association results for 880 SNPs in the SLC2A9 gene region. SLC2A9 SNP association with gene expression Left Y-axis:–log10(p-value) from association between SNPs and SLC2A9 gene expression, accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs2240724 (purple diamond). Fig 2. SLC2A9 gene expression association results for 880 SNPs in the SLC2A9 gene region. Left Y-axis:–log10(p-value) from association between SNPs and SLC2A9 gene expression, accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs2240724 (purple diamond). doi:10.1371/journal.pone.0128593.g002 doi:10.1371/journal.pone.0128593.g002 genotype was not associated with gene expression for SLC22A12 (β = 0.04, p-value = 1), SLC17A1 (β = 0.02, p-value = 1), SLC17A3 (β = 0.05, p-value = 0.75), or ABCG2 (β = 0.00, p-value = 1), corrected for multiple testing. Gene-by-dietary protein interactions Models based upon urinary UA excretion and rs12509955-by-dietary protein interactions showed no significant interactions between dietary protein source and rs12509955 genotype, though the interaction between total protein and rs12509955 genotype was marginally signifi- cant (total protein: β = 0.50, p-value = 0.11; animal protein: β = 0.01, p-value = 0.98; vegetable protein β = 1.00, p-value = 0.27). Though not significant, those with two copies of the T allele had increased predicted urinary UA compared to those with one or those with no copies of the T allele at every level of dietary protein consumption for each type of dietary protein. Interaction models (Table 2, serum UA S7 Table) examining the associations between gene expression and dietary protein, or the cumulative measure of variation in the gene and dietary protein, on urinary uric acid showed a marginally significant association between SLC2A9 gene expression and dietary vegetable protein intake (β = 3.21, p-value = 0.07) and non-significant interactions with the other two protein sources (total protein: β = 1.00, p-value = 0.18; animal protein: β = 0.63, p-value = 0.52). No other genes showed significant or marginally significant interactions between gene expression and any of the dietary protein measures. According to the literature, there are 13 splice variants in the SLC2A9 gene region, four of which code for proteins [50]. In our data we observed three protein coding transcripts and two of them appeared at sufficient frequency (<5% missing data) to include in linear mixed models (SLC2A9-001 ENST00000506583, bp = 1927, protein 511 aa, n = 515; SLC2A9-002 ENST00000264784, bp = 1850, protein 540 aa, n = 541). These two transcripts are not signifi- cantly associated with urinary UA in bivariate models (SLC2A9-001 β = 1.72, p-value = 0.82; SLC2A9-002 β = -3.85, p-value = 0.59) or in models adjusted for BMI and urinary sodium (SLC2A9-001 β = 3.95, p-value = 0.50; SLC2A9-002 β = -2.24, p-value = 0.69). However, the SLC2A9-001 gene expression transcript significantly interacts with dietary animal protein (β = 0.65, p-value = 0.01, S3 Fig) and has a marginally significant interaction with total dietary pro- tein (β = 0.38, p-value = 0.07) in urinary UA models. Gene expression association with urinary UA SLC2A9 gene expression level, representing SLC2A9 gene variation across subjects, was not sig- nificantly associated with urinary UA excretion in either a bivariate model (Table 1, β = 18.0, p-value = 0.51) or a model adjusted for BMI and urinary sodium concentration (β = 15.5, p- value = 0.47). In bivariate models, none of the gene expressions were significantly associated with urinary uric acid (ABCG2 (β = -13.4, p-value = 0.68); SLC17A1 (β = -31.9, p-value = 0.34); SLC17A3 (β = 6.59, p-value = 0.77); SLC22A12 (β = -39.2, p-value = 0.16)), nor was any gene’s PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 8 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid expression significantly associated with urinary uric acid after adjustment for BMI and urinary sodium (ABCG2 (β = -47.3, p-value = 0.06); SLC17A1 (β = -39.2, p-value = 0.14); SLC17A3 (β = -12.6, p-value = 0.48); SLC22A12 (β = -40.7, p-value = 0.06)), though both ABCG2 and SLC22A12 may be considered to have marginally significant effects. We did not adjust for age or sex in these models since gene expression had previously been adjusted for age, sex, and batch during quality control. Based on these marginally significant effects, indicating possible gene expression associations with urinary UA, we went on to investigate genetic interactions for all genes of interest (including transcript-based gene expression for the SLC2A9 gene) with dietary protein. Equivalent analyses for serum uric acid can be found in supplemental material (S1 Text). Gene-by-dietary protein interactions The SLC2A9-002 gene expression tran- script significantly interacts with dietary total protein (β = -0.004, p-value = 0.04, S7 Table) and has marginally significant interactions with dietary animal and vegetable protein (β = -0.005, p-value = 0.07; β = -0.01, p-value = 0.06, respectively) in serum UA models (S7 Table). SLC2A9 Genotype, Gene Expression, and Uric Acid Table 2. Gene expression by dietary protein intake interaction associations for urinary uric acid. Urinary uric acid (mg/day) Total protein Animal protein Vegetable protein B P B P B P ABCG2 Protein -1.78 0.71 -65.39 0.33 -11.95 0.34 Gene expression -60.81 0.51 -2.78 0.59 -67.56 0.42 Protein x Gene expression 0.57 0.58 0.70 0.54 2.75 0.31 SLC17A1 Protein 0.59 0.92 -39.93 0.57 12.30 0.32 Gene expression -26.19 0.79 -1.92 0.76 61.32 0.45 Protein x Gene expression 0.05 0.97 0.46 0.72 -2.27 0.35 SLC17A3 Protein 0.74 0.75 6.15 0.88 2.40 0.72 Gene expression 14.31 0.81 0.86 0.72 38.94 0.50 Protein x Gene expression 0.02 0.98 -0.13 0.84 -0.46 0.80 SLC22A12 Protein 8.90 0.11 48.47 0.38 2.10 0.88 Gene expression 55.42 0.46 10.06 0.10 -14.43 0.82 Protein x Gene expression -1.20 0.14 -1.45 0.11 -0.18 0.93 SLC2A9 Protein -5.17 0.20 -20.45 0.71 -17.33 0.07 Gene expression -52.36 0.40 -4.76 0.38 -69.28 0.20 Protein x Gene expression 1.00 0.18 0.96 0.34 3.21 0.07 SLC2A9-001 Protein -22.31 0.19 -25.51 0.08 1.67 0.92 Gene expression -1.96 0.10 -3.41 0.02 -0.24 0.94 Protein x Gene expression 0.38 0.07 0.65 0.01 0.06 0.92 SLC2A9-002 Protein -5.88 0.70 2.70 0.83 -21.24 0.18 Gene expression -0.01 1.00 0.99 0.47 -3.80 0.24 Protein x Gene expression 0.04 0.82 -0.11 0.64 0.67 0.21 *statistically signiﬁcant association at α = 0.05. ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. Table 2. Gene expression by dietary protein intake interaction associations for urinary uric acid. y g ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. Our analysis demonstrated that the SNP most strongly associated with urinary UA excre- tion, rs12509955, also significantly associated with SLC2A9 gene expression. This association between SNP-level variation within the SLC2A9 region and SLC2A9 gene expression has not previously been reported. The analysis also revealed other SNPs within the SLC2A9 region that were even more significantly associated with SLC2A9 gene expression than rs12509955. One isoform of the SLCA29 transporter is expressed the apical membrane of the proximal tubule, while another is expressed in the basolateral membrane. The transporter exchanges UA for glucose and fructose resulting in net reabsorption of filtered UA. Genetic variation within SLC2A9 has previously been associated with serum UA concentrations and fractional erection of UA into the urine. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 Discussion SNP-level genetic variation within SLC2A9 has previously been demonstrated to associate with urinary and serum UA phenotypes (30–36). Serum UA concentration and urinary UA excre- tion tend to be inversely related, since both are determined by the combination of UA reab- sorption and secretion in the proximal tubule, and the SLC2A9 urate transporter is a key mediator of this process. Indeed the top SNP we identified, rs1209955, was previously identi- fied by Doring and colleagues as significantly associated with serum UA levels [22]. 9 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 In the current study we confirm and extend these observations, and dem- onstrate that total urinary UA excretion is also influenced by genetic variation in SLC2A9. This 10 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid observation implies that changes in UA generation occur when urinary fractional excretion of UA decreases [51]. Glut9, the protein product of SLC2A9, is required for hepatocyte UA uptake. In mice, live specific knockout of SLC2A9 results in hyperuricemia and hyperuricosuria due to decreased degradation of UA by hepatic uricase. Humans lack uricase, but the current study suggests that genetic variability in SLCA29 alters overall UA production, perhaps via feedback at the level of the liver on enzymatic pathways that result in its generation. It is also conceivable that changes in availability of fructose could alter UA generation [52]. We previously hypothesized that an association between SLC2A9 genotype and UA pheno- types was mediated by SLC2A9 gene expression. SLC2A9 gene expression was not a significant predictor of urinary UA in our study. However, the gene expression data was obtained from transformed Beta-lymphocytes, and this cell type may not reflect gene expression levels present in the liver and kidneys, organs more directly responsible for UA metabolism and concentra- tion. Nevertheless, the gene expression measured from transformed Beta-lymphocytes is thought to better reflect the proximal influences of genetic variation, since environmental mod- ulators of gene expression are standardized by the uniform culture medium in which the cells are grown. It is quite possible that SLC2A9 gene expression in kidney cells would have more variability, since both environmental and genetic variability would be contributing to gene expression. Specifically, environmental conditions, such as dietary protein consumption, could have profound effects on tissue level gene expression, but have limited impact on gene expres- sion levels measured in transformed Beta-lymphocytes, except perhaps through epigenetic DNA modifications that persist post-transformation. Overall, gene expression measurements are more reflective of the cumulative effects of multiple variants in the gene effecting gene expression. This may explain why highly significant associations between SNP genotypes and gene-level gene expression were observed in the current study. However, numerous steps between gene/gene expression and urinary UA excretion substantially weaken the pathway of statistical association. The current study identified one significant transcript-level gene expression-by-diet interac- tion and several marginally significant interactions influencing urinary UA excretion, and sev- eral significant gene expression-by-dietary protein interactions in the context of serum UA. Since so many SNPs influenced SLC2A9 gene expression (Fig 2), it may be that gene expression values provide a better overall summary of the contribution common genetic variation in SLC2A9 makes to UA excretion, and could explain why gene expression displays a greater interaction with diet than rs12509955 genotype. In the future, identifying means to capture the cumulative effects of many genetic variations in the SLC2A9 region and their interactions with environmental factors such as dietary protein could facilitate efforts to assess the relative impact of genetic variability on serum and urinary UA levels. In summary, the current study confirms that genetic variation within SLC2A9 plays an important role in determining urinary UA excretion. The most significant SNP associated with urinary UA (rs12509955), and many other SNPs in the gene region, were associated with SLC2A9 gene-level gene expression. Future investigations of mechanisms underlying associa- tions between SLC2A9 genetic variation and urinary UA should include efforts to understand the influence of diet on SLC2A9 gene expression, urinary UA excretion, and NL risk. Supporting Information S1 Fig. Serum UA association results for 880 SNPs in the SLC2A9 gene region. Left Y-axis:– log10(p-value) from association between SNPs and serum UA, adjusted for age, sex, BMI, and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of 11 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid linkage disequilibrium with index (most strongly associated) SNP, rs11723439 (purple dia- mond). (PDF) S2 Fig. Serum UA association results for 879 SNPs in the SLC2A9 gene region, conditioned on rs11723439. Left Y-axis:–log10(p-value) from association between SNPs and serum UA, adjusted for age, sex, BMI, rs11723439 and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs16894555 (purple diamond). Corrected p-value for rs16894555 is p = 0.044, indicating a second, independently SNP associated with serum uric acid in this sample. (PDF) S3 Fig. SLC2A9-001 gene expression transcript interaction plot with dietary animal protein on urinary uric acid. (PDF) S1 Table. Correlations between gene-level gene expression for uricosuric genes in the Genetic Epidemiology Network of Arteriopathy study. ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9-001 ENST00000506583, SLC2A9-201 ENST00000309065, SLC2A9-002 ENST00000264784. (PDF) S2 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models association with urinary uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S3 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models, conditioned on rs12509955, association with urinary uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S4 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models association with serum uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S5 Table. Supporting Information Forty SLC2A9 SNPs with the lowest p-values in linear mixed models, conditioned on rs11723439, association with serum uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S6 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models association with SLC2A9 gene expression in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for BMI and urinary sodium and adjusted for sibships. (PDF) linkage disequilibrium with index (most strongly associated) SNP, rs11723439 (purple dia- mond). (PDF) linkage disequilibrium with index (most strongly associated) SNP, rs11723439 (purple dia- mond). (PDF) S2 Fig. Serum UA association results for 879 SNPs in the SLC2A9 gene region, conditioned on rs11723439. Left Y-axis:–log10(p-value) from association between SNPs and serum UA, adjusted for age, sex, BMI, rs11723439 and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs16894555 (purple diamond). Corrected p-value for rs16894555 is p = 0.044, indicating a second, independently SNP associated with serum uric acid in this sample. (PDF) S2 Fig. Serum UA association results for 879 SNPs in the SLC2A9 gene region, conditioned on rs11723439. Left Y-axis:–log10(p-value) from association between SNPs and serum UA, adjusted for age, sex, BMI, rs11723439 and urinary sodium, and accounting for sibship; Right Y-axis: SNP recombination rate based on HapMap hg18 CEU; X-axis: chromosomal location and gene regions; r2 color code: degree of linkage disequilibrium with index (most strongly associated) SNP, rs16894555 (purple diamond). Corrected p-value for rs16894555 is p = 0.044, indicating a second, independently SNP associated with serum uric acid in this sample. (PDF) S3 Fig. SLC2A9-001 gene expression transcript interaction plot with dietary animal protein on urinary uric acid. (PDF) S3 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models, conditioned on rs12509955, association with urinary uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S4 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models association with serum uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. References 1. Sierakowski R, Finlayson B, Landes RR, Finlayson CD, Sierakowski N. The frequency of urolithiasis in hospital discharge diagnoses in the United States. Investigative urology. 1978; 15(6):438–41. Epub 1978/05/01. PMID: 649290. 1. Sierakowski R, Finlayson B, Landes RR, Finlayson CD, Sierakowski N. The frequency of urolithiasis in hospital discharge diagnoses in the United States. Investigative urology. 1978; 15(6):438–41. Epub 1978/05/01. PMID: 649290. 1. Sierakowski R, Finlayson B, Landes RR, Finlayson CD, Sierakowski N. The frequency of urolithiasis in hospital discharge diagnoses in the United States. Investigative urology. 1978; 15(6):438–41. Epub 1978/05/01. PMID: 649290. 2. Scales CD Jr, Smith AC, Hanley JM, Saigal CS, Urologic Diseases in America P. Prevalence of kidney stones in the United States. European urology. 2012; 62(1):160–5. Epub 2012/04/14. doi: 10.1016/j. eururo.2012.03.052 PMID: 22498635; PubMed Central PMCID: PMC3362665. 2. Scales CD Jr, Smith AC, Hanley JM, Saigal CS, Urologic Diseases in America P. Prevalence of kidney stones in the United States. European urology. 2012; 62(1):160–5. Epub 2012/04/14. doi: 10.1016/j. eururo.2012.03.052 PMID: 22498635; PubMed Central PMCID: PMC3362665. 3. Mandel NS, Mandel GS. Urinary tract stone disease in the United States veteran population. I. Geo- graphical frequency of occurrence. The Journal of urology. 1989; 142(6):1513–5. Epub 1989/12/01. PMID: 2585626. 3. Mandel NS, Mandel GS. Urinary tract stone disease in the United States veteran population. I. Geo- graphical frequency of occurrence. The Journal of urology. 1989; 142(6):1513–5. Epub 1989/12/01. PMID: 2585626. 4. Curhan GC, Willett WC, Rimm EB, Stampfer MJ. Family history and risk of kidney stones. Journal of the American Society of Nephrology: JASN. 1997; 8(10):1568–73. Epub 1997/10/23. PMID: 9335385. 5. Griffin DG. A review of the heritability of idiopathic nephrolithiasis. Journal of clinical pathology. 2004; 57(8):793–6. Epub 2004/07/29. doi: 10.1136/jcp.2003.014886 PMID: 15280397; PubMed Central PMCID: PMC1770398. 6. Kuo CF, Grainge MJ, See LC, Yu KH, Luo SF, Valdes AM, et al. Familial aggregation of gout and rela- tive genetic and environmental contributions: a nationwide population study in Taiwan. Annals of the rheumatic diseases. 2013. Epub 2013/11/23. doi: 10.1136/annrheumdis-2013-204067 PMID: 24265412. 7. McGeown MG. Heredity in renal stone disease. Clinical science. 1960; 19:465–71. Epub 1960/08/01. PMID: 13773937. 8. Resnick M, Pridgen DB, Goodman HO. Genetic predisposition to formation of calcium oxalate renal cal- culi. The New England journal of medicine. 1968; 278(24):1313–8. Epub 1968/06/13. doi: 10.1056/ NEJM196806132782403 PMID: 5648597. 9. Trinchieri A, Mandressi A, Luongo P, Coppi F, Pisani E. Familial aggregation of renal calcium stone dis- ease. Supporting Information Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S5 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models, conditioned on rs11723439, association with serum uric acid in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for age, sex, BMI and urinary sodium and adjusted for sibships. (PDF) S6 Table. Forty SLC2A9 SNPs with the lowest p-values in linear mixed models association with SLC2A9 gene expression in the Genetic Epidemiology Network of Arteriopathy study. corrected for multiple testing. Linear mixed models adjusted for BMI and urinary sodium and adjusted for sibships. (PDF) 12 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 SLC2A9 Genotype, Gene Expression, and Uric Acid S7 Table. Gene expression by dietary protein intake interaction associations for serum uric acid in the Genetic Epidemiology Network of Arteriopathy study. statistically significant association at α = 0.05. ABCG2 ENSG00000118777, SLC17A1 ENSG00000124568, SLC17A3 ENSG00000124564, SLC22A12 ENSG00000197891, SLC2A9 ENSG00000109667, SLC2A9- 001 ENST00000506583, SLC2A9-002 ENST00000264784. (PDF) Author Contributions Conceived and designed the experiments: STT JCL. Analyzed the data: EBW ER WZ. Contrib- uted reagents/materials/analysis tools: SLRK JCL STT. Wrote the paper: EBW ER JAS SLRK JCL STT. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 References Research and reports in urology. 2013; 5:11–5. Epub 2014/01/09. doi: 10.2147/RRU. S39581 PMID: 24400230; PubMed Central PMCID: PMC3826902. 21. Augustin R, Carayannopoulos MO, Dowd LO, Phay JE, Moley JF, Moley KH. Identification and charac- terization of human glucose transporter-like protein-9 (GLUT9): alternative splicing alters trafficking. The Journal of biological chemistry. 2004; 279(16):16229–36. Epub 2004/01/24. doi: 10.1074/jbc. M312226200 PMID: 14739288. 22. Doring A, Gieger C, Mehta D, Gohlke H, Prokisch H, Coassin S, et al. SLC2A9 influences uric acid con- centrations with pronounced sex-specific effects. Nature genetics. 2008; 40(4):430–6. Epub 2008/03/ 11. doi: 10.1038/ng.107 PMID: 18327256. 23. Witkowska K, Smith KM, Yao SY, Ng AM, O'Neill D, Karpinski E, et al. Human SLC2A9a and SLC2A9b isoforms mediate electrogenic transport of urate with different characteristics in the presence of hex- oses. American journal of physiology Renal physiology. 2012; 303(4):F527–39. Epub 2012/06/01. doi: 10.1152/ajprenal.00134.2012 PMID: 22647630; PubMed Central PMCID: PMC3423118. 24. Anzai N, Jutabha P, Amonpatumrat-Takahashi S, Sakurai H. Recent advances in renal urate transport: characterization of candidate transporters indicated by genome-wide association studies. Clinical and experimental nephrology. 2012; 16(1):89–95. Epub 2011/11/01. doi: 10.1007/s10157-011-0532-z PMID: 22038265. 25. Roch-Ramel F, Guisan B. Renal Transport of Urate in Humans. News in physiological sciences: an international journal of physiology produced jointly by the International Union of Physiological Sciences and the American Physiological Society. 1999; 14:80–4. Epub 2001/06/08. PMID: 11390825. 26. Dinour D, Gray NK, Ganon L, Knox AJ, Shalev H, Sela BA, et al. Two novel homozygous SLC2A9 mutations cause renal hypouricemia type 2. Nephrology, dialysis, transplantation: official publication of the European Dialysis and Transplant Association—European Renal Association. 2012; 27(3):1035– 41. Epub 2011/08/04. doi: 10.1093/ndt/gfr419 PMID: 21810765. 27. Kawamura Y, Matsuo H, Chiba T, Nagamori S, Nakayama A, Inoue H, et al. Pathogenic GLUT9 muta- tions causing renal hypouricemia type 2 (RHUC2). Nucleosides, nucleotides & nucleic acids. 2011; 30 (12):1105–11. Epub 2011/12/03. doi: 10.1080/15257770.2011.623685 PMID: 22132964. 28. Stiburkova B, Ichida K, Sebesta I. Novel homozygous insertion in SLC2A9 gene caused renal hypouri- cemia. Molecular genetics and metabolism. 2011; 102(4):430–5. Epub 2011/01/25. doi: 10.1016/j. ymgme.2010.12.016 PMID: 21256783. 29. Sull JW, Park EJ, Lee M, Jee SH. Effects of SLC2A9 variants on uric acid levels in a Korean population. Rheumatology international. 2013; 33(1):19–23. Epub 2012/01/04. doi: 10.1007/s00296-011-2303-2 PMID: 22212408. 30. Dehghan A, Kottgen A, Yang Q, Hwang SJ, Kao WL, Rivadeneira F, et al. References The Journal of urology. 1988; 139(3):478–81. Epub 1988/03/01. PMID: 3343730. 10. Park C, Ha YS, Kim YJ, Yun SJ, Lee SC, Kim WJ. Comparison of Metabolic Risk Factors in Urolithiasis Patients according to Family History. Korean journal of urology. 2010; 51(1):50–3. Epub 2010/04/24. doi: 10.4111/kju.2010.51.1.50 PMID: 20414411; PubMed Central PMCID: PMC2855461. 11. Leonetti F, Dussol B, Berthezene P, Thirion X, Berland Y. Dietary and urinary risk factors for stones in idiopathic calcium stone formers compared with healthy subjects. Nephrology, dialysis, transplantation: official publication of the European Dialysis and Transplant Association—European Renal Association. 1998; 13(3):617–22. Epub 1998/04/29. PMID: 9550636. 12. Gutman AB, Yu TF. Uric acid nephrolithiasis. The American journal of medicine. 1968; 45(5):756–79. Epub 1968/11/01. PMID: 4879835. 13. Gault MH, Chafe L. Relationship of frequency, age, sex, stone weight and composition in 15,624 stones: comparison of resutls for 1980 to 1983 and 1995 to 1998. The Journal of urology. 2000; 164 (2):302–7. Epub 2000/07/14. PMID: 10893570. 14. Grover PK, Ryall RL, Marshall VR. Dissolved urate promotes calcium oxalate crystallization: epitaxy is not the cause. Clin Sci (Lond). 1993; 85(3):303–7. Epub 1993/09/01. PMID: 8403802. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 13 / 15 SLC2A9 Genotype, Gene Expression, and Uric Acid 15. Coe FL. Hyperuricosuric calcium oxalate nephrolithiasis. Kidney international. 1978; 13(5):418–26. Epub 1978/05/01. PMID: 661071. 16. Kirejczyk JK, Porowski T, Filonowicz R, Kazberuk A, Stefanowicz M, Wasilewska A, et al. An associa- tion between kidney stone composition and urinary metabolic disturbances in children. Journal of pedi- atric urology. 2014; 10(1):130–5. Epub 2013/08/21. doi: 10.1016/j.jpurol.2013.07.010 PMID: 23953243. 17. Pais VM Jr, Holmes RP, Assimos DG. Effect of dietary control of urinary uric acid excretion in calcium oxalate stone formers and non-stone-forming controls. Journal of endourology / Endourological Soci- ety. 2007; 21(2):232–5. Epub 2007/03/07. doi: 10.1089/end.2007.2218 PMID: 17338626; PubMed Central PMCID: PMC2705619. 18. Pak CY, Waters O, Arnold L, Holt K, Cox C, Barilla D. Mechanism for calcium urolithiasis among patients with hyperuricosuria: supersaturation of urine with respect to monosodium urate. The Journal of clinical investigation. 1977; 59(3):426–31. Epub 1977/03/01. doi: 10.1172/JCI108656 PMID: 14173; PubMed Central PMCID: PMC333378. 19. Robertson WG, Peacock M, Heyburn PJ, Marshall DH, Clark PB. Risk factors in calcium stone disease of the urinary tract. British journal of urology. 1978; 50(7):449–54. Epub 1978/12/01. PMID: 37975. 20. Mosli HA, Mosli HH, Kamal WK. Kidney stone composition in overweight and obese patients: a prelimi- nary report. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 References Association of three genetic loci with uric acid concentration and risk of gout: a genome-wide association study. Lancet. 2008; 372 (9654):1953–61. Epub 2008/10/07. doi: 10.1016/S0140-6736(08)61343-4 PMID: 18834626; PubMed Central PMCID: PMC2803340. 31. Hamajima N, Okada R, Kawai S, Hishida A, Morita E, Yin G, et al. Significant association of serum uric acid levels with SLC2A9 rs11722228 among a Japanese population. Molecular genetics and metabo- lism. 2011; 103(4):378–82. Epub 2011/04/23. doi: 10.1016/j.ymgme.2011.04.001 PMID: 21511506. 32. Kottgen A, Albrecht E, Teumer A, Vitart V, Krumsiek J, Hundertmark C, et al. Genome-wide association analyses identify 18 new loci associated with serum urate concentrations. Nature genetics. 2013; 45 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 14 / 15 SLC2A9 Genotype, Gene Expression, and Uric Acid (2):145–54. Epub 2012/12/25. doi: 10.1038/ng.2500 PMID: 23263486; PubMed Central PMCID: PMC3663712. 33. 33. Polasek O, Gunjaca G, Kolcic I, Zgaga L, Dzijan S, Smolic R, et al. Association of nephrolithiasis and gene for glucose transporter type 9 (SLC2A9): study of 145 patients. Croatian medical journal. 2010; 51 (1):48–53. Epub 2010/02/18. PMID: 20162745; PubMed Central PMCID: PMC2829176. 34. Rule AD, de Andrade M, Matsumoto M, Mosley TH, Kardia S, Turner ST. Association between SLC2A9 transporter gene variants and uric acid phenotypes in African American and white families. Rheumatol- ogy. 2011; 50(5):871–8. Epub 2010/12/28. doi: 10.1093/rheumatology/keq425 PMID: 21186168; PubMed Central PMCID: PMC3077913. 35. Voruganti VS, Franceschini N, Haack K, Laston S, Maccluer JW, Umans JG, et al. Replication of the effect of SLC2A9 genetic variation on serum uric acid levels in American Indians. European journal of human genetics: EJHG. 2013. Epub 2013/12/05. doi: 10.1038/ejhg.2013.264 PMID: 24301058. 36. Voruganti VS, Kent JW Jr, Debnath S, Cole SA, Haack K, Goring HH, et al. Genome-wide association analysis confirms and extends the association of SLC2A9 with serum uric acid levels to Mexican Ameri- cans. Frontiers in genetics. 2013; 4:279. Epub 2014/01/01. doi: 10.3389/fgene.2013.00279 PMID: 24379826; PubMed Central PMCID: PMC3863993. 37. Enomoto A, Kimura H, Chairoungdua A, Shigeta Y, Jutabha P, Cha SH, et al. Molecular identification of a renal urate anion exchanger that regulates blood urate levels. Nature. 2002; 417(6887):447–52. doi: 10.1038/nature742 PMID: 12024214. 38. Lipkowitz MS. Regulation of uric acid excretion by the kidney. Current rheumatology reports. 2012; 14 (2):179–88. doi: 10.1007/s11926-012-0240-z PMID: 22359229. 39. Reginato AM, Mount DB, Yang I, Choi HK. The genetics of hyperuricaemia and gout. Nature reviews Rheumatology. 2012; 8(10):610–21. doi: 10.1038/nrrheum.2012.144 PMID: 22945592; PubMed Cen- tral PMCID: PMC3645862. References 40. Wright AF, Rudan I, Hastie ND, Campbell H. A 'complexity' of urate transporters. Kidney international. 2010; 78(5):446–52. doi: 10.1038/ki.2010.206 PMID: 20613716. 41. Giannini S, Nobile M, Sartori L, Dalle Carbonare L, Ciuffreda M, Corro P, et al. Acute effects of moder- ate dietary protein restriction in patients with idiopathic hypercalciuria and calcium nephrolithiasis. The American journal of clinical nutrition. 1999; 69(2):267–71. Epub 1999/02/16. PMID: 9989691. 42. Tracy CR, Best S, Bagrodia A, Poindexter JR, Adams-Huet B, Sakhaee K, et al. Animal Protein and the Risk of Kidney Stones: A Comparative Metabolic Study of Animal Protein Sources. The Journal of urol- ogy. 2014. Epub 2014/02/13. doi: 10.1016/j.juro.2014.01.093 PMID: 24518789. 43. Heilberg IP, Goldfarb DS. Optimum nutrition for kidney stone disease. Advances in chronic kidney dis- ease. 2013; 20(2):165–74. Epub 2013/02/27. doi: 10.1053/j.ackd.2012.12.001 PMID: 23439376. 44. Reddy ST, Wang CY, Sakhaee K, Brinkley L, Pak CY. Effect of low-carbohydrate high-protein diets on acid-base balance, stone-forming propensity, and calcium metabolism. American journal of kidney dis- eases: the official journal of the National Kidney Foundation. 2002; 40(2):265–74. Epub 2002/07/31. doi: 10.1053/ajkd.2002.34504 PMID: 12148098. 45. Lieske JC, Turner ST, Edeh SN, Smith JA, Kardia SL. Heritability of Urinary Traits That Contribute to Nephrolithiasis. Clinical journal of the American Society of Nephrology: CJASN. 2014. Epub 2014/03/ 01. doi: 10.2215/CJN.08210813 PMID: 24578335. 46. Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins T. Measurement characteris- tics of the Women's Health Initiative food frequency questionnaire. Annals of epidemiology. 1999; 9 (3):178–87. PMID: 10192650. 47. Base SAS(R) Cary, NC: SAS Institute Inc.; 2011. 48. Pruim RJ, Welch RP, Sanna S, Teslovich TM, Chines PS, Gliedt TP, et al. LocusZoom: regional visuali- zation of genome-wide association scan results. Bioinformatics. 2010; 26(18):2336–7. doi: 10.1093/ bioinformatics/btq419 PMID: 20634204; PubMed Central PMCID: PMC2935401. 49. Gao X, Starmer J, Martin ER. A multiple testing correction method for genetic association studies using correlated single nucleotide polymorphisms. Genetic epidemiology. 2008; 32(4):361–9. doi: 10.1002/ gepi.20310 PMID: 18271029. 50. Cunningham F, Amode MR, Barrell D, Beal K, Billis K, Brent S, et al. Ensembl 2015. Nucleic Acids Research. 2014. doi: 10.1093/nar/gku1010 51. Preitner F, Laverriere-Loss A, Metref S, Da Costa A, Moret C, Rotman S, et al. Urate-induced acute renal failure and chronic inflammation in liver-specific Glut9 knockout mice. American journal of physiol- ogy Renal physiology. 2013; 305(5):F786–95. doi: 10.1152/ajprenal.00083.2013 PMID: 23804456. 52. Khitan Z, Kim DH. PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015 References Fructose: a key factor in the development of metabolic syndrome and hypertension. Journal of nutrition and metabolism. 2013; 2013:682673. doi: 10.1155/2013/682673 PMID: 23762544; PubMed Central PMCID: PMC3677638. 15 / 15 PLOS ONE \| DOI:10.1371/journal.pone.0128593 July 13, 2015
https://openalex.org/W2605761454	https://ijpds.org/article/download/286/268	English	null	Data linkage in social care: a pilot project	International journal of population data science	2,017	cc-by	432	International Journal of Population Data Science (2017) 1:266 International Journal of Population Data Science (2017) 1:266 Results Data relating to well over 20,000 referrals generated by 17,000+ social services clients in Gwynedd Local Authority from the pe- riod 2008 to 2015 were anonymised into the SAIL databank in Swansea, and linked to records from primary and secondary care. We will present results on the success of this process and on the emerging ﬁndings from the linked datasets. Objectives The main aims of this research are to: 1. Test the feasibility of linking datasets from Local Author- ity, the NHS and third sector organisations. 2. Build a more complete picture of service provision using adults who have been referred to social services in order to avoid admission to hospital or to facilitate their discharge from hospital. 3. Assess the range and quality of data available in each of the relevant organisations providing services to those indi- viduals. The research outcome is a better understanding of the utility of data linkage across statutory and third party organisations. tivecommons.org/licenses/by-nc-nd/4.0/deed.en) nd third party organisations. arch team partnered with the plore the Governance Issues and c.uk (A. Orrell) ds.v1i1.286 Access under CC BY-NC-ND 4.0 (https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en) Data linkage in social care: a pilot project Orrell, Alison1*, Heaven, Martin2, Seddon, Diane1, and Robinson, Catherine1 1Bangor University 2Swansea University Objectives practicalities of providing an anonymised dataset to the SAIL databank at Swansea. Two third sector agencies were also ap- proached. With the various required Service Level Agreements in place, data were put through the tried and trusted SAIL process for analysis. To build the complete picture of service provision there is a need to broaden the linked data available to include health, social ser- vice provision by Local Authorities, and provision of support by third sector organisations. Data Linkage in Social Care is a pi- lot project to test the feasibility of linking datasets from a local authority, the NHS and third sector organisations. The focus of this work is individual level data from adults who are referred to social services in order to avoid admission to hospital or to facilitate their discharge from hospital. The data linkage will include data from statutory and third sector organisations and services which provide interventions and support in community settings. ∗Corresponding Author: Email Address: a.orrell@bangor.ac.uk (A. Orrell) http://dx.doi.org/10.23889/ijpds.v1i1.286 August 2016 c⃝The Authors. Open Access under CC BY-NC-ND 4.0 (https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en) Journal Website: www.ijpds.org Journal Website: www.ijpds.org Approach A Bangor University led research team partnered with the Gwynedd Local authority to explore the Governance Issues and ∗Corresponding Author: Email Address: a.orrell@bangor.ac.uk (A. Orrell) ∗Corresponding Author:
https://openalex.org/W4367841708	https://www.researchsquare.com/article/rs-2868607/latest.pdf	English	null	A Novel CircRNA Circ_0001722 Regulates Proliferation and Invasion of Osteosarcoma Cells Through Targeting miR-204-5p/RUNX2 Axis	Research Square (Research Square)	2,023	cc-by	6,922	Results A novel circRNA, circ_001722, is significantly upregulated in OS tissues and cells. Downregulation of circ_0001722 can suppress proliferation and invasion of human OS cells in vitro and in vivo. Computational algorithms predict miR-204-5p can bind with circ_0001722 and RUNX2 mRNA 3’UTR, which is verified by Dual-luciferase assay and RNA immunoprecipitation assay. Further functional experiments show that circ_0001722 competitively binds to miR-204-5p and prevents it to decrease the level of RUNX2, which upregulates proliferation and invasion of human OS cells. Conclusion Circ_001722 is a novel tumor promotor in OS, and promotes the progression of OS via miR-204- 5p/RUNX2 axis. Background Osteosarcoma (OS) is the most prevalent primary fatal bone neoplasm in adolescents and children owing to limited therapeutic methods. Circular RNAs (circRNAs) are identified as vital regulators in a variety of cancers. However, the roles of circRNAs in OS are still unclear. Methods Firstly, we evaluate the differentially expressed circRNAs in 3 paired OS and corresponding adjacent nontumor tissue samples by circRNA microarray assay, finding a novel circRNA, circ_001722, significantly upregulated in OS tissues and cells. The circular structure of candidate circRNA was confirmed through Sanger sequencing, divergent primer PCR, and RNase R treatments. Proliferation of OS cells was evaluated in vitro and in vivo. The microRNA (miRNA) sponge mechanism of circRNAs was verified by dual-luciferase assay and RNA immunoprecipitation assay. Keywords: Posted Date: May 3rd, 2023 DOI: https://doi.org/10.21203/rs.3.rs-2868607/v1 DOI: https://doi.org/10.21203/rs.3.rs-2868607/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Journal of Cancer Research and Clinical Oncology on July 15th, 2023. See the published version at https://doi.org/10.1007/s00432-023-05166-3. Page 1/19 Patients and OS samples A series of 20 surgically resected fresh human OS and corresponding adjacent nontumor tissue samples were collected at the First Affiliated Hospital of Zhengzhou University (Zhengzhou, China) and snap- frozen in liquid nitrogen from July 2018 to January 2019. Among them, 3 pairs were used for circRNA microarray analysis. No patients had received any preoperative treatment. Clinical data of patients included in this study are detailed in Supplementary Table 1. Samples used in this study were approved by the Committees for Ethical Review of the First Affiliated Hospital of Zhengzhou University. Background Page 2/19 Osteosarcoma (OS) is the most prevalent primary malignant bone neoplasm causing substantial morbidity in adolescents and children [1]. It originates from mesenchymal cells and is characterized by rapid infiltrating growth, early lung metastasis and a high recurrence rate [2]. Studies have shown that the overall 5-year survival rate of patients with localized OS ranges between 65 ~ 75% and is only 20% for those with recurrent and metastatic tumors [3]. Despite advances in OS treatment approaches such as adjuvant chemotherapy and surgical resection, the survival rates have plateaued in the last 3 decades and are less than satisfactory [4]. Indeed, no specific diagnostic and prognostic biomarkers for OS have Page 2/19 Page 2/19 been found. Consequently, molecular studies aiming to identify promising therapeutic targets for OS are urgently needed. Circular RNAs (circRNAs) regulate various functions in eukaryotic cells [5]. Based on the order of splicing events and different intermediates, two mechanisms exist for the biogenesis of circRNAs: canonical spliceosome induced splicing and noncanonical lariat splicing [6, 7]. Accumulating studies have shown that circRNAs modulate diverse physiological and pathophysiological processes by sponging microRNAs (miRNAs), interacting with RNA binding proteins, and modulating epigenetic, transcriptional, or translational alterations in target genes as well [8–11]. Abnormal circRNA expression has been found to correlate with the pathogenesis of various cancers and to exert essential regulatory effects on gene expression, cell invasion, cell cycle progression, migration, apoptosis, and proliferation [12–14]. Moreover, circRNAs are thought to possess high diagnostic and therapeutic potential given their structural stability, evolutionary conservation, abundance and organ specificity [15, 16]. However, to date, the roles of circRNAs in OS are not clearly known. This study evaluated the expression profiles of circRNAs in OS tissues using high-throughput sequencing. We found a novel circRNA, designated circ_001722, significantly upregulated in OS tissues and cells. Our experimental results indicated that circ_001722 exerted pro-oncogenic effects on OS proliferation and invasion through circ_001722/miR-204-5p/RUNX2 Axis. This preliminary study revealed that circ_001722 is a potential therapeutic target for OS. CircRNA microarray analysis Three pairs of human OS and corresponding adjacent nontumor tissue samples were used for the circRNA microarray assay to determine differentially expressed circRNAs. The microarray hybridization was performed based on the manufacturer’s standard protocols (Agilent Technologies, USA), which included purifying the RNA, transcribing it into fluorescent cRNA, and then hybridizing it onto the Human circRNA Arrays (Agilent Technologies, USA). Finally, the hybridized slides were washed, fixed and scanned to images by an Agilent Scanner G2505C. The data collection was performed using Agilent Feature Extraction software (version 11.0.1.1). The raw data were quantile normalized, and further data analysis was performed with the R software package, GeneSpring GX (Agilent Technologies, USA) and gene Page 3/19 Page 3/19 expression dynamics inspector (GEDI). The statistical significance of differentially regulated circRNAs between OS tissue (T) and adjacent nontumor tissue (N) was identified through p-values and fold changes. Significantly differentially expressed transcripts were retained by screening for a fold change ≥ 2.0 and P < 0.05. Hierarchical clustering was performed to generate an overview of the characteristics of expression profiles based on the values of all expressed transcripts and significant differentially expressed transcripts. Circular structure confirmation The circular structure of circ_0001722 was confirmed by RNase R treatment and Sanger sequencing by divergent primer PCR. For RNase R treatment, 3 µg total RNA extracted from OS tissues and cell lines were incubated with 20 U RNAse R (Epicentre Biotechnologies, USA) in a 10 µl volume at 37°C for 45 min, followed by 70°C for 10 min to deactivate the RNase R. The treated RNAs were used for RT-PCR. For Sanger sequencing, PCR products amplified by divergent primers of circ_0001722 were inserted into the T vector and sequenced by Tsingke Biotechnology (Beijing) Co., Ltd. The result was crosschecked with the back-splicing junction sites of circ_0001722 supplied by circBASE [17]. Cell culture, transfection, and lentiviral infection Two human OS cell lines (MG63 and U2OS), a normal human fetal osteoblastic cell line (hFOB1.19) and a human embryonic kidney cell line (HEK293T) used in this study were purchased from Chinese National Collection of Authenticated Cell Cultures. MG63 cells were cultured in DMEM medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. U2OS cells were cultured in McCoy's 5a medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. hFOB1.19 cells were cultured in DMEM/F12 (1:1) medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. HEK293T cells were cultured in DMEM/high glucose medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. The authenticity of cell lines was verified by DNA fingerprinting before use. MiR-204-5p and its negative control (miR-NC), RUNX2 eukaryotic expression recombinant pIRESpuro2-RUNX2 were transfected transiently into OS cells using Lipofectamine 3000 (Invitrogen, USA) according to the manufacturer’s instructions. To prepare sh-circ_0001722 lentiviral particles, the lentiviral vector harboring sh-circ_0001722 full hairpin sequence and packaging vectors were transfected into HEK293T cells using iMFectin Poly DNA Transfection Reagent (GenDEPOT, USA) following the manufacturer’s suggested protocols. The transfection medium was changed at 8 h after transfection and then cells were cultured for 36 h. The lentiviral particles were harvested by filtration using a 0.45 µm sodium acetate syringe filter and then combined with 8 µg/ml of polybrane (Millipore, USA) and infected Two human OS cell lines (MG63 and U2OS), a normal human fetal osteoblastic cell line (hFOB1.19) and a human embryonic kidney cell line (HEK293T) used in this study were purchased from Chinese National Collection of Authenticated Cell Cultures. MG63 cells were cultured in DMEM medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. U2OS cells were cultured in McCoy's 5a medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% ( ) ( ) penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. Cell culture, transfection, and lentiviral infection hFOB1.19 cells were cultured in DMEM/F12 (1:1) medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. HEK293T cells were cultured in DMEM/high glucose medium (Hyclone, USA) supplemented with 10% fetal bovine serum (HyClone, USA) and 1% penicillin/streptomycin (Invitrogen, USA) at 37°C under 5% CO2 and saturated moisture. The authenticity of cell lines was verified by DNA fingerprinting before use. MiR-204-5p and its negative control (miR-NC), RUNX2 eukaryotic expression recombinant pIRESpuro2-RUNX2 were transfected transiently into OS cells using Lipofectamine 3000 (Invitrogen, USA) according to the manufacturer’s instructions. To prepare sh-circ_0001722 lentiviral particles, the lentiviral vector harboring sh-circ_0001722 full hairpin sequence and packaging vectors were transfected into HEK293T cells using iMFectin Poly DNA Transfection Reagent (GenDEPOT, USA) following the manufacturer’s suggested protocols. The transfection medium was changed at 8 h after transfection and then cells were cultured for 36 h. The lentiviral particles were harvested by filtration using a 0.45 µm sodium acetate syringe filter and then combined with 8 µg/ml of polybrane (Millipore, USA) and infected Page 4/19 Page 4/19 overnight into 60% confluent OS cells. The cell culture medium was replaced with fresh complete growth medium and after 24 h, cells were selected with 2 µg/ml of puromycine for an additional 24 h. The selected cells were used for experiments. MiR-204-5p mimic and its negative control (miR-NC), pIRESpuro2-RUNX2 vectors, lentiviral vectors harboring sh-circ_0001722 full hairpin sequence and mock sequence, were purchased from Shanghai GenePharma Co., Ltd. RNA extraction and qRT-PCR analysis Total RNA derived from human OS tissues and cells was isolated using TRIzol reagent (TAKARA, CHN) according to the manufacturer’s instructions. RNA was reverse transcribed into cDNA using a Primer- Script one step RT-PCR kit (TAKARA, CHN). Quantitative real-time PCR experiments were performed using a SYBR Premix Dimmer Eraser kit (TAKARA, CHN) on an ABI 7500 Real-Time PCR System (Applied Biosystems, USA). The fold change in relative expression level was calculated using the 2−ΔΔCt method. Relative circ_0001722 expression was normalized to GAPDH expression, and miR-204-5p expression was normalized to U6 small nuclear RNA (U6 snRNA). The primer sequences used in our study are purchased from Tsingke Biotechnology (Beijing) Co., Ltd and shown in Supplementary Table 2. Cell proliferation assay The proliferation of human OS cells was evaluated by the CellTiter 96 AQueous One Solution cell proliferation assay. human OS cells were plated in 96-well culture plates (3 × 103 per well). After 24 h of incubation, the cells were transfected with 30 pmol of target gene (sh-circ_0001722 or miR-204-5p or negative control) for 24, 48, 72, and 96 h. Then 20 µl Cell Titer 96 Aqueous One Solution (Promega, USA) were added and cells incubated for another 1 h. Absorbance was read at 492 nm. Matrigel invasion assay The invasion abilities of OS cells were evaluated using Transwell invasion chambers precoated with 50 µl of 2 mg/ml Matrigel (BD Biosciences, USA). In brief, 5 × 104 transfected cells suspended in 200 µl of serum-free DMEM were seeded into the upper chambers. A 600 µl volume of DMEM supplemented with 10% FBS was used as the attractant and was added into the lower chambers. After culture for 24 h, cells adhering to the lower surface of the membrane were fixed with paraformaldehyde (4%) and stained using crystal violet (0.1%), whereas cells on the upper surface of the membrane were removed by wiping with cotton swabs. At least three random fields of view containing cells that had migrated or invaded to the lower surface were imaged under an inverted light microscope. Western blot analysis Protein concentrations of cell lysates were determined using a protein assay kit (Bio-Rad Laboratories, USA). Total proteins (20 µg) were separated by SDS-PAGE and transferred onto a polyvinylidene difluoride membrane (Millipore, USA). After blocking in 5% non-fat milk, the membranes were probed with primary antibodies (RUNX2: sc-101145, GAPDH: sc-47724, Santa Cruz Biotechnology, USA) overnight at 4°C, then washed 3 times with TBS-Tween 20 followed by incubation at room temperature 1 h with a horseradish peroxidase (HRP)-conjugated secondary antibody. The protein bands were visualized with an Immobilon Western Chemiluminescent HRP Substrate (Millipore, USA). Dual-luciferase assay Page 5/19 Page 5/19 HEK293T cells were spread to 96 well plates at the concentration of 1×104 cells per well. After 24 hours, HEK293T cells were co-transfected with dual-luciferase reporter vector (pmirGLO-Wt-circ or pmirGLO-Mt- circ, pmirGLO-Wt-3’UTR or pmirGLO-Mt-3’UTR) and miR-204-5p mimics or negative control (miR-NC) using the Lipo-fectomine 3000 transfection reagent (Invitrogen, USA), respectively. After 48h of incubation, firefly and Renilla luciferase activities were measured using a dual-luciferase reporter assay system (Promega, USA) according to the manufacturer’s instructions. RNA immunoprecipitation (RIP) assay The RIP assay was performed using a Magna RIP RNA Binding Protein Immunoprecipitation Kit (Bersinbio, China) according to the manufacturer’s protocol. 2×107 MG63 or U2OS cells were lysed in complete RIP lysis buffer and the cell lysates were divided into two equal parts and incubated with either 5 µg human anti-Argonaute2 (AGO2) antibody (Millipore, USA) with rotation at 4°C overnight. Magnetic beads were added to the cell lysates and incubation was continued at 4°C for 1h. The samples were then incubated with Proteinase K at 55°C for 1h. The enriched RNA was obtained using RNA Extraction Reagent (Solarbio, CHN). The purified RNA was used to detect the expression levels of the genes of interest by qRT-PCR. Xenograft nude mouse model Six-week-old male BALB/C nude mice (Vital River Laboratory Animal Technology, Beijing, CHN) were maintained under specific pathogen-free conditions with a 12h light/dark cycle. All animal experiments were performed in accordance with the guide lines for the Care and Use of Laboratory Animals of Zhengzhou University. MG63 cells stably transfected with sh-circ_0001722 lentivirus or control lentivirus were subcutaneously injected into the right upper back of the nude mice (5×106 cells per mouse). Four weeks later, the mice were sacrificed and tumor tissues were collected for examination of the parameters of interest. Immunohistochemistry staining Page 6/19 Immunohistochemical analysis for Ki67 was performed on 4-µm sections. The Envision Plus detection system (Dako, USA) was used for the detection of immunostaining. Tissue sections were pretreated with 10 mM sodium citrate buffer for antigen unmasking (pH 6.0) after deparaffinized in xylene. Endogenous peroxidase activity was blocked by incubation with 0.03% hydrogen peroxide in methanol for 15 min. Then sections were incubated with Ki67 primary antibody (MA5-14520, Thermo Scientific, USA) at 4°C overnight after blocked in normal serum for 30min. Next, Sections were incubated with secondary antibody at room temperature for 60 min before staining for 5 min with 3'3-diaminobenzidine tetrahydrochloride, counterstained by hematoxylin, and observed by microscope (200×). The results of circRNA identification and annotation The reads were aligned to the reference genome using bwa mem software. The read distribution of OS and corresponding adjacent nontumor tissues on the chromosomes were shown in Supplementary Fig. 1a and 1b, separately. After bwa mem alignment, sam files were subjected to CIRI2 processing twice. Firstly, the junction reads were detected by paired chiastic clipping (PCC) signals. Basing on paired-end mapping (PEM) and GT-AG sequence feature, the reads were preliminary filtered to obtain candidate circRNAs. Secondly, the junction reads were detected again to filter out the false-positive candidate circRNAs (Supplementary Fig. 1c). Totally, 6646 circRNAs were identified in OS and nontumor tissues, and in each type of samples, with 1–3 circRNAs on most genes. Among these 6646 circRNAs, 6014 circRNAs were distributed on exon region, 459 circRNAs were distributed on intron region, and the rest 173 circRNAs were distributed on intergenic region. Moreover, 4275 circRNAs have been annotated before, and the remaining 2371 were newly annotated. The lengths of 6646 circRNAs were mainly distributed in the range of 150–1000 bp (Supplementary Fig. 1d). Statistical analysis Data for continuous variables are presented as means ± standard deviations. All analyses were performed using SPSS 21.0 software (IBM, USA). All experiments were performed with three technical replicates, and at least three biological replicates were performed. Differences between groups were analyzed using unpaired Student’s t-test or one-way analysis of variance (ANOVA) with Tukey’s test. A P value of < 0.05 was considered to be statistically significant. Expression of circ_0001722 is significantly upregulated in OS tissues and cell lines Firstly, we determined whether circ_0001722 is a closed circular RNA that is resistant to RNase R digestion. We investigated its expression level in OS tissues and cell lines. Our result showed that RNase R digestion could decrease the RNA level of linear GAPDH, but could not affect the level of circ_0001722 significantly (Fig. 1d), indicating that circ_0001722 was resistant to RNase R digestion. To confirm circ_0001722 is a closed circRNA, we ran PCR amplification of circ_0001722 with specific divergent primers. PCR products amplified by divergent primers were detected by Sanger sequencing to confirm the circular structure of circ_0001722. Our data showed that the back splicing junction sites of circ_0001722 were consistent with the sequence from circBase (Fig. 1e). Then, we measured the expression level of circ_0001722 in a series of 20 surgically removed fresh OS tissues and their corresponding adjacent nontumor tissues by qRT-PCR. Comparing with nontumor tissues, the expression level of circ_0001722 in OS tissues was significantly upregulated (P < 0.01; Fig. 1f). Similarly, the expression of circ_0001722 in OS cell lines (MG63 and U2OS) was significantly higher than in normal human fetal osteoblastic cell line (hFOB 1.19) (P < 0.01; Fig. 1g). Downregulation of circ_0001722 suppresses cell proliferation of human OS cells in vitro and in vivo. Differentially expressed circRNAs and circRNA/miRNA/mRNA regulatory network Additionally, we have also identified the differentially expressed circRNAs between OS and corresponding adjacent nontumor tissue samples. Compared with nontumor samples, there were totally 1088 significantly differentially expressed circRNAs in OS samples, including 1052 upregulated circRNAs and Page 7/19 Page 7/19 36 downregulated circRNAs (Fig. 1a). The expression levels of differentially expressed circRNAs were significantly different (Fig. 1b). Furthermore, we have estimated the circRNA/miRNA interactions using miRanda and TargetScan. After the cross analysis of predicted results of two software, there were 36 differentially expressed circRNAs and their corresponding miRNAs in regulatory network (Supplementary Fig. 2a). Among them, circ_0001722 with small expression standard deviations in three OS samples and three nontumor samples (standard deviation values were 3.1 and 0.7, respectively) was selected for subsequent analysis. Regarding circ_0001722, there were 11 overlapped regulatory pairs between miRanda and TargetScan (Supplementary Fig. 2b), including miR-365b-5p, miR-3122, miR-211-5p, miR-208b-5p, miR-3913-5p, miR- 204-5p, miR-365a-5p, miR-5006-3p, miR-208a-5p, miR-3137 and miR-6875-3p. Then, the target genes of these 11 miRNAs were also estimated utilizing miRTarBase database. We found that miR-204-5p and its corresponding 76 target genes, comprising RUNX2, were supported by most experimental validation (469 experimental results). The circRNA/miRNA/mRNA regulatory network, based on circ_0001722, 11 miRNAs, and target mRNAs, was shown in Fig. 1c. Downregulation of circ_0001722 suppresses cell proliferation of human O We noticed that the expression of circ_0001722 was higher in MG63 cells and U2OS cells. Based on the observation, exogenous shRNA (sh-circ_0001722) was used to knock down circ_0001722 expression in two OS cell lines (Fig. 2a). Based on the high silencing efficiency, we carried out cell proliferation assay and matrigel invasion assay. The results showed that the viability level of MG63 cells and U2OS cells transfected with sh-circ_0001722 was lower than that of the cells transfected with sh-Mock (Fig. 2b). Consistently, sh-circ_0001722 significantly decreased the number of OS cells invaded through the Page 8/19 Page 8/19 matrigel. Quantitative analysis of cell numbers revealed that, in the negative control (sh-Mock) and blank control groups, the number of cells invaded through the matrigel was almost 5 times higher than sh- circ_0001722 group (Fig. 2c). The above experiments verified to some extent that silencing circ_0001722 can inhibit the proliferation and invasion of OS cells in vitro. To investigate whether circ_0001722 regulates the tumorigenesis of OS in vivo, we established OS xenograft mouse models. MG63 cells transfected with sh-circ_0001722 or sh-Mock were inoculated subcutaneously in the right flank of athymic nude mice (5 × 106 cells per mouse, 5 mice per group). After 4 weeks, all experimental mice were euthanized and tumor tissues were collected (Fig. 2d). Comparing with sh-Mock group, smaller tumor volume and lower tumor weight were observed in sh-circ_0001722 group. The same conclusion was reached with the in vitro experimental results. Downregulation of circ_0001722 can suppress tumorigenesis of OS cells in vivo. In addition, compare with the control group, knocking down of circ_0001722 led to the decrease of Ki67 expression, which implied the cell proliferation of sh- circ_0001722 group is slower than sh-Mock group. Taken altogether, downregulation of circ_0001722 can restrain the growth of OS cells in vitro and in vivo. Circ_0001722 acts as a sponge for miR-204-5p. The sequences of circRNAs are highly conservative, and circRNAs can function as miRNA sponge to regulate gene expression. Through bioinformatical analysis, we discovered that complementary pairing sites (80–89 nt) on circ_0001722 that could bind to miR-204-5p (Fig. 3a). To validate binding capability of the miR-204-5p to circ_0001722, we constructed the circ_0001722 luciferase reporter system. In the dual-luciferase reporter assay, comparing with miR-NC, miR-204-5p mimics could significantly suppressed the luciferase activity of pmirGLO-Wt-circ in HEK293T cells, while the luciferase activity of pmirGLO-Mt-circ could not be suppressed (Fig. 3b). We next performed Ago2 immunoprecipitation to determine whether circ_0001722 served as a platform for Ago2 and miR-204-5p. As shown in Fig. 3c, circ_0001722 and miR-204-5p were abundantly enriched more in Ago2 protein than in IgG, suggesting that expression of miR-204-5p could be affected by circ_0001722. Moreover, the expression of miR-204- 5p was elevated in circ_0001722 silenced MG63 and U2OS cells (Fig. 3d). Furthermore, the results of Pearson’s correlation analysis showed the circ_0001722 level was inversely to correlate with miR-204-5p level in 20 human OS tissues (Fig. 3e), which providing evidence of the potential correlation between circ_0001722 and miR-204-5p. Given all of these data, circ_0001722 not only targeted miR-204-5p but also acted as a sponge for miR-204-5p in OS cells. Sh-circ_0001722 inhibits the proliferation and invasion of human OS cells by upregulating miR-204-5p. Sh-circ_0001722 inhibits the proliferation and invasion of human OS cells by upregulating miR-204-5p. Sh-circ_0001722 mediates OS cells suppression through the miR-204-5p/RUNX2 axis. Computational algorithms predicted that miR-204-5p could specifically bind to complementary sequence (268–274 nt) of RUNX2 3’UTR (Fig. 5a). We carried out the 3’UTR luciferase reporter assay to verify whether miR-204-5p can bind to RUNX2 3’UTR and inhibit its expression. The result showed that miR-204- 5p could significantly reduce the luciferase activity of pmirGLO-Wt-3’UTR clone, but could not affect the luciferase activity of pmirGLO-Mt-3’UTR clone (Fig. 5b), indicating that miR-204-5p can directly target the 3’UTR of RUNX2 mRNA, leading to the inhibition of its translation. In addition, Western blotting analysis showed that, both miR-204-5p and sh-circ_0001722 could induce a significantly reduction of endogenous RUNX2 expression in OS cells (Fig. 5c). Moreover, rescue assays showed that the cell proliferation ability inhibited by sh-circ_0001722 could be rescued by exogenous RUNX2 (Fig. 5d). The inhibitory effect of sh- circ_0001722 on cell invasion could also be rescued by exogenous RUNX2 as well (Fig. 5e). Our results verified that miR-204-5p contributed to proliferation and invasion of OS cells through targeting RUNX2 mRNA, and sh-circ_0001722 reduced tumorigenesis of OS cells by directly binding and upregulating miR- 204-5p/RUNX2 axis. h-circ_0001722 inhibits the proliferation and invasion of human OS cells by In order to further investigate whether circ_0001722 plays a promotion role in OS cells through binding with miR-204-5p, we conducted the following experiments. We transfected MG63 and U2OS cells with miR-204-5p, making miR-204-5p expression increased remarkably, which also could be observed after sh- circ_0001722 transfection as well (Fig. 4a). The cell proliferation assays showed that both upregulation of miR-204-5p and downregulation of circ_0001722 could suppress OS cell proliferation ability in vitro (Fig. 4b). Transwell assay showed that cell invasion was remarkably restrained after sh-circ_0001722 or Page 9/19 Page 9/19 miR-204-5p transfection (Fig. 4c). These findings demonstrated that miR-204-5p contributed to proliferation and invasion of OS cells, and sh-circ_0001722 reduced tumorigenesis in OS cells by directly binding and upregulating miR-204-5p. Discussion OS is the most prevalent malignant bone tumor. It is highly metastatic, resulting in a very poor survival rate [2]. Approximately 80% of OS patients exhibit subclinical pulmonary micro metastases at the time of diagnosis [18]. The lack of accurate biomarkers has further hindered efforts to improve the clinical outcome of OS. Recently, the dysregulation of ncRNAs in OS has generated significant interest from scientific communities. Being different from the other miRNAs or lncRNAs, circRNAs have emerged as more reliable and promising tumor biomarkers owing to their exceptionally stable structure. Advanced genome sequencing techniques have validated the roles of circRNAs in multiple cancers, including hepatocellular carcinoma [19], gastric cancer [20], colorectal cancer [21] and lung squamous cell carcinoma [22]. However, to date, the expression profiles and roles of circRNAs in OS are not well understood. Our study provides the first evidence that circ_001722 contributes to the malignant progression of OS. CircRNAs are widely accepted to be an unorthodox RNA species generated by alternative splicing of pre- mRNAs [23]. There are three main classes of circRNAs: exonic circRNAs, exon-intron circRNAs and intronic circRNAs [24]. Herein, we revealed upregulated expression of circ_001722 in OS tissues and cells using high-throughput sequencing and qRT-PCR. Functional analyses further validated the role of circ_001722 in promoting the proliferation and metastasis of OS cells both in vivo and in vitro. Page 10/19 The subcellular distribution of RNAs is intimately tied to their biological functions [25]. Accumulating evidence shows that cytoplasmic circRNAs sponge miRNAs, which represses the translation or induces the degradation of the target mRNAs. Herein, through bioinformatical analysis, we discovered that complementary pairing sites on (82–89 nt) circ_0001722 that can bind to miR-204-5p. Despite this new finding, the involvement of miR-204-5p in the pathogenesis of multiple tumors is not a new phenomenon. Reports on the interactions between miR-204-5p and circRNAs in cancer are scarce. Herein, we found that miR-204-5p expression was inversely correlated with circ_001722. Functional rescue experiments further revealed that the miR-204-5p inhibitor substantially reversed the suppressive effects of circ_001722 depletion on proliferation and metastasis of OS cells, whereas miR-204-5p could abolish the promotive effects of circ_001722 overexpression. Moreover, we found that RUNX2 is a downstream target of miR-204-5p in OS cells. Functional experiments revealed that circ_001722 upregulated RUNX2 expression by sponging miR-204-5p. Discussion But, RUNX2 triggered which pathway activation to accelerate OS progression via mechanisms including suppression of apoptosis and promotion of cell proliferation, migration and invasion, still need to explore. Evidence indicates that RUNX2 is implicated in diverse biological processes, including bone development, tumor invasion and metastasis [26, 27]. Initial findings indicate the carcinogenesis mediated by the circ_001722/miR-204-5p/RUNX2 axis in OS. Conclusions In summary, our research firstly showed that circ_001722 promotes the progression and metastasis of OS via the circ_001722/miR-204-5p/RUNX2 axis. Our findings elucidate a novel regulatory network that may offer new insight into the identification of potential biomarkers or therapeutic targets for OS. Ethics approval and consent to participate All aspects of this study were approved by Institutional Research Ethics Committee of the First Affiliated Hospital of Zhengzhou University. Written informed consents were obtained from all participants Consent for publication Not applicable. Availability of data and material All data are fully available without restrictions. Data Availability All data are fully available without restrictions. Acknowledgements Not applicable. Acknowledgements This work was supported by National Natural Science Foundation of China (82102715), the Foundation of Henan Educational Committee (21A320030) and the Henan Provincial Science and Technology Research Project (212102310126). Contributions HW contributed to the conception of the study; ZY collected clinical cases; GS and PLN performed the experiment; GS contributed to analysis and manuscript preparation; GS and WLY performed the data analyses; HW and GS wrote the manuscript; HW made contributions to quality control of the study and to critical revision of the manuscript; all authors have read and approved the final manuscript. Competing interests The authors declare no competing interests. Page 11/19 Page 11/19 This work was supported by National Natural Science Foundation of China (82102715), the Foundation of Henan Educational Committee (21A320030) and the Henan Provincial Science and Technology Research Project (212102310126). Authors' contributions WH contributed to the conception of the study; YZ collected clinical cases; SG and LNP performed the experiment; SG contributed to analysis and manuscript preparation; SG and LYW performed the data analyses; WH and SG wrote the manuscript; WH made contributions to quality control of the study and to critical revision of the manuscript; all authors have read and approved the final manuscript. References 1. Ritter J, Bielack SS. Osteosarcoma. Ann Oncol. 2010;21(Suppl 7):i320–5. 1. Ritter J, Bielack SS. Osteosarcoma. Ann Oncol. 2010;21(Suppl 7):i320–5 2. Ottaviani G, Jaffe N. The epidemiology of osteosarcoma. Cancer Treat Res. 2009;152:3–13. https://doi.org/10.1007/978-1-4419-0284-9_1. 2. Ottaviani G, Jaffe N. The epidemiology of osteosarcoma. Cancer Treat Res. 2009;152:3–13. https://doi.org/10.1007/978-1-4419-0284-9_1. 3. Miwa S, Shirai T, Yamamoto N, Hayashi K, Takeuchi A, Igarashi K, et al. Current and emerging targets in immunotherapy for osteosarcoma. J Oncol. 2019;2019:7035045. 4. Bielack SS, Kempf-Bielack B, Delling G, Exner GU, Flege S, Helmke K, et al. Prognostic factors in high- grade osteosarcoma of the extremities or trunk: an analysis of 1,702 patients treated on neoadjuvant cooperative osteosarcoma study group protocols. J Clin Oncol. 2002;20(3):776–90. https://doi.org/10.1200/JCO.2002.20.3.776. Page 12/19 Page 12/19 5. Meng S, Zhou H, Feng Z, Xu Z, Tang Y, Li P, et al. CircRNA: functions and properties of a novel potential biomarker for cancer. Mol Cancer. 2017;16(1): 94. https://doi.org/10.1186/s12943-017- 0663-2. 6. Chen LL, Yang L. Regulation of circRNA biogenesis. RNA Biol. 2015;12(4): 381–8. https://doi.org/10.1080/15476286.2015.1020271. 7. Zhou WY, Cai ZR, Liu J, Wang DS, Ju HQ, Xu RH. Circular RNA: metabolism, functions and interactions with proteins. Mol Cancer. 2020;19(1):172. https://doi.org/10.1186/s12943-020-01286- 7. Zhou WY, Cai ZR, Liu J, Wang DS, Ju HQ, Xu RH. Circular RNA: metabolism, functions and interactions with proteins. Mol Cancer. 2020;19(1):172. https://doi.org/10.1186/s12943-020-01286- 3. 8. Ashwal-Fluss R, Meyer M, Pamudurti NR, Ivanov A, Bartok O, Hanan M, et al. Kadener S: circRNA biogenesis competes with pre-mRNA splicing. Mol Cell. 2014;56(1):55–66. https://doi.org/10.1016/j.molcel.2014.08.019. 9. Du WW, Yang W, Liu E, Yang Z, Dhaliwal P, Yang BB. Foxo3 circular RNA retards cell cycle progression via forming ternary complexes with p21 and CDK2. Nucleic Acids Res. 2016;44(6):2846–58. https://doi.org/10.1093/nar/gkw027. 10. Su M, Xiao Y, Ma J, Tang Y, Tian B, Zhang Y, et al. Circular RNAs in Cancer: emerging functions in hallmarks, stemness, resistance and roles as potential biomarkers. Mol Cancer. 2019;18(1):90. https://doi.org/10.1186/s12943-019-1 002–6. 11. Li Z, Huang C, Bao C, Chen L, Lin M, Wang X, et al. Exon-intron circular RNAs regulate transcription in the nucleus. Nat Struct Mol Biol. 2015; 22(3): 256–64. https://doi.org/10.1038/nsmb.2959. 12. Wang D, Yang S, Wang H, Wang J, Zhang Q, Zhou S, et al. The progress of circular RNAs in various tumors. Am J Transl Res. 2018;10(6):1571–82. 13. Li Z, Yanfang W, Li J, Jiang P, Peng T, Chen K, et al. References Tumor-released exosomal circular RNA PDE8A promotes invasive growth via the miR-338/MACC1/MET pathway in pancreatic cancer. Cancer Lett. 2018;432:237–50. https://doi.org/10.1016/j.canlet.2018.04.035. 14. Zhao ZJ, Shen J. Circular RNA participates in the carcinogenesis and the malignant behavior of cancer. RNA Biol. 2017;14(5):514–21. https://doi.org/1 0.1080/15476286.2015.1122162. 14. Zhao ZJ, Shen J. Circular RNA participates in the carcinogenesis and the malignant behavior of cancer. RNA Biol. 2017;14(5):514–21. https://doi.org/1 0.1080/15476286.2015.1122162. 15. Rybak-Wolf A, Stottmeister C, Glažar P, Jens M, Pino N, Giusti S, et al. Circular RNAs in the mammalian brain are highly abundant, conserved, and dynamically expressed. Mol Cell. 2015;58(5):870–85. https://doi.org/10.1016/j. molcel.2015.03.027. 15. Rybak-Wolf A, Stottmeister C, Glažar P, Jens M, Pino N, Giusti S, et al. Circular RNAs in the mammalian brain are highly abundant, conserved, and dynamically expressed. Mol Cell. 2015;58(5):870–85. https://doi.org/10.1016/j. molcel.2015.03.027. 16. Cui X, Wang J, Guo Z, Li M, Li M, Liu S, et al. Emerging function and potential diagnostic value of circular RNAs in cancer. Mol Cancer. 2018;17(1): 123. https://doi.org/10.1186/s12943-018-0877-y. 16. Cui X, Wang J, Guo Z, Li M, Li M, Liu S, et al. Emerging function and potential diagnostic value of circular RNAs in cancer. Mol Cancer. 2018;17(1): 123. https://doi.org/10.1186/s12943-018-0877-y. 17. Chen X, Zhou Y, Liu S, Zhang D, Yang X, Zhou Q, et al. LncRNA TP73-AS1 predicts poor prognosis and functions AS oncogenic lncRNA in osteosarcoma. J Cell Biochem. 2018. 17. Chen X, Zhou Y, Liu S, Zhang D, Yang X, Zhou Q, et al. LncRNA TP73-AS1 predicts poor prognosis and functions AS oncogenic lncRNA in osteosarcoma. J Cell Biochem. 2018. 18. Jaffe N. Osteosarcoma: review of the past, impact on the future. The American experience. Cancer Treat Res. 2009;152:239–62. https://doi.org/10.1007/978-1-4419-0284-9_12. 18. Jaffe N. Osteosarcoma: review of the past, impact on the future. The American experience. Cancer Treat Res. 2009;152:239–62. https://doi.org/10.1007/978-1-4419-0284-9_12. 19. Xu J, Ji L, Liang Y, Wan Z, Zheng W, Song X, et al. CircRNA-SORE mediates sorafenib resistance in hepatocellular carcinoma by stabilizing YBX1. Signal Transduct Target Ther. 2020;5(1):298. https://doi.org/10.1038/s41392-020-00375-5. 19. Xu J, Ji L, Liang Y, Wan Z, Zheng W, Song X, et al. CircRNA-SORE mediates sorafenib resistance in hepatocellular carcinoma by stabilizing YBX1. Signal Transduct Target Ther. 2020;5(1):298. https://doi.org/10.1038/s41392-020-00375-5. Page 13/19 Page 13/19 20. Niu Q, Dong Z, Liang M, Luo Y, Lin H, Lin M, et al. Circular RNA hsa_circ_0001829 promotes gastric cancer progression through miR-155-5p/SMAD2 axis. J Exp Clin Cancer Res. 2020;39(1):280. https://doi.org/10.1186/s13046-020-01790-w. 21. References Wu M, Kong C, Cai M, Huang W, Chen Y, Wang B, et al. Hsa_circRNA_002144 promotes growth and metastasis of colorectal cancer through regulating miR-615-5p/LARP1/mTOR pathway. Carcinogenesis. 2020. 22. Harrison EB, Porrello A, Bowman BM, Belanger AR, Yacovone G, Azam SH, et al. A circle RNA regulatory Axis promotes lung squamous metastasis via CDR1-mediated regulation of Golgi trafficking. Cancer Res. 2020;80(22):4972–85. https://doi.org/10.1158/0008-5472.CAN-20-1162. 23. Jia GY, Wang DL, Xue MZ, Liu YW, Pei YC, Yang YQ, et al. CircRNAFisher: a systematic computational approach for de novo circular RNA identification. Acta Pharmacol Sin. 2019;40(1):55–63. https://doi.org/10.1038/s41401-018-0063-1. 24. Ruan Y, Li Z, Shen Y, Li T, Zhang H, Guo J. Functions of circular RNAs and their potential applications in gastric cancer. Expert Rev Gastroenterol Hepatol. 2020;14(2):85–92. https://doi.org/10.1080/17474124.2020.1715211. 24. Ruan Y, Li Z, Shen Y, Li T, Zhang H, Guo J. Functions of circular RNAs and their potential applications in gastric cancer. Expert Rev Gastroenterol Hepatol. 2020;14(2):85–92. https://doi.org/10.1080/17474124.2020.1715211. 25. Buxbaum AR, Haimovich G, Singer RH. In the right place at the right time: visualizing and understanding mRNA localization. Nat Rev Mol Cell Biol. 2015;16(2):95–109. https://doi.org/10.1038/nrm3918. 25. Buxbaum AR, Haimovich G, Singer RH. In the right place at the right time: visualizing and understanding mRNA localization. Nat Rev Mol Cell Biol. 2015;16(2):95–109. https://doi.org/10.1038/nrm3918. 26. Komori T: Molecular Mechanism of Runx2-Dependent Bone Development. Mol Cells. 2020 Feb 29;43(2):168–175. doi: 10.14348/molcells.2019.0244. 26. Komori T: Molecular Mechanism of Runx2-Dependent Bone Development. Mol Cells. 2020 Feb 29;43(2):168–175. doi: 10.14348/molcells.2019.0244. 27. Zhao W, Yang H, Chai J, Xing L. RUNX2 as a promising therapeutic target for malignant tumors. Cancer Manag Res. 2021 Mar 16;13:2539–2548. doi: 10.2147/CMAR.S302173. eCollection 2021. 27. Zhao W, Yang H, Chai J, Xing L. RUNX2 as a promising therapeutic target for malignant tumors. Cancer Manag Res. 2021 Mar 16;13:2539–2548. doi: 10.2147/CMAR.S302173. eCollection 2021. Figures Page 14/19 Page 14/19 Page 15/19 Figure 1 Identification and validation of differential expression of circ_0001722 in OS tissues and correspondin adjacent nontumor tissues. a: CircRNA expression profiling between two groups is showed with volcan plot. The vertical lines refer to a 2.0-fold (log2 scaled) up-regulation and down-regulation, respectively. The horizontal line corresponds to a P value of 0.05 (−log10 scaled). The red points represent up regulated circRNAs with statistical significance, and green points represent down regulated ones. b: Figure 1 Identification and validation of differential expression of circ_0001722 in OS tissues and corresponding adjacent nontumor tissues. a: CircRNA expression profiling between two groups is showed with volcano plot. The vertical lines refer to a 2.0-fold (log2 scaled) up-regulation and down-regulation, respectively. The horizontal line corresponds to a P value of 0.05 (−log10 scaled). The red points represent up regulated circRNAs with statistical significance, and green points represent down regulated ones. b: Page 15/19 Page 15/19 Hierarchical clustering heatmap indicates differences in circRNA expression profiling between the two groups. c: 11 target miRNAs of circ_0001722 are predicted by miRanda and TargetScan, as well as their corresponding target cancer related mRNA, estimated by miRTarBase database. Green: circ-0001722; red: miRNAs; blue: mRNAs. d: Relative levels of circ_0001722 and GAPDH in OS cell line MG63 after their RNAs are treated with or without RNase R digestion. e: The validation of back-splicing junction sites of circ_0001722 by Sanger sequencing. f: The relative expression of circ_0001722 was detected in 20 OS tissues and corresponding adjacent nontumor tissues. (b) The relative expression of circ_0001722 was detected in two OS cell lines and an immortalized human fetal osteoblastic cell line. ØØP<0.01. Figure 2 Figure 2 Downregulation of circ_0001722 suppresses cell proliferation of human OS cells. a: circ_0001722 is down-regulated by exogenous shRNA (sh-circ_0001722) in two human OS cell lines. b: Down-regulation of circ_0001722 can suppress cell proliferation of both human OS cell lines in vitro. c: Downregulation of circ_0001722 can suppress the invasion ability of both human OS cell lines. d: Downregulation of circ_0001722 can suppress cell proliferation of MG63 cells in vivo. ØP<0.05. ØØP<0.01. Figure 2 Figure 2 Downregulation of circ_0001722 suppresses cell proliferation of human OS cells. a: circ_0001722 is down-regulated by exogenous shRNA (sh-circ_0001722) in two human OS cell lines. b: Down-regulation of circ_0001722 can suppress cell proliferation of both human OS cell lines in vitro. c: Downregulation of circ_0001722 can suppress the invasion ability of both human OS cell lines. d: Downregulation of circ_0001722 can suppress cell proliferation of MG63 cells in vivo. ØP<0.05. ØØP<0.01. Page 16/19 Page 16/19 Figure 3 circ_0001722 binds to miR-204-5p to suppress its expression in human OS cells. a: Computational algorithms predicts complementary sequences of the circ_0001722 and miR-204-5p binding sequence. b: Compared with miR-NC, miR-204-5p mimic can significantly reduce the luciferase activity of pmirGLO-Wt- circ clone in HEK293T cells. However, neither miR-NC nor miR-204-5p mimic can affect the luciferase activity of pmirGLO-Mt-circ clone. c: Ago2 RIP assay shows that Ago2 can significantly enrich circ_0001722 and miR-204-5p. d: Downregulation of circ_0001722 can promote miR-204-5p expression in both human OS cell lines. e: The expression of miR-204-5p was negatively associated with circ_0001722 in human OS tissues. NC: negative control. Wt: wild type. Mt: mutant type. ØØP<0.01. Figure 5 sh-circ_0001722 suppresses proliferation and invasion of human OS cells through miR-204-5p/RUNX2 axis. a: Computational algorithms predicts that miR-204-5p targets RUNX2 mRNA 3’UTR. b: Comparing with miR-NC, miR-204-5p can significantly reduce the luciferase activity of pmirGLO-Wt-3’UTR clone in HEK293T cells. However, neither miR-NC nor miR-204-5p can affect the luciferase activity of pmirGLO-Mt- 3’UTR clone. c: Both miR-204-5p and sh-circ_0001722 can suppress RUNX2 expression in human OS cells. d: Suppression affect by downregulating circ_0001722 on cell proliferation can be rescued by exogenous RUNX2. e: Suppression affect by downregulating circ_0001722 on cell invasion can be rescued by exogenous RUNX2. NC: negative control. Wt: wild type. Mt: mutant type. ØP<0.05. ØØP<0.01. Figure 3 circ_0001722 binds to miR-204-5p to suppress its expression in human OS cells. a: Computational algorithms predicts complementary sequences of the circ_0001722 and miR-204-5p binding sequence. b: Compared with miR-NC, miR-204-5p mimic can significantly reduce the luciferase activity of pmirGLO-Wt- circ clone in HEK293T cells. However, neither miR-NC nor miR-204-5p mimic can affect the luciferase activity of pmirGLO-Mt-circ clone. c: Ago2 RIP assay shows that Ago2 can significantly enrich circ_0001722 and miR-204-5p. d: Downregulation of circ_0001722 can promote miR-204-5p expression in both human OS cell lines. e: The expression of miR-204-5p was negatively associated with circ_0001722 in human OS tissues. NC: negative control. Wt: wild type. Mt: mutant type. ØØP<0.01. Page 17/19 Page 17/19 Page 17/19 Figure 4 Sh-circ_0001722 suppresses proliferation and invasion of human OS cells by upregulating miR-204-5p. a: Both miR-204-5p and sh-circ_0001722 can upregulate miR-204-5p expression level in two human OS cell lines. b: Both miR-204-5p and sh-circ_0001722 can suppress cell proliferation of human OS cell lines in vitro. c: Both miR-204-5p and sh-circ_0001722 can suppress the invasion ability of human OS cell lines in vitro. NC: negative control. ØP<0.05. ØØP<0.01. Sh-circ_0001722 suppresses proliferation and invasion of human OS cells by upregulating miR-204-5p. a: Both miR-204-5p and sh-circ_0001722 can upregulate miR-204-5p expression level in two human OS cell lines. b: Both miR-204-5p and sh-circ_0001722 can suppress cell proliferation of human OS cell lines in vitro. c: Both miR-204-5p and sh-circ_0001722 can suppress the invasion ability of human OS cell lines in vitro. NC: negative control. ØP<0.05. ØØP<0.01. Figure 5 sh-circ_0001722 suppresses proliferation and invasion of human OS cells through miR-204-5p/RUNX2 axis. a: Computational algorithms predicts that miR-204-5p targets RUNX2 mRNA 3’UTR. b: Comparing with miR-NC, miR-204-5p can significantly reduce the luciferase activity of pmirGLO-Wt-3’UTR clone in HEK293T cells. However, neither miR-NC nor miR-204-5p can affect the luciferase activity of pmirGLO-Mt- 3’UTR clone. c: Both miR-204-5p and sh-circ_0001722 can suppress RUNX2 expression in human OS cells. d: Suppression affect by downregulating circ_0001722 on cell proliferation can be rescued by exogenous RUNX2. e: Suppression affect by downregulating circ_0001722 on cell invasion can be rescued by exogenous RUNX2. NC: negative control. Wt: wild type. Mt: mutant type. ØP<0.05. ØØP<0.01. SupplementalTable2.docx Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. sfig1.tif sfig2.tif SupplementalTable1.docx Page 18/19 SupplementalTable2.docx SupplementalTable2.docx Page 19/19
https://openalex.org/W2899275110	https://europepmc.org/articles/pmc6290729?pdf=render	English	null	Postoperative management of patients undergoing cardiac surgery in Austria	Wiener klinische Wochenschrift	2,018	cc-by	4,643	original article original article Wien Klin Wochenschr (2018) 130:716–721 https://doi.org/10.1007/s00508-018-1403-3 A national survey on current clinical practice in hemodynamic monitoring and postoperative management A national survey on current clinical practice in hemodynamic monitoring and postoperative management Johannes Menger · Maximilian Edlinger-Stanger · Martin Dworschak · Barbara Steinlechner Received: 13 August 2018 / Accepted: 12 October 2018 / Published online: 29 October 2018 © The Author(s) 2018 Conclusions This study provides insights into the cur- rent state of postoperative management of cardiac surgical patients in Austria. Standard monitoring as proposed by international guidelines is well estab- lished in Austrian intensive care units. Echocardio- graphy is widely seen as a very important tool in the postoperative care of cardiac surgical patients. Knowl- edge about the status quo of postoperative intensive care management of cardiac surgical patients enables further development of patient care. Summary Background No data are currently available regarding the current clinical practice in postoperative care of cardiac surgical patients in Austria. Objective The study investigated the current intensive care management concerning hemodynamic moni- toring and strategies to treat common perioperative disorders of patients after cardiac surgery in Austria. Objective The study investigated the current intensive care management concerning hemodynamic moni- toring and strategies to treat common perioperative disorders of patients after cardiac surgery in Austria. Methods A survey consisting of 31 questions was sent to intensivists at all 9 hospitals offering cardiac surgery in Austria. Methods A survey consisting of 31 questions was sent to intensivists at all 9 hospitals offering cardiac surgery in Austria. Keywords Survey · Cardiac surgery · Hemodynamic monitoring · Inotropic drugs · Volume therapy Results The response rate was 100%. The mean num- ber of procedures on cardiopulmonary bypass per centre was 722± 223. In the majority of cases postop- erative critical care is performed by anesthesiologists. Blood gas analysis, pulse oximetry, electrocardiogram, temperature, central venous pressure, arterial pres- sure and hourly urine output are de facto standard monitoring in all centers. Transesophageal echocar- diography is available in all centers and is frequently used. Crystalloids are the ﬁrst choice for volume replacement, whereas levosimendan and adrenaline are employed for the treatment of low cardiac output syndrome. Introduction There are nine centers in Austria serving patients requiring cardiac surgery. Postoperative care of the great majority of these patients is provided on in- tensive care units (ICU). Perioperative disorders in patients undergoing cardiac surgery often evolve very dynamically and need immediate response. A prereq- uisite is, however, that these disorders are promptly detected to be able to respond. Adequate monitoring of hemodynamic parameters and suitable treatment strategies are considered to be key factors in high quality care of patients after cardiac surgery. Com- mon treatment strategies comprise the differentiated use of inotropes, vasopressors and adequate ﬂuid management [1]. The technological progress during the last decades gave the critical care physicians mul- tiple different methods and devices for hemodynamic monitoring to choose from. The choice also includes more sophisticated hemodynamic monitoring de- vices measuring multiple hemodynamic parameters; this includes pulmonary artery catheter (PAC), trans- Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00508-018-1403-3) contains supplementary material, which is available to authorized users. J. Menger () · M. Edlinger-Stanger · M. Dworschak · B. Steinlechner Division of Cardiac Thoracic Vascular Anaesthesia and Intensive Care Medicine, Department of Anaesthesia, Intensive Care Medicine and Pain Medicine, Medical University of Vienna, Spitalgasse 23, 1090 Vienna, Austria johannes.menger@meduniwien.ac.at K 716 Postoperative management of patients undergoing cardiac surgery in Austria original article esophageal echocardiography (TEE) and newer less invasive devices. Innsbruck, Klagenfurt, Linz, Salzburg, Sankt Pölten, Wels and two centers in Vienna (Krankenhaus Hietz- ing and Allgemeines Krankenhaus). The participants were asked to complete the questionnaire regarding the practice at their centre during the year 2016 on the primary ICU responsible for cardiac surgical patients. International guidelines provide relatively non- speciﬁc recommendations concerning hemodynamic monitoring and management of patients after cardiac surgery, particularly with respect to hemodynamic and clinical endpoints [1, 2]. Therefore, based on these published guidelines, highly divergent strate- gies may be employed in the management of the postoperative cardiac surgical patient. Indeed, Italian and German studies reported considerable variations in the postoperative care after cardiac surgery among different institutions [3–5]. Moreover, there is an on- going scientiﬁc debate about the appropriate type of ﬂuids for volume replacement in the critically ill [6]. The template of the questionnaire has already been used in national surveys in Germany in 2005 and 2011 [3, 4] and slightly altered in Italy in 2013 [5]. Results All nine centres (100%) returned the questionnaire by September 2017. Data from one center were in- complete regarding the volume of performed cases. With the exception of this single questionnaire there were no missing data. All institutions were public hospitals and 67% were university hospitals. Struc- tural data of the participating centres are shown in Table 1. Isolated coronary artery bypass graft (CABG) was the most common procedure performed account- ing for 31% (2346/7485) of all reported procedures among 8 reporting centers, followed by combined pro- cedures, such as multiple valve surgery/valve surgery plus CABG (19%), aortic valve surgery (19%) and mi- tral valve surgery (8%). An average of 722± 223 pro- cedures were performed on cardiopulmonary bypass in adults at each of the 8 reporting centers in 2016. Introduction The ques- tionnaire consisted of 31 questions covering different aspects of perioperative care in adult patients under- going cardiac surgery. Main aspects of this question- naire are on hemodynamic monitoring and strategies in the use of drugs and ﬂuids in common disorders in patients after cardiac surgery. The questionnaire was in German and can be found in the Electronic Supplementary Material. For some questions multi- ple selections could be made. Prompted by the German report in 2005 [4], the German Society of Anesthesiology and Intensive Care Medicine and the German Society of Thoracic and Cardiovascular Surgery published clinical practice guidelines for the postoperative care of the adult car- diac surgical patient [7]. These guidelines have been updated in 2010 [8] and 2018 [1]. So far, no data have been published of the clinical management of cardiac surgery patients in Austrian intensive care units. Final analysis was done after all questionnaires were returned. Only descriptive statistics were used for the analysis of the collected data. Data are pre- sented as a percentage of all Austrian centers offering cardiac surgery. Missing data are also reported for every item whenever this occurred. The aim of this study was to collect information on the current clinical practice regarding postoperative critical care of these patients in Austria. In partic- ular, this survey intended to address hemodynamic monitoring and management of common postopera- tive problems. Postoperative management of patients undergoing cardiac surgery in Austria 717 ECG ABP CVP LAP SaO2 ScO2 SpO2 eCO2 EEG NIRS HUP Temp Monitoring technique % 0 20 40 60 80 100 Always Sometimes Case by case Never Fig. 1 Routine use of postoperative monitoring. ECG continuous electrocar- diogram, ABP arterial blood pressure, CVP central ve- nous pressure, LAP left atrial pressure, SaO2 ar- terial oxygen saturation, ScO2 central venous oxy- gen saturation, SpO2 pe- ripheral oxygen saturation, EEG processed electroen- cephalogram, NIRS near- infrared spectroscopy, HUP hourly urine portions, Temp body temperature Always Sometimes Case by case Never Always Sometimes Case by case Never left ventricular ejection fraction (i.e. <32± 3%) that necessitate the use of PAC. Availability and usage of advanced hemodynamic monitoring techniques are given in Fig. 2. left ventricular ejection fraction (i.e. <32± 3%) that necessitate the use of PAC. Availability and usage of advanced hemodynamic monitoring techniques are given in Fig. 2. A heart transplant program was active in 3 centers in 2016 (Graz, Innsbruck and Vienna—Allgemeines Krankenhaus). Of the 8 reporting cardiac surgery cen- ters 2 routinely cared for pediatric patients in 2016 (Linz and Vienna—Allgemeines Krankenhaus). All centers employ transesophageal echocardiog- raphy (TEE) and have 24/7 in-house availability of a physician experienced with this technique. Among the three most important reasons for the postopera- tive use of TEE were: hemodynamic instability (100%), suspected cardiac tamponade (100%), and evaluation of valve function (78%). Blood gas analysis, pulse oximetry, electrocardio- gram, patient temperature, central venous and arterial pressure and hourly urine output were monitored in all nine centres. The use of postoperative monitoring is shown in Fig. 1. Pulmonary artery catheters (PAC) are available in all centers. Of the centers three reported frequent use of PAC, three hospitals occasional and another three centers isolated use. Of the centers six preferred continuous and three centers bolus thermodilution cardiac output measurements with PAC. Among the three most important reasons for the usage of PAC were monitoring of pulmonary hypertension (89%), measurement of cardiac output (78%) and moni- toring of hemodynamic instability (67%). Of the centers four reported having a speciﬁc threshold for the treatment of systolic pulmonary artery pressure (i.e. >50± 10mmHg) and three centers a threshold for As ﬁrst choice for volume therapy all centers use balanced crystalloid solutions (56% Ringer’s acetate solution, 44% Ringer’s lactate solution). As second choice, 6 centers reported the use of 4% succinylated gelatine, 2 centers use human serum albumin so- lutions and 1 center uses blood products. Gelatine solutions are never used by 3 centers and 6 centres never use hydroxyethyl starch solutions. Methods This study was a national survey using a question- naire. It was moderated by the Austrian Society of Anesthesiology, Resuscitation and Intensive Care (ÖGARI) working group for Cardiothoracic and Vas- cular Anesthesia (ARGE Herz/Thorax/Gefäß-Anästhe- sie). Due to the design of the study, approval from the local ethics committee was waived. The survey was e-mailed in March 2017 to member anesthesiologists at all nine hospitals providing cardiac surgery in Aus- tria. In 2016 cardiac surgery was performed in Graz, Table 1 Structural data of centers offering cardiac surgery to patients in Austria in 2016 Number of centers (N) Number of cardiac surgical procedures for cardiopulmonary bypass per year (centers) <500 2 500–750 3 750–1000 2 (3)a >1000 1 Specialists mainly effectuating postoperative critical care (centers) Anesthesiologists 8 (89%) Anesthesiologists and surgeons 1 (11%) Department that manages postoperative intensive care unit (centers) Anesthesiology 8 (89%) Surgery 1 (11%) Intensive care unit dedicated specially to cardiac surgery patients (centers) 3 (33%) aOne answer is missing. Missing center has older publicly available data ranging from 750 to 1000 procedures on cardiopulmonary bypass per year in the decade 2000–2009 [25, 26] K Postoperative management of patients undergoing cardiac surgery in Austria 717 Table 1 Structural data of centers offering cardiac surgery to patients in Austria in 2016 Postoperative management of patients undergoing cardiac surgery in Austria 717 K K original article 718 Postoperative management of patients undergoing cardiac surgery in Austria Fig. 3 Availability of me- chanical cardiac assist de- vices. IABP intra-aortic balloon pump, LVAD left ventricular assist device, RVAD right ventricular as- sist device, BiVAD biven- tricular assist devices, ECMO veno-arterial and veno- venous extracorporeal mem- brane oxygenation IABP LVAD RVAD BiVAD ECMO Technique % 0 20 40 60 80 100 The 3 most reported vasodilators were nitroglycerin (100%), ura- pidil (100%) and clonidine (56%). All centers use in- halative vasodilators to treat pulmonary hypertension and six centres have the possibility to use inhala- tive nitric oxide. The 3 most common drugs to treat pulmonary hypertension were inhalative prostacyclin (78%), inhalative nitric oxide (67%) and intravenous levosimendan (44%). The availability of mechanical cardiac assist devices is shown in Fig. 3. Protocols for the use of vasopressor were established in four centes and for transfusions of blood products in ﬁve centres. Interestingly, no center offers preoperative autologous blood donation. used drugs for the treatment of postoperative low cardiac output syndrome were levosimendan (89%), adrenaline (56%) and dobutamine (56%). For the treatment of systemic inﬂammatory response syn- drome noradrenaline (100%), vasopressin (100%) and hydrocortisone (78%) were used. The 3 most reported vasodilators were nitroglycerin (100%), ura- pidil (100%) and clonidine (56%). All centers use in- halative vasodilators to treat pulmonary hypertension and six centres have the possibility to use inhala- tive nitric oxide. The 3 most common drugs to treat pulmonary hypertension were inhalative prostacyclin (78%), inhalative nitric oxide (67%) and intravenous levosimendan (44%). The availability of mechanical cardiac assist devices is shown in Fig. 3. Protocols for the use of vasopressor were established in four centes and for transfusions of blood products in ﬁve centres. Interestingly, no center offers preoperative autologous blood donation. Fig. 3 Availability of me- chanical cardiac assist de- vices. IABP intra-aortic balloon pump, LVAD left ventricular assist device, RVAD right ventricular as- sist device, BiVAD biven- tricular assist devices, ECMO veno-arterial and veno- venous extracorporeal mem- brane oxygenation IABP LVAD RVAD BiVAD ECMO Technique % 0 20 40 60 80 100 Fig. 3 Availability of me- chanical cardiac assist de- vices. IABP intra-aortic balloon pump, LVAD left ventricular assist device, RVAD right ventricular as- sist device, BiVAD biven- tricular assist devices, ECMO veno-arterial and veno- venous extracorporeal mem- brane oxygenation Fig. 3 Availability of me- chanical cardiac assist de- vices. IABP intra-aortic balloon pump, LVAD left ventricular assist device, RVAD right ventricular as- sist device, BiVAD biven- tricular assist devices, ECMO veno-arterial and veno- venous extracorporeal mem- brane oxygenation According to current German S3 guidelines [1] electrocardiogram (ECG), pulse oximetry, continu- ous invasive arterial and central venous blood pres- sure, assessment of the ﬂuid balance and blood gas analysis of arterial and central venous blood are considered standard monitoring for patients after cardiac surgery. Likewise, all nine Austrian centers implemented respective postoperative standards in monitoring. Merely two centers reported not mea- suring central venous oxygen saturation routinely despite a recommendation in the current German guidelines. The use of neurological monitoring con- tinued to the postoperative period is heterogeneous in Austria. Guidelines recommend the intraoperative use only in certain subgroups, such as heart trans- plantation, whereas several studies report favorable outcome in general cardiac surgery [9, 10]. Whatever the indication for neuromonitoring during surgery is, continuing intraoperatively instituted and frequently expensive neuromonitoring postoperatively seems to make sense as it may detect potentially hazardous situations in critical patients that can not yet be neu- rologically evaluated [11]. Although PAC is available in all centers it is employed heterogeneously. Despite the fact that PAC was the only advanced monitoring technique that was associated with reduced mortality in goal-directed therapy [12], its use has been criti- cized for being related to various risks without actually improving outcome in coronary artery bypass surgery [13]. The current German S3 guidelines [1] name pre- operative right heart dysfunction and patients at risk for low cardiac output syndrome and/or pulmonary hypertension as indications for PAC-guided hemody- namic management. This corresponds to the reported indications in this survey. Different frequency in its usage in this study may partly be explained by differ- ent risk proﬁles of patients in the nine centres. i f h hi h f l i used drugs for the treatment of postoperative low cardiac output syndrome were levosimendan (89%), adrenaline (56%) and dobutamine (56%). For the treatment of systemic inﬂammatory response syn- drome noradrenaline (100%), vasopressin (100%) and hydrocortisone (78%) were used. ECG ABP CVP LAP SaO2 ScO2 SpO2 eCO2 EEG NIRS HUP Temp Monitoring technique % 0 20 40 60 80 100 Always Sometimes Case by case Never The three most important targets for hemodynamic stabiliza- tion were optimization of central venous pressure (67%), arterial blood pressure (56%) and echocar- diographic cardiac ﬁlling (33%). The 3 most often Fig. 2 Availability of ad- vanced hemodynamic mon- itoring and usage. TTE transthoracic echocardio- graphy, TEE transesopha- geal echocardiography, PAC pulmonary artery catheter, TPTD transpulmonary ther- modilution, LD lithium di- lution cardiac output mea- surement, UPWA uncali- brated pulse wave analysis TTE TEE PAC TPTD LD UPWA Monitoring technique % 0 20 40 60 80 100 Frequently Sometimes Case by case Never 718 Postoperative management of patients undergoing cardiac surgery in Austria K Fig. 2 Availability of ad- vanced hemodynamic mon- itoring and usage. TTE transthoracic echocardio- graphy, TEE transesopha- geal echocardiography, PAC pulmonary artery catheter, TPTD transpulmonary ther- modilution, LD lithium di- lution cardiac output mea- surement, UPWA uncali- brated pulse wave analysis TTE TEE PAC TPTD LD UPWA Monitoring technique % 0 20 40 60 80 100 Frequently Sometimes Case by case Never 718 Postoperative management of patients undergoing cardiac surgery in Austria K 718 Postoperative management of patients undergoing cardiac surgery in Austria original article Postoperative management of patients undergoing cardiac surgery in Austria 719 Discussion Data from this survey that included all Austrian car- diac surgery centers provide insights into the current state of the postoperative management of these pa- tients. The survey covered the extent of monitoring on the ICU as well as treatment strategies that are com- pletely in line with current recommendations [3–5]. In contrast to Germany and Italy all Austrian cen- ters offering cardiac surgery are public hospitals [4, 5]. The average annual case volume per center is slightly higher than that reported for Italy (722 vs. 617 cardiac procedures) but smaller than in Germany (1460 car- diac procedures) [4, 5]. As in Italy, postoperative in- tensive care of patients following cardiac surgery in Austria is mainly allocated to anesthesiologists. In contrast, in 29% of all German centers postoperative intensive care is managed solely by cardiac surgeons [4, 5]. In spite of the high expenses for personnel, train- ing and technical support, all centers can offer trans- esophageal echocardiography by an experienced user K original article at all times. Kastrup et al. [4] showed that on Ger- man intensive care units caring for patients under- going cardiac surgery in 2005, only 65% had a 24h availability of an experienced user. This lack in man- power or insufﬁcient training may partly be explained by technical and medical progress as in Italy in 2013 [5] all centers had a 24h availability of an experienced user. This survey conﬁrms once more the importance of this technique in the care of patients following car- diac surgery [14]. this survey did not cover the full scope of postopera- tive intensive care but only hemodynamic monitoring and treatment of common postoperative disorders. p p In conclusion, this is the ﬁrst investigation provid- ing insights into the current state of postoperative care of cardiac surgical patients in Austria. As pro- posed by international guidelines standard monitor- ing is well established and routinely used on Austrian ICUs. Echocardiography is widely seen as a very im- portant tool in the care of cardiac surgical patients and is always available in every center. Use of PAC is heterogeneous with only 33% of centers reporting frequent use. Crystalloids are the ﬂuids of choice in all centers for volume replacement. If needed gela- tine solution is the most frequently used colloid. Low cardiac output syndrome is most often treated with levosimendan. Discussion Knowledge about the status quo of postoperative intensive care management of cardiac surgical patients allows for further development of pa- tient care. The deﬁnition of speciﬁc indications for implementation of certain treatment measures, clini- cal endpoints, and their regular evaluation would be the next steps to systematically improve care after car- diac surgery in Austria on a national base. Despite of the low absolute number of centers in Austria the results of this survey showed remarkable differences in postoperative clinical management of patients undergoing cardiac surgery. In Austria crys- talloids are the ﬁrst choice for volume expansion for all centers. Recent surveys on ﬂuid resuscitation in cardiac surgery in the USA [15] and Europe [6] showed a more diverse picture. For colloids the Austrian re- sults are in line with the European results naming gelatine as the preferred colloid in the majority of the centers [6]. Although controversial, [16] central venous pres- sure together with arterial blood pressure are in Aus- tria, as in Italy and Germany, the preferred parame- ters to guide ﬂuid therapy. Echocardiographic ﬁlling as additional parameter underlines the importance of echocardiography. In Austria too, dynamic indices derived from additional devices are only infrequently used in routine care to guide ﬂuid management de- spite evidence of feasibility and accuracy of these de- vices in this setting [17, 18]. Acknowledgements We thank the Cardiothoracic and Vas- cular Anesthesia working group (ARGE Herz-Thorax-Gefäß Anästhesie) of the Austrian Society of Anesthesia, Resusci- tation and Intensive Care Medicine (ÖGARI) for their kind support. We acknowledge all participating centers (in alpha- betical order): Landeskrankenhaus Graz (Medizinische Uni- versität Graz), Landeskrankenhaus Innsbruck (Medizinische Universität Innsbruck), Klinikum Klagenfurt, Kepler Univer- sitätsklinikum Linz (Johannes Kepler Universität Linz), Lan- deskrankenhaus Salzburg (Paracelsus Medizinische Privat- universität), Universitätsklinikum St. Pölten (Karl Land- steiner Privatuniversität für Gesundheitswissenschaften), Klinikum Wels-Grieskirchen, Krankenhaus Wien Hietzing, Allgemeines Krankenhaus der Stadt Wien (Medizinische Uni- versität Wien) The most important drug to treat low cardiac out- put syndrome was levosimendan. Shortly after this survey three major trials [19–21] put the expected ef- fect of levosimendan in cardiac surgery into perspec- tive, which prompted a more conservative expert con- sensus regarding its use in this context [22]. This may have changed the centers’ attitude towards use of this inotrope. There is high accordance regarding employ- ment of systemic vasoconstrictors and vasodilators among all centers. Discussion The beneﬁcial effect of aerosolized vasodilators and inhaled nitric oxide for the treatment of pulmonary hypertension has been proven [23]. Al- though inhalative nitric oxide needs expensive tech- nical equipment it is favored by all centers where it is available as compared to aerosolized vasodilators as potential alternatives. Funding Open access funding provided by Medical Univer- sity of Vienna. Funding Open access funding provided by Medical Univer- sity of Vienna. Conﬂict of interest J. Menger, M. Edlinger-Stanger, M. Dwor- schak, and B. Steinlechner declare that they have no compet- ing interests. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which per- mits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the origi- nal author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. A limitation of this study is that it is not known if the answers of the selected anesthesiologists truly reﬂect the actual state at their institution. In this respect, dif- ferences in care between different intensivists of the same institution must also be taken in consideration [24]; however, due to local protocols clinical manage- ment should be more or less homogeneous within one center. A further limitation of this study is that besides the primary ICU responsible for cardiac surgical pa- tients, some patients go on ICUs managed by other disciplines (e.g. internal medicine) at the same insti- tution that are not covered by this survey. Additionally, 720 Postoperative management of patients undergoing cardiac surgery in Austria References 1. Habicher M, Zajonz T, Heringlake M, Böning A, Treskatsch S,SchirmerU,etal. S3-Leitliniezurintensivmedizinischen Versorgung herzchirurgischer Patienten [S3 guidelines on intensive medical care of cardiac surgery patients: Hemody- namic monitoring and cardiovascular system - an update. Anaesthesist. 2018;67:375–9. namic monitoring and cardiovascular system - an update. Anaesthesist. 2018;67:375–9. K 720 Postoperative management of patients undergoing cardiac surgery in Austria K original article 13. Schwann NM, Hillel Z, Hoeft A, Barash P, Möhnle P, Miao Y, etal. Lackofeffectivenessofthepulmonaryarterycatheter incardiacsurgery. AnesthAnalg. 2011;113:994–1002. 2. Vincent J-L, Pelosi P, Pearse R, Payen D, Perel A, Hoeft A, et al. Perioperative cardiovascular monitoring of high-risk patients: aconsensusof12. CritCare. 2015;19:224. 14. Renner J, Bein B, Broch O. Kardioanästhesie: anästhesi- ologisches Management [Cardiac Anaesthesia: Anaesthe- siological Management]. Anasthesiol Intensivmed Not- fallmedSchmerzther. 2018;53:346–62. 3. Kastrup M, Carl M, Spies C, Sander M, Markewitz A, Schirmer U. Clinical impact of the publication of S3 guide- lines for intensive care in cardiac surgery patients in Ger- many: results from a postal survey. Acta Anaesthesiol Scand. 2013;57:206–13. 15. AronsonS, Nisbet P, Bunke M. Fluid resuscitation practices in cardiac surgery patients in the USA: a survey of health careproviders. PerioperMed(Lond). 2017;6:15. 4. Kastrup M, Markewitz A, Spies C, Carl M, Erb J, Große J, et al. Current practice of hemodynamic monitoring and vasopressor and inotropic therapy in post-operative cardiac surgery patients in Germany: results from a postal survey. ActaAnaesthesiolScand. 2007;51:347–58. 16. Marik PE, Cavallazzi R. Does the central venous pressure predict ﬂuid responsiveness? An updated meta-analysis and a plea for some common sense. Crit Care Med. 2013;41:1774–81. 5. Bignami E, Belletti A, Moliterni P, Frati E, Guarnieri M, Tritapepe L. Clinical practice in perioperative monitoring in adult cardiac surgery: is there a standard of care? Results from an national survey. J Clin Monit Comput. 2016;30:347–65. 17. Mora B, Ince I, Birkenberg B, Skhirtladze K, Pernicka E, Ankersmit HJ, et al. Validation of cardiac output measure- ment with the LiDCOTM pulse contour system in patients withimpairedleftventricularfunctionaftercardiacsurgery. Anaesthesia. 2011;66:675–81. 6. Protsyk V, Rasmussen BS, Guarracino F, Erb J, Turton E, Ender J. Fluid management in cardiac surgery: results of a survey in European cardiac anesthesia departments. JCardiothoracVascAnesth. 2017;31:1624–9. 18. Menger J, Mora B, Skhirtladze K, Fischer A, Ankersmit JH, DworschakM.Accuracyofcontinuouscardiacoutputmea- surement with the LiDCOplus system during intra-aortic counterpulsationaftercardiacsurgery. JCardiothoracVasc Anesth. 2016;30:592–8. 7. Carl M, Alms A, Braun J, Dongas A, Erb J, Goetz A, et al. DieintensivmedizinischeVersorgungherzchirurgischer Patienten: HämodynamischesMonitoringundHerz-Kreis- lauf-Therapie. References S3-Leitlinie der Deutschen Gesellschaft für Thorax-, Herz- und Gefäßchirurgie (DGTHG) und der Deutschen Gesellschaft für Anästhesiologie und Inten- sivmedizin (DGAI) [Guidelines for intensivecarein cardiac surgery patients: haemodynamic monitoring and cardio circulatory treatment guidelines of the German Society for Thoracic and Cardiovascular Surgery and the German Society of Anaesthesiology and Intensive Care Medicine]. ThoracCardiovascSurg. 2007;55:130–48. 19. Mehta RH, Leimberger JD, van Diepen S, Meza J, Wang A, Jankowich R, et al. Levosimendan in patients with left ventriculardysfunctionundergoingcardiacsurgery.NEngl JMed. 2017;376:2032–42. 20. LandoniG,LomivorotovVV,AlvaroG,LobreglioR,PisanoA, Guarracino F, et al. Levosimendan for hemodynamic sup- portaftercardiacsurgery. NEnglJMed. 2017;376:2021–31. 21. Cholley B, Caruba T, Grosjean S, Amour J, Ouattara A, Villa- corta J, etal. Effectof Levosimendan on lowcardiac output syndrome in patients with low ejection fraction undergo- ing coronary artery bypass grafting with cardiopulmonary bypass: the LICORN randomized clinical trial. JAMA. 2017;318:548–56. 8. Carl M, Alms A, Braun J, Dongas A, Erb J, Goetz A, et al. S3-Leitlinie zur intensivmedizinischen Versorgung herzchirurgischer Patienten [S3 guidelines on intensive medical care of cardiac surgery patients]. Z Herz- Thorax- Gefäßchir. 2010;24:294–310. 22. Guarracino F, Heringlake M, Cholley B, Bettex D, Bouchez S, Lomivorotov VV, et al. Use of Levosimendan in cardiac surgery: an update after the LEVO-CTS, CHEETAH, and LICORN trials in the light of clinical practice. J Cardiovasc Pharmacol. 2018;71:1–9. 9. Zheng F, Sheinberg R, Yee M-S, Ono M, Zheng Y, Hogue CW. Cerebral near-infrared spectroscopy monitoring and neurologic outcomes in adult cardiac surgery patients: asystematicreview. AnesthAnalg. 2013;116:663–76. 23. Elmi-Sarabi M, Deschamps A, Delisle S, Ased H, Haddad F, LamarcheY,etal. Aerosolizedvasodilatorsforthetreatment of pulmonary hypertension in cardiac surgical patients: a systematic review and meta-analysis. Anesth Analg. 2017;125:393–402. y g 10. Palmbergen WAC, van Sonderen A, Keyhan-FalsaﬁAM, Keunen RWM, Wolterbeek R. Improved perioperative neu- rological monitoring of coronary artery bypass graft pa- tients reduces the incidence of postoperative delirium: the Haga Brain Care Strategy. Interact Cardiovasc Thorac Surg. 2012;15:671–7. 24. Glance LG, Kellermann AL, Hannan EL, Fleisher LA, Eaton MP, Dutton RP, et al. The impact of anesthesiologists on coronary artery bypass graft surgery outcomes. Anesth Analg. 2015;120:526–33. 11. Sanchez-de-Toledo J, Chrysostomou C, Munoz R, Lichten- stein S, Sao-Avilés CA, Wearden PD, et al. Cerebral regional oxygen saturation and serum neuromarkers for the pre- diction of adverse neurologic outcome in pediatric cardiac surgery. NeurocritCare. 2014;21:133–9. 25. Klinische Abteilung für Herzchirurgie – LKH Universitäts- klinikum Graz. Jahresbericht 2006 [Internet]. 2007 [cited 2018 Oct 24]. p. 29. Available from: http://chirurgie. uniklinikumgraz.at/Documents/2006%20JH.pdf 12. Postoperative management of patients undergoing cardiac surgery in Austria 721 References Hamilton MA, Cecconi M, Rhodes A. A systematic review and meta-analysis on the use of preemptive hemody- namic intervention to improve postoperative outcomes in moderate and high-risk surgical patients. Anesth Analg. 2011;112:1392–402. g 26. Klinische Abteilung für Herzchirurgie – LKH Universitäts- klinikum Graz. Jahresbericht 2009 [Internet]. 2010 [cited 2018 Oct 24]. p. 18. Available from: http:// chirurgie.uniklinikumgraz.at/herzchirurgie/Abteilung/ Tätigkeitsbericht%20Herz%202009.pdf K K
https://openalex.org/W1556721782	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0106505&type=printable	English	null	Sliding-Scale versus Basal-Bolus Insulin in the Management of Severe or Acute Hyperglycemia in Type 2 Diabetes Patients: A Retrospective Study	PloS one	2,014	cc-by	5,721	Abstract Sliding-scale and basal-bolus insulin regimens are two options available for the treatment of severe or acute hyperglycemia in type 2 diabetes mellitus patients. Although its use is not recommended, sliding-scale insulin therapy is still being used widely. The aims of the study were to compare the glycemic control achieved by using sliding-scale or basal-bolus regimens for the management of severe or acute hyperglycemia in patients with type 2 diabetes and to analyze factors associated with the types of insulin therapy used in the management of severe or acute hyperglycemia. This retrospective study was conducted using the medical records of patients with acute or severe hyperglycemia admitted to a hospital in Malaysia from January 2008 to December 2012. A total of 202 patients and 247 admissions were included. Patients treated with the basal- bolus insulin regimen attained lower fasting blood glucose (10.862.3 versus 11.663.5 mmol/L; p = 0.028) and mean glucose levels throughout severe/acute hyperglycemia (12.361.9 versus 12.862.2; p = 0.021) compared with sliding-scale insulin regimens. Diabetic ketoacidosis (p = 0.043), cardiovascular diseases (p = 0.005), acute exacerbation of bronchial asthma (p = 0.010), and the use of corticosteroids (p = 0.037) and loop diuretics (p = 0.016) were significantly associated with the type of insulin regimen used. In conclusion, type 2 diabetes patients with severe and acute hyperglycemia achieved better glycemic control with the basal-bolus regimen than with sliding-scale insulin, and factors associated with the insulin regimen used could be identified. Citation: Zaman Huri H, Permalu V, Vethakkan SR (2014) Sliding-Scale versus Basal-Bolus Insulin in the Management of Severe or Acute Hyperglycemia in Type 2 Diabetes Patients: A Retrospective Study. PLoS ONE 9(9): e106505. doi:10.1371/journal.pone.0106505 Editor: Massimo Pietropaolo, University of Michigan Medical School, United States of America Editor: Massimo Pietropaolo, University of Michigan Medical School, United States of America Received December 7, 2013; Accepted August 7, 2014; Published September 2, 2014 Received December 7, 2013; Accepted August 7, 2014; Published September 2, 2014 Copyright: 2014 Zaman Huri et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: 2014 Zaman Huri et al. This is an open-access article distributed under the terms of the Creative Commons Att unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Abstract Funding: The authors would like to thank the Ministry of Science, Technology and Innovation, Malaysia (Science fund: 12-02-03-2097) and University of Malaya, Malaysia (RG428/12HTM, RP024/14HTM, RP024A/14HTM, RP024B/14HTM and RP024C/14HTM) for financial and technical support. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * Email: hasnizazh@um.edu.my Sliding-Scale versus Basal-Bolus Insulin in the Management of Severe or Acute Hyperglycemia in Type 2 Diabetes Patients: A Retrospective Study Hasniza Zaman Huri1,2, Vishaaliny Permalu1, Shireene Ratna Vethakkan3 1 Department of Pharmacy, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia, 2 Clinical Investigation Centre, University Malaya Medical Centre, Lembah Pantai, Kuala Lumpur, Malaysia, 3 Endocrinology Unit, Department of Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia Hasniza Zaman Huri1,2, Vishaaliny Permalu1, Shireene Ratna Vethakkan3 1 Department of Pharmacy, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia, 2 Clinical Investigation Centre, University Malaya Medical Centre, Lembah Pantai, Kuala Lumpur, Malaysia, 3 Endocrinology Unit, Department of Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia 1 Department of Pharmacy, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia, 2 Clinical Investigation Centre, Unive Pantai, Kuala Lumpur, Malaysia, 3 Endocrinology Unit, Department of Medicine, Faculty of Medicine, University of Malaya, Kuala Lu Introduction T2DM patients with severe or acute hyperglycemia based on the type of insulin regimen used. Therefore, this study was conducted to identify the treatment approach and the achievement of glycemic control among hospitalized T2DM patients with severe or acute hyperglycemia. The specific objectives were twofold: (i) to compare the glycemic control achieved by using sliding-scale (Actrapid or basal-bolus (Actrapid and Insulatard) regimens for the management of severe or acute hyperglycemia in T2DM patients; and (ii) to analyze factors associated with the types of insulin therapy used in the management of severe or acute hyperglycemia. Diabetes mellitus is a significant global health disorder. Type 2 diabetes mellitus (T2DM) is becoming more common in almost every population, accounting for approximately 90% of all cases of diabetes in adults in Malaysia in 2008 [1]. Severe or acute hyperglycemia is an acute manifestation of diabetes that commonly occurs in T2DM patients, and requires intensive treatment and hospitalization [2]. According to a prospective cohort study, the causes of admission to hospital in T2DM patients with hyperglycemia include diabetic ketoacidosis (DKA), hyper- osmolar hyperglycemic state and serious infections [3]. In addition, the concurrent use of blood-glucose altering medications such as corticosteroids, antipsychotics and diuretics tend to worsen severe or acute hyperglycemia by increasing hepatic gluconeo- genesis as well as impairing peripheral glucose uptake [2]. Exclusion Criteria 1) Patients with other types of diabetes mellitus 2) Patient with incomplete data Sliding-Scale versus Basal-Bolus Insulin Therapy 3) Admitted to general medical units T2DM according to International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) codes E11.0–E11.9 were identified via the Hospital Information System. Of these 1167 patients, medical records for 602 patients were successfully traced. Using the convenient sampling method, 202 patients who fulfilled the inclusion criteria (see below) were included in this study. An overview of the study methodology is shown in Figure 1. 4) Treated only with insulin during the hospitalization Study Population This retrospective study consisted of T2DM patients with severe or acute hyperglycemia admitted to the University of Malaya Medical Centre (UMMC), a principal 1000-bed teaching hospital in Kuala Lumpur, Malaysia, from January 2008 to December 2012. The study was conducted in accordance with the Declaration of Helsinki and was approved by the medical ethics committee of the UMMC (reference number 956.32). The committee waived the need for written informed consent from participants. The registration numbers of 1167 patients with Despite the available treatment options for severe or acute hyperglycemia in T2DM patients, glycemic control in this population remains suboptimal [4]. This is partly attributable to the continued use of sliding-scale insulin regimens to manage severe or acute hyperglycemia, despite many treatment guidelines [5], recommending against its use. In addition, there are limited local and global data on the level of glycemic control achieved in PLOS ONE \| www.plosone.org 1 September 2014 \| Volume 9 \| Issue 9 \| e106505 Sliding-Scale versus Basal-Bolus Insulin Therapy Statistical Techniques Collected data were pooled and analyzed using IBM SPSS Statistics Version 20.0 (Armonk, New York, USA). Kolmogorov- Smirnov was used to test for normality of continuous data. Normally distributed data was expressed as mean 6 standard deviation whereas data that was not normally distributed was expressed as median (interquartile range). Continuous data were expressed as mean 6standard deviation while categorical data were expressed as percentages. The association between categor- ical variables was examined using the Pearson Chi Square test with Continuity correction and Fisher’s exact adjustment when necessary. The t-test was used to evaluate the differences in means between groups of continuous data. Significance was set at p, 0.05. The minimum sample size was calculated using Epi InfoTM Program Version 7.0 (Centers for Disease Control and Prevention, Atlanta, USA). A minimum of 108 patients were needed to detect Demographic Characteristics A total of 202 T2DM patients with severe or acute hypergly- cemia on admission were included in this study of a total of 247 hospital admissions. There were slightly more female than male patients, and the most common ethnicity was Malay (42.6%), followed by Indians (38.6%), Chinese (17.3%) and others (1.5%). A total of 73.8% and 26.2% of the study population was non- elderly ($18 years) and elderly ($65 years), respectively. Data were available on BMI for 28.7% of patients, 12.9% of whom had BMI in the normal range, followed by pre-obese (7.9%), obese (5.9%) and underweight (2%) (see Table 1). Sliding-Scale versus Basal-Bolus Insulin Therapy Sliding-Scale versus Basal-Bolus Insulin Therapy Table 1. Demographic characteristics of the patients (N = 202). Demographic characteristics Number of patients (Percentage, %) Gender Male 96 (47.5) Female 106 (52.5) Age Non-elderly ($18 years) 149 (73.8) Elderly ($65 years) 53 (26.2) Ethnicity Malay 86 (42.6) Indian 78 (38.6) Chinese 35 (17.3) Others 3 (1.5) BMI (kg/m2)a Underweight (,18.5) 4 (2.0) Normal range (18.5–24.9) 26 (12.9) Pre-obese (25–29.9) 16 (7.9) Obese ($30) 12 (5.9) Unknown 144 (71.3) aBased on data available for 28.7% of patients. BMI = body mass index. doi:10.1371/journal.pone.0106505.t001 a minimum difference of 1 mmol/L, a power of b = 0.8 and a confidence level of 95%. a minimum difference of 1 mmol/L, a power of b = 0.8 and a confidence level of 95%. Glycemic Control achieved with Insulin Regimens Table 3 shows the level of glycemic control achieved with each regimen. Of the 338 cases, 159 were treated using basal-bolus insulin, and 179 cases were treated using sliding-scale insulin. Patients treated with the basal-bolus insulin regimen attained lower fasting blood glucose (10.862.3 versus 11.663.5 mmol/L; p = 0.028) and mean glucose levels (12.361.9 versus 12.862.2; p = 0.021) throughout severe or acute hyperglycemia compared with sliding-scale insulin regimens. Insulin Regimens Used during Severe or Acute Hyperglycemia Patients were monitored to evaluate glycemic control through- out the severe or acute hyperglycemia phase. Assessment of glycemic control was based on the glucose readings measured during treatment. Glycemic targets were defined according to American Diabetes Association (ADA) recommendations (Amer- ican Diabetes Association, 2013), i.e., fasting plasma glucose , 7.0 mmol/L; pre-meal plasma glucose and overall blood glucose ,10 mmol/L. Admissions were evaluated based on the insulin regimen used to manage the severe or acute hyperglycemia. A total of 338 cases were evaluated for insulin use. Sliding-scale insulin and basal-bolus insulin was used in 53% and 47% of admissions, respectively. Results Demographic characteristics Number of patients (Percentage, %) Clinical Characteristics Clinical characteristics of the patients are shown in Table 2. Of the 202 patients, more than 50% of the patients were hospitalized for #7 days, with a minimum stay of 2 days. The mean duration of the 247 admissions was 7.966.3 days. Blood glucose levels on admission were normally distributed with a mean of 24.469.3 mmol/L. Almost half of the patients (48.5%) were admitted to hospital with a blood glucose level .22.3 mmol/ L, with a maximum of 65.3 mmol/L. Overall, mean hemoglobin (Hb) A1c was 11.7%62.6% (104 mmol/mol 628.4 mmol/mol). The most common cause of severe or acute hyperglycemia among the admitted patients was infection, accounting for 44.9% of admissions, followed by DKA (13.4%), uncontrolled diabetes secondary to non-compliance (13.4%), and cardiovascular disease (13%). The majority of patients (72.5%) had more than one comorbidity; only 27.5% of patients had no comorbidities. Hypertension was the most frequent comorbidity, reported in 61.9% of patients, followed by ischemic heart disease (18.8%) and renal impairment (16.8%). iii. the use of concurrent medications; iv. blood glucose levels on admission and throughout the severe or acute hyperglycemia phase; and v. laboratory results and other monitoring parameters as stated in case notes. iv. blood glucose levels on admission and throughout the severe or acute hyperglycemia phase; and Data Collection 1) Adult T2DM patients who are equal or more than 18 years old The following patient data were collected: 1) Adult T2DM patients who are equal or more than 18 years old The following patient data were collected: i. demographic characteristics (age, sex, ethnicity, and body mass index [BMI]). 2) Hospitalized with severe or acute hyperglycemia with blood glucose level over 13.9 mmol/L ii. comorbidities. e 1. Flow chart of methodology. ICD-10 = International Statistical Classification of Diseases and Related Problems 10th Revision; UMM rsity of Malaya Medical Clinic. 0.1371/journal.pone.0106505.g001 Figure 1. Flow chart of methodology. ICD-10 = International Statistical Classification of Diseases and Related Problems 10th Revision; UMMC = University of Malaya Medical Clinic. doi:10.1371/journal.pone.0106505.g001 September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 2 Factors Associated with the Management of Severe or Acute Hyperglycemia Causes of Severe or Acute Hyperglycemia. DKA, cardio- vascular diseases and acute exacerbation of bronchial asthma were significantly associated with the insulin regimen used (Table 4). The use of sliding-scale insulin (67.3%) was more common than that of basal-bolus insulin (32.7%) among patients with DKA. In contrast, when compared with sliding-scale insulin, basal-bolus insulin was more frequently used in managing severe or acute September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org September 2014 \| Volume 9 \| Issue 9 \| e106505 3 Sliding-Scale versus Basal-Bolus Insulin Therapy Table 2. Clinical characteristics of the patients (N = 202). Clinical characteristics N Number of patients (Percentage, %) Duration of hospital stay 202 Not more than 7 days 125 (61.9) 8 to 14 days 50 (24.8) More than 15 days 27 (13.4) Admission blood glucose levels (mmol/L) 202 12.3–14.4 5 (2.5) 14.5–16.7 27 (13.4) 16.8–19.4 35 (17.3) 19.5–22.2 37 (18.3) .22.3 98 (48.5) HbA1c 202 Achieve target (,6.5) 0 (0) Not achieve target ($6.5 92 (45.5) Unknown 110 (54.5) qHyperglycemia occurred secondary to or was caused by 247 Infection 111 (44.9) Diabetic ketoacidosis 33 (13.4) Uncontrolled diabetes (non-compliance) 33 (13.4) Cardiovascular disease 32 (13.0) Cerebrovascular accident 9 (3.6) Renal impairment 8 (3.2) Hyperosmolar hyperglycemia 7 (2.8) Acute exacerbation of bronchial asthma 6 (2.4) Others 8 (3.2) Comorbidities 202 Hypertension 125 (61.9) Ischemic heart disease 38 (18.8) Renal Impairment 34 (16.8) Dyslipidemia 19 (9.4) Stroke 12 (5.9) Bronchial Asthma 9 (4.5) Liver Impairment 3 (1.5) Infection 2 (1.0) Pneumonia 1 (0.5) Others 8 (4.0) qOne patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin; One patient may have more than one comorbidity. doi:10.1371/journal.pone.0106505.t002 Table 2. Clinical characteristics of the patients (N = 202). qOne patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin; * One patient may have more than one comorbidity. doi:10.1371/journal.pone.0106505.t002 hyperglycemia secondary to cardiovascular diseases and acute exacerbation of bronchial asthma (15 and 6 cases, respectively). prednisolone 30 mg was the most common corticosteroid dosing regimen among the sliding-scale insulin and basal-bolus insulin- treated cases, comprising 66.7% and 53.3% of cases, respectively. Concomitant Drug Use during Severe or Acute Hyperglycemia Factors not associated with the Management of Severe or Acute Hyperglycemia Factors that had no significant association with the management of severe or acute hyperglycemia are shown in Tables 6–8. Regarding concomitant drug use, corticosteroids (p = 0.037), and loop diuretics (p = 0.016) appeared to be significantly associated with basal-bolus and sliding-scale insulin regimens (Table 5). Discussion Figure 2 shows the common dosing regimens of corticosteroids administered throughout the severe or acute hyperglycemia phase stratified by insulin regimen (15 cases using basal-bolus insulin and six cases using sliding-scale insulin). Oral administration of Demographic Characteristics Of the 202 patients, majority were female. Malays comprised the highest, followed by the Indian patients. The observed September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org 4 Sliding-Scale versus Basal-Bolus Insulin Therapy Table 3. Level of glycemic control achieved with the two insulin regimens. Basal-bolus Insulin Sliding-scale insulin p-value Number of casesq 159 179 Number of blood glucose readings 2766 4015 Mean fasting blood glucose (mmol/L) 10.862.3 11.663.5 0.028* Mean blood glucose throughout severe or acute hyperglycemia (mmol/L) 12.361.9 12.862.2 0.021* Mean insulin doses 12.5165.5 units 3.1460.9 units/h Blood glucose ,3.3 mmol/L Number of cases (%) 4 (2.5) 18 (10.1) 0.005* Number of readings (%) 5 (0.18) 21 (0.52) Blood glucose ,3.9 mmol/L Number of cases (%) 9 (5.7) 17 (9.5) 0.186 Number of readings (%) 14 (0.51) 18 (0.45) qOne patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin; * Statistically significant (p,0.05). doi:10.1371/journal.pone.0106505.t003 qOne patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin; * Statistically significant (p,0.05). doi:10.1371/journal.pone.0106505.t003 is presumably associated with the development of severe or acute hyperglycemia leading to hospitalization. difference in ethnic distribution might be attributable to the stringent inclusion criteria of this study, where only T2DM patients with severe or acute hyperglycemia on admission were considered for analysis. The most cause of severe or acute hyperglycemia in this study was infection, comprising 44.9% of admissions. Other reasons for admission included DKA, uncontrolled diabetes secondary to non- compliance, and cardiovascular diseases. Cardiovascular diseases and non-compliance to diabetes medications were found to be common among the 156 patients, comprising 7.1% and 8.3% of cases, respectively. The proportions of obese and pre-obese patients were higher in a study conducted by Zaman Huri et al. [3], in which 46.2% and 37.2% of the patients were obese and pre-obese, respectively. However, this is in contrast to the current study where, in the 28.7% of the study population for whom data were available, 12.9% of the study population had a BMI in the normal range, and 7.9% and 5.9% were classified as pre-obese and obese, respectively. This study also revealed that 72.5% of the studied patients had more than one comorbidity, with hypertension being the most common (61.9%), followed by ischemic heart disease (18.8%) and renal impairment (16.8%). Zaman Huri et al. Demographic Characteristics [3] reported a similar pattern, where hypertension was the most common comorbidity (82.7%), followed by renal impairment (39.7%) and ischemic heart disease (27.5%). Insulin Regimens Used during Severe or Acute Hyperglycemia Glycemic Control achieved with Insulin Regimens In this study, the mean insulin dose used in sliding-scale insulin regimens was low (3.1460.9 units/hour), which is possibly attributable to the insulin titration algorithms used, whereby insulin was administered on hourly basis. On the other hand, a higher mean insulin dose (12.5165.5 units) was achieved with the basal-bolus insulin regimen. This is primarily because in basal- bolus-treated patients the insulin units were calculated based on the patient’s body weight and adjusted appropriately based on the blood glucose levels throughout hospitalization. Clinical Characteristics The mean duration of the 247 hospital admissions was 7.9 days, similar to that in a retrospective study involving T2DM patients in which 71.5% of patients stayed in hospital for not more than seven days and 9.5% were hospitalized for more than 15 days [3]. In this study, the use of sliding-scale insulin regimen was common among the study population. Of 338 cases, 53% involved the use of sliding-scale insulin regimen, despite its use not recommended by the ADA and published journals [5,7,8,9]. The use of sliding-scale insulin regimen is discouraged as it only attempts to treat severe or acute hyperglycemia after it has occurred [10]. According to a published local study, 12% and 83% Data on HbA1c were available for 45.5% of patients in our study. The mean HbA1c value was 11.7% (104 mmol/mol). However, a mean HbA1c of 7.7% (61 mmol/mol) was reported in a study by Umpierrez et al. [6]. The higher HbA1c in the current study reflects the poor glycemic control among study subjects and Table 4. Causes of severe or acute hyperglycemia based on the insulin regimens. Table 4. Causes of severe or acute hyperglycemia based on the insulin regimens. Table 4. Causes of severe or acute hyperglycemia based on the insulin regimens. Causes of severe or acute hyperglycemia Insulin regimens p-value Basal-bolus (n = 159) Sliding-scale (n = 179) Diabetic ketoacidosis Present 16 (32.7%) 33 (67.3%) 0.043* Not Present 143 (49.5%) 146 (50.5%) Cardiovascular diseases Present 27 (69.2%) 12 (30.8%) 0.005* Not Present 132 (44.1%) 167 (55.9%) Acute exacerbation of bronchial asthma Present 6 (100%) 0 (0%) 0.010* Not Present 153 (46.1%) 179 (53.9%) * Statistically significant (p,0.05). doi:10.1371/journal.pone.0106505.t004 PLOS ONE \| www.plosone.org 5 September 2014 \| Volume 9 \| Issue 9 \| e106505 September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org Sliding-Scale versus Basal-Bolus Insulin Therapy Table 5. Comparison of drug use based on the insulin regimen used. Table 5. Comparison of drug use based on the insulin regimen used. Drug Class Insulin regimens p-value Basal-bolus (n = 159) Sliding-scale (n = 179) Corticosteroids Present 15 (71.4%) 6 (28.6%) 0.037* Not Present 144 (45.4%) 173 (54.6%) Loop diuretics Present 22 (68.8%) 10 (31.3%) 0.016* Not Present 137 (44.8%) 169 (55.2%) *Statistically significant (p,0.05). doi:10.1371/journal.pone.0106505.t005 of admitted patients treated with sliding-scale insulin had at least one episode of hypoglycemia and hyperglycemia, respectively [7]. Clinical Characteristics This finding may relate to the fact that sliding-scale insulin was used in more admissions than basal-bolus insulin in this study population; the number of blood glucose readings where the sliding-scale insulin regimen was used was double that of basal- bolus insulin. Factors Associated with the Management of Severe or Acute Hyperglycemia Causes of Severe or Acute Hyperglycemia. This study demonstrated a significant association between DKA and the use of insulin regimens throughout the severe or acute hyperglycemia phase (p = 0.043), with DKA more common in cases in which sliding-scale insulin was used. Sliding-scale insulin use remained common among the DKA patients, despite recommendations urging its discontinuation [5]. The results of this study also demonstrate significant differences in cases of hypoglycemia (defined as blood glucose ,3.3 mmol/L) between basal-bolus insulin and sliding-scale insulin regimens (p = 0.005). The use of sliding-scale insulin and basal-bolus insulin resulted in 10.1% and 2.5% cases of hypoglycemia, respectively. Conversely, cardiovascular disease was also significantly associ- ated with the insulin regimen used (p = 0.005), but the number of cardiovascular disease cases in which basal-bolus insulin was used Figure 2. Corticosteroid dosing regimens used. BBI = basal-bolus insulin, SSI = sliding-scale insulin, OD = once daily, BD = twice daily, TDS = three times daily, IV = intravenous. doi:10.1371/journal.pone.0106505.g002 Figure 2. Corticosteroid dosing regimens used. BBI = basal-bolus insulin, SSI = sliding-scale insulin, OD = once daily, BD = twice daily, TDS = three times daily, IV = intravenous. doi:10.1371/journal.pone.0106505.g002 Figure 2. Corticosteroid dosing regimens used. BBI = basal-bolus insulin, SSI = sliding-scale insulin, OD = once daily, BD = twice daily, TDS = three times daily, IV = intravenous. doi:10.1371/journal.pone.0106505.g002 September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 6 Sliding-Scale versus Basal-Bolus Insulin Therapy Table 6. Comparison of demographic and clinical characteristics based on insulin regimens. Factors Associated with the Management of Severe or Acute Hyperglycemia Insulin regimensa p-value Basal-bolus (n = 159) Sliding-scale (n = 179) Demographic characteristics Gender Male 75 (47.2%) 82 (45.8%) 0.802 Female 84 (52.8%) 97 (54.2%) Age Non-elderly 116 (73%) 134 (74.9%) 0.690 Elderly 43 (27%) 45 (25.1%) Causes of severe or acute hyperglycemia Cerebrovascular accident Present 5 (3.1%) 5 (2.8%) 1.000 Not present 154 (96.9%) 174 (97.2%) Uncontrolled diabetes Present 19 (11.9%) 28 (15.6%) 0.411 Not present 140 (88.1%) 151 (84.4%) Renal impairment Present 5 (3.1%) 6 (3.4%) 1.000 Not present 154 (96.9%) 173 (96.6%) Others Present 6 (3.8%) 4 (2.2%) 0.525b Not present 153 (96.2%) 175 (97.8%) Comorbidities Hypertension Present 103 (64.8%) 109 (60.9%) 0.532 Not present 56 (35.2%) 70 (39.1%) Ischemic heart disease Present 38 (23.9%) 27 (15.1%) 0.056 Not present 121 (76.1%) 152 (84.9%) Stroke Present 11 (6.9%) 13 (7.3%) 1.000 Not present 148 (93.1%) 166 (92.7%) Pneumonia Present 1 (0.6%) 0 (0%) 0.470b Not present 158 (99.4%) 179 (100%) Infection Present 2 (1.3%) 2 (1.1%) 1.000b Not present 157 (98.7%) 177 (98.9%) Dyslipidemia Present 15 (9.4%) 13 (7.3%) 0.599 Not present 144 (90.6%) 166 (92.7%) Renal Impairment Present 27 (17%) 27 (15.1%) 0.744 Not present 132 (83%) 152 (84.9%) Liver Impairment Present 5 (3.1%) 2 (0.6%) 0.260b Not present 154 (96.9%) 177 (98.9%) Bronchial Asthma Present 9 (5.7%) 8 (4.5%) 0.802 Not present 150 (94.3%) 171 (95.5%) Others Present 6 (3.8%) 8 (4.5%) 0.963 Not present 153 (96.2%) 171 (95.5%) a One patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. b Computed using Fisher’s Exact Test. doi:10.1371/journal.pone.0106505.t006 atient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. uted using Fisher’s Exact Test. 371/journal.pone.0106505.t006 because of the increase in stress hormones such as cortisol and catecholamines [12]. According to Dungan et al. [13], a subcutaneous basal-bolus insulin regimen is a better approach than sliding-scale insulin for achieving effective glycemic control in stress hyperglycemia following an acute illness, which is similar to the findings reported by [8]. was approximately double that of sliding-scale insulin. A study focusing on cardiovascular disease reported that the strict control of preprandial and postprandial hyperglycemia resulted in the reduction of cardiovascular disease among T2DM patients [11]. Factors Associated with the Management of Severe or Acute Hyperglycemia Thus, treatment of severe or acute hyperglycemia secondary to cardiovascular diseases with a basal-bolus insulin regimen is reasonable, where the bolus doses are administered to control the excessive rise of postprandial blood glucose levels. g Use during the Severe or Acute Hyperglycem Six patients were admitted with hyperglycemia secondary to or caused by acute exacerbation of bronchial asthma. All were treated with basal-bolus insulin. The development of severe or acute hyperglycemia following an acute asthma attack could be The use of certain classes of medications including corticoste- roids (p = 0.037), and loop diuretics (p = 0.016) appeared to have a September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 7 Sliding-Scale versus Basal-Bolus Insulin Therapy Table 7. Comparison of drug use according to insulin regimen used. Insulin regimensa p-value Basal-bolus (n = 159) Sliding-scale (n = 179) Beta blockers Present 23 (14.5%) 19 (10.6%) 0.365 Not present 136 (85.5%) 160 (89.4%) Calcium channel blockers Present 37 (23.3%) 29 (16.2%) 0.134 Not present 122 (76.7%) 150 (83.8%) Alpha-1 adrenergic blockers Present 6 (3.8%) 2 (1.1%) 0.155b Not present 153 (96.2%) 177 (98.9%) Angiotensin-converting enzyme inhibitors Present 32 (20.1%) 22 (12.3%) 0.070 Not present 127 (79.9%) 157 (87.7%) Cephalosporins Present 38 (23.9%) 48 (26.8%) 0.625 Not present 121 (76.1%) 131 (73.2%) Penicillins Present 63 (39.6%) 74 (41.3%) 0.834 Not present 96 (60.4%) 105 (58.7%) Macrolides Present 13 (8.2%) 13 (7.3%) 0.912 Not present 146 (91.8%) 166 (92.7%) Fluoroquinolones Present 0 (0%) 1 (0.6%) 1.000b Not present 159 (100%) 178 (99.4%) Opiod analgesics Present 18 (11.32%) 19 (10.6%) 0.974 Not present 141 (88.7%) 160 (89.4%) Atypical antipsychotics Present 1 (0.6%) 1 (0.6%) 1.000b Not present 158 (99.4%) 178 (99.4%) Proton-pump inhibitors Present 14 (8.8%) 11 (6.1%) 0.469 Not present 145 (91.2%) 168 (93.9%) Histamine-2 receptor antagonist Present 23 (14.5%) 29 (16.2%) 0.771 Not present 136 (85.5%) 150 (83.8%) a One patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. b Computed using Fisher’s Exact Test. doi:10.1371/journal.pone.0106505.t007 a One patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. b Computed using Fisher’s Exact Test. a One patient may have more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. b Computed using Fisher’s Exact Test. doi:10.1371/journal.pone.0106505.t007 Table 8. Comparison of corticosteroid dosing regimen used according to insulin regimen used. more than one hospital admission and been treated with both basal-bolus insulin and sliding-scale insulin. E T Conclusion The use of a sliding-scale insulin regimen among T2DM patients with severe or acute hyperglycemia admitted to our institution was common. In addition, we found that DKA, cardiovascular diseases and acute exacerbation of bronchial asthma appeared to have a significant association with the insulin regimens used in glycemic control. Several concurrent medica- tions, including corticosteroids, and loop diuretics were also found to be significantly associated with the insulin regimen used. Overall, this study revealed that lower blood glucose levels were achieved with a basal-bolus insulin regimen compared with sliding-scale insulin in the population studied. Furthermore, significant association was observed between the use of loop diuretics and the use of the basal-bolus insulin regimen (p = 0.016). A recent study by Zaman Huri et al. [3] revealed that the use of loop diuretics was found to have significant association with insulin resistance in T2DM patients during severe or acute hyperglycemia. The study reported that more patients receiving loop diuretics were insulin resistant (26 patients) compared with those who were insulin sensitive (19 patients). The authors concluded that this may indicate that loop diuretics might increase insulin resistance in T2DM patients during severe or acute hyperglycemia [3]. The identification of factors associated with the insulin regimens used in managing severe or acute hyperglycemia may contribute towards achieving optimal glycemic control in T2DM patients. There is currently a lack of published studies on the factors associated with the management of severe or acute hyperglycemia, and further investigation of this is warranted. A limitation of this study relates to its retrospective nature, whereby assessment of glycemic control in the patients studied could be based only on the data available in medical records. A patient’s condition throughout the severe or acute hyperglycemia phase could not be assessed, and it was not possible to further 1. Mafauzy M, Hussein Z, Chan SP (2011) The status of diabetes control in Malaysia: Results of DiabCare 2008. The Medical Journal of Malaysia 66(3): 175–181. g Use during the Severe or Acute Hyperglycem Corticosteroids Insulin regimens p-value Basal-bolus (n = 15) Sliding-scale (n = 6) PO Prednisolone 30 mg OD Present 8 (53.3%) 4 (66.7%) 0.659a Not present 7 (46.7%) 2 (33.3%) PO Prednisolone 50 mg OD Present 2 (13.3%) 0 (0%) 1.000a Not present 13 (86.7%) 6 (100%) PO Prednisolone 60 mg OD Present 2 (13.3%) 1 (16.7%) 1.000a Not present 13 (86.7%) 5 (83.3%) PO Hydrocortisone 20 mg BD Present 1 (6.67%) 0 (0%) 1.000a Not present 14 (93.3%) 6 (100%) IV Hydrocortisone 50 mg TDS Present 1 (6.67%) 0 (0%) 1.000a Not present 14 (93.3%) 6 (100%) IV Hydrocortisone 100 mg TDS Present 1 (6.67%) 1 (16.7%) 0.500a Not present 14 (93.3%) 5 (83.3%) a Computed using Fisher’s Exact Test. doi:10.1371/journal.pone.0106505.t008 PLOS ONE \| www.plosone.org 8 September 2014 \| Volume 9 \| Issue 9 \| e106505 September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 8 Sliding-Scale versus Basal-Bolus Insulin Therapy significant influence on the management of severe or acute hyperglycemia. investigate decisions on the approach taken by clinicians regarding glycemic control. Corticosteroids are used widely in hospital setting and are known to provoke new-onset hyperglycemia in non-diabetic patients or exacerbate severely uncontrolled hyperglycemia in patients with diabetes [14]. The development of severe or acute hyperglycemia resulting from the administration of corticosteroids occurs primarily because of a decrease in insulin secretion and insulin sensitivity [15]. In this study, the most common cortico- steroid dosing regimen encountered was oral prednisolone 30 mg administered once daily. Author Contributions Conceived and designed the experiments: HZH SRV. Performed the experiments: HZH VP. Analyzed the data: HZH VP. Wrote the paper: HZH. 7. Huri HZ, Yeap SM, Pendek R (2007) Episodes of hypoglycemia and hyperglycemia during the use of sliding scale insulin in hospitalized diabetes patients. Asian Biomedicine, 1(3): 307–311. 6. Umpierrez GE, Smiley D, Jacobs S, Peng L, Temponi A, et al. (2011) Randomised study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes undergoing general surgery (RABBIT 2 surgery). Diabetes Care 34(2): 256–261. y y ( ) 5. American Diabetes Association (2013) Standard of medical care in diabetes. Diabetes Care 36(1): S11–S66. 2. Dombrowski NC, Karounos DG (2012) Pathophysiology and management strategies for hyperglycemia for patients with acute illness during and following a hospital stay. Metabolism. Retrieved October 20, 2012, from http://ezproxy. um.edu.my:2095/science/article/pii/S0026049512002831. 3. Huri HZ, Makmor-Bakry M, Hashim R, Mustafa N, Wan NWZ (2013) Demographic and clinical predictors for insulin resistance in type 2 diabetes mellitus patients during severe/acute hyperglycemia phase. Latin American Journal of Pharmacy 32(1): 120–127. 4. Nau KC, Lorenzetti RC, Cucuzzella, Devine T, Kline J (2010) Glycemic control in hospitalized patients not in intensive care: Beyond sliding-scale insulin. American Family Physician 81(9): 1130–1135. 13. Dungan KM, Braithwaite SS, Preiser JC (2009) Stress hyperglycaemia. The Lancet 373(9677): 1798–1807. References 8. Umpierrez GE, Smiley D, Zisman A, Prieto LM, Palacio A, et al. (2007) Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes (RABBIT 2 trial). Diabetes Care 30(9): 2181–2186. 2. Dombrowski NC, Karounos DG (2012) Pathophysiology and management strategies for hyperglycemia for patients with acute illness during and following a hospital stay. Metabolism. Retrieved October 20, 2012, from http://ezproxy. um.edu.my:2095/science/article/pii/S0026049512002831. 9. Shaw JE, Sicree RA, Zimmet PZ (2010) Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Research and Clinical Practice 87(1): 4–14. 10. Magaji V, Johnston JM (2011) Inpatient management of hyperglycemia and diabetes. Clinical Diabetes 29(1): 3–9. 3. Huri HZ, Makmor-Bakry M, Hashim R, Mustafa N, Wan NWZ (2013) Demographic and clinical predictors for insulin resistance in type 2 diabetes mellitus patients during severe/acute hyperglycemia phase. Latin American Journal of Pharmacy 32(1): 120–127. 11. Bonora E, Muggeo M (2001) Postprandial blood glucose as a risk factor for cardiovascular disease in type II diabetes: The epidemiological evidence. Diabetologia 44(12): 2107–2114. 12. Lutfi MF, Sukkar MY (2011) The hyperglycemic effect of bronchial asthma. Suddan Medical Journal 47(2): 69–74. 4. Nau KC, Lorenzetti RC, Cucuzzella, Devine T, Kline J (2010) Glycemic control in hospitalized patients not in intensive care: Beyond sliding-scale insulin. American Family Physician 81(9): 1130–1135. 13. Dungan KM, Braithwaite SS, Preiser JC (2009) Stress hyperglycaemia. The Lancet 373(9677): 1798–1807. y y ( ) 5. American Diabetes Association (2013) Standard of medical care in diabetes. Diabetes Care 36(1): S11–S66. 14. Baldwin D, Apel J (2013) Management of hyperglycemia in hospitalized patients with renal insufficiency or steroid-induced diabetes. Current Diabestes Reports 13(1): 114–120. 6. Umpierrez GE, Smiley D, Jacobs S, Peng L, Temponi A, et al. (2011) Randomised study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes undergoing general surgery (RABBIT 2 surgery). Diabetes Care 34(2): 256–261. 15. Umpierrez GE, Hellman R, Korytkowski MT, Kosiborod M, Maynard GA, et al. (2012) Management of hyperglycemia in hospitalized patients in non- critical care setting: An Endocrine Society Clinical Practice Guideline. The Journal of Clinical Endocrinology & Metabolism 97(1): 16–38. September 2014 \| Volume 9 \| Issue 9 \| e106505 PLOS ONE \| www.plosone.org 9
https://openalex.org/W1573416599	https://hal.inrae.fr/hal-02757188/document	English	null	39th Annual European Brain and Behaviour Society Abstracts	Neural plasticity	2,007	cc-by	2,152	Subdivisions of the arcopallium are differentially involved in the control of fear behaviour in the japanese quail ne Saint-Dizier, Christine Leterrier, Frédéric Lévy, Sabine Richard To cite this version: Hélène Saint-Dizier, Christine Leterrier, Frédéric Lévy, Sabine Richard. Subdivisions of the arco- pallium are differentially involved in the control of fear behaviour in the japanese quail. 39. Annual Meeting of the European Brain and Behaviour Society, Sep 2007, Trieste, Italy. 10.1155/2007/23250. hal-02757188 Distributed under a Creative Commons Attribution 4.0 International License NON-CONSCIOUS EMOTIONAL CONTAGION IN BLINDSIGHT M. Tamietto,1 L. Castelli,1 S. Vighetti, L. Latini Corazzini,1 G. Geminiani,1 L. Weiskrantz,2 and B. de Gelder3 M. Tamietto,1 L. Castelli,1 S. Vighetti, L. Latini Corazzini,1 G. Geminiani,1 L. Weiskrantz,2 and B. de Gelder3 1Depatment of Psychology, University of Torino, Italy 2University of Oxford, UK 3University of Tilburg and MGH Harvard Medical School, The Netherlands Observing facial expressions prompts imitation as can be typically observed with facial electromyography (EMG). Here we explored whether this automatic reaction occurs even in the absence of visual awareness for the stimulus, and whether this can be elicited also by bodily expressions. Facial and bodily expressions of happiness and fear were presented either in the intact visual ﬁeld or in the blind ﬁeld of two well-known hemianope patients (DB and GY) with striate cortex lesions but residual vision (blindsight). The patients were required to judge the emotional expression of the pic- tures presented in their intact visual ﬁeld, and “to guess” the expression of the unseen pictures shown in their blind ﬁeld. During the task we recorded emotion-speciﬁc facial muscle activity (zygomaticus major for happy, corrugator supercilii for fear). Despite both patients reported no visual awareness for stimuli projected in the blind ﬁeld and commented their performance as “at chance”, their evaluation of the emotional expressions was signiﬁcantly above chance-level for faces and bodies alike. Most notably, unseen facial as well as bodily ex- pressions produced a congruent emotional reaction in pa- tients’ face, comparable to that observed in response to con- sciously perceived pictures. Our ﬁndings provide evidence that facial expressions in the observer may unfold as an au- tomatic reaction that results from emotional contagion. This expressive response seems insensitive to visual awareness and to the speciﬁc perceptual features of the stimuli. Rather, it ap- pears to be modulated by the emotional valence of external events. HAL Id: hal-02757188 https://hal.inrae.fr/hal-02757188v1 Submitted on 3 Jun 2020 de Gelder3 1Depatment of Psychology, University of Torino, Italy 2University of Oxford, UK 3University of Tilburg and MGH Harvard Medical School, The Netherlands M. C. Schweizer,1 M. S. H. Henniger,1 F. Schleicher,1 T. Pohl,1 M. B. M¨uller,2 and I. Sillaber1 1Affectis Pharmaceuticals AG, Munich, Germany 2Max-Planck-Institute of Psychiatry, Munich, Germany Chronic mild stress (CMS) as an animal model of depression enjoys a certain popularity in psychiatric research - not least due to its face validity and comprehensive readout. However, the results are diﬃcult to replicate in diﬀerent labs. The goal of the present study was to examine the inﬂuence of stress- induced changes in general activity on the behavioural read- out. We particularly focussed on the inﬂuence of light as a widely used stressor in CMS protocols on subsequent mea- sures. A weekly CMS schedule consisting of common mild stressors was applied for at least 4 weeks to diﬀerent strains of mice. During this stress period saccharin intake and pref- erence over water was acquired twice a week, each time dur- ing the ﬁrst 2 hours of the dark phase. To exclude inter- ventions with a putatively high impact on consuming be- haviour per se, food and water deprivation was omitted. Pa- rameters were assessed using tests like open ﬁeld, modiﬁed holeboard and long-term home cage observation. Indepen- dent of the illumination conditions in the behavioural tests, an apparently paradox decrease in anxiety-related behaviour after CMS was observed. This could be explained by a gen- erally increased stress-induced activity, which in turn ap- peared as reduced risk assessment behaviour. Preceding ap- plication of a single footshock normalised the latter in CMS mice. While no enduring decrease in saccharin intake (‘an- hedonia’) due to CMS was observed, over-night illumina- HAL Id: hal-02757188 https://hal.inrae.fr/hal-02757188v1 Submitted on 3 Jun 2020 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License Neural Plasticity 72 upon rat turning behaviour and hippocampal levels of no- radrenaline and corticosterone, in order to minimize stress, limitation of movement and other confounding factors in microdialysis studies. Three groups of rat setups were exam- ined: fast to slow turning, slow to fast turning and a liquid swivel. Rat turning behaviour was recorded digitally over a period of forty-ﬁve hours. Brain samples were taken twice for two hours in twenty-four hours and analysed for corti- costerone and noradrenalin with radio immuno assay and liquid chromatography coupled to electrochemical detection respectively. The results show a signiﬁcant diﬀerence in to- tal turning time between the ﬁrst day/night and the second day/night in both turning setups, meaning that it takes et least twelve hours before the rats are adapted to the turn- ing apparatus. These results are conﬁrmed by a decrease of hippocampal corticosterone and noradrenalin levels during the second day compared to the ﬁrst day in both turning se- tups. The slow to fast turning setup showed enhanced slow movement and therefore a higher total turning compared to the fast to slow turning setup. Therefore, these results reveal that the adaptation period to the turning apparatus and the turning speed in microdialysis studies require careful consid- eration when interpreting data. tion as particular stressor of the weekly paradigm turned out to be associated with a signiﬁcant decrease in saccharin in- take during the measurement period the day after. Taken to- gether, these results suggest that in mice the CMS regimen as a whole causes hyperreactivity to a novel environment rep- resented by the test situation and therefore requires careful interpretation of behaviour. Shifts in circadian rhythms due to light cycle changes may mimic an ‘anhedonic’ eﬀect of CMS. NON-CONSCIOUS EMOTIONAL CONTAGION IN BLINDSIGHT M. Tamietto,1 L. Castelli,1 S. Vighetti, L. Latini Corazzini,1 G. Geminiani,1 L. Weiskrantz,2 and B. SUBDIVISIONS OF THE ARCOPALLIUM ARE DIFFERENTIALLY INVOLVED IN THE CONTROL OF FEAR BEHAVIOUR IN THE JAPANESE QUAIL H. Saint-Dizier,1 C. Leterrier,1 F. L´evy,1 and S. Richard1 1Equipe Comportement, Neurobiologie et Adaptation, Unit`e de Physiologie de la Reproduction et des Comportements, UMR85 INRA – CNRS – Universit´e de Tours Haras Nationaux, F-37380 Nouzilly, France 1Equipe Comportement, Neurobiologie et Adaptation, Unit`e de Physiologie de la Reproduction et des Comportements, UMR85 INRA – CNRS – Universit´e de Tours Haras Nationaux, F-37380 Nouzilly, France 73 Luciana Biecker Growing interest in the phylogeny of emotions within ver- tebrates has motivated research on the neurobiology of fear reactions in birds. In the avian brain, the arcopallium has been suspected to play a major role in the control of fear re- actions. This structure is considered as a partial homologue of the mammalian amygdale, on the basis of developmen- tal and anatomical data. Moreover, lesions or stimulations of the arcopallium induce respectively a decrease or an increase in fear reactions. However, the arcopallium is a large and het- erogeneous structure and the speciﬁc roles of its subdivisions are unknown. The present study aimed at investigating the respective implications of diﬀerent subdivisions of the arco- pallium in the control of fear behaviour. Adult Japanese quail were given bilateral electrolytic lesions of the arcopallium or sham-operation, and were subsequently placed in several tests of fear: open-ﬁeld, hole-in the-wall box, tonic immo- bility and novel object tests. Quail with lesions of the ante- rior part of the arcopallium exhibited reduced fear behaviour when compared to sham-operated quail. By contrast, quail with lesions in the caudal part of the arcopallium tended to show more pronounced fear behaviour than sham-operated quail. The behaviour of quail with combined lesions of the anterior and caudal parts of the arcopallium was not signif- icantly diﬀerent from that of shamoperated quail. Those re- sults are the ﬁrst to show a diﬀerential involvement of sub- divisions of the arcopallium in the control of fear behaviour in birds. The results will be discussed in the light of current knowledge regarding the neuroanatomical characteristics of the arcopallium. isolation on the females, because when sexes were analysed separately, individually housed females showed higher activ- ity than those socially housed, however, housing conditions had no eﬀect on males. Furthermore, isolated females dis- played higher activity in the actimeter than isolated males. Individual housing also increased the general activity of mice on the elevated plus-maze, but equally in both sexes. SUBDIVISIONS OF THE ARCOPALLIUM ARE DIFFERENTIALLY INVOLVED IN THE CONTROL OF FEAR BEHAVIOUR IN THE JAPANESE QUAIL The fe- males displayed more anxiety than males, spending less time on the central square of the open-ﬁeld. This eﬀect was not due to sex diﬀerences in general locomotion, because it was precisely the females which presented more activity in this behavioural test. Nevertheless, neither the individual hous- ing condition nor the sex produced signiﬁcant diﬀerences in anxiety on the plus-maze. These results indicate that so- cial housing can reduce the hyperactive response to novelty in females, whereas in males, the housing environment did not have a signiﬁcant eﬀect on activity or performance in the anxiety test. WHAT CAN NAPLES HIGH EXCITABILITY AND SPONTANEOUSLY HYPERTENSIVE RATS TELL US ABOUT DIFFERENT VARIANTS OF ATTENTION-DEFICIT HYPERACTIVITY DISORDER? L. A. Ruocco,1 D. Vallone,2 D. Viggiano,3 U. A. Gironi Carnevale,1 and A. G. Sadile1 1Lab Neurophysiol. Behavior & Neural Networks, Department Exptl. Med, Second Univ. Naples, Naples, Italy 2Institut f¨ur Toxikologie und Genetik, Forschungszentrum Karlsruhe, Germany 3Department Neurosci., University of Naples “Federico II”, Naples, Italy A. Puigcerver,2 A. Valls-Rios,1 C. I. Navarro-Frances,1 and M. C. Arenas1 A. Puigcerver,2 A. Valls-Rios,1 C. I. Navarro-Frances,1 and M. C. Arenas1 A. Puigcerver,2 A. Valls-Rios,1 C. I. Navarro-Frances,1 and M. C. Arenas1 1Department of Psychobiology, University of Valencia, Valencia, Spain 2Department of Psychobiology, University of Malaga, Malaga, Spain 1Department of Psychobiology, University of Valencia, Valencia, Spain 2Department of Psychobiology, University of Malaga, Malaga, Spain EFFECTS OF HOUSING ENVIRONMENT ON ACTIVITY AND ANXIETY LEVEL IN MALE AND FEMALE MICE Attention Deﬁcit Hyperactivity Disorder (ADHD) is a neu- rodevelopmental problem aﬀecting 1–3% of school children, mainly boys (4 : 1 ratio). It is characterized by inatten- tion, hyperactivity and impulsivity. An altered mesocorti- colimbic dopamine (DA) system is thought to be associ- ated to diﬀerent variants of ADHD. The Naples High Ex- citability (NHE) and the Spontaneously Hypertensive (SHR) rats model the variant with altered executive functions and response inhibition respectively. The NHE show hyperac- tive mesocortical DA branch by hypertrophic DA neurons, high expression of tyrosine hidroxylase (TH), high DA D2 receptor density and overexpression of DA-related phos- phoprotein (DARP32) in the mesencephalon. Conversely they show in the prefrontal cortex (PFC) more axonal vari- cosities, high DA transporter (DAT) density and lower DA D1 and D2 receptors. Moreover the mesotriatal branch is not altered as shown by TH, DAT, DA D1 and D2 recep- tors and DARP32. Treatment with methylphenidate (MPH; 3mg/Kg i.p. for 14days) reverses the basal proﬁle. In con- trast, the SHR show altered mesolimbic and mesocortical branches associated with no main changes in the mesen- cephalon but with a high responsiveness for TH expression and no responsiveness for DA D2 autoreceptors to MPH treatment. Conversely in the PFC a higher basal DA tone (Carboni et al. 2003, 2004) is associated with high DAT 1Department of Psychobiology, University of Valencia, Valencia, Spain 2Department of Psychobiology, University of Malaga, Malaga, Spain The present study was designed to determine the eﬀect of individual or social housing on various tests of anxiety in mice of both sexes. After an 18-day isolation period or group- housing, general activity of each mouse in the actimeter and in the open-ﬁeld test was recorded for 5 min. Afterwards, the animals were individually placed onto the central square of the elevated plus-maze and video recorded for 5 min. The number of counts in the actimeter and number of crossings from one square to another in the openﬁeld were registered as measures of activity, and also the number of closed arm entries in the elevated plus-maze. The percentage of time on the central square of the openﬁeld, together with the percent- age of time spent on the open arms and the percentage of open arm entries were scored as measures of anxiety level. Individual housing increased the activity on both open-ﬁeld and actimeter, this increase being really due to the eﬀect of
https://openalex.org/W4213242658	https://www.frontiersin.org/articles/10.3389/fvets.2022.762771/pdf	English	null	Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases in Omusati and Kunene Regions of Namibia	Frontiers in veterinary science	2,022	cc-by	6,248	Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases in Omusati and Kunene Regions of Namibia Ndahambelela Eiki 1, Matome Maake 1, Sogolo Lebelo 2, Bellonah Sakong 1, Nthabiseng Sebola 1 and Monnye Mabelebele 1 Ndahambelela Eiki 1, Matome Maake 1, Sogolo Lebelo 2, Bellonah Sakong 1, Nthabiseng Sebola 1 and Monnye Mabelebele 1 1 Department of Agriculture and Animal Health, College of Agriculture and Environmental Sciences, University of South Africa, Pretoria, South Africa, 2 Department of Life and Consumer Sciences, College of Agriculture and Environmental Sciences, University of South Africa, Pretoria, South Africa The aim of this study was to ﬁnd, evaluate, and document ethnoveterinary medications used to treat livestock ailments in Namibia’s Omusati and Kunene regions. A semi- structured questionnaire was used to interview a total of 100 people. The results of the survey will be useful in creating the groundwork for future scientiﬁc research and validation. There were 15 veterinary medicinal plant species identiﬁed, which were divided into 10 families. The only types of growth that were utilized were trees, herbs, and bushes. Leaves (71%) were the most widely used plant parts for ethnoveterinary medicine (EVM), followed by bark (14%), stem (8%), and root (7%). Fresh components were frequently preferred in medical compositions. Oral administration was the most common (42.76%), followed by cutaneous (topical) administration (36.18%). Indigenous knowledge was largely passed down through the generations by word of mouth, indicating that it was vulnerable to fragmentation and loss. EVMs were crushed, soaked in water, and administered orally or topically. Farmers who were older had greater EVM knowledge than those who were younger. Ziziphus mucronate, Combretum collinum, and Colophospermum mopane were used in the treatment of diarrhea. Z. mucronate was also used in the treatment of mastitis. Skin infections were treated using Aloe esculenta and Salvadora persica. Ximenia americana and C. imberbe were used to treat eye infections in cattle, goats, and sheep. Retained placentas were treated using Acacia nilotica, A. erioloba, and Grewia ﬂavescens. Roots from Fockea angustifolia were used in treating anthrax. A. esculenta Leach placed best with a ﬁdelity level (FL) value of 90%, followed by A. littoralis Baker in second place (56%), and Combretum collinum Fresen in third place (54%). The majority of EVM recipes took 2–3 days to recover. Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases in Omusati and Kunene Regions of Namibia More research is needed to determine the minimum inhibitory concentrations, biological activities, and toxicities, as well as characterize the chemical components of the plants and determine whether there is a plausible mechanism by which plant chemicals or possible physiological effects could achieve the results described by the respondents. Edited by: Farhana Ijaz, Hazara University, Pakistan Reviewed by: Muzammil Shah, King Abdulaziz University, Saudi Arabia Naila Hameed, University of the Punjab, Pakistan Correspondence: Ndahambelela Eiki ndahambelelaeiki@yahoo.com M M b l b l Reviewed by: Muzammil Shah, King Abdulaziz University, Saudi Arabia Naila Hameed, University of the Punjab, Pakistan Correspondence: Ndahambelela Eiki ndahambelelaeiki@yahoo.com Monnye Mabelebele mabelm@unisa.ac.za Specialty section: This article was submitted to Veterinary Pharmacology and Toxicology, a section of the journal Frontiers in Veterinary Science Received: 22 August 2021 Accepted: 05 January 2022 Published: 22 February 2022 Specialty section: This article was submitted to Veterinary Pharmacology and Toxicology, Specialty section: This article was submitted to Veterinary Pharmacology and Toxicology, Specialty section: This article was submitted to Veterinary Pharmacology and Toxicology, a section of the journal Frontiers in Veterinary Science Received: 22 August 2021 Accepted: 05 January 2022 Published: 22 February 2022 ORIGINAL RESEARCH published: 22 February 2022 doi: 10.3389/fvets.2022.762771 Keywords: ethnoveterinary, livestock, health management, diseases, knowledge, Omusati, Kunene, Namibia Edited by: Farhana Ijaz, Hazara University, Pakistan Abbreviations: EVM, ethnoveterinary medicines; S, shrub; T, tree; H, herb; R, root; B, bark; L, leaves; St, stem; W, wild; ∧FL, ﬁdelity level. Citation: Eiki N, Maake M, Lebelo S, Sakong B, Sebola N and Mabelebele M (2022) Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases in Omusati and Kunene Regions of Namibia. Front. Vet. Sci. 9:762771. doi: 10.3389/fvets.2022.762771 February 2022 \| Volume 9 \| Article 762771 Frontiers in Veterinary Science \| www.frontiersin.org Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. Study Area The research was carried out in Namibia’s Omusati and Kunene areas (see Figure 1). Omusati and Kunene regions border the Kunene River along the Angolan border in northern Namibia. Otjiherero and Nama/Damara languages are spoken by 46 and 36% of people in Kunene, respectively, while Oshiwambo is spoken by 96% of people in Omusati (9). People who speak Otjiherero are generally from the Herero tribe, whereas those who speak Oshiwambo are from the Ovambo tribe. The Bantu tribes found in the study region include the Herero and Ovambo, whereas the Nama/Damara (commonly known as Hottentots or San) are non-Bantu. Aawambo (Ovambo), Ovaherero-speaking pastoralists (Ovahimba, Ovatjimba, and Ovazemba), and Nama/Damara are some of the ethnic groups found in the regions. p y Ethnoveterinary therapies provide the best choices to farmers in Namibia’s Omusati and Kunene areas during challenging economic times, when their purchasing power is insuﬃcient to aﬀord ever-increasing veterinary prices. Furthermore, ethnoveterinary medical knowledge is in peril because it is reliant primarily on the collective memories of a few community practitioners (7). The situation is even more dire when it comes to ethnoveterinary medical knowledge, which is limited to a small number of livestock owners. As a result, it is critical to chronicle this ethnoveterinary medical knowledge to pass it along to future generations. Furthermore, identifying therapeutic plants would contribute to the development of ways for protecting and conserving endangered species. It will also help with the creation of herbal gardens, which will help to preserve biodiversity. Ethnoveterinary medical knowledge must be documented to aid in the ﬁnding of innovative drug sources (8). The people’s culture in the study area is diverse due to the tribal diversity. Men are recognized as the head of the household in Ovambo and Nama/Damara cultures, and they are responsible for cattle and household decision-making (10). In the Ovambo and Nama/Damara tribal groups, men also hunt game animals for meat, build huts for family living, cultivate agricultural land, and provide water for the family. In addition, Damara men oversee agriculture planting and harvesting. Damara and Ovambo women, on the other hand, are responsible for cooking and other homework [Ambunda & de Klerk (10). Pastoralists make up a large portion of the Damara population. The most common religions in the study region are Christianity and African traditional religions. INTRODUCTION approaches for treating livestock infections in the Omusati and Kunene regions of Namibia. One of the most important income sources in the Omusati and Kunene regions of Namibia is livestock rearing. Livestock as widely known provides local people with calories in the form of meat, milk, and derivatives, as well as a source of income (1). Moreover, it is also a source of employment, manure, and draft power for the cultivation and transport of goods (2) in majority of the developing countries. Farmers in the Omusati and Kunene regions regard cattle farming as a symbol of riches and honor, as well as a precaution against crop failure during droughts (3). These farmers also keep cattle on hand for special occasions like weddings, funerals, and christenings (4). Despite livestock production’s contribution to the livelihoods of people in these regions and the world over, its development is inhibited by diﬀerent constraints (5). Diseases are one of the most signiﬁcant constraints to cattle productivity (6). Study Area The Omusati and Kunene regions are largely rural, with crop and animal farming serving as the primary source of income for most residents. The Omusati and Kunene aboriginal peoples rely on animal husbandry as their primary source of income. As a result, the two zones are overgrazed and degraded, with a low diversity of plant species (2). The Omusati and Kunene regions are mostly impoverished, and many villages are isolated due to a lack of proper roads, making it diﬃcult for residents to get modern veterinary services (11). Furthermore, modern veterinary facilities and services are primarily found in cities far from farming villages. Importantly, forests are necessary for the survival of ethnoveterinary medicine (EVM) therapeutic plant species in the Omusati and Kunene regions. Therefore, documenting the state of indigenous ﬂora would aid in public awareness campaigns about endangered species and sustainable plant- collecting methods. Keeping these facts in view, the study was initiated, to document ethnoveterinary practices used for treating livestock diseases in two selected regions of Namibia. Thus, the study’s ﬁndings will be utilized to educate communities on the importance of EVM in providing primary livestock healthcare. It will also aid in disseminating the study’s livestock healthcare expertise to potentially inﬂuence policy change in favor of incorporating favorable EVM practices into national livestock healthcare systems. Further to this, the outcome will also help in raising a local awareness of the importance of medicinal plant conservation and participation in the propagation of threatened species. Agricultural extension oﬃcials and veterinary oﬃcers will make the entire study available to the public. Hence, the purpose of this study was launched to describe ethnoveterinary As a result, the Omusati and Kunene peoples have created their own treatment method for most diseases that plague their domestic animals. As a result, these two regions constitute an excellent research model for documenting and distributing EVM knowledge to assist and exchange information to improve basic animal healthcare. The most frequent livestock in the research region is cattle, goats, and sheep. RESULTS The varieties of plants used for EVM, as well as the preparation and treatment techniques for EVM, were all gathered. All types of EVM plants were also collected for identiﬁcation and documentation. Information such as habitat data, a general description of the plant, the collection’s geographical location, and the collector’s initials were recorded during the data collection procedure. For taxonomic identiﬁcation, Data Analysis y Codes, themes, and indicators were used to analyze the qualitative data from open-ended questions that required respondents to articulate themselves. The quantitative data collected were entered on a Microsoft Excel database and analyzed using IBM Statistical Package for the Social Sciences (SPSS) version 27. In SPSS, descriptive statistics, mainly frequencies, were used to analyze categorical data. The data for types of plant species collected were analyzed using literature for identiﬁcation purpose. The voucher specimens were identiﬁed after conferring with experts at the Department of Forestry in Windhoek and studying the Namibian Plants Red Data Book (12). The ﬁdelity level (FL) was used to measure the importance of ethnoveterinary medicinal species for a certain purpose. The following formula was used to calculate the FL index: The Aawambo ethnic group is made up most of the participants, followed by Herero-speaking pastoralists (Ovahimba, Ovatjimba, and Ovazemba), San, and Damara/Nama. Only ﬁve women were interviewed among the hundred EVM experts [Omusati (n = 45 men and 5 females) and Kunene (n = 50 males)], which could be due to men’s dominance in livestock raising in the study area. All the EVM experts interviewed were Christians. Three percent (3%) of participants were between the ages of 41 and 50, followed by 69% of those between 51 and 60 and 28% of those over 60 (>60). Forty-nine percent (49%) of the participants had no formal education, 41% had basic education (primary and secondary education), and 7 and 3% had tertiary education and other educational attainments, respectively. All participants raised cattle, with 86% raising chickens, 83% goats, 4% pigs, and 45% raising other livestock. To collect data for the current study, face-to-face interviews were done utilizing a semi- structured questionnaire. FL(%) = NP/N×100 N is the total number of uses cited for every given species, and NP is the number of cited species for each condition group. N is the total number of uses cited for every given species, and NP is the number of cited species for each condition group. Frontiers in Veterinary Science \| www.frontiersin.org Sampling Procedures and Data Collection Sampling Procedures and Data Collection Despite EVM’s popularity among rural communities in Omusati and Kunene, the number of farmers who use it in either region has not been formally veriﬁed. As a result, the researchers had no notion who the study population (N) was before data February 2022 \| Volume 9 \| Article 762771 Frontiers in Veterinary Science \| www.frontiersin.org 2 Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. Eiki et al. FIGURE 1 \| Map of the study area. FIGURE 1 \| Map of the study area. collection. Snowball sampling was utilized as a result, with key informants included. With the cooperation of extension oﬃcials, farmers, and community leaders in both regions, 100 farmers with EVM expertise were identiﬁed. Fifty (50) EVM experts were therefore identiﬁed in the Omusati region. Despite the diﬃculty in obtaining the requisite sample size due to geographical limits, the researcher also chose 50 EVM specialists in the Kunene region. Therefore, 50 EVM experts were chosen from each region (Omusati and Kunene). plant specimens were taken from the vegetative component, leaves, ﬂoral, fruiting, and/or seed sections, as appropriate. The specimens were labeled with their vernacular names and transported in plastic bags to avoid drying. Plant Families Used as Ethnoveterinary Medicine According to the ﬁndings, 15 medicinal herbs were utilized for EVM in the Omusati and Kunene regions (see Table 1). The therapeutic plants identiﬁed belonged to 10 diﬀerent plant families, with Fabaceae being the most common, followed by February 2022 \| Volume 9 \| Article 762771 3 ber *Habit #Habitat †Parts used Preparation used Disease treated ∧FL (%) Application method S W L Fresh leaves infusion Skin infections and 90.0 Topical coughs 28.0 Oral H W R Root powder Anthrax 18.0 Oral T W L Leaf Infusion Unthriftiness in poultry 56.0 Oral E T W S & B Stem/bark infusion Constipation, diarrhea, and colic 54.0 Oral E T W L/R/St Leaf powder decoction Eye infection 6.0 Topical Diarrhea 9.0 Oral E B/T W T/L/R Decoction Helminths and lung and liver infections 31.0 Topical T W B Bark infusion Retained placenta 26.0 Oral E T W B Branch infusion Retained placenta 22.0 Oral T W R Root infusion Coughs 34.0 Topical Eye inﬂammation 26.0 Topical E T W B and L Bark infusion and decoction of leaves Diarrhea 14.0 Oral S/T W L and R decoction Retained placenta 44.0 Oral E T W L Leaf infusion Eye infections 26.0 Topical R and B Roots/bark powder Wounds 50.0 Topical T T B Bark infusion Calf weakness 48.0 Oral E T W L Leaf paste Diarrhea 17.0 Oral Mastitis 9.0 Topical T W B and St Bark/stem infusion Skin infections 45.0 Topical ant habitat: W, wild; T, terrestrial ∧FL, ﬁdelity level. Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. FIGURE 2 \| Plant families and number of plant species used in EVM in Omusati and Kunene regions of Namibia. FIGURE 3 \| EVM preparation methods used in Kunene and Omusati regions, Namibia (n = 100). FIGURE 2 \| Plant families and number of plant species used in EVM in Omusati and Kunene regions of Namibia. FIGURE 2 \| Plant families and number of plant species used in EVM in Omusati and Kunene regions of Namibia. FIGURE 3 \| EVM preparation methods used in Kunene and Omusati regions, Namibia (n = 100). FIGURE 3 \| EVM preparation methods used in Kunene and Omusati regions, Namibia (n = 100). Combretaceae and Rhamnaceae (see Figure 2). It demonstrates that in the research region, livestock diseases are treated with a variety of plant species. in airtight containers. Frontiers in Veterinary Science \| www.frontiersin.org DISCUSSION TABLE 2 \| status of EVM plant species. In the Kunene region, all EVM practitioners were men (100%), whereas in Omusati, 45% of EVM practitioners were men and 5% were women. Similarly, EVM is mostly performed by men in Pakistan’s Kohat district, according to (13). This observation was attributed by Taliq et al. (13) to the fact that men are more favored in a shift of knowledge, while women are consigned to family care in most cultures. The fact that males made up the majority of ethnoveterinary practitioners in the study area contradicts (14), who found that female herbal practitioners predominate in Northwest (NW) Yunnan, China. The current research identiﬁed 15 plant species from 10 families that have veterinary use. However, this number is signiﬁcantly less than the 46 species reported by (8) in Southern Ethiopia and the 31 plant species reported by (20) in north central Nigeria. Most of the species in EVM came from the Fabaceae family, which contributed four species. The dominance of the Fabaceae family shown in this study is consistent with previous research (23–26). The outcomes of the study revealed that most respondents had a lower level of education, supporting the idea that less educated people are less acculturated and hence have a greater understanding of conventional remedies (15). This situation is unsustainable since it may be diﬃcult for less educated persons to document the adoption of such methods for future reference. Ethnoveterinary practitioners require some education to acquire specialized skills that may be valuable, such as natural resource conservation and management. It is critical that educated members of the community support ethnoveterinary medicines in the future. Ethnoveterinary practitioners make EVM medicines with components from trees, bushes, and herbs. This could be owing to their relative abundance in the research region compared to other habits. The current study’s ﬁndings, however, contrast with those of (24) who reported shrubs and climbers as suppliers of EVM materials in their study. The respondents gathered medicinal substances from the wild, except for Berchemia discolor (Klotzsch) Hemsl., which has a terrestrial habitat. Collecting medicinal materials from the wild is problematic, according to (25), because two-thirds of the world’s 50,000 medicinal plants are harvested from the wild, and one-ﬁfth of them are currently threatened. Plant Families Used as Ethnoveterinary Medicine The recovery time for majority of EVM treatments was 2–3 days as shown by 59% of the respondents, followed by 4–5 days (41%). Trees, shrubs, and herbs were among the ethnoveterinary medicinal plant species. Most of the therapeutic plants were found in the wild, with only a few species found on farms (terrestrially). There were no medicinal plant species cultivated. The most part of the plant used for EVM was leaves (71%), followed by bark (14%), stem (8%), and root (7%). With regards to preparation methods for EVM, crushing was mostly utilized (89.3%), followed by boiling (72.3%), juice (51.2%), drying (35.1%), and latex (11.4%) (Figure 3). FL values ranged from 6 to 90%. Aloe esculenta leach had the greatest FL value (90%), followed by A. littoralis Baker in second place (56%), Combretum collinum Fresen in third place (54%), and Ximenia americana L. in fourth place (50%) (Table 1). Moreover, Combretum imberbe Wawra had a lowest FL value of 6%. Generally, the results showed that most plants had an FL value of <50%. Status of Ethnoveterinary Medicinal Plants Majority of respondents (59%) feel that medicinal plant populations are dwindling, while 35% believe that species are still scarce and 6% believe that nothing has changed (Table 2). The therapeutic plant species are declining, according to the respondents, because they are harder to get by than in previous years. The most prevalent treatment for EVM was for wounds (22%), followed by dermatological issues (20%) and parasitic diseases (14%). Moreover, EVM was also used to treat respiratory diseases (12%), ocular infections (11%), gastrointestinal diseases (7%), and infectious (anthrax and mastitis) diseases (4%) (Figure 4). The majority of the EVM plant materials were utilized fresh, while dried plants were pulverized into a powder and stored February 2022 \| Volume 9 \| Article 762771 5 Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. FIGURE 4 \| Proportion of veterinary diseases treated with EVM in Omusati and Kunene regions of Namibia (n = 100). FIGURE 4 \| Proportion of veterinary diseases treated with EVM in Omusati and Kunene regions of Namibia (n = 100). TABLE 2 \| status of EVM plant species. Items Regional counts (%) Combination (%) Omusati Kunene Declining 29 30 59 Sparsely available 19 16 35 No change 2 4 6 Frontiers in Veterinary Science \| www.frontiersin.org ETHICS STATEMENT The studies involving human participants were reviewed, approved, and conducted in accordance with the University of South Africa’s (UNISA) Ethics code for people’s participation and plant specimens in research with ethics reference number 2020/CAES_HREC/025.. The patients/participants provided their written informed consent to participate in this study. EVM practitioners noted that Salvadora persica L., B. albitrunca Burch., F. angustifolia K. Schum, and G. ﬂavescens have become scarce in the Omusati and Kunene regions, and some of them are no longer found in their natural habitats. It is unclear whether these four species were simply uncommon or were kept secret by EVM practitioners. More research is needed, particularly ecological investigations in areas of Omusati and Kunene that were inaccessible during this study. To maintain long-term supplies, the conservation status of these species must be reviewed further. CONCLUSION As a result, 15 species having veterinary use have been identiﬁed. Plant parts, the methods of preparation, and the sources of such plants were all presented in detail. Because ethnoveterinary medicinal plants are not owned by individual farmers, they may not be properly conserved in the wild, and as a result, they may be lost to deforestation and overexploitation. Therefore, the cultivation of awareness among ethnoveterinary practitioners is one of the most critical approaches for the preservation of these indigenous medicinal plant species. Plants with a high FL should be investigated further for phytochemical analysis and pharmacological activity. In Omusati and Kunene, ethnoveterinary medicinal prescriptions should be examined further to see whether there are any active ingredient interactions and what clinical implications they have. DATA AVAILABILITY STATEMENT The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s. High ﬁdelity (FL) values are highly important in the selection of speciﬁc plants for further search of bioactive chemicals, according to (34). The maximum FL is always achieved by widely utilized medicinal plant species. Diﬀerent plants, such as A. esculenta Leach, A. littoralis Baker, C. collinum Fresen, and X. americana L., scored the highest ﬁdelity values and should be subjected to further phytochemical and pharmacological investigation to prove their medicinal eﬃcacy, according to the current study. DISCUSSION The results also demonstrate that most medicines (89.34%) were manufactured by crushing, a process that is widely used in the preparation of remedies throughout Africa (28–31). It is possible that the ease with which crushing is used in the production of remedies has something to do with it. Oral and topical administration were the only modes of administration described in this study. This corresponds to the results of a study done by (17). physiologically active secondary metabolites that the plant uses to protect itself against herbivores, pests, and diseases (20). Moreover, the fact that leaves were the most used plant parts in this study is a more viable approach than using roots or whole plant, which can endanger the plants’ existence (27). However, the ﬁndings of this study contrast those of a study conducted by (24) in which roots were reported to be the most often-requested resources. The results also demonstrate that most medicines (89.34%) were manufactured by crushing, a process that is widely used in the preparation of remedies throughout Africa (28–31). It is possible that the ease with which crushing is used in the production of remedies has something to do with it. Oral and topical administration were the only modes of administration described in this study. This corresponds to the results of a study done by (17). When it came to respondents’ knowledge of diminishing medicinal plants, the majority of EVM practitioners were aware that wild medicinal plants are dwindling. Some EVM practitioners, however, stated that medicinal plants are not endangered if it rains. This could indicate that rainfall plays an essential role in the survival of plants rather than exploitation. To ensure their protection, future wider usage, and the preservation of information about their use, it is necessary to properly identify the most endangered and threatened species in Namibia’s Omusati and Kunene districts. As a result, the ﬁndings of this study can be used to develop threatened species conservation strategies. DISCUSSION The results of the current study will be critical in raising the knowledge of optimal propagation procedures for these plant species to ensure their long-term survival for sustainable use. The present study showed that ethnoveterinary medicines were used more by people in the range of 51–60 years old than by the younger generation. These ﬁndings corroborate the ﬁndings of other researchers who found that knowledge of ethnoveterinary medicine is primarily limited to the elderly in communities (16–19). This narrative where the elderly people are the guardian of ethnoveterinary medicine and at the forefront of its use presents a risk to EVM (20) contended that it is not sustainable to place EVM knowledge in the elderly because it is susceptible to and maybe threatened by death. As a result, immediate documenting of ethnoveterinary medicine practices in Namibia’s Omusati and Kunene areas is essential to preserve such knowledge before it is lost forever. Leaves (71%) were the most used plant materials, followed by bark (14%), stem (8%), and root (7%). This can be explained by the fact that leaves, unlike other parts, such as subterranean organs and plant exudates, take less work to collect when compared to underground parts, stems, bark, and complete plants (26). Many arguments may be made for why leaves are preferred in EVM. They are the part of the plant where photosynthesis takes place, and they contain many The majority of ethnoveterinary practitioners learned EVM from their parents or grandparents, according to the ﬁndings of this study. These ﬁndings agreed with those of (21–23). This is unsustainable because undocumented information may be lost when these individuals pass away. The goal of this survey was to gather and preserve EVM knowledge before it becomes obsolete. February 2022 \| Volume 9 \| Article 762771 6 Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. physiologically active secondary metabolites that the plant uses to protect itself against herbivores, pests, and diseases (20). Moreover, the fact that leaves were the most used plant parts in this study is a more viable approach than using roots or whole plant, which can endanger the plants’ existence (27). However, the ﬁndings of this study contrast those of a study conducted by (24) in which roots were reported to be the most often-requested resources. CONCLUSION done by (17). The therapeutic indication of ethnoveterinary remedies in the current study area covered all livestock species. However, EVM were more used for cattle, goat, and chicken diseases. This discrepancy may be due to the richness and how livestock species are valued in the study area rather than the therapeutic range of medicinal plants themselves. Most therapies were used for the treatment of wounds, followed by a gastrointestinal disease characterized by diarrhea. Three plants (Acacia nilotica, A. erioloba I, and Grewia ﬂavescens Juss.) from the same family were used in the treatment of retained placenta. Ziziphus mucronate and A. karroo were used in the treatment of diarrhea. Z. mucronate was also used in the treatment of mastitis. These ﬁndings agree with a study conducted in Ethiopia by (32). EVM practitioners reported using roots from Fockea angustifolia in treating cattle suﬀering from anthrax. However, a study conducted by (18) noted the use of fresh roots from F. angustifolia in drawing out poison to snakebites and stings. X. americana, and C. imberbe were used to relieve eye infections in cattle, goats, and sheep. The use of X. americana and C. imberbe was also reported by (8, 17, 33). Leaves from Boscia albitrunca were used as a cold infusion for treating inﬂamed eyes of cattle. The recovery time for majority of EVM treatments in the current study was 2–3 days. Similar ﬁndings were also reported by (8, 33). Farmers in the Omusati and Kunene regions, according to this study, have sound ethnoveterinary knowledge and practices. As a result, 15 species having veterinary use have been identiﬁed. Plant parts, the methods of preparation, and the sources of such plants were all presented in detail. Because ethnoveterinary medicinal plants are not owned by individual farmers, they may not be properly conserved in the wild, and as a result, they may be lost to deforestation and overexploitation. Therefore, the cultivation of awareness among ethnoveterinary practitioners is one of the most critical approaches for the preservation of these indigenous medicinal plant species. Plants with a high FL should be investigated further for phytochemical analysis and pharmacological activity. In Omusati and Kunene, ethnoveterinary medicinal prescriptions should be examined further to see whether there are any active ingredient interactions and what clinical implications they have. Farmers in the Omusati and Kunene regions, according to this study, have sound ethnoveterinary knowledge and practices. REFERENCES Ethnoveterinary study of medicinal plants in a tribal society of Sulaiman range. Sci World J. (2014) 170:275–83. doi: 10.1155/2014/127 526 33. Oﬃah NV, Dawurung CJ, Oladipo OO, Makoshi MS, Makama S, Elisha IL, et al. Survey of herbal remedies used by Fulani herdsmen in the management of animal diarrhoea in Plateau State, Nigeria. J Med Plant Res. (2012) 6:4625–32. doi: 10.5897/JMPR1 1.1301 14. Shen S, Qian J, Ren J. Ethnoveterinary plant remedies used by NU people in NW Yunnan of China. J Ethnobiol Ethnomed. (2010) 6:24. doi: 10.1186/1746-4269-6-24 34. Ahmed MJ, Murtaza G. A study of medicinal plants used as ethnoveterinary: harnessing potential phytotherapy in Bheri, district Muzaﬀarabad (Pakistan). J Ethnopharmacol. (2015) 159:209–14. doi: 10.1016/j.jep.2014.11.016 34. Ahmed MJ, Murtaza G. A study of medicinal plants used as ethnoveterinary: harnessing potential phytotherapy in Bheri, district Muzaﬀarabad (Pakistan). 15. Tesfaye M, Erena MG. Indigenous ethnozoological and ethnoveterinary medicinal practices in leka dullecha district, Western Ethiopia. Glob Vet. (2020) 22:286–97. doi: 10.2337/dc10-1303 Conﬂict of Interest: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. 16. Bartha SG, Quave CL, Balogh L, Papp, N. Ethnoveterinary practices of covasna county, transylvania, Romania. Ethnobiol Ethnomed. (2015) 11:35. doi: 10.1186/s13002-015-0020-8 17. Ahmad SUHEEL, Gangoo SA, Sultan SM, Dar MD. Ethnoveterinary practices and use of herbal medicine by pastoralists of Himalaya: a review. Annals of biology. (2016) 32:260–8. doi: 10.21010/ajtcam.v13i5.19.145 17. Ahmad SUHEEL, Gangoo SA, Sultan SM, Dar MD. Ethnoveterinary practices and use of herbal medicine by pastoralists of Himalaya: a review. Annals of biology. (2016) 32:260–8. doi: 10.21010/ajtcam.v13i5.19.145 Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their aﬃliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. 18. Jambwa P, Nyahangare ET. Ethnoveterinary Medicine: a zimbabwean perspective. In Ethnovet Med. (2020) 12:269– 83. doi: 10.1007/978-3-030-32270-0_12 18. Jambwa P, Nyahangare ET. Ethnoveterinary Medicine: a zimbabwean perspective. In Ethnovet Med. (2020) 12:269– 83. doi: 10.1007/978-3-030-32270-0_12 19. Menale B, Muoio R. Use of medicinal plants in the southeastern area of the partenio regional park (Campania, Southern Italy). J Ethnopharmacol. (2014) 153:297–307. REFERENCES relationship with traditional human herbal medicine. J Ethnopharmacol. (2019) 244:112164. doi: 10.1016/j.jep.2019.112164 relationship with traditional human herbal medicine. J Ethnopharmacol. (2019) 244:112164. doi: 10.1016/j.jep.2019.112164 1. Herrero M, Grace D, Njuki J, Johnson N, Enahoro D, Silvestri S, et al. The roles of livestock in developing countries. Animal. (2013) 7(Suppl. 1):3– 18. doi: 10.1017/S1751731112001954 1. Herrero M, Grace D, Njuki J, Johnson N, Enahoro D, Silvestri S, et al. The roles of livestock in developing countries. Animal. (2013) 7(Suppl. 1):3– 18. doi: 10.1017/S1751731112001954 22. Kidane B, Van Der Maesen LJG, van Andel T, Asfaw Z. Ethnoveterinary medicinal plants used by the maale and ari ethnic communities in southern Ethiopia. J Ethnopharmacol. (2014) 153:274–82. doi: 10.1016/j.jep.2014.02.031 2. Banda LJ, Tanganyika J. Livestock provide more than food in smallholder production systems of developing countries. Animal Front. (2021) 11:7– 14. doi: 10.1093/af/vfab001 23. Markos S. Ethnoveterinary medicine and common diseases of chicken producers in western zone of tigray, northern Ethiopia. J Agric Ecol. (2019) 1–17. doi: 10.9734/jaeri/2019/v20i130096 2. Banda LJ, Tanganyika J. Livestock provide more than food in smallholder production systems of developing countries. Animal Front. (2021) 11:7– 14. doi: 10.1093/af/vfab001 24. Birhanu T, Abera D. Survey of ethno-veterinary medicinal plants at selected horro gudurru districts, Western Ethiopia. Afr J Plant Sci. (2015) 9:185– 92. doi: 10.5897/AJPS2014.1229 3. Swanepoel FJC, Stroebel A, Moyo S. The Role of Livestock in Developing Communities: Enhancing Multifunctionality. Free State: University of the Free State/CTA (2010). 25. Tasheva K, Kosturkova G. Role of biotechnology for protection of endangered medicinal plants. Environ Biotechnol N Appro Prospect Applic. (2013) 235– 8. doi: 10.5772/55024 4. Togarepi C, Thomas B, Kankono M. Cattle marketing constraints and opportunities in north-central communal areas of namibia, ohangwena region. Livestock Res Rural Dev. (2016) 28. doi: 10.1007/s11250-020-02468-3 5. Rich KM, Perry BD. The economic and poverty impacts of animal diseases in developing countries: new roles, new demands for economics and epidemiology. Prev Vet Med. (2011) 101:133–47. doi: 10.1016/j.prevetmed.2010.08.002 5. Rich KM, Perry BD. The economic and poverty impacts of animal diseases in developing countries: new roles, new demands for economics and epidemiology. Prev Vet Med. (2011) 101:133–47. doi: 10.1016/j.prevetmed.2010.08.002 26. Bhatt A, Singh P, Kumar V, Baunthiya M. Documentation of ethnoveterinary practices used for treatments of diﬀerent ailments in Garhwal Himalayan Region. J Environ Nanotechnol. (2013) 2:22–9. doi: 10.13074/jent.2013.02.nciset34 j p 6. Aziz MA, Khan AH, Adnan M, Ullah H. REFERENCES Traditional uses of medicinal plants used by indigenous communities for veterinary practices at Bajaur Agency, Pakistan. J. Ethnobiol Ethnomed. (2018) 14:1–18. doi: 10.1186/s13002-018-0212-0 27. Lulekal E, Asfaw Z, Kelbessa E, Van Damme P. Ethnoveterinary plants of ankober district, north shewa zone, amhara region, Ethiopia. J Ethnobiol Ethnomed. (2014) 10:21. doi: 10.1186/1746-4269-10-21 28. Gabalebatse M, Ngwenya BN, Teketay D, Kolawole OD. Ethno- veterinary practices amongst livestock farmers in Ngamiland District, Botswana. Afr J Tradit Complement Altern Med. (2013) 10:490–502. doi: 10.4314/ajtcam.v10i3.16 7. Wanzala W, Zessinb KH, Kyule NM, Baumann MPO, Mathiasd E, Hassanali A. Ethnoveterinary medicine: a critical review of its evolution, perception, understanding and the way. Livest Res Rural Dev. (2005) 17:11. 29. Mthi S, Rust JM, Morgenthal TL, Moyo B. An ethno-veterinary survey of medicinal plants used to treat bacterial diseases of livestock in three geographical areas of the Eastern Cape Province, South Africa. J Med Plant Res. (2018) 12:240–7. doi: 10.5897/JMPR2017.6444 8. Dinbiso TD, Tolosa TT, Begna FD. Ethnoveterinary Practices of Medicinal Plants and Non-plant Remedies Used in Animal Health Management in Dawuro Zone, Southern Ethiopia. 9. Namibia Statistics Agency (NSA). Namibia Inter-Censal Demographic Survey 2016 Report. Windhoek: NSA (2016). 30. Giday M, Asfaw Z, Woldu Z. Ethnomedicinal study of plants used by Sheko ethnic group of Ethiopia. J Ethnopharmacol. (2010) 132:75– 85. doi: 10.1016/j.jep.2010.07.046 10. Ambunda L, de Klerk S. Women custom in Namibia. A research overview. In Ruppel OC, editor Windhoek, Namibia. Windhoek: Macmillan Education Publishers (2008). 31. Luseba D, Tshisikhawe M. Medicinal plants used in the treatment of livestock diseases in vhembe region, Limpopo province, South Africa. J Med Plants Res. (2013) 7:593–601. doi: 10.5897/JMPR012.1213 11. Haakuria VM, Pyatt AZ, Mansbridge SC. Exploration of veterinary service supply to rural farmers in Namibia: a one health perspective. PAMJ One Health. (2020) 2:17. doi: 10.11604/pamj-oh.2020.2.17.24658 32. Temeche MA, Asnakew AT. A review on status of ethnoveterinary medicine and challenges it faces in Ethiopia. Int J Vet Sci Anim Husb. (2020) 5:39–48. doi: 10.4172/2161-0525.1000215 12. Loots, Sonja. Red Data Book of Namibian Plants. Vol. 38. Pretoria: Sabonet (2005). 33. Oﬃah NV, Dawurung CJ, Oladipo OO, Makoshi MS, Makama S, Elisha IL, et al. Survey of herbal remedies used by Fulani herdsmen in the management of animal diarrhoea in Plateau State, Nigeria. J Med Plant Res. (2012) 6:4625–32. doi: 10.5897/JMPR1 1.1301 13. Tariq A, Mussarat S, Adnan M, AbdElsalam NM, Ullah R, Khan AL. AUTHOR CONTRIBUTIONS NE conceived and designed the study, conducted the survey, and ﬁnalized the submitted version. MaM, SL, BS, NS, and MoM were involved in data analysis and the drafting of manuscripts. All authors contributed to the article and approved the submitted version. February 2022 \| Volume 9 \| Article 762771 Frontiers in Veterinary Science \| www.frontiersin.org 7 Survey of Ethnoveterinary Medicines Used to Treat Livestock Diseases Eiki et al. REFERENCES doi: 10.1016/j.jep.2014.02.039 Copyright © 2022 Eiki, Maake, Lebelo, Sakong, Sebola and Mabelebele. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. 20. Adeniran LA, Okpi S, Anjorin TS, Ajagbonna OP. Medicinal plants used in ethnoveterinary practices in the federal capital territory, north-central Nigeria. J Med Plant Res. (2020) 14:377–88. doi: 10.5897/JMPR2020.6975 21. Miara MD, Bendif H, Ouabed A, Rebbas K, Hammou MA, Amirat M, et al. Ethnoveterinary remedies used in the Algerian steppe: exploring the 21. Miara MD, Bendif H, Ouabed A, Rebbas K, Hammou MA, Amirat M, et al. Ethnoveterinary remedies used in the Algerian steppe: exploring the February 2022 \| Volume 9 \| Article 762771 Frontiers in Veterinary Science \| www.frontiersin.org 8
https://openalex.org/W4308442304	https://qmro.qmul.ac.uk/xmlui/bitstream/123456789/82533/2/He%20A%20Town-Level%20Comprehensive%20Intervention%20Study%20to%20Reduce%20Salt%20Intake%20in%20China%3a%20Cluster%20Randomized%20Controlled%20Trial%202022%20Published.pdf	English	null	A Town-Level Comprehensive Intervention Study to Reduce Salt Intake in China: Cluster Randomized Controlled Trial	Nutrients	2,022	cc-by	9,166	† These authors contributed equally to this work. Citation: Liu, M.; Xu, J.; Li, Y.; He, F.J.; Zhang, P.; Song, J.; Gao, Y.; Yan, S.; Yan, W.; Jin, D.; et al. A Town-Level Comprehensive Intervention Study to Reduce Salt Intake in China: Cluster Randomized Controlled Trial. Nutrients 2022, 14, 4698. https://doi.org/10.3390/ nu14214698 Academic Editor: Josep A. Tur Received: 2 September 2022 Accepted: 2 November 2022 Published: 7 November 2022 Citation: Liu, M.; Xu, J.; Li, Y.; He, F.J.; Zhang, P.; Song, J.; Gao, Y.; Yan, S.; Yan, W.; Jin, D.; et al. A Town-Level Comprehensive Intervention Study to Reduce Salt Intake in China: Cluster Randomized Controlled Trial. Nutrients 2022, 14, 4698. https://doi.org/10.3390/ nu14214698 Academic Editor: Josep A. Tur Received: 2 September 2022 Accepted: 2 November 2022 Published: 7 November 2022 Citation: Liu, M.; Xu, J.; Li, Y.; He, F.J.; Zhang, P.; Song, J.; Gao, Y.; Yan, S.; Yan, W.; Jin, D.; et al. A Town-Level Comprehensive Intervention Study to Reduce Salt Intake in China: Cluster Randomized Controlled Trial. Nutrients 2022, 14, 4698. https://doi.org/10.3390/ nu14214698 Abstract: We determined whether a town-level comprehensive intervention program could lower the salt intake of a population. The parallel, cluster randomized controlled trial was carried out between October 2018 and January 2020 in 48 towns from 12 counties across 6 provinces in China. All participants were asked to complete the 24 h urine collections, anthropometric measurements and questionnaires at the baseline and one-year post-intervention survey. A total of 2693 participants aged 18 to 75 years were recruited at the baseline. A total of 1347 individuals in 24 towns were allocated to the intervention group and the others were allocated to the control group. Valid information from 2335 respondents was collected in the follow-up survey. The 24-h urinary sodium excretion was 189.7 mmol/24 h for the intervention group and 196.1 mmol/24 h for the control group at baseline. At a one-year follow-up, the mean effect of salt intake did not show a signiﬁcant change (p = 0.31) in the intervention group compared to the control group. However, the mean result of potassium excretion in the intervention group increased by 2.18 mmol/24 h (85.03 mg/24 h) (p = 0.004) and systolic blood pressure decreased by 2.95 mmHg (p < 0.001). The salt-related knowledge and attitude toward salt reduction improved signiﬁcantly in the intervention group (p < 0.05). A longer period of intervention and follow-up assessment might be needed to evaluate the long-term effectiveness of the program on salt reduction. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Keywords: twenty-four hour urinary sodium; salt reduction; randomized trial; sodium; China Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Article A Town-Level Comprehensive Intervention Study to Reduce Salt Intake in China: Cluster Randomized Controlled Trial Min Liu 1,† , Jianwei Xu 1,†, Yuan Li 2 , Feng J He 3 , Puhong Zhang 2 , Jing Song 3, Yifu Gao 4, Shichun Yan 5 Wei Yan 6, Donghui Jin 7, Xiaoyu Chang 8, Zhihua Xu 9, Yamin Bai 1, Ning Ji 1 and Jing Wu 1,* Min Liu 1,† , Jianwei Xu 1,†, Yuan Li 2 , Feng J He 3 , Puhong Zhang 2 , Jing Song 3, Yifu Gao 4, Shichun Yan 5, Wei Yan 6, Donghui Jin 7, Xiaoyu Chang 8, Zhihua Xu 9, Yamin Bai 1, Ning Ji 1 and Jing Wu 1,* 1 National Center for Chronic and Non-Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention (China CDC), Beijing 100050, China Disease Control and Prevention (China CDC), Beijing 100050, China 2 George Institute for Global Health, Peking University Health Science Center, Beijing 100600, China 3 Wolfson Institute of Population Health, Barts and The London School of Medicine & Dentistry, 2 George Institute for Global Health, Peking University Health Science Center, Beijing 100600, Ch 3 2 George Institute for Global Health, Peking University Health Science Center, Beijing 100600, China 3 Wolfson Institute of Population Health, Barts and The London School of Medicine & Dentistry, 3 Wolfson Institute of Population Health, Barts and The Londo Queen Mary University of London, London E1 4NS, UK Queen Mary University of London, London E1 4NS, UK 4 Department for Chronic and Non-Communicable Disease Control and Prevention, Hebei Provincial Center for Disease Control and Prevention, Shijiazhuang 050024, China j g 5 Department for Chronic and Non-Communicable Disease Control and Prevention, Heilongjiang Provincial Center for Disease Control and Prevention Harbin 150030 China j g 5 Department for Chronic and Non-Communicable Disease Control and Prevention, Heilongjiang Provincial j g 5 Department for Chronic and Non-Communicable Disease Control and Prev Center for Disease Control and Prevention, Harbin 150030, China p gj g Center for Disease Control and Prevention, Harbin 150030, China nter for Disease Control and Prevention, Harbin 150030, Chi 6 Department for Chronic and Non-Communicable Disease Control and Prevention, Jiangxi Provincial Center for Disease Control and Prevention, Nanchang 330029, China g 7 Department for Chronic and Non-Communicable Disease Control and Prevention, Hunan Provincial Center for Disease Control and Prevention, Changsha 410028, China 8 Department for Chronic and Non-Communicable Disease Control and Prevention, Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610044, China 9 Department for Chronic and Non-Communicable Disease Control a for Disease Control and Prevention, Xining 810007, China 9 Department for Chronic and Non-Communicable Disease Control and Prevention, Qinghai Provincial Center for Disease Control and Prevention, Xining 810007, China g * Correspondence: wujing@chinacdc.cn; Tel.: +86-10-8313-6485 nutrients nutrients nutrients nutrients nutrients nutrients 2.3. Randomization and Masking Towns (clusters) which were stratiﬁed by the province were randomized 1:1 to inter- vention and control groups using a computer-generated randomization sequence. Random- ization took place after the baseline survey. The local participants and investigators were unaware of the assignment until the intervention began. 2. Materials and Methods 2.1. Design A detailed description of the methods has been published elsewhere [16]. The study conducted a cluster RCT in 6 provinces, including Hebei, Jiangxi, Hunan, Sichuan, Hei- longjiang and Qinghai, which accounted for the diversity in geographical distribution, economic level and dietary habits of the Chinese population. In total, 48 towns (clusters) were selected from 12 counties across 6 provinces. Two counties were chosen from each province. From each county, we selected four towns that were similar in development levels and population size and randomly divided them into the intervention group and control group. Twenty-eight eligible participants were selected from each village. In order to mini- mize contamination between the intervention group and control group, most towns were selected from rural and suburban zones where living environments are relatively isolated. g y The baseline survey was carried out between October and December 2018. The one- year intervention procedures started after the baseline investigation was completed and the one-year follow-up evaluation was conducted between November 2019 and January 2020. The program was approved by the Queen Mary Research Ethics Committee (QMERC 2018/16) and the Institutional Review Board of the National Center for Chronic and Noncommunicable Disease Control and Prevention (201807). All participants signed written informed consent forms and they could withdraw from the study at any time. 2.2. Study Participants Individuals aged 18 to 75 years and who had been local residents for over 6 months were eligible for inclusion in the outcome evaluation. Only one person could be selected from each family. We excluded women who were pregnant or in the lactation period, participants who were not suitable for 24 h urine collection, and patients with severe psychiatric and physical diseases. 1. Introduction Studies have shown that excessive salt intake is associated with high blood pressure (BP), which is a major cause of cardiovascular disease (CVD), such as stroke and ischemic heart disease [1–4]. In China, the prevalence rate of hypertension in adults was 27.5% [5], and high blood pressure contributed to 2.33 million CVD deaths [6]. In order to reduce the harm caused by a high-salt diet, the World Health Organization (WHO) recommended that https://www.mdpi.com/journal/nutrients Nutrients 2022, 14, 4698. https://doi.org/10.3390/nu14214698 Nutrients 2022, 14, 4698 2 of 13 adults reduce their salt intake to lower than 5 g/d (87 mmol/d) [7]. However, the average salt intake of Chinese residents was 11.0 g per day per capita in 2020, more than twice the WHO’s recommended amount [8]. Salt reduction is one of the most cost-effective means of preventing cardiovascular disease and it was recommended as one of the best strategies to solve the global crisis in noncommunicable diseases [9]. Actions to reduce salt have been initiated in many countries [9–12]. For China, which is a country with the largest share of the world’s CVDs, more intervention studies on salt reduction are urgently needed. In order to achieve the goal of reducing salt intake, the Action on Salt China (ASC) program was launched in 2017 to implement a series of salt reduction projects targeting various settings (e.g., local restaurants, schools, hospitals and communities) and salt intake sources [13–15]. The Comprehensive Intervention Study (CIS) [16] was a community-based Randomized Controlled Trial (RCT) included in the ASC program [17–19] aimed at evaluating the acceptability, scalability and effectiveness of the comprehensive intervention and its components. In this study, we evaluate the one-year effectiveness of implementing the comprehensive salt reduction intervention, which provides evidence for a national promotion. 2.4. Data Collection The mean of the last two BP measurements was included in the analysis. Height, weight, and outdoor temperature were recorded by qualiﬁed investigators using calibrated equipment. Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m2). Information on salt reduction knowledge, attitude and practice (KAP), and lifestyle factors (e.g., physical activity, and alcohol drinking) were collected in face-to-face question- naires administered by trained researchers. KAP questions related to salt intake recom- mendations, low-sodium salt, the identiﬁcation of sodium content on food labels, dietary tastes and consumption of processed foods. Physical activity was deﬁned as participants self-reporting their participation in moderate physical activity for 30 min or more at least three times a week. Alcohol drinking status was classiﬁed as non-drinker (the respondents answered “no” to the question “do you drink alcohol”), occasional drinker (the respon- dents answered “sometimes”), and regular drinker (the respondents answered “always” or “addicted”). All of the baseline and 12-month data were managed within a mobile electronic data collection system [20]. 2.5. Intervention The intervention package was designed by the CIS national project ofﬁce, and imple- mented by the local Centers for Disease Control and Prevention (CDC). The researchers trained local health educators in a three-day workshop. Detailed intervention workbooks and lecture courses were offered. The salt reduction education included three parts: salt and health; salt reduction target; and how to reduce salt intake and use the salt substitute. The salt substitute courses introduced the beneﬁts of low-sodium salt and how to choose low-sodium salt. The details of the intervention procedures and resources have been described before [16]. To achieve better intervention effects, a multifaceted comprehensive salt reduction strategy of proposed based on the existing evidence from other countries [1,3,11,21], and was implemented by the county, township and village local governments, respectively. Other major stakeholders, such as hospitals, schools, restaurants, and publicity departments were also engaged in the development of the intervention. 2.4. Data Collection All of the participants were asked to complete the 24-h urine collections, anthropo- metric measurements and questionnaires at the baseline and evaluation survey. The 24-h urine samples were collected following the instruction from the trained research staff. Nutrients 2022, 14, 4698 3 of 13 The participants were asked to empty their bladders, note down the time at the start of collecting urine on the ﬁrst day, collect all subsequent urine voids over the next 24-h period, and return the urine collection equipment with all of the 24-h urine samples on the second day with the time of the last urine collection recorded. If the participants reported that they forgot to gather or splashed more than 10% of their total amount of urine, or if blood, excrement, or other impurities had contaminated the urine sample, the 24-h urine samples needed to be collected again following the abovementioned procedures. The urine samples were tested for the concentrates of sodium, potassium and creatinine, and the 24-h urinary excretions of sodium, potassium and creatinine were calculated as the product of urine volume and concentrate. The BP was measured three times by trained researchers using a validated automatic blood pressure monitor (OMRON: HEM-7120) following the standard protocol [16,17]. The mean of the last two BP measurements was included in the analysis. Height, weight, and outdoor temperature were recorded by qualiﬁed investigators using calibrated equipment. Body mass index (BMI) was calculated as weight (kg) divided by the square of height (m2). Information on salt reduction knowledge, attitude and practice (KAP), and lifestyle factors (e.g., physical activity, and alcohol drinking) were collected in face-to-face question- naires administered by trained researchers. KAP questions related to salt intake recom- mendations, low-sodium salt, the identiﬁcation of sodium content on food labels, dietary tastes and consumption of processed foods. Physical activity was deﬁned as participants self-reporting their participation in moderate physical activity for 30 min or more at least three times a week. Alcohol drinking status was classiﬁed as non-drinker (the respondents answered “no” to the question “do you drink alcohol”), occasional drinker (the respon- dents answered “sometimes”), and regular drinker (the respondents answered “always” or “addicted”). All of the baseline and 12-month data were managed within a mobile electronic data collection system [20]. The BP was measured three times by trained researchers using a validated automatic blood pressure monitor (OMRON: HEM-7120) following the standard protocol [16,17]. 2.5.3. Salt Reduction Interventions in Schools At all of the schools in the intervention group, publicity posters were put up on bulletin boards or school canteens. Health education activities related to salt reduction were carried out at least once a year in teacher training and school–parent meetings. Salt reduction education courses were carried out to promote the harm of a high-salt diet to students. Public activities, such as making salt-related handwritten art and other patterns were conducted in schools. 2.6. Outcomes The primary outcome included the difference in the change in 24-h urinary sodium, urinary potassium, the sodium–potassium ratio and BP from the baseline to the end of the trial between the intervention group and the control group. Secondary outcomes were the differences in the changes from KAPs on salt reduction from the baseline to the end of the trial between the intervention group and the control group. 2.5.4. Salt Reduction Interventions in Restaurants 2.5.4. Salt Reduction Interventions in Restaurants Information about salt reduction, health and salt was displayed through posters, videos and table stickers in restaurants to create an environment conducive to salt reduction. The chef and waiters in the restaurants were offered standardized training at least four times a year on how to reduce salt usage during cooking and how to guide customers to choose lower-salt dishes. Lower-salt dishes were marked on the menu to make it easier for the customers to choose. Participants in the control group carried on with the usual health education (Health Action for All, Basic National Public Health Service, and so on) and no additional interven- tions on salt reduction were conducted during the intervention period. 2.5.2. Salt Reduction Interventions in Primary Health Centers 2.5.2. Salt Reduction Interventions in Primary Health Centers 2.5.2. Salt Reduction Interventions in Primary Health Centers Besides public education, CIS emphasized the proactive role of primary health centers in the implementation of the salt reduction intervention. The county-level CDC integrated the training with the National Basic Public Health Service and conducted training at least twice a year for all primary health care providers in the intervention group. Each primary health center held at least two salt reduction lectures and activities every year using stan- dardized teaching materials. Primary health care providers would impart knowledge and tips on salt reduction during routine outpatient visits to improve patients’ KAP toward salt reduction, thereby reducing the burden of hypertension and CVDs attributed to excessive salt intake. In order to achieve a good effect in rural areas, salt reduction knowledge was also publicized in the form of broadcasting among villagers. Salt-restricting spoons (2 g per spoon) for measuring salt during cooking were distributed to the family chef. 2.5.3. Salt Reduction Interventions in Schools 2.5.1. Salt Reduction Campaign Various kinds of salt reduction-related activities were carried out in the intervention group. Community residents were supplied with educational materials including posters, brochures, leaﬂets and signs. Salt reduction videos were broadcasted in public places such as parks and buses. Salt reduction publicity was carried out on at least two publicity days or important holidays every year, such as world salt reduction week and national hypertension day. Mass culture and publicity activities related to “salt and health” were organized at least once a year, such as knowledge competitions, family healthy cooking competitions, or other activities, to create a better salt reduction environment. In order to better cover young and middle-aged people, we also promoted salt and health knowledge and skills through social media, such as WeChat public accounts. Nutrients 2022, 14, 4698 4 of 13 3.1. Baseline Characteristics of Participants A total of 2981 adult participants were recruited in 48 towns (8 towns in each province, 2 communities in each town) from six provinces. 288 respondents were excluded be- cause they did not meet the inclusion criteria (n = 192) or refused to participate (n = 96). 2693 participants completed a baseline assessment. After randomization, 1347 individuals (from 24 towns) were allocated to the intervention group and 1346 persons (from 24 towns) were allocated to the control group. During the trial, 237 (8.8%) persons were lost before the 12-month follow-up evaluation, due to moving to other places, a long time out of work, or being unable to attend the follow-up assessment. After excluding 121 incomplete urine samples in the evaluation survey, the sample size was reduced to 2335 participants. Figure 1 shows the baseline and follow-up numbers for the intervention and control groups. IEW 6 of 14 Figure 1. Flowchart of participants in the baseline and follow-up surveys. Table 1 shows the baseline characteristics of the participants in the control and intervention groups. The mean age of the 2693 participants was 48.0 years, and 49.5% of them were men. The mean BMI was 24.7 kg/m2. The two groups were well balanced in most parameters except age, education status, self-reported hypertension and blood pressure treatment in the self-reported Figure 1. Flowchart of participants in the baseline and follow-up surveys. Table 1 shows the baseline characteristics of the participants in the control and inter- vention groups. The mean age of the 2693 participants was 48.0 years, and 49.5% of them were men. The mean BMI was 24.7 kg/m2. The two groups were well balanced in most Figure 1. Flowchart of participants in the baseline and follow-up surveys. Figure 1. Flowchart of participants in the baseline and follow-up surveys. Table 1 shows the baseline characteristics of the participants in the control and intervention groups. The mean age of the 2693 participants was 48.0 years, and 49.5% of them were men. The mean BMI was 24.7 kg/m2. The two groups were well balanced in most parameters except age, education status, self-reported hypertension and blood pressure treatment in the self-reported Table 1 shows the baseline characteristics of the participants in the control and inter- vention groups. The mean age of the 2693 participants was 48.0 years, and 49.5% of them were men. The mean BMI was 24.7 kg/m2. 2.7. Data Analysis The effect of the town-level comprehensive intervention on the outcomes was analyzed using a general linear mixed model with a random intercept assessing the 3-level clusters (individual-level data were nested at the village level, and the village-level data were nested at the county level). The independent variables included group (control and intervention), time (baseline and 12-month), and time × group interaction. The time × group interaction term means the difference in the change in outcome measurement over the 12 months from the baseline between the intervention group and the control group. Stratiﬁcation variables at randomization (towns) and potential confounding variables including age group (<40 = 1, 40~60 = 2, ≥60 = 3), sex (male = 0, female = 1), education level (primary education or less = 1, secondary school = 2, high school = 3, university or college = 4), BMI, outdoor temperature, physical activity and alcohol drinking status were adjusted in the general linear mixed models. g The Intent-To-Treat (ITT) analyses were used, but possibly incomplete 24-h urine collections were excluded in the primary analyses of urinary outcome measures. We Nutrients 2022, 14, 4698 5 of 13 deﬁned the possibly incomplete 24-h urine samples as urine volume <500 mL/24-h, or creatinine <6.0 mmol/24-h in men or <4.0 mmol/24-h in women. Urine samples with collection times <20 h or >28 h were also excluded. If the 24-h urine samples were deﬁned as incomplete at either baseline or 12 months, we used only the complete samples. In total, we excluded 409 urine collections from 5386 for the primary analyses of urinary outcome measures. To examine the robustness of the conclusions from the primary analysis, we also carried out two sensitivity analyses of urinary outcome measures: (1) in all the participants who attended the two surveys, and (2) in participants who completed both baseline and 12-month assessments (named as completers). ( p ) We used SAS (version 9.4) for data analysis. Continuous variables were described as means and standard deviations (SDs) and presented as mean estimates and 95% conﬁdence intervals (CI) in the inferential analyses. Categorical variables were described as the frequencies and percentages, odds ratios (ORs) and 95% CI in the inferential analyses. The t-test and chi-square test were used to describe the difference in the characteristics of the participants between the intervention group and the control group. All analyses were two-sided, and p < 0.05 was considered signiﬁcant. 3.1. Baseline Characteristics of Participants The two groups were well balanced in most Nutrients 2022, 14, 4698 6 of 13 parameters except age, education status, self-reported hypertension and blood pressure treatment in the self-reported hypertensive. The age, self-reported hypertension and blood pressure treatment were higher in the intervention group and the control group had a higher level of education. Table 1. Participants’ baseline characteristics. Table 1. Participants’ baseline characteristics. Characteristics Control (n = 1347) Intervention (n = 1346) p-Value Age (year), Mean (SD) 47.2 (13.0) 48.8 (12.6) <0.001 Men, n (%) 667 (49.5) 667 (49.6) 0.98 Weight (kg), Mean (SD) 63.3 (12.1) 63.1 (11.4) 0.72 BMI (kg/m2), Mean (SD) 24.7 (3.7) 24.8 (3.5) 0.39 Physical activity: active n (%) 567 (42.1) 519 (38.6) 0.06 Education status (n, %) Primary education or less 540 (40.1) 596 (44.3) 0.02 Secondary school 536 (39.8) 503 (37.4) High school 159 (11.8) 161 (12.0) University or college 112 (8.3) 86 (6.4) Alcohol drinkers a (n, %) 0.82 Non-drinkers 792 (58.8) 800 (59.4) Occasional drinkers 426 (31.7) 415 (30.8) Regular drinkers 128 (9.5) 131 (9.7) Urine creatinine (mmol/24 h) 10.7 (3.3) 10.5 (3.5) 0.13 Self-reported hypertension b (n, %) 233 (17.4) 300 (22.3) <0.001 BP treatment in self-reported hypertensives, (n, %) c 157 (66.8) 247 (82.3) <0.001 a Number of participants with missing value = 1; b Number of participants with missing values = 5; c Number of self-reported hypertensive patients = 533. a Number of participants with missing value = 1; b Number of participants with missing values = 5; c Number of self-reported hypertensive patients = 533. 3.2. Primary Outcome Table 2 shows the covariates-adjusted mixed linear model result of the urinary outcomes and blood pressure. The mean baseline 24-h urinary sodium excretion was 196.1 mmol/24 h (equivalent to 11.5 g/d of salt) in the control group and 189.7 mmol/24 h (equivalent to 11.1 g/d of salt) in the intervention group. After one-year follow-up, sodium excretion does not signiﬁcantly change in either the intervention group or control group (change from baseline in the intervention group: −0.27 mmol/24 h, p = 0.91; change from baseline in the control group: −4.00 mmol/24 h, p = 0.12). Comparing the intervention with the control group, the mean effect on salt intake did not show a signiﬁcant change (p = 0.31). The 24-h urinary potassium excretion decreased in the control group after the one-year follow-up while no changes were observed in the intervention group (change from baseline in the intervention group: 0.53 mmol/24 h, p = 0.32; change from baseline in the control group: −1.65 mmol/24 h, p = 0.002). The comparison of the change in urinary potas- sium excretion between the intervention group and the control group shows a signiﬁcant intervention effect on increasing the urinary potassium (2.18 mmol/24 h, equivalent to 85.03 mg/24 h) (p = 0.004). Table 2 shows the covariates-adjusted mixed linear model result of the urinary outcomes and blood pressure. The mean baseline 24-h urinary sodium excretion was 196.1 mmol/24 h (equivalent to 11.5 g/d of salt) in the control group and 189.7 mmol/24 h (equivalent to 11.1 g/d of salt) in the intervention group. After one-year follow-up, sodium excretion does not signiﬁcantly change in either the intervention group or control group (change from baseline in the intervention group: −0.27 mmol/24 h, p = 0.91; change from baseline in the control group: −4.00 mmol/24 h, p = 0.12). Comparing the intervention with the control group the mean effect on salt intake did not show a signiﬁcant change (p = 0 31) The 24-h urinary potassium excretion decreased in the control group after the one-year follow-up while no changes were observed in the intervention group (change from baseline in the intervention group: 0.53 mmol/24 h, p = 0.32; change from baseline in the control group: −1.65 mmol/24 h, p = 0.002). 3.2. Primary Outcome The comparison of the change in urinary potas- sium excretion between the intervention group and the control group shows a signiﬁcant intervention effect on increasing the urinary potassium (2.18 mmol/24 h, equivalent to 85.03 mg/24 h) (p = 0.004). The 24-h urinary potassium excretion decreased in the control group after the one-year follow-up while no changes were observed in the intervention group (change from baseline in the intervention group: 0.53 mmol/24 h, p = 0.32; change from baseline in the control group: −1.65 mmol/24 h, p = 0.002). The comparison of the change in urinary potas- sium excretion between the intervention group and the control group shows a signiﬁcant intervention effect on increasing the urinary potassium (2.18 mmol/24 h, equivalent to 85.03 mg/24 h) (p = 0.004). Nutrients 2022, 14, 4698 7 of 13 Table 2. Results for 24-h urinary sodium excretion, other urinary measurements, and blood pressure from the covariates-adjusted mixed linear model. Control Intervention Adjust Difference * (Intervention vs. * Model-based change for urinary outcomes were adjusted for age categories, sex, education level and BMI at baseline and follow-up. Model-based changes for blood pressure were further adjusted for outdoor temperature at baseline and follow-up, physical activity and alcohol drinker status. r age categories, sex, education level and BMI at baseline and follow-up. Model-based changes for blood pressure were w-up, physical activity and alcohol drinker status. 3.2. Primary Outcome Control) (95% CI) p-Value Baseline N, Mean (SD) 12-Month N, Mean (SD) Model-Based Change from Baseline * (95% CI) Baseline N, Mean (SD) 12-Month N, Mean (SD) Model-Based Change from Baseline * (95% CI) Salt intake (g/d) 1327 1159 −0.23 1315 1176 −0.02 0.22 0.31 11.5 (4.8) 11.3 (4.7) (−0.53 to 0.06) 11.1 (4.5) 11.1 (4.8) (−0.31 to 0.27) (−0.20 to 0.64) Urinary sodium (mmol/24 h) 1327 1159 −4.00 1315 1176 −0.27 3.72 0.31 196.1 (81.3) 192.6 (80.9) (−9.06 to 1.07) 189.7 (77.0) 189.1 (82.4) (−5.33 to 4.78) (−3.43 to 10.87) Urinary sodium (mg/24 h) 1327 1159 −91.99 1315 1176 4349.8 −6.40 85.58 0.31 4510.6 (187.0) 4430.9 (1860.8) (−208.47 to 24.50) 4363.0 (1770.9) (1894.2) (−122.63 to 109.83) (−78.82 to 249.99) Urinary potassium (mmol/24 h) 1327 1159 −1.65 1315 1176 0.53 2.18 0.004 40.6 (16.8) 39.0 (15.1) (−2.70 to −0.61) 39.6 (16.3) 40.3 (16.3) (−0.52 to 1.57) (0.70 to 3.66) Urinary potassium (mg/24 h) 1327 1159 1519.9 −64.45 1315 1176 20.58 85.03 0.004 1584.6 (654.0) (589.4) (−105.30 to −23.60) 1546.2 (637.0) 1573.4 (634.6) (−20.18 to 61.34) (27.38 to 142.68) sodium-to-potassium ratio 1327 1159 0.09 1315 1176 −0.07 −0.17 0.11 5.2 (2.3) 5.3 (2.3) (−0.05 to 0.24) 5.1 (2.1) 5.0 (2.2) (−0.22 to 0.07) (−0.37 to−0.04) Systolic blood pressure (mm Hg) 1347 1242 0.71 1346 1231 −2.24 −2.95 <0.001 125.6 (19.3) 128.0 (20.0) (−0.21 to 1.64) 127.2 (19.3) 126.8 (18.8) (−3.21 to −1.27) (−4.08 to −1.83) Diastolic blood pressure (mm Hg) 1347 1242 −0.89 1346 1231 −1.34 −0.45 0.22 79.0 (11.8) 79.1 (12.3) (−1.49 to −0.29) 80.1 (11.4) 80.0 (11.5) (−1.97 to −0.71) (−1.17 to 0.27) Urinary creatinine (mmol/24 h) 1327 1159 −0.57 1315 1176 −0.27 0.30 0.01 10.8 (3.2) 10.1 (3.2) (−0.74 to −0.41) 10.6 (3.4) 10.3 (3.4) (−0.44 to −0.11) (0.06 to 0.53) Urine volume (mL/24 h) 1327 1159 −43.01 1315 1176 78.55 121.57 <0.001 1608.3 (643.0) 1565.7 (631.1) (−79.71 to −6.32) 1623.2 (652.4) 1711.9 (699.2) (41.95 to 115.16) (66.79 to 173.34) * Model-based change for urinary outcomes were adjusted for age categories, sex, education level and BMI at baseline and follow-up. Model-based changes for blood pressure were further adjusted for outdoor temperature at baseline and follow-up, physical activity and alcohol drinker status. measurements, and blood pressure from the covariates-adjusted mixed linear model. 3.2. Primary Outcome Nutrients 2022, 14, 4698 8 of 13 8 of 13 After adjusting for the stratiﬁcation variables at randomization and the confounding factors, the systolic blood pressure was decreased in the intervention group from base- line, but does not change in the control group (change from baseline in the intervention group: −2.24 mmHg, p < 0.001; change from baseline in the control group: 0.71 mmHg, p = 0.13), and the intervention was estimated to have lowered the systolic blood pressure by −2.95 mmHg (95%CI: −4.08 mmHg to −1.83 mmHg, p < 0.001) after the one-year follow-up. Comparing the 12-month changes in other outcomes in the intervention group with those estimated in the control group, there was no observed intervention effect on the diastolic blood pressure and sodium-to-potassium ratio after the intervention (p > 0.05), but a signiﬁcant effect in increasing the 24-h urine volume (121.57 mL/24 h, p < 0.001). g g p Supplementary Table S1 shows the results of sensitivity analyses. The results were similar to those from the primary analyses. The mean effect on 24-h urinary sodium excretion was unchanged when the data included possible incomplete 24-h urine collections or only included the participants who completed the baseline and end trial assessment with complete 24-h urine collections. Supplementary Table S2 shows the subgroup results. The 24-h urinary sodium excretion does not signiﬁcantly change in any of the subgroups between the intervention and control groups. 3.3. Secondary Outcomes Table 3 shows the results of the knowledge, attitude and behaviors of salt reduction in the baseline and 12-month survey. Comparing the intervention with the control group, there is a signiﬁcant intervention effect on increasing the proportion of participants with the knowledge of salt intake recommended by Chinese nutrition guidelines after the one-year follow-up (OR = 9.43 (7.28–12.21), p < 0.001), having heard about the low-sodium salt substitute (OR = 2.20 (1.73–2.80), p < 0.001), having the ability to identify the salt content on nutrition labels (OR = 3.50 (2.76–4.43), p < 0.001) and the willingness to choose a low- sodium diet (OR = 1.99 (1.48–2.66), p < 0.001). No signiﬁcant difference was seen in the proportion of participants who prefer a less salty taste and who use a low-sodium salt substitute. After the intervention, the frequency of eating processed foods once per week or less increased in the intervention group (OR = 1.34 (1.14–1.58), p < 0.001)), but there was no signiﬁcant change in the control group (OR = 1.10 (0.94–1.29), p = 0.22). The mean effect of eating processed food frequency does not signiﬁcantly change when comparing the intervention group with the control group (OR = 1.22 (0.97–1.52), p = 0.09). Nutrients 2022, 14, 4698 9 of 13 Table 3. Results for knowledge, attitude and behaviors of salt reduction in baseline and 12-month surveys from the covariates-adjusted mixed linear model. Control Intervention Intervention Effect * (Intervention Group vs. a Questions about using low-sodium salt substitutes were surveyed among people who have heard about low-sodium salt substitutes; * Model-based change for urinary outcomes were adjusted for age categories, sex, education level, and BMI at baseline and follow-up. a Questions about using low-sodium salt substitutes were surveyed among people who have heard about low-sodium s adjusted for age categories, sex, education level, and BMI at baseline and follow-up. 4. Discussion The study was a large-scale town-level comprehensive intervention study designed to reduce salt intake in China. Although the 24-h urinary sodium excretion did not change with the one-year comprehensive intervention, the ﬁndings showed that the interven- tion did increase the 24-h potassium excretion, and signiﬁcantly reduced systolic blood pressure. Sodium-related knowledge and attitude improved signiﬁcantly following the comprehensive intervention. We used 24-h urine collection which was the most accurate method to estimate the salt intake levels in the community adults [22]. The baseline data showed that the 24-h sodium excretion was excessively high (>4300 mg/d sodium or >11.0 g/d salt), whereas the potassium excretion was insufﬁciently low (<1600 mg/d). These ﬁndings were consistent with the result of the latest meta-analysis which demonstrated that the published 24-h urinary sodium and potassium levels in China over the past 40 years were 11.06 g/d and 1.42 g/d, respectively [23], highlighting the importance of reducing the salt intake and increasing the potassium intake in the Chinese population to reduce the disease burden attributed to the high-sodium and low-potassium diet. To tackle this issue, the Chinese government set a target of a 20% reduction in salt consumption in adults by 2030 as the key component of “Health China 2030” [24]. Our study was set up to develop an evidence-based and comprehensive salt reduction intervention to help achieve China’s salt reduction goal. p p g Various regional multifaceted salt reduction programs have been undertaken in China, such as SMASH (Shandong Ministry of Health Action on Salt Reduction and Hypertension) and the Resolve to Save Lives project [25]. Different from the ﬁndings of our study, the SMASH program which was a ﬁve-year intervention to reduce sodium consumption in Shandong province indicated that the government-led and population-based intervention led to a decline in dietary sodium intake, as well as in blood pressure [25]. However, there were some other previous studies on comprehensive salt reduction interventions reporting non-signiﬁcant intervention effects on the population’s salt intake, which were similar to our results [26–28]. The conﬂicting in-study ﬁndings might be related to several reasons. First, the duration of the salt reduction intervention (e.g., one year) might only be enough to improve people’s knowledge and understanding of salt intake, but not sufﬁcient to modify the dietary behavior and taste for foods in the community setting. 3.3. Secondary Outcomes Control Group, OR, 95% CI) p-Value Baseline n (%) 12-Month n (%) Model-Based Change (OR, 95% CI) Baseline n (%) 12-Month n (%) Model-Based Change (OR, 95% CI) Knowledge Knowledge of the salt intake recommended by the Chinese nutrition guidelines (6 g/d) 330 375 1.38 233 889 12.99 9.43 (7.28,12.21) <0.001 (24.5) (30.4) (1.16, 1.64) (17.3) (71.6) (10.73, 15.72) Having heard about a low-sodium salt substitute 342 458 1.81 371 725 3.99 2.20 (1.73, 2.80) <0.001 (25.4) (37.1) (1.52, 2.15) (27.6) (58.4) (3.37, 4.72) Having the ability to identify salt content on a food label 577 505 0.97 504 784 3.40 3.50 (2.76, 4.43) <0.001 (42.8) (40.9) (0.83, 1.15) (37.4) (63.2) (2.87, 4.03) Attitude Willingness to choose a low-sodium diet 1075 991 1.02 1067 1099 2.03 1.99 (1.48, 2.66) <0.001 (79.8) (80.3) (0.84, 1.24) (79.3) (88.6) (1.63, 2.53) Preferring a less salty taste 380 372 1.08 366 369 1.12 1.04 (0.81, 1.32) 0.77 (28.2) (30.2) (0.91, 1.28) (27.2) (29.7) (0.94, 1.33) Behaviors Using a low-sodium salt substitute a 109 133 0.91 130 239 1.01 1.11 (0.74, 1.67) 0.62 (31.9) (29.0) (0.67, 1.24) (35.0) (33.0) (0.77, 1.32) Eating processed food once per week or less 746 714 1.10 788 814 1.34 1.22 (0.97, 1.52) 0.09 (55.4) (57.9) (0.94, 1.29) (58.5) (65.6) (1.14, 1.58) a Questions about using low-sodium salt substitutes were surveyed among people who have heard about low-sodium salt substitutes; * Model-based change for urinary outcomes were adjusted for age categories, sex, education level, and BMI at baseline and follow-up. ors of salt reduction in baseline and 12-month surveys from the covariates-adjusted mixed linear model. n baseline and 12-month surveys from the covariates-adjusted mixed linear model. 10 of 13 Nutrients 2022, 14, 4698 10 of 13 4. Discussion Some of the studies that have successfully reduced salt intake in community populations lasted for more than three years [25,29,30]. This is partially supported by our results of the KAP outcomes that the one- year intervention managed to improve the knowledge and attitudes toward salt reduction, but no intervention effects were observed for either the frequency of having salty processed foods or the use of low-sodium salt substitutes. Compared with knowledge and attitude, behavior change is complex, long-term and slow [31,32]. The one-year intervention time was limited, and it was difﬁcult for the intervention population to transform the knowledge and attitude of salt reduction into behaviors. Secondly, our study showed that the urine sodium excretion in both the intervention group and the control group decreased after one year, which might be partially explained by the ongoing salt reduction initiatives nationwide. For example, the Chinese government launched a nationwide campaign called “Healthy Lifestyle for All”, including calling for the reduction of the population’s salt intake [33]. Thirdly, there were differences in age, education level, prevalence and the treatment of hypertension, especially in the 24-h urinary sodium between the intervention group and the control group at baseline. It might take a long time for the intervention group with an older age group, lower education level, a higher prevalence of hypertension and a lower salt intake to obtain positive effects. In addition to the 24-h urine sodium excretion, our study also showed an increase in urine potassium excretion and a fall in systolic blood pressure after the intervention. The comparison results within each group showed that the urine potassium excretion decreased in the control group and increased in the intervention group after the intervention, but the change in the intervention group was not statistically signiﬁcant. This phenomenon may be related to various factors. Firstly, the time of follow-up surveys between the two Nutrients 2022, 14, 4698 11 of 13 11 of 13 groups was inconsistent. The time of the control group was one month later than the intervention group, which was closer to winter when the intake of fruits and vegetables was less. Many studies have shown that an adequate intake of vegetables and fruits could increase potassium excretion and help lower blood pressure [34,35]. 4. Discussion Secondly, maybe due to the promotion of a healthy diet (including eating more vegetables and fruits and using a salt substitute), the intervention group had a slight increase in urinary potassium. Because of the decrease in urinary sodium and increase in urinary potassium, there was a decrease in systolic blood pressure in the intervention group compared with the control group. After the intervention, there was a signiﬁcant difference in urine volume between the two groups. This may be related to the fact that the health education course included an introduction to the recommended daily water intake per person (no less than 1500 mL). Strengths of our study include the large sample size from six provinces, cluster ran- domized controlled trial design, and comprehensive salt reduction interventions. Another strength is the collection of 24-h urine samples, which was widely acknowledged as the “gold standard” for measuring individual sodium intake [22]. In addition, careful efforts were made to standardize the collection of 24-h urine under the support of a bespoke electronic data capture system. p y This study also had some limitations. First, a single year of intervention is not enough to modify the dietary behaviors in the community setting. Our research will continue to collect follow-up information for the second and third years after the completion of the intervention. Second, only one 24-h urine sample was collected from all participants, which cannot reﬂect the day-to-day variation in sodium and potassium excretion. This may have reduced the stability and reliability of the result. Third, data on the food consumed and the actual use of low-sodium salt substitutes was not collected, thus it would be difﬁcult to understand the reasons for the changes in potassium levels in the study. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: Data are available upon request. Acknowledgments: We thank all local and provincial Centers for Disease Control and Prevention (CDC) in Hebei, Heilongjiang, Jiangxi, Hunan, Sichuan, and Qinghai provinces for their partic- ipation and contribution; the participants who were involved in the project; all members of the project team; the Chinese Center for Health Education for their help with the development of the educational materials. Conﬂicts of Interest: The authors declare no conﬂict of interest. References Global Status Report on Noncommunicable Diseases 2014; World Switzerland, 2015. 7. World Health Organization. Global Status Report on Noncommunicable Diseases 2014; World Health Organization: Geneva, Switzerland, 2015. 8. Li, Y.; Zhang, P.; Wu, J.; Ma, J.; Xu, J.; Zhang, X.; Luo, R.; Liu, M.; Sun, Y.; Li, X.; et al. Twenty-Four-Hour Urinary Sodium and Potassium Excretion and Their Associations with Blood Pressure Among Adults in China: Baseline Survey of Action on Salt China. Hypertension 2020, 76, 1580–1588. [CrossRef] yp , , [ ] 9. Beaglehole, R.; Bonita, R.; Horton, R.; Adams, C.; Alleyne, G.; Asaria, P.; Baugh, V.; Bekedam, H.; Billo, N.; Casswell, S.; et al. Priority actions for the non-communicable disease crisis. Lancet 2011, 377, 1438–1447. [CrossRef] 10. Cobb, L.; Appel, L.; Anderson, C. Strategies to reduce dietary sodium intake. Curr. Treat. Options Cardiovasc. Med. 2012, 14, 425–434. [CrossRef] 11. Trieu, K.; Neal, B.; Hawkes, C.; Dunford, E.; Campbell, N.R.C.; Rodriguez-Fernandez, R.; Legetic, B.; McLaren, L.; Barberio, A.; Webster, J. Salt reduction initiatives around the world-a systematic review of progress towards the global target. PLoS ONE 2015, 10, e0130247. [CrossRef] 12. Chen, X.; Wang, Y. Tracking of blood pressure from childhood to adulthood: A systematic review and meta-regression analysis. Circulation 2008, 117, 3171–3780. [CrossRef] [PubMed] He, F.J.; Zhang, P.; Li, Y.; MacGregor, G.A. Action on 14. Zhang, P.; He, F.J.; Li, Y.; Li, C.; Wu, J.; Ma, J.; Zhang, B.; Wang, H.; Li, Y.; Hanbury, M.; et al. Reducing Salt Intake in China with “Action on Salt China” (ASC): Protocol for Campaigns and Randomized Controlled Trials. JMIR Res. Protoc. 2020, 9, e15933. [CrossRef] [PubMed] 15. He, F.J.; Zhang, P.; Luo, R.; Li, Y.; Sun, Y.; Chen, F.; Zhao, Y.; Zhao, W.; Li, D.; Chen, H.; et al. App based education programme to reduce salt intake (AppSalt) in schoolchildren and their families in China: Parallel, cluster randomized controlled trial. BMJ 2022, 376, e066982. [CrossRef] 16. Xu, J.; Tang, B.; Liu, M.; Bai, Y.; Yan, W.; Zhou, X.; Xu, Z.; He, F.J.; Jin, D.; Sun, J.; et al. A town level com study to reduce salt intake in China: Protocol for a cluster randomised controlled trial. BMJ Open 2020 16. Xu, J.; Tang, B.; Liu, M.; Bai, Y.; Yan, W.; Zhou, X.; Xu, Z.; He, F.J.; Jin, D.; Sun, J.; et al. A town level comprehensive intervention study to reduce salt intake in China: Protocol for a cluster randomised controlled trial. References 1. He, F.J.; Macgregor, G.A. A comprehensive review on salt and health and current experience of worldwide salt reduction programmes. J. Hum. Hypertens. 2009, 23, 363–384. [CrossRef] [PubMed] 1. He, F.J.; Macgregor, G.A. A comprehensive review on salt and health and current experience of worldwide salt reduction programmes. J. Hum. Hypertens. 2009, 23, 363–384. [CrossRef] [PubMed] 2. Wang, C.; Wang, C.; Liu, M.; Chen, Z.; Liu, S. Temporal and spatial trends of ischemic heart disease burden in Chinese and subgroup populations from 1990 to 2016: Data from the 2016 global burden of disease study. BMC Cardiovasc. Disord. 2020, 20, 243. [CrossRef] [PubMed] 2. Wang, C.; Wang, C.; Liu, M.; Chen, Z.; Liu, S. Temporal and spatial trends of ischemic heart disease burden in Chinese and subgroup populations from 1990 to 2016: Data from the 2016 global burden of disease study. BMC Cardiovasc. Disord. 2020, 20, 243. [CrossRef] [PubMed] , [ ] [ ] 3. He, F.J.; Tan, M.; Ma, Y.; MacGregor, G.A. Salt Reduction to Prevent Hypertension and Cardiovascular Disease: JACC State-of-the- Art Review. J. Am. Coll. Cardiol. 2020, 75, 632–647. [CrossRef] [PubMed] 3. He, F.J.; Tan, M.; Ma, Y.; MacGregor, G.A. Salt Reduction to Prevent Hypertension and Cardiovascular Disease: JACC State-of-the- Art Review. J. Am. Coll. Cardiol. 2020, 75, 632–647. [CrossRef] [PubMed] 4. NCD Risk Factor Collaboration. Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: A pooled analysis of 1201 population-representative studies with 104 million participants. Lancet 2021, 398, 957–980. [CrossRef] 4. NCD Risk Factor Collaboration. Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: A pooled analysis of 1201 population-representative studies with 104 million participants. Lancet 2021, 398, 957–980. [CrossRef] 5. Zhang, M.; Wu, J.; Zhang, X.; Hu, C.; Zhao, Z.; Li, C.; Huang, Z.; Zhou, M.; Wang, L. Prevalence and control of hypertension in adults in China, 2018. Chin. J. Epidemiol. 2021, 42, 1780–1789. 5. Zhang, M.; Wu, J.; Zhang, X.; Hu, C.; Zhao, Z.; Li, C.; Huang, Z.; Zhou, M.; Wang, L. Prevalence and control of hypertension in adults in China, 2018. Chin. J. Epidemiol. 2021, 42, 1780–1789. p 6. He, F.J.; Gu, D.; Chen, J.; Wu, X.; Kelly, T.; Huang, J.-F.; Chen, J.-C.; Chen, C.-S.; Bazzano, L.A.; Reynolds, K.; et al. Premature deaths attributable to blood pressure in China: A prospective cohort study. Lancet 2009, 374, 1765–1772. [CrossRef] 7. World Health Organization. 5. Conclusions Although the 24-h urinary sodium excretion level did not change after one year of salt reduction intervention, the population potassium intake was observed to have in- creased and the systolic blood pressure had signiﬁcantly decreased. Furthermore, people’s knowledge and attitudes signiﬁcantly improved over the one-year intervention. A longer period of intervention and follow-up assessment might be needed to evaluate the long-term effectiveness of the intervention program on salt reduction in the communities. In order to achieve the “Health China 2030” target, effective public health policies and targeted interventions for salt reduction are urgently needed. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14214698/s1, Table S1: Sensitivity analysis for 24-h urinary measurements from the covariates-adjusted mixed linear model in the baseline and 12-month surveys; Table S2: Salt intake (g/day) as measured by the 24-h urinary sodium excretion from a subgroup in the baseline and 12-month surveys. Author Contributions: Conceptualization, J.W., Y.L., F.J.H. and P.Z.; methodology, M.L., J.X. and J.S.; formal analysis, M.L. and Y.L.; investigation, M.L., J.X., W.Y., X.C., S.Y., D.J., Z.X., Y.G., Y.B. and N.J.; data curation, M.L., J.X. and Y.L.; writing—original draft preparation, M.L. and J.X.; writing—review and editing, M.L., J.X., Y.L., P.Z., J.W. and F.J.H.; supervision, J.W. and F.J.H.; project administration, J.W., F.J.H. and P.Z.; funding acquisition, F.J.H. and P.Z. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the National Institute of Health Research (NIHR, NIHR Global Health Research Unit Action on Salt China at Queen Mary University of London), grant number 16/136/77. The study is registered with the Chinese Clinical Trial Registry ChiCTR1800018119. Funding: This research was funded by the National Institute of Health Research (NIHR, NIHR Global Health Research Unit Action on Salt China at Queen Mary University of London), grant number 16/136/77. The study is registered with the Chinese Clinical Trial Registry ChiCTR1800018119. Institutional Review Board Statement: The study was conducted in accordance with the Decla- ration of Helsinki and approved by the Ethics Committee of the National Center for Chronic and Noncommunicable Disease Control and Prevention, the Chinese Center for Disease Control and Prevention (201807), and the Queen Mary Research Ethics Committee (QMERC2018/16). Nutrients 2022, 14, 4698 12 of 13 12 of 13 Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Data Availability Statement: Data are available upon request. References BMJ Open 2020, 10, e032976. [CrossRef] 17. Sun, Y.; Luo, R.; Li, Y.; He, F.J.; Tan, M.; MacGregor, G.A.; Lui, H.; Zhang, P. App-Based Salt Reduction Intervention in School Children and Their Families (AppSalt) in China: Protocol for a Mixed Method Process Evaluation. JMIR Res. Protoc. 2021, 10, e19430. [CrossRef] 18. Zhang, X.; Hu, X.; Ma, J.; Zhang, P.; Li, Y.; Luo, R.; He, F.J.; MacGregor, G.A.; Wang, J.; Yin, Z. Cluster randomised controlled trial y J p , , [ ] 17. Sun, Y.; Luo, R.; Li, Y.; He, F.J.; Tan, M.; MacGregor, G.A.; Lui, H.; Zhang, P. App-Based Salt Reduction Intervention in School Children and Their Families (AppSalt) in China: Protocol for a Mixed Method Process Evaluation. JMIR Res. Protoc. 2021, 10, e19430. [CrossRef] 18. Zhang, X.; Hu, X.; Ma, J.; Zhang, P.; Li, Y.; Luo, R.; He, F.J.; MacGregor, G.A.; Wang, J.; Yin, Z. Cluster randomised controlled trial of home cook intervention to reduce salt intake in China: A protocol study. BMJ Open 2020, 10, e033842. [CrossRef] 19. Du, W.; Zhang, J.; Li, Y.; He, F.J.; Zhou, X.; Xu, Z.; Gao, Y.; Yin, L.; Chang, X.; Yan, W.; et al. Restaur reduction in China: Protocol for a randomised controlled trial. BMJ Open 2020, 10, e038744. [CrossRef] 20. Zhang, J.; Sun, L.; Liu, Y.; Wang, H.; Sun, N.; Zhang, P. Mobile Device-Based electronic data capture system used in a clinical randomized controlled trial: Advantages and challenges. J. Med. Internet Res. 2017, 19, e6978. [CrossRef] 13 of 13 13 of 13 Nutrients 2022, 14, 4698 21. He, F.J.; Wu, Y.; Feng, X.X.; Ma, J.; Ma, Y.; Wang, H.; Zhang, J.; Yuan, J.; Lin, C.-P.; Nowson, C.; et al. School based education programme to reduce salt intake in children and their families (School-EduSalt): Cluster randomised controlled trial. BMJ 2015, 350, h770. [CrossRef] , [ ] 22. McLean, R.M. Measuring population sodium intake: A review of methods. Nutrients 2014, 6, 4651–466 23. Tan, M.; He, F.J.; Wang, C.; MacGregor, G.A. Twenty-Four-Hour Urinary Sodium and Potassium Excretion in China: A Systematic Review and Meta-Analysis. J. Am. Heart Assoc. 2019, 8, e012923. [CrossRef] [PubMed] y 24. Chen, P.; Li, F.; Harmer, P. Healthy China 2030: Moving from blueprint to action with a new focus on public health. Lancet Public Health 2019, 4, e447. [CrossRef] 25. References Xu, A.; Ma, J.; Guo, X.; Wang, L.; Wu, J.; Zhang, J.; Bai, Y.; Xu, J.; Lu, Z.; Xu, Z.; et al. Association of a Province-Wide Intervention With Salt Intake and Hypertension in Shandong Province, China, 2011–2016. JAMA Intern. Med. 2020, 180, 877–886. [CrossRef] [PubMed] [ ] 26. Tian, H.G.; Guo, Z.Y.; Hu, G.; Yu, S.J.; Sun, W.; Pietinen, P.; Nissinen, A. Changes in sodium intake and blood pressure in a community-based intervention project in China. J. Hum. Hypertens. 1995, 9, 959–968. [PubMed] y p j yp 27. Lin, P.H.; Yancy, W.S., Jr.; Pollak, K.I.; Dolor, R.J.; Marcello, J.; Samsa, G.P.; Batch, B.C.; Svetkey, L.P. The and patient intervention program on dietary intake. J. Acad. Nutr. Diet. 2013, 113, 1465–1475. [CrossRe 27. Lin, P.H.; Yancy, W.S., Jr.; Pollak, K.I.; Dolor, R.J.; Marcello, J.; Samsa, G.P.; Batch, B.C.; Svetkey, L.P. The inﬂuence of a physician and patient intervention program on dietary intake. J. Acad. Nutr. Diet. 2013, 113, 1465–1475. [CrossRef] , ; cy, S , J ; o , ; o o , J ; ce o, J ; S s , G ; c , C ; S e ey, e e ce o p ys c and patient intervention program on dietary intake. J. Acad. Nutr. Diet. 2013, 113, 1465–1475. [CrossRef] 28 S tk L P P ll k K I Y W S J D l R J B t h B C S G M t h D B Li PH H t i i t y y atient intervention program on dietary intake. J. Acad. Nutr. Diet. 2013, 113, 1465–1475. [CrossRef] 28. Svetkey, L.P.; Pollak, K.I.; Yancy, W.S., Jr.; Dolor, R.J.; Batch, B.C.; Samsa, G.; Matchar, D.B.; Lin, P.H. Hypertension improvement project: Randomized trial of quality improvement for physicians and lifestyle modiﬁcation for patients. Hypertension 2009, 54, 1226–1233. [CrossRef] 29. Cappuccio, F.P.; Kerry, S.M.; Micah, F.B.; Plange-Rhule, J.; Eastwood, J.B. A community programme to reduce salt intake and blood pressure in Ghana (ISRCTN88789643]. BMC Public Health 2006, 6, 13. [CrossRef] 30. Li, N.; Yan, L.L.; Niu, W.; Labarthe, D.; Feng, X.; Shi, J.; Zhang, J.; Zhang, R.; Zhang, Y.; Chu, H.; et al. A large-scale cluster randomized trial to determine the effects of community-based dietary sodium reduction—The China Rural Health Initiative Sodium Reduction Study. Am. Heart J. 2013, 166, 815–822. [CrossRef] 31. Bhana, N.; Utter, J.; Eyles, H. References Knowledge, Attitudes and Behaviors Related to Dietary Salt Intake in High-Income Countries: A Systematic Review. Curr. Nutr. Rep. 2018, 7, 183–197. [CrossRef] 32. Bi, Z.; Liang, X.; Xu, A.; Wang, L.; Shi, X.; Zhao, W.; Ma, J.; Guo, X.; Zhang, X.; Zhang, J.; et al. Hypertension Prevalence, Awareness, Treatment, and Control and Sodium Intake in Shandong Province, China: Baseline Results from Shandong–Ministry of Health Action on Salt Reduction and Hypertension (SMASH), 2011. Prev. Chronic Dis. 2014, 11, E88. [CrossRef] [PubMed] 32. Bi, Z.; Liang, X.; Xu, A.; Wang, L.; Shi, X.; Zhao, W.; Ma, J.; Guo, X.; Zhang, X.; Zhang, J.; et al. Hypertension Prevalence, Awareness, Treatment, and Control and Sodium Intake in Shandong Province, China: Baseline Results from Shandong–Ministry of Health Action on Salt Reduction and Hypertension (SMASH), 2011. Prev. Chronic Dis. 2014, 11, E88. [CrossRef] [PubMed] 33. Li, Y.; Zhang, J.; Shi, X.-M.; Liang, X.-F. A cross-sectional survey on the efﬁcacy of ‘China Healthy Lifestyle for All’ in 2012. Chin. J. Epidemiol. 2013, 34, 869–873. (In Chinese) of Health Action on Salt Reduction and Hypertension (SMASH), 2011. Prev. Chronic Dis. 2014, 11, E88. [CrossRef] [PubMed] 33. Li, Y.; Zhang, J.; Shi, X.-M.; Liang, X.-F. A cross-sectional survey on the efﬁcacy of ‘China Healthy Lifestyle for All’ in 2012. Chin. J. Epidemiol. 2013, 34, 869–873. (In Chinese) 34. He, F.J.; MacGregor, G.A. Fortnightly review: Beneﬁcial effects of potassium. BMJ 2001, 323, 497–501. [CrossRef] [PubMed] 35. He, F.J.; Nowson, C.A.; MacGregor, G.A. Fruit and vegetable consumption and stroke: Meta-analysis of cohort studies. Lancet 2006, 367, 320–326. [CrossRef] 34. He, F.J.; MacGregor, G.A. Fortnightly review: Beneﬁcial effects of potassium. BMJ 2001, 323, 497–501. [ g g y p 35. He, F.J.; Nowson, C.A.; MacGregor, G.A. Fruit and vegetable consumption and stroke: Meta-analysis of cohort studies. Lancet 2006, 367, 320–326. [CrossRef]
https://openalex.org/W2991502893	https://europepmc.org/articles/pmc6928968?pdf=render	English	null	Computer-Assisted Design of Environmentally Friendly and Light-Stable Fluorescent Dyes for Textile Applications	International journal of molecular sciences	2,019	cc-by	11,602	Computer-Assisted Design of Environmentally Friendly and Light-Stable Fluorescent Dyes for Textile Applications Songsong Tang 1 , Guoqiang Chen 1,* and Gang Sun 2,* Songsong Tang 1 , Guoqiang Chen 1,* and Gang Sun 2,* 1 National Engineering Laboratory for Modern Silk, College of Textile and Clothing Engineering, Soochow University, Suzhou 215123, China; songsongtang@outlook.com 2 Division of Textiles and Clothing, University of California, Davis, CA 95616, USA * Correspondence: chenguojiang@suda.edu.cn (G.C.); gysun@ucdavis.edu (G.S.) R i d 12 O t b 2019 A t d 25 N b 2019 P bli h d 27 N b 2019 Songsong Tang 1 , Guoqiang Chen 1,* and Gang Sun 2,* 1 National Engineering Laboratory for Modern Silk, College of Textile and Clothing Engineering, Soochow University, Suzhou 215123, China; songsongtang@outlook.com 2 Division of Textiles and Clothing, University of California, Davis, CA 95616, USA * Correspondence: chenguojiang@suda.edu.cn (G.C.); gysun@ucdavis.edu (G.S.) Received: 12 October 2019; Accepted: 25 November 2019; Published: 27 November 2019 Received: 12 October 2019; Accepted: 25 November 2019; Published: 27 November 2019 Received: 12 October 2019; Accepted: 25 November 2019; Published: 27 November 2019 Abstract: Five potentially environmentally friendly and light-stable hemicyanine dyes were designed based on integrated consideration of photo, environmental, and computational chemistry as well as textile applications. Two of them were synthesized and applied in dyeing polyacrylonitrile (PAN), cotton, and nylon fabrics, and demonstrated the desired properties speculated by the programs. The computer-assisted analytical processes includes estimation of the maximum absorption and emission wavelengths, aquatic environmental toxicity, aﬃnity to ﬁbers, and photo-stability. This procedure could eﬀectively narrow down discovery of new potential dye structures, greatly reduce and prevent complex and expensive preparation processes, and signiﬁcantly improve the development eﬃciency of novel environmentally friendly dyes. Keywords: ﬂuorescence; hemicyanine; photo-stability; Gaussian calculations Int. J. Mol. Sci. 2019, 20, 5971; doi:10.3390/ijms20235971 International Journal of Molecular Sciences International Journal of Molecular Sciences 1. Introduction Fluorescent dyes, an important kind of functional dyes, have been used in many ﬁelds such as solar cells [1,2], biological imaging [3,4], probes [5,6], and optical nanoscopy [7,8]. However, many ﬂuorescent dyes have been banned and will be phased out in commercial uses due to the concerns of human and environmental health as well as increasing regulations. For example, many azo dyes are prohibited to be used in textiles in European countries since aromatic amines, as potential metabolized products, are harmful to humans and the environment [9,10]. Hemicyanine ﬂuorescent dyes usually have high molar absorption coeﬃcients, broad spectrum width of absorption, and high-quantum yields, because of good coplanarity and conjugated systems with suitable electron donating (EDG) and electron withdrawing groups (EWG) [11–14]. However, these dyes may not have good light stability, especially when applying them in textile materials. Taking compound Z1 (Figure 1a,b) as an example, it possesses perfect chemical structural features and great ﬂuorescence properties (Figure 1c), emitting yellow–red ﬂuorescence [15–18], but the light reﬂectivity of polyacrylonitrile (PAN) fabric dyed by the compound Z1 decreased greatly when it was exposed to the light for 5 h (see Figure 1d). Thus, it was hardly used in dyeing fabrics. Fortunately, there are some diﬀerent ways to improve the photo-stability of the dyes, for example improving the photo-stability of dyes using TiO2 [19], while computational chemistry could estimate behaviors of dyes under light according to quantum chemistry, which could provide theoretical speculations of photo-chemical/light fastness of the dyes [20–25]. Int. J. Mol. Sci. 2019, 20, 5971; doi:10.3390/ijms20235971 www.mdpi.com/journal/ijms www.mdpi.com/journal/ijms 2 of 16 Int. J. Mol. Sci. 2019, 20, 5971 (a) (b) (c) (d) 380 475 570 665 760 35 70 105 140 175 Reflectance (%) Wavelength (nm) blank 0h 1h 2h 3h 4h 5h Figure 1. (a), (b) Chemical and geometry structure of the hemicyanine–DYE BD; (c) maximum reflectance of dyed polyacrylonitrile (PAN) fabrics (% owf :0.01; 0.05; 0.1; 0.2; 0.4; 0.6); (d) reflectivity of PAN fabrics dyed by DYE BD when it was exposed to light for different hours (0.4% owf). 0.00 0.15 0.30 0.45 0.60 130 143 156 169 Maximum reflectance (%) Dye concentration (owf%) Figure 1. 1. Introduction (a,b) Chemical and geometry structure of the hemicyanine–DYE BD; (c) maximum reﬂectance of dyed polyacrylonitrile (PAN) fabrics (% owf: 0.01; 0.05; 0.1; 0.2; 0.4; 0.6); (d) reﬂectivity of PAN fabrics dyed by DYE BD when it was exposed to light for diﬀerent hours (0.4% owf). (b) (d) 380 475 570 665 760 35 70 105 140 175 Reflectance (%) Wavelength (nm) blank 0h 1h 2h 3h 4h 5h (a) (b) (c) 0.00 0.15 0.30 0.45 0.60 130 143 156 169 Maximum reflectance (%) Dye concentration (owf%) (c) Figure 1. (a), (b) Chemical and geometry structure of the hemicyanine–DYE BD; (c) maximum reflectance of dyed polyacrylonitrile (PAN) fabrics (% owf :0.01; 0.05; 0.1; 0.2; 0.4; 0.6); (d) reflectivity of PAN fabrics dyed by DYE BD when it was exposed to light for different hours (0.4% owf). Figure 1. (a,b) Chemical and geometry structure of the hemicyanine–DYE BD; (c) maximum reﬂectance of dyed polyacrylonitrile (PAN) fabrics (% owf: 0.01; 0.05; 0.1; 0.2; 0.4; 0.6); (d) reﬂectivity of PAN fabrics dyed by DYE BD when it was exposed to light for diﬀerent hours (0.4% owf). Structural design of fluorescent dyes could be conducted by using computational programs, which could avoid unnecessary work in searching for dyes with improved light-stability. Potential environmental impacts of any designed dye could be estimated by using an ecological structure activity relationships (ECOSAR) predictive model, which is maintained by US Environmental Protection Agency (US-EPA) for predicting aquatic toxicity of chemicals [26,27], while suitability of the designed dyes on different fibers can be estimated by using a Hansen Solubility parameter theory (HSP) [27–30]. In this manuscript, five hemicyanine dyes were designed by using similar starting chemicals and reactions as environmentally friendly fluorescent dyes with improved photo-stability for textile applications. Gaussian 09, a computational program, was employed in analysis of chemical structures of the dyes, while the ECOSAR was adopted to estimate the aquatic-toxicity of raw materials, intermediates, and the dyes. Furthermore, the photo-stability of the dyes were analyzed as well after the affinities of the dyes to fibers were estimated by using Hansen solubility parameters (HSP). In addition, two of the dyes with expected good fluorescent features were prepared, and basic properties of the dyed materials were tested. Structural design of ﬂuorescent dyes could be conducted by using computational programs, which could avoid unnecessary work in searching for dyes with improved light-stability. 1. Introduction Potential environmental impacts of any designed dye could be estimated by using an ecological structure activity relationships (ECOSAR) predictive model, which is maintained by US Environmental Protection Agency (US-EPA) for predicting aquatic toxicity of chemicals [26,27], while suitability of the designed dyes on diﬀerent ﬁbers can be estimated by using a Hansen Solubility parameter theory (HSP) [27–30]. In this manuscript, ﬁve hemicyanine dyes were designed by using similar starting chemicals and reactions as environmentally friendly ﬂuorescent dyes with improved photo-stability for textile applications. Gaussian 09, a computational program, was employed in analysis of chemical structures of the dyes, while the ECOSAR was adopted to estimate the aquatic-toxicity of raw materials, intermediates, and the dyes. Furthermore, the photo-stability of the dyes were analyzed as well after the aﬃnities of the dyes to ﬁbers were estimated by using Hansen solubility parameters (HSP). In addition, two of the dyes with expected good ﬂuorescent features were prepared, and basic properties of the dyed materials were tested. 2.1. Biosafety Analysis of Synthetic Processes 2.1. Biosafety Analysis of Synthetic Processes With the results of analysis on the synthesis of the proposed dye molecules, the potential toxicities of the raw materials, all intermediates, and the dyes were estimated by using the ECOSAR. Among all raw materials, compounds A2, A3, A4, and A5 showed LC50 and EC50 values higher than 1000, and especially the LC50 (fish, 96h) and LC50 (daphnid, 48h) values of compounds A2, A3, and A4 are greater than 10 million. It means that the starting compounds A2, A3, A4, and A5 are more environmentally friendly than the compounds A1 and A6 With the results of analysis on the synthesis of the proposed dye molecules, the potential toxicities of the raw materials, all intermediates, and the dyes were estimated by using the ECOSAR. Among all raw materials, compounds A2, A3, A4, and A5 showed LC50 and EC50 values higher than 1000, and especially the LC50 (ﬁsh, 96h) and LC50 (daphnid, 48h) values of compounds A2, A3, and A4 are greater than 10 million. It means that the starting compounds A2, A3, A4, and A5 are more environmentally friendly than the compounds A1 and A6. environmentally friendly than the compounds A1 and A6. Compounds A1, A2, and A5 have similar structures but very different toxicity values due to the different substituents. The parent chemical (A1) has LC50 (fish, 96 h), LC50 (daphnid, 48 h), and EC50 Compounds A1, A2, and A5 have similar structures but very diﬀerent toxicity values due to the diﬀerent substituents. The parent chemical (A1) has LC50 (ﬁsh, 96 h), LC50 (daphnid, 48 h), and EC50 Int. J. Mol. Sci. 2019, 20, 5971 3 of 16 (green algae, 96 h) values of 292.3, 156.33, and 90.92, respectively. Thus, the A1 is toxic to aquatic lives—especially to the green algae. The corresponding values of the A2 are about 1000 times greater than those of the A1, while the values of LC50 and EC50 of the A5 are about 20–38 times larger than those of the A1. The results reveal that both sulfonic acid (–SO3H) and carboxyl (–COOH) groups could increase LC50 and EC50 values and decrease the toxicity of the chemicals, while the sulfonic acid group is more eﬀective than carboxyl group to decrease the toxicity of chemicals. The positions of diﬀerent substituents on the pyridinium ring have diﬀerent inﬂuence on their toxicities. 2.1. Biosafety Analysis of Synthetic Processes 2.1. Biosafety Analysis of Synthetic Processes The carboxyl group at the 3 position (compound A5) of the ring increased the values of LC50 and EC50 of the chemicals, but the carboxyl group at the 2 position (compound A6) increases the toxicity of the chemical. Diﬀerent to the carboxyl group, the sulfonic acid group (see compounds A2, A3, and A4 in Figure S1 and Table 1) decreases the toxicity of the chemicals on any positions. Table 1. Estimated bio-toxicity of raw materials, intermediates, and designed dyes. Item Name Fish 96h-LC50 (mg/L) Daphnid 48h-LC50 (mg/L) Green Algae 96h-EC50 (mg/L) Raw materials A1 292.3 156.33 90.92 A2 3,727,904 1,489,483.88 259,458.53 A3 3,727,904 1,489,483.88 259,458.53 A4 3,727,904 1,489,483.88 259,458.53 A5 11,169.25 5724.55 2792.02 A6 91.92 5.36 8.56 Intermediates B1 15,283,593 5,209,308.5 470,385.34 B2 11,033,896,960 3,153,353,728 137,435,888 B3 11,033,896,960 3,153,353,728 137,435,888 B4 11,033,896,960 3,153,353,728 137,435,888 B5 482,951,392 157,749,760 11,945,150 B6 391,939.94 2527.1 61,436.25 Designed dyes Z1 3094.48 1582.45 764.68 Z2 25,924,188 9,904,030 1,433,391.12 Z3 25,924,188 9,904,030 1,433,391.12 Z4 25,924,188 9,904,030 1,433,391.12 Z5 90,077.77 44,143.83 17,888.29 Z6 588.73 27.57 57.77 Table 1. Estimated bio-toxicity of raw materials, intermediates, and designed dyes. All pyridinium salts (compounds B1 to B6) are less toxic than the starting pyridine derivatives based on their higher values of LC50 and EC50. However, the toxicities of the designed dyes, except Z6, are relatively higher than the intermediates; they are signiﬁcantly lower than the starting compounds and Z1. The increased values of LC50 and EC50 of chemicals are possibly due to the increased water solubility. 2.2. Fluorescence Properties of Designed Dyes Fluorescent properties of these designed dyes were speculated by using the computational program as well. The dye Z1 (see Figures S1 and S2) still possess coplanar conjugated structures even though diﬀerent substituent groups are not in the same plane. It would emit ﬂuorescence after excited by relevant light. Table S1 shows the emission and absorption data, calculated by the Gaussian 09 based on the b3lyp/6-31G(d) level of N-methylacridinium chloride in water. The calculated maximum emission wavelength was 482.73 nm (2.5648 eV), which was close to the data (498 nm) in reference [31], while the maximum absorption wavelength was 412.93 nm—well matched with the reference data (415 nm) [31]. The maximum absorption and emission wavelengths of designed dyes (shown in Table 2) are varied from that of Z1 but still can emit visible lights after excited by relevant light. Int. J. Mol. Sci. 2019, 20, 5971 4 of 16 Table 2. Fluorescence properties calculated by b3lyp/6-31G(d) for designed dyes in water. Item Max Absorption Wavelength Max Emission Wavelength Stokes (nm) 1E (ev) Wavelength (nm) 2f E (ev) Wavelength (nm) f Z1 2.6105 474.94 1.4785 2.2869 542.15 1.5166 67.21 Z2 2.5576 484.77 1.4978 2.2167 559.32 1.5085 74.55 Z3 3.3454 370.61 0.1305 - 419.61 1.5554 49 Z4 2.6517 467.57 1.1095 2.4054 515.45 0.9673 47.88 Z5 2.4842 499.09 1.3926 2.0643 600.60 1.2270 101.51 Z6 2.3827 520.35 1.2081 1.7730 699.31 0.8955 178.96 1E: the energy of the light. 2f: oscillator strength. Table 2. Fluorescence properties calculated by b3lyp/6-31G(d) for designed dyes in water. 2.3. Dyeing Properties Analysis of Designed Dyes The aﬃnity between the dyes and ﬁbers could be analyzed by using Hansen solubility parameter theory (HSP), which has been employed in estimating aﬃnity and interactions between molecules [32,33]. HSP distances (Ra values) between the diﬀerent molecules in Hansen space can be calculated by Equation (1) [34,35]. The smaller Ra values represent better aﬃnity between two molecules. q Ra = q 4∆δ2 D + ∆δ2 P + ∆δ2 H (1) (1) δD: The energy from dispersion forces between molecules. δP: The energy from dipolar forces between molecules. δH: The energy from hydrogen bonds. δH: The energy from hydrogen bonds. Dye HSP (MPa1/2) Ra1 (MPa1/2) Ra2 (MPa1/2) Ra3 (MPa1/2) Ra4 (MPa1/2) Ra5 (MPa1/2) Ra6 (MPa1/2) δD δP δH Z1 18.4 3.0 1.7 43.02 23.69 14.48 9.22 9.90 9.99 Z2 18.9 8.4 10.4 33.49 13.64 9.47 7.69 4.47 4.53 Z3 18.9 8.4 10.4 33.49 13.64 9.47 7.69 4.47 4.53 Z4 18.9 8.4 10.4 33.49 13.64 9.47 7.69 4.47 4.53 Z5 18.5 4.3 4.6 39.93 20.51 12.57 7.78 7.43 7.53 Z6 18.5 4.3 4.6 39.93 20.51 12.57 7.78 7.43 7.53 CI Disperse Yellow 11 20.2 5.3 8.7 36.49 16.64 12.53 8.27 7.08 7.16 H2O 15.5 16 42.3 0 20.26 36.09 39.78 36.51 36 Cellulose (cellobiose) 18.7 12.5 23.4 20.26 - - - - - PAN 17.9 16.7 6.3 36.09 - - - - - PET 19.6 11.7 3.6 39.78 - - - - - Nylon 6 18.5 11.3 7.1 36.51 - - - - - Nylon 66 18.5 11.4 7.1 36 - - - - - Comparing the dyes Z1–Z6, they have similar conjugated structures but diﬀerent dyeing properties due to varied substitutes. Both of the carboxyl and sulfonic acid groups can improve the aﬃnity between dyes and certain polymers, but the sulfonic acid group showed stronger eﬀect than that of the carboxyl group. It should be pointed that the disperse dyes, such like CI Disperse Yellow 11, were hardly used to color PAN ﬁber, while some water-soluble dyes are diﬃcult to dye PET ﬁber in practice. Thus, we should analyze the dyeing properties not only based on the HSP distance, but also need to consider the practical/commercial knowledge of dyeing ﬁbers. δH: The energy from hydrogen bonds. If a dye has high aﬃnity to a ﬁber, it will be more attractive to the ﬁber [36], showing better dyeing properties. Five popular natural and chemical ﬁbers were selected to estimate the aﬃnities to the designed dyes. CI Disperse Yellow 11, one of the earliest commercial ﬂuorescent dyes and widely employed in coloration of synthetic ﬁbers (i.e., nylon and PET) and plastics [37,38], was selected randomly to show the practicality of this calculation model. As shown in Table 3, Ra1 to Ra6 values reﬂect aﬃnities of a dye to water, cellulose, acrylics (PAN), polyester (PET), Nylon 6, and Nylon 66, respectively. Ra4, Ra5, and Ra6 of CI Disperse Yellow 11 are 8.27, 7.08, and 7.16, which are much lower than Ra1 (36.49) and Ra2 (16.64), indicating that the CI Disperse Yellow has better aﬃnity with PET, Nylon 6, and Nylon 66 than water and cellulose—consistent with the existing data. The designed dye Z1 possess low but close Ra3, Ra4, Ra5, and Ra6 values, which are lower than Ra2 (23.69) and Ra1 (43.02). It means the Z1 could dye PAN, PET, Nylon 6, and Nylon 66 with potential good wash fastness. The Z1 could color cellulose, however due to the small diﬀerence of Ra2 and Ra1 values, the wash fastness of the Z1 on cellulose may not be so good. Based on the HSP results, the dye Z2 should have the same dyeing properties to the dyes Z3 and Z4, while the dye Z5 should have same dyeing properties as the dye Z6. Similarly, the dyes Z2, Z3, and Z4 could dye all ﬁve ﬁbers but may have diﬀerent color fastness on the products. 5 of 16 Int. J. Mol. Sci. 2019, 20, 5971 Table 3. Hansen solubility parameter (HSP) values of the diﬀerent dyes and Hansen distances to water (Ra1), cellobiose (Ra2), polyacrylonitrile (Ra3), poly (ethylene terephthalate) (Ra4), Nylon 6 (Ra5), and Nylon 66 (Ra6). (Ra1), cellobiose (Ra2), polyacrylonitrile (Ra3), poly (ethylene terephthalate) (Ra4), Nylon 6 (Ra5), and Nylon 66 (Ra6). 2.4. Photo-Stability Analysis of Designed Dyes Photo-stability, an important parameter of dyes, is the main concern when selecting dyes for textile application. According to the literature [20,24], photo-degradation of hemicyanine dyes could be caused by reactions between the dyes and reactive oxygen species-such as singlet oxygen (1O2) and superoxide anion (O2−) existing in the air. Thus, the reactions between the designed dyes and two kinds of reactive oxygen were analyzed by the computational and quantum chemistry method to estimate the photo-stability of the designed dyes. As shown in Figure 2, the highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbitals (LUMO) of the designed dyes were calculated by the Gaussian 09. The atoms were numbered in Figure S6. Comparing the HOMOs and LUMOs of the dyes, the diﬀerent substituents and their positions on the pyridinium ring have varied inﬂuence, especially for the dyes of Z1 and Z4. As shown in Table 4 and Table S2, the HOMO and LUMO orbitals of 1O2, O2−, 3O2, and Z1–Z6 were calculated based on hf/6-31 + g (d). Int. J. Mol. Sci. 2019, 20, 5971 p ( two kinds of reactive ox 6 of 16 method 6 of 16 method Figure 2. Highest occupied molecular orbitals (HOMO) (first six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. Figure 2. Highest occupied molecular orbitals (HOMO) (ﬁrst six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. Figure 2. Highest occupied molecular orbitals (HOMO) (first six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. Figure 2. Highest occupied molecular orbitals (HOMO) (ﬁrst six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. Figure 2. Highest occupied molecular orbitals (HOMO) (first six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. Figure 2. Highest occupied molecular orbitals (HOMO) (ﬁrst six) and lowest unoccupied molecular orbitals (LUMO) (last six) orbitals of Z1–Z6. own in Figure 2, the highest occupied molecular orbitals (HOMO) and lowest un Table 4. HOMO and LUMO orbital levels of 3O2, O2−, and 1O2 based on hf/6-31 + g(d). As shown in Figure 2, the highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbitals (LUMO) of the designed dyes were calculated by the Gaussian 09. The atoms were numbered in Figure S6. 2.4. Photo-Stability Analysis of Designed Dyes Comparing the HOMOs and LUMOs of the dyes, the different substituents and their positions on the pyridinium ring have varied influence, especially for the dyes of Z1 and Z4. As shown in Table 4 and Table S2, the HOMO and LUMO orbitals of 1O2, O2−, 3O2, and Z1–Z6 were calculated based on hf/6-31 + g (d). Table 4. HOMO and LUMO orbital levels of 3O2, O2−, and 1O2 based on hf/6-31 + g(d). Eigenvalues (Hartree) 3O2 1O2 O2− HOMO LUMO HOMO LUMO HOMO LUMO −0.56418 0.16997 −0.47706 0.01129 −0.12516 0.37411 Atomic orbital coeﬃcients O1 2S 0 0.12445 0 0 0 0.11664 2PX 0 0 0 0.4592 0.49969 0 2PY 0.54995 0 0.53402 0 0 0 2PZ 0 −0.06966 0 0 0 −0.07467 O2 2S 0 −0.12445 0 0 0 −0.11664 2PX 0 0 0 −0.4592 −0.49969 0 2PY −0.54995 0 −0.53402 0 0 0 2PZ 0 −0.06966 0 0 0 −0.07467 According to the basic principles of linear combination of atomic orbitals (LCAO)-molecular orbital theory (MO), HOMO of the dyes and LUMO of the reactive oxygen species should have similar symmetry, energy levels, as well as the maximum overlap in the orbitals. Obviously, the HOMO orbitals energy of the dyes is close to that of the LUMO orbital of 1O2, while the energy of HOMO orbital of O2−is approximate to that of the LUMO orbitals of the dyes. Thus, the 1O2 LUMO orbitals could react with the HOMO orbitals of the dyes, while the O2−HOMO orbitals could react with the LUMO orbitals of the dyes—causing photo-oxidation. The O2−1 could be more powerful than 1O2 in photo-oxidation process since the energy gap between the LUMO orbitals of the dyes and O2−HOMO orbitals are lower than energy gap between the HOMO orbital of the dyes and 1O2 LUMO orbital (see Table S3). The active positions of HOMO and LUMO orbitals of the dyes are summarized from Figure 2 and Table S2, and listed in Table 5. As shown in Figure 2, the highest occupied molecular orbitals (HOMO) and lowest unoccupied molecular orbitals (LUMO) of the designed dyes were calculated by the Gaussian 09. The atoms were numbered in Figure S6. Comparing the HOMOs and LUMOs of the dyes, the different substituents and their positions on the pyridinium ring have varied influence, especially for the dyes of Z1 and Z4. 2.4. Photo-Stability Analysis of Designed Dyes As shown in Table 4 and Table S2, the HOMO and LUMO orbitals of 1O2, O2−, 3O2, and Z1–Z6 were calculated based on hf/6-31 + g (d). Table 4. HOMO and LUMO orbital levels of 3O2, O2−, and 1O2 based on hf/6-31 + g(d). Eigenvalues (Hartree) 3O2 1O2 O2− HOMO LUMO HOMO LUMO HOMO LUMO −0.56418 0.16997 −0.47706 0.01129 −0.12516 0.37411 Atomic orbital coeﬃcients O1 2S 0 0.12445 0 0 0 0.11664 2PX 0 0 0 0.4592 0.49969 0 2PY 0.54995 0 0.53402 0 0 0 2PZ 0 −0.06966 0 0 0 −0.07467 O2 2S 0 −0.12445 0 0 0 −0.11664 2PX 0 0 0 −0.4592 −0.49969 0 2PY −0.54995 0 −0.53402 0 0 0 2PZ 0 −0.06966 0 0 0 −0.07467 According to the basic principles of linear combination of atomic orbitals (LCAO)-molecular orbital theory (MO), HOMO of the dyes and LUMO of the reactive oxygen species should have similar symmetry, energy levels, as well as the maximum overlap in the orbitals. Obviously, the HOMO orbitals energy of the dyes is close to that of the LUMO orbital of 1O2, while the energy of HOMO orbital of O2−is approximate to that of the LUMO orbitals of the dyes. Thus, the 1O2 LUMO orbitals could react with the HOMO orbitals of the dyes, while the O2−HOMO orbitals could react with the LUMO orbitals of the dyes—causing photo-oxidation. The O2−1 could be more powerful than 1O2 in photo-oxidation process since the energy gap between the LUMO orbitals of the dyes and O2−HOMO orbitals are lower than energy gap between the HOMO orbital of the dyes and 1O2 LUMO orbital (see Table S3). The active positions of HOMO and LUMO orbitals of the dyes are summarized from Figure 2 and Table S2, and listed in Table 5. According to the basic principles of linear combination of atomic orbitals (LCAO)-molecular orbital theory (MO), HOMO of the dyes and LUMO of the reactive oxygen species should have similar symmetry, energy levels, as well as the maximum overlap in the orbitals. Obviously, the HOMO orbitals energy of the dyes is close to that of the LUMO orbital of 1O2, while the energy of HOMO orbital of O2−is approximate to that of the LUMO orbitals of the dyes. Thus, the 1O2 LUMO orbitals could react with the HOMO orbitals of the dyes, while the O2−HOMO orbitals could react with the LUMO orbitals of the dyes—causing photo-oxidation. 2.4. Photo-Stability Analysis of Designed Dyes The O2−1 could be more powerful than 1O2 in photo-oxidation process since the energy gap between the LUMO orbitals of the dyes and O2−HOMO orbitals are lower than energy gap between the HOMO orbital of the dyes and 1O2 LUMO orbital (see Table S3). The active positions of HOMO and LUMO orbitals of the dyes are summarized from Figure 2 and Table S2, and listed in Table 5. According to the basic principles of linear combination of atomic orbitals (LCAO)-molecular orbital theory (MO), HOMO of the dyes and LUMO of the reactive oxygen species should have similar symmetry, energy levels, as well as the maximum overlap in the orbitals. Obviously, the HOMO orbitals energy of the dyes is close to that of the LUMO orbital of 1O2, while the energy of HOMO orbital of O2−is approximate to that of the LUMO orbitals of the dyes. Thus, the 1O2 LUMO orbitals could react with the HOMO orbitals of the dyes, while the O2−HOMO orbitals could react with the LUMO orbitals of the dyes—causing photo-oxidation. The O2−1 could be more powerful than 1O2 in photo-oxidation process since the energy gap between the LUMO orbitals of the dyes and O2−HOMO orbitals are lower than energy gap between the HOMO orbital of the dyes and 1O2 LUMO orbital (see Table S3). The active positions of HOMO and LUMO orbitals of the dyes are summarized from Figure 2 and Table S2, and listed in Table 5. 7 of 16 Int. J. Mol. Sci. 2019, 20, 5971 Table 5. HOMO/LUMO orbitals of active positions of dyes. Table 5. HOMO/LUMO orbitals of active positions of dyes. Table 5. HOMO/LUMO orbitals of active positions of dyes. Item HOMO 8 LUMO Active Position Atomic Orbital Coeﬃcients Active Position Atomic Orbital Coeﬃcients Z1 N34-C17 0.27263, −0.13943 N46-C2 0.22269, −0.18852 N46-C3 0.22269, −0.15797 N34-C17 0.10212, −0.13939 Z2 N34-C17 0.26495, −0.1272 N45-C2 0.19891, −0.14321 N45-C3 0.19891, −0.17116 N34-C17 0.11526, −0.14978 Z3 N34-C17 0.26631, −0.13001 N44-C2 0.22645, −0.20734 N44-C3 0.22645, −0.15098 N34-C17 0.10354, −0.13363 Z4 N39-C18 0.28064, −0.14857 N43-C41 0.28690, −0.16810 N43-C4 0.28690, −0.25597 C2-C1 0.30619, −0.15157 Z5 N33-C16 0.26451, −0.13186 N45-C2 0.18462, −0.20217 N45-C3 0.18462, −0.09251 N33-C16 0.10986, −0.14634 Z6 N34-C17 0.26918, −0.1317 N44-C2 0.22579, −0.20415 N44-C3 0.22579, −0.14513 N34-C17 0.10063, −0.13091 The dyes Z1–Z6 have similar active positions in HOMO and LUMO orbitals due to the similar chemical structures. 2.4. Photo-Stability Analysis of Designed Dyes The atomic orbital coeﬃcient of the dye Z4 in HOMO orbital was higher than that of the Z1, while the energies of HOMO orbitals of the dyes Z2, Z3, Z5, and Z6 are close to each other and lower than that of Z1, indicating that the dye Z4 would be more active than other dyes. Also, the dyes Z2, Z3, Z5, and Z6 would be more stable than the dye Z1 during the photo-oxidation process in terms of the lower atomic orbital coeﬃcients in HOMO orbitals. The LUMO orbitals of the dyes are more complicated since they have three active positions in LUMO orbitals. Three atomic orbital coeﬃcients (green values) of the dyes Z2, Z5, and Z6 were lower than that of the dye Z1, while the other atomic orbital coeﬃcients of the dyes are increased (red values) or decreased compared to the dye Z1. However, the atomic orbital coeﬃcients of N45-C3 in the dye Z5 are lower than that of N46-C3 in dye Z1, and the atomic orbital coeﬃcients of the N33-C16 in dye Z5 are higher than that of N34-C17 in dye Z1. The same situation could be found in dye Z2, meaning that dyes Z2, Z5, and Z1 have similar reactive LUMO orbitals. Based on the above analysis and Table S4, dyes Z2 and Z5 would have better photo-stability than the dyes Z3, Z4, and Z6, but it is hard to determine whether dyes Z2 and Z5 have better photo-stability than dye Z1. 2.5. Synthesis and Applications of Dyes Z2 and Z5 (b) (d) (a) (b) 3500 3000 2500 2000 1500 1000 500 wave number(cm-1) Z2 Z5 (a) (b) 3500 3000 2500 2000 1500 1000 500 wave number(cm-1) Z2 Z5 (c) 300 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0 Fluorescence intesnsity Absorption intensity Wavelength(nm) Z5-UV/Vis Z2-UV/Vis Z5-Fluorescent Z2-Fluorescent 0 5 10 15 20 (d) 300 400 500 600 700 800 900 30 60 90 120 150 Z2-Cellouse Z2-PAN Z5-Nylon Z5-Cellouse PAN Cellouse Nylon Z2-Cellouse Z2-Nylon Reflectance (%) Wavelength(nm) (d) (c) Absorption intensity Fluorescence intesnsity Reflectance (%) Figure 3. (a) FTIR spectral of dyes Z2 and Z5; (b) ethanol solutions of dyes Z2 (left) and Z5 (right) under D65 (upper) and UV (bottom) light; (c) absorption and emission spectra of dyes Z2 and Z5 in water; (d) reflectivity of different fabrics dyed by dyes Z2 and Z5. Figure 3. (a) FTIR spectral of dyes Z2 and Z5; (b) ethanol solutions of dyes Z2 (left) and Z5 (right) under D65 (upper) and UV (bottom) light; (c) absorption and emission spectra of dyes Z2 and Z5 in water; (d) reﬂectivity of diﬀerent fabrics dyed by dyes Z2 and Z5. Ethanol solutions of Z2 and Z5 under UV and D65 light display diﬀerent colors (Figure 3b). According to Figure 3c, the experimental maximum absorbance wavelengths of dyes Z2 and Z5 were around 520 nm and 490 nm, and the measured maximum emission wavelengths of them were about 622 nm and 612 nm, respectively. The results indicated that the measured maximum absorption and emission wavelengths are close to these estimated data, especially for the dye Z5 (see Tables S5–S7). The synthesized dyes (Z2 and Z5, Figures S7 and S9) and commercial Rhodamine B (Figure S8) were used to dye PAN, cellulose, and nylon (2 g/piece, woven, Shanghai Textile Industry Institute of Technical Supervision, Figure 3d and Figure S7a) following Figures S10 and S11 in an X-5 DYEING machine (Foshan HUANGJU, China) with liquor ratio 50:1 at pH 4.5–5.0 by an acetic acid–sodium acetate buﬀer solution. The dye solutions were prepared with dyes (1% owf), sodium sulphate (3 g/L), and surfactant (0.5 g/L). The temperature was raised from room temperature to 100 ◦C at the rate of 1 ◦C/min after the fabrics were immersed into the dye solutions. 2.5. Synthesis and Applications of Dyes Z2 and Z5 Following the theoretical speculations on ideal ﬂuorescent dyes for textile applications, dyes Z2 and Z5 were selected to prove the practicality of the theory. The prepared dyes Z2 and Z5 were conﬁrmed by FTIR, UV-vis, and ﬂuorescent spectra (Figure 3). As shown in Figure 3a, the FTIR spectrum of dye Z2 is similar to that of dye Z5 due to the similar chemical structures. The diﬀerences in the FTIR of dyes Z2 and Z5 are marked in diﬀerent colors in Figure 3a. Obviously, there are characteristic absorption peaks (1743 cm−1, 1243 cm−1) of carboxyl group in Z5, while the 632 cm−1 is the absorption peak of S=O in sulfonic group of Z2. The absorption peaks of carbon double bonds (–C=C–) of dye Z2 and Z5 are about 1630 cm−1 and 1650 cm−1, respectively, while the absorption peaks of pyridine (Py) of dyes Z2 (1574 cm−1) and Z5 (1577 cm−1) are slightly diﬀerent. 8 of 16 ) pyridine Int. J. Mol. Sci. 2019, 20, 5971 dye Z2 and Z5 are abou (a) (b) (c) (d) Figure 3. (a) FTIR spectral of dyes Z2 and Z5; (b) ethanol solutions of dyes Z2 (left) and Z5 (right) under D65 (upper) and UV (bottom) light; (c) absorption and emission spectra of dyes Z2 and Z5 in water; (d) reflectivity of different fabrics dyed by dyes Z2 and Z5. 3500 3000 2500 2000 1500 1000 500 wave number(cm-1) Z2 Z5 300 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0 Fluorescence intesnsity Absorption intensity Wavelength(nm) Z5-UV/Vis Z2-UV/Vis Z5-Fluorescent Z2-Fluorescent 0 5 10 15 20 300 400 500 600 700 800 900 30 60 90 120 150 Z2-Cellouse Z2-PAN Z5-Nylon Z5-Cellouse PAN Cellouse Nylon Z2-Cellouse Z2-Nylon Reflectance (%) Wavelength(nm) Figure 3. (a) FTIR spectral of dyes Z2 and Z5; (b) ethanol solutions of dyes Z2 (left) and Z5 (right) under D65 (upper) and UV (bottom) light; (c) absorption and emission spectra of dyes Z2 and Z5 in water; (d) reﬂectivity of diﬀerent fabrics dyed by dyes Z2 and Z5. 2.5. Synthesis and Applications of Dyes Z2 and Z5 After that, dyeing took place at this temperature and continued for a further 60 min; the dye solutions were then cooled to 70 ◦C at 1.25 ◦C/min, and the dyed fabrics were washed thoroughly in distilled water and dried in the open air [39,40]. The three blank fabrics have no obvious absorption and emission peaks under visible light, while the six dyed fabrics displayed obvious absorption and emission peaks under the visible light, Int. J. Mol. Sci. 2019, 20, 5971 Int. J. Mol. Sci. 2019, 20, x 9 of 16 8 of 15 indicating that dyes Z2 and Z5 can color PAN, nylon, and cotton fabrics. However, diﬀerent to the dyed PAN and Nylon, the cotton fabrics dyed by Z2 and Z5 displayed small adsorption and emission intensity. Similar to the water solutions of dyes, the maximum absorption wavelengths of dye Z2 on PAN and Nylon were around 510 nm, and the maximum emission wavelengths of them were around 625 nm. The maximum absorption wavelengths of dye Z5 on PAN and Nylon were about 475 nm and 505 nm, while the maximum emission wavelengths of dye Z5 on PAN and nylon were around 610 nm and 615 nm, respectively. Ethanol solutions of Z2 and Z5 under UV and D65 light display different colors (Figure 3b) According to Figure 3c, the experimental maximum absorbance wavelengths of dyes Z2 and Z5 were around 520 nm and 490 nm, and the measured maximum emission wavelengths of them were about 622 nm and 612 nm, respectively. The results indicated that the measured maximum absorption and emission wavelengths are close to these estimated data, especially for the dye Z5 (see Tables S5–7). 2.5. Synthesis and Applications of Dyes Z2 and Z5 The synthesized dyes (Z2 and Z5, Figure S7, Figure S9) and commercial Rhodamine B (Figure S8) were used to dye PAN, cellulose, and nylon (2 g/piece, woven, Shanghai Textile Industry Institute of Technical Supervision Figures 5d and S7a) following Figure S10 and Figure S11 p y As shown in Figure 2, the energy of HOMO orbitals of Z2 and Z5 were −0.36080 hartree and −0.35420 hartree, so the energy gap between the HOMO orbital of dye Z2 and 1O2 LUMO (0.01129 hartree) orbital was greater than the energy gap between the HOMO orbital of dye Z5 orbitals and 1O2 LUMO orbital, which means that dye Z5 was more reactive than dye Z2 in terms of HOMO orbitals; similarly but diﬀerent to the HOMO orbital, the energy gap between the LUMO (−0.10124 hartree) orbital of dye Z5 and O2−HOMO (−0.12516 hartree) orbital was smaller than the energy gap between the LUMO (−0.09660 hartree) orbitals of dye Z2 and O2−HOMO orbital. Thus, dye Z2 would be more stable than dye Z5 when it reacts with reactive oxygen species. However, when compared with Z1, the energy of the HOMO and LUMO orbitals of dye Z2 and Z5 were lower than that of dye Z1, so it is hard to determine whether dyes Z2 and Z5 have better photo-stability than dye Z1. According to Figure 4, the maximum reﬂectivity of the Z1 dyed fabric lost about 20% of initial color intensity after 5 h light exposure. The Z2 dyed fabric only displayed about 16% loss in color intensity after the same duration of light exposure, while the Z5 dyed fabric showed about 24% loss in color intensity, meaning that dye Z2 had better photo-stability than dye Z1, but dye Z5 showed slightly lower photo-stability than dye Z1 in dyed PAN fabrics. The results are a little disappointing but will be addressed in future work. Industry Institute of Technical Supervision, Figures 5d and S7a) following Figure S10 and Figure S11 in an X-5 DYEING machine (Foshan HUANGJU, China) with liquor ratio 50:1 at pH 4.5–5.0 by an acetic acid–sodium acetate buffer solution. The dye solutions were prepared with dyes (1% owf), sodium sulphate (3 g/L), and surfactant (0.5 g/L). The temperature was raised from room temperature to 100 °C at the rate of 1 °C/min after the fabrics were immersed into the dye solutions. 2.5. Synthesis and Applications of Dyes Z2 and Z5 After that, dyeing took place at this temperature and continued for a further 60 min; the dye solutions were then cooled to 70 °C at 1.25 °C/min, and the dyed fabrics were washed thoroughly in distilled water and dried in the open air [39,40]. The three blank fabrics have no obvious absorption and emission peaks under visible light, while the six dyed fabrics displayed obvious absorption and emission peaks under the visible light, indicating that dyes Z2 and Z5 can color PAN, nylon, and cotton fabrics. However, different to the dyed PAN and Nylon, the cotton fabrics dyed by Z2 and Z5 displayed small adsorption and emission intensity. Similar to the water solutions of dyes, the maximum absorption wavelengths of dye Z2 on PAN and Nylon were around 510 nm, and the maximum emission wavelengths of them were around 625 nm. The maximum absorption wavelengths of dye Z5 on PAN and Nylon were about 475 nm and 505 nm, while the maximum emission wavelengths of dye Z5 on PAN and nylon were around 610 nm and 615 nm, respectively. 400 500 600 700 800 900 0 30 60 90 120 150 Z5-5h Z2-5h Z1-5h Blank Z1-0h Z2-0h Z5-0h Reflectance (%) Wavelength(nm) Figure 4. Light stability of the Z2 and Z5 dyed PAN fabrics exposed to the light for 0 and 5 h. Figure 4. Light stability of the Z2 and Z5 dyed PAN fabrics exposed to the light for 0 and 5 h. Reflectance (%) Figure 4 Light stability of the Z2 and Z5 dyed PAN fabrics exposed to the light for 0 and 5 h Figure 4. Light stability of the Z2 and Z5 dyed PAN fabrics exposed to the light for 0 and 5 h. 3.2.1. Route Design of Hemicyanine Dyes yp free energies were estimated at b3lyp/6-31G i i l h f h di As shown in Figure 5, the hemciayanine dyes could be prepared by 2 steps; the chemical structures of raw materials, intermediates, and dyes are shown in Figure S1, while the geometry structures of the designed dyes are shown in Figure S2. Z1 (DYE-BD) could be easily synthesized in two steps following the references [15–18], and the route of the synthesis can serve as an example to illustrate the designed routes. The nitrogen center of pyridine features a basic lone pair of electrons; consequently, pyridine is a strong nucleophile. Thus, pyridine could react with 1-bromoethane to form pyridinium, while the second reaction is a typical Knoevenagel condensation reaction between the methyl group on the pyridinium and the carbonyl group in 4-diethylaminobenzaldehyde. emission wavelengths of them were predicted by using time-dependent density functional theory (TD-DFT) at the b3lyp/6-31G(d) level [47–52]. ECOSAR was used to predict aquatic toxicity of the chemicals by using the database based on the quantitative structure–activity relationship (QSAR) [53,54]. The Hansen distances of the dyes to water and fiber materials were calculated based on HSPiP 4.1.07 and used to estimate the different affinities between dyes, water, and fiber materials. 3.2. Design and Feasibility Analysis of Dyes 3.2.1. Route Design of Hemicyanine Dyes Figure 5. General method for prepare designed dyes. (a) formation of pyridinium; (b) preparation of designed dye. Figure 5. General method for prepare designed dyes. (a) formation of pyridinium; (b) preparation of designed dye. Figure 5. General method for prepare designed dyes. (a) formation of pyridinium; (b) preparation of designed dye. Figure 5. General method for prepare designed dyes. (a) formation of pyridinium; (b) preparation of designed dye. As shown in Figure 5, the hemciayanine dyes could be prepared by 2 steps; the chemical structures of raw materials, intermediates, and dyes are shown in Figure S1, while the geometry structures of the designed dyes are shown in Figure S2. Z1 (DYE-BD) could be easily synthesized in two steps following the references [15–18], and the route of the synthesis can serve as an example to illustrate the designed routes. The nitrogen center of pyridine features a basic lone pair of electrons; Although the Z1 in Figure 5 has been reported and investigated by researchers [15–18], the properties of other chemicals could be changed by diﬀerent substituent groups, and the reactivity of intermediates would be changed as well. 3.2.1. Route Design of Hemicyanine Dyes yp free energies were estimated at b3lyp/6-31G i i l h f h di Thus, analysis and design of proper synthesis routes are important and necessary. The feasibilities of the routes were analyzed by using the computational methods. 3.2. Design and Feasibility Analysis of Dyes Geometry structures of the designed b3lyp method with 6-31G(d) basis sets [41 3.2.1. Route Design of Hemicyanine Dyes yp free energies were estimated at b3lyp/6-31G i i l th f th di As shown in Figure 2, the ene −0.35420 hartree, so the energy gap 3.1. Software and Calculation Methods hartree) orbital was greater than the energy gap between the HOMO orbital of dye Z5 orbitals and 1O2 LUMO orbital, which means that dye Z5 was more reactive than dye Z2 in terms of HOMO orbitals; similarly but different to the HOMO orbital, the energy gap between the LUMO (−0.10124 hartree) orbital of dye Z5 and O2− HOMO (−0.12516 hartree) orbital was smaller than the energy gap between the LUMO (−0.09660 hartree) orbitals of dye Z2 and O2− HOMO orbital. Thus, dye Z2 would be more stable than dye Z5 when it reacts with reactive oxygen species. However, when compared with Z1, the energy of the HOMO and LUMO orbitals of dye Z2 and Z5 were lower than that of dye Z1, so it is hard to determine whether dyes Z2 and Z5 have better photo-stability than dye Z1 A o di to Fi u e 4 the a i u efle ti ity of the Z1 dyed fab i lo t about 20% of i itial olo Geometry structures of the designed dye were optimized by using Gaussian 09 based on the b3lyp method with 6-31G(d) basis sets [41–43], and their charge distributions, enthalpies, and Gibbs free energies were estimated at b3lyp/6-31G(d) level [41,44–46], while the maximum absorption and emission wavelengths of them were predicted by using time-dependent density functional theory (TD-DFT) at the b3lyp/6-31G(d) level [47–52]. ECOSAR was used to predict aquatic toxicity of the chemicals by using the database based on the quantitative structure–activity relationship (QSAR) [53,54]. The Hansen distances of the dyes to water and ﬁber materials were calculated based on HSPiP 4.1.07 and used to estimate the diﬀerent aﬃnities between dyes, water, and ﬁber materials. Int. J. Mol. Sci. 2019, 20, 5971 3 1 Software and Calculat 10 of 16 3.2. Design and Feasibility Analysis of Dyes Geometry structures of the designed b3lyp method with 6-31G(d) basis sets [41– consequently, pyridine is a strong nucleophile. Th form pyridinium, while the second reaction is a typ 3.2.2. Analysis of Charge Density of Intermediates the methyl group on the pyridinium and the carbonyl group in 4-diethylaminobenzaldehyde. Although the Z1 in Figure 5 has been reported and investigated by researchers [15–18], the properties of other chemicals could be changed by different substituent groups, and the reactivity of intermediates would be changed as well. Thus, analysis and design of proper synthesis routes are The ﬁrst step of the preparation route of designed dyes in Figure 5 is a formation of quaternary pyridinium salts on pyridine derivatives. Due to the existence of unconjugated lone pair electrons on N in pyridine rings, all six derivatives of pyridine can form the corresponding pyridinium salts easily (see Figures S3–S5). The second step reaction, a nucleophilic reaction, is between the pyridinium salts and 4-diethylaminobenzaldehyde, while the reactive sites on the pyridinium salts are carbons bearing more electrons and more negative charges. Figure 6 displays the charge distributions of the pyridinium intermediates. As a result of calculation, for the pyridinium salt (g) in Figure 6, the carbon, with an electron density of −0.391, in the methyl group should be the reactive site when it reacts with 4-diethylaminobenzaldehyde, while the 11 of 16 11 of 16 Int. J. Mol. Sci. 2019, 20, 5971 methyl group in compound B1 is more reactive than methyl group in compound A1, which can be conﬁrmed in the references [55–61]. Thus, the more negative charges of carbon in methyl group, the more reactive the methyl group will be. Sulfonic acid and carboxyl groups have a complex inﬂuence on the charge of the atoms in pyridinium salts. The carboxyl group (2 position) increases the electronegativity of the carbon in methyl group slightly, but the carboxyl (3 position) decreases the charge on the carbon of the methyl group. Thus, the methyl group of compounds B2, B3, B4, and B6 should have the similar reactivity to that of the compound B1, while the reactivity of methyl group in compound B5 should be lower than that of the compound B1. Thus, compounds B2, B3, B4, and B6 could react with 4-diethylaminobenzaldehyde, but the compound B5 may be diﬃcult to or may not react with it. Int. J. Mol. Sci. 2019, 20, x 10 of 15 important and necessary. The feasibilities of the routes were analyzed by using the computational methods. 3 2 2 Analysis of Charge Density of Intermediates .2.2. Analysis of Charge Density of Intermediates Figure 6. consequently, pyridine is a strong nucleophile. Th form pyridinium, while the second reaction is a typ 3.2.2. Analysis of Charge Density of Intermediates Electronegativity of raw materials and intermediates calculated by Gaussian 09. (A1) 4-Picoline; (A2) 4-Methylpyridine-3-sulfonic acid; (A3) 4-Methylpyridine-2-sulfonic acid; (A4) 5-Methyl-3-pyridinesulfonic acid; (A5) 4-Methylnicotinic acid; (A6) 4-Methyl-pyridine- 2-carboxyl acid; (B1) 1-ethyl-4-methylpyridin- 1-ium bromide; (B2) 1-ethyl-4-methyl-3-sulfopyridin-1-ium bromide; (B3) 1-ethyl-4-methyl-2-sulfopyridin-1-ium bromide; (B4) 1-ethyl-3-methyl-5-sulfopyridin-1-ium bromide; (B5) 3-carboxy-1-ethyl-4-methylpyridin-1-ium bromide; (B6) 2-carboxy-1-ethyl-4-methylpyridin-1-ium bromide. Figure 6. Electronegativity of raw materials and intermediates calculated by Gaussian 09. (A1) 4-Picoline; (A2) 4-Methylpyridine-3-sulfonic acid; (A3) 4-Methylpyridine-2-sulfonic acid; (A4) 5-Methyl-3-pyridinesulfonic acid; (A5) 4-Methylnicotinic acid; (A6) 4-Methyl-pyridine- 2-carboxyl acid; (B1) 1-ethyl-4-methylpyridin- 1-ium bromide; (B2) 1-ethyl-4-methyl-3-sulfopyridin-1-ium bromide; (B3) 1-ethyl-4-methyl-2-sulfopyridin-1-ium bromide; (B4) 1-ethyl-3-methyl-5-sulfopyridin-1-ium bromide; (B5) 3-carboxy-1-ethyl-4-methylpyridin-1-ium bromide; (B6) 2-carboxy-1-ethyl-4-methylpyridin-1-ium bromide. Figure 6. Electronegativity of raw materials and intermediates calculated by Gaussian 09. (A1) 4-Picoline; (A2) 4-Methylpyridine-3-sulfonic acid; (A3) 4-Methylpyridine-2-sulfonic acid; (A4) 5-Methyl-3-pyridinesulfonic acid; (A5) 4-Methylnicotinic acid; (A6) 4-Methyl-pyridine- 2-carboxyl acid; (B1) 1-ethyl-4-methylpyridin- 1-ium bromide; (B2) 1-ethyl-4-methyl-3-sulfopyridin-1-ium bromide; (B3) 1-ethyl-4-methyl-2-sulfopyridin-1-ium bromide; (B4) 1-ethyl-3-methyl-5-sulfopyridin-1-ium bromide; (B5) 3-carboxy-1-ethyl-4-methylpyridin-1-ium bromide; (B6) 2-carboxy-1-ethyl-4-methylpyridin-1-ium bromide. Figure 6. Electronegativity of raw materials and intermediates calculated by Gaussian 09. (A1) 4-Picoline; (A2) 4-Methylpyridine-3-sulfonic acid; (A3) 4-Methylpyridine-2-sulfonic acid; (A4) 5-Methyl-3-pyridinesulfonic acid; (A5) 4-Methylnicotinic acid; (A6) 4-Methyl-pyridine- 2-carboxyl acid; (B1) 1-ethyl-4-methylpyridin- 1-ium bromide; (B2) 1-ethyl-4-methyl-3-sulfopyridin-1-ium bromide; (B3) 1-ethyl-4-methyl-2-sulfopyridin-1-ium bromide; (B4) 1-ethyl-3-methyl-5-sulfopyridin-1-ium bromide; (B5) 3-carboxy-1-ethyl-4-methylpyridin-1-ium bromide; (B6) 2-carboxy-1-ethyl-4-methylpyridin-1-ium bromide. The first step of the preparation route of desig 3.2.3. Analysis of Enthalpy and Gibbs Free Energy The first step of the preparation route of desig 3.2.3. Analysis of Enthalpy and Gibbs Free Energy pyridinium salts on pyridine derivatives. Due to the existence of unconjugated lo N in pyridine rings all six derivatives of pyridine can form the correspondin The enthalpy of a thermodynamic system is deﬁned in Equation (2) [62,63]. f pyridine can form the corresponding pyridinium salts step reaction, a nucleophilic reaction, is between the aldehyde while the reactive sites on the pyridinium salts H = U + pV (2) the (2) pyridinium salts and 4 diethylaminobenzaldehyde, while the reactive sites on the pyridinium salts are carbons bearing more electrons and more negative charges. Figure 6 displays the charge distributions of the pyridinium intermediates. As a result of H is the enthalpy (SI unit: Joule); U is the internal energy (SI unit: Joule); p is pressure (SI unit: Pascal); V is volume (SI unit: m3). Int. J. Mol. Sci. 2019, 20, 5971 12 of 16 The standard enthalpy of formation (∆fH) of a compound is the change of enthalpy during the formation of 1 mole of the substance from its constituent elements in their standard states. The standard enthalpy change (∆rH) of any reaction can be calculated from the standard enthalpies of formation of reactants and products using Hess’s law. It could be illustrated in Equation (3): ∆rH = X ∆fH(products) − X ∆fH(reactants) (3) (3) The Gibbs free energy is deﬁned in Equation (4): G = U + pV −TS = H −TS (4) (4) T is the temperature (SI unit: Kelvin); S is the entropy (SI unit: Joule per kelvin). The Gibbs free energy change (∆rG) of any reaction can be calculated based on Equation (5): T is the temperature (SI unit: Kelvin); S is the entropy (SI unit: Joule per kelvin). The Gibbs free energy change (∆rG) of any reaction can be calculated based on Equation (5): T is the temperature (SI unit: Kelvin); S is the entropy (SI unit: Joule per kelvin). ∆rG = X ∆fG(products) − X ∆fG(reactants) (5) (5) According to the ﬁrst step reaction between derivatives of pyridine and 1-bromoethane (Figure 5a), the reaction could be illustrated as A + D→B, so the ∆rH could be calculated by ∆rH = ∆fH(B) −∆fH(A) −∆fH(D), while the ∆G could be illustrated as ∆rG = ∆fG(B) −∆fG(A) −∆fG(D). 1Hartree = 627.509 kcal mol−1 = 27.2116 eV. The first step of the preparation route of desig 3.2.3. Analysis of Enthalpy and Gibbs Free Energy The changes of enthalpy and Gibbs free energies in the ﬁrst step reactions are calculated by Gaussian 09 in ethanol at 80 ◦C and shown in Table 6; similarly, the ∆rH values of the second reaction in Figure 5b could be illustrated as ∆rH = ∆fH(Z) + ∆fH(F) −∆fH(B) −∆fH(E), while the ∆G values in Table 7 could be calculated by ∆rG = ∆fG(Z) + ∆fG(F) −∆fG(B) −∆fG(E). Meanwhile, the bromide ion is not involved in the second step reactions. It seems all designed ﬂuorescent dyes could be produced according to the thermodynamic analysis. Table 6. The enthalpy and Gibbs free energy values of chemicals in ﬁrst step and changes of them in ethanol under 353K based on Gaussian 09. Item ∆fH (Hartree) ∆fG (Hartree) Item ∆fH (Hartree) ∆fG (Hartree) ∆rH (Hartree) ∆G (Hartree) A1 −287.48332 −287.53033 B1 −2938.370434 −2938.438338 −0.022028 −0.002863 A2 −911.256057 −911.316383 B2 −3562.148829 −3562.223358 −0.027686 −0.00183 A3 −911.264174 −911.326257 B3 −3562.139176 −3562.218212 −0.009916 0.01319 A4 −911.258738 −911.321063 B4 −3562.13095 −3562.212119 −0.007126 0.014089 A5 −476.036178 −476.091818 B5 −3126.917758 −3126.99308 −0.016494 0.003883 A6 −476.039513 −476.096557 B6 −3126.905813 −3126.980969 −0.001214 0.020733 D −2650.865086 −2650.905145 - - - - - Table 7. The enthalpy and Gibbs free energy values of chemicals in second step and changes of them in ethanol under 353K based on Gaussian 09. Item ∆fH (1Hartree) ∆fG (Hartree) Item ∆fH (Hartree) ∆fG (Hartree) ∆H (Hartree) ∆G (Hartree) B1 −366.48955 −366.545969 Z1 −848.007794 −848.103322 0.015431 0.019032 B2 −990.24633 −990.315512 Z2 −1471.772855 −1471.88174 0.00715 0.010157 B3 −990.238486 −990.308045 Z3 −1471.762939 −1471.871835 0.009222 0.012595 B4 −990.249985 −990.320103 Z4 −1471.761824 −1471.871864 0.021836 0.024624 B5 −555.034403 −555.097776 Z5 −1036.548901 −1036.654302 0.019177 0.019859 B6 −555.02708 −555.092909 Z6 −1036.547848 −1036.651849 0.012907 0.017445 E −557.924048 −557.992916 F −76.390373 −76.416531 - - 1Hartree = 627.509 kcal mol−1 = 27.2116 eV. Table 6. The enthalpy and Gibbs free energy values of chemicals in ﬁrst step and changes of them in ethanol under 353K based on Gaussian 09. Item ∆fH (Hartree) ∆fG (Hartree) Item ∆fH (Hartree) ∆fG (Hartree) ∆rH (Hartree) ∆G (Hartree) A1 −287.48332 −287.53033 B1 −2938.370434 −2938.438338 −0.022028 −0.002863 A2 −911.256057 −911.316383 B2 −3562.148829 −3562.223358 −0.027686 −0.00183 A3 −911.264174 −911.326257 B3 −3562.139176 −3562.218212 −0.009916 0.01319 A4 −911.258738 −911.321063 B4 −3562.13095 −3562.212119 −0.007126 0.014089 A5 −476.036178 −476.091818 B5 −3126.917758 −3126.99308 −0.016494 0.003883 A6 −476.039513 −476.096557 B6 −3126.905813 −3126.980969 −0.001214 0.020733 D −2650.865086 −2650.905145 - - - - - Table 7. 3.3. Sample Synthesis Route Preparation of Z2: 4-Methylpyridine-3-sulfonicacid (A2) (1.7319 g, 0.01 mol) and sodium bicarbonate (NaHCO3, 0.84 g, 0.01 mol) were dissolved in a mixture solution (ethanol/water = 2:1) and stirred at room temperature for several minutes. Then, 1-bromoethane (1.19 g, 0.011 mol) was added into the mixture solution, and the mixture was reﬂuxed under 80 ◦C for 6 h. Afterward, 1.77 g of 4-diethylaminobenzaldehyde (0.01 mol) and several drops of pyridine were added into the mixture; keeping the temperature for 6 h, a crude product was obtained after the solvent was removed by vacuum distillation. Preparation of Z5: 4-Methylpyridine-3-carboxylic acid (A5) (1.3714 g, 0.01 mol) and sodium bicarbonate (NaHCO3, 0.84 g, 0.01 mol) were dissolved in a mixture solution (ethanol/water = 2:1) and stirred at room temperature for several minutes. Then, 1-bromoethane (1.19 g, 0.011 mol) was added into the mixture solution, and the mixture was reﬂuxed under 80 ◦C for 6 h. Afterward, 1.77 g of 4-diethylaminobenzaldehyde (0.01 mol) and several drops of pyridine were added into the mixture; keeping the temperature for 6 h, a crude product was obtained after the solvent was removed by vacuum distillation. The first step of the preparation route of desig 3.2.3. Analysis of Enthalpy and Gibbs Free Energy The enthalpy and Gibbs free energy values of chemicals in second step and changes of them in ethanol under 353K based on Gaussian 09. Table 6. The enthalpy and Gibbs free energy values of chemicals in ﬁrst step and changes of them in ethanol under 353K based on Gaussian 09. Table 7. The enthalpy and Gibbs free energy values of chemicals in second step and changes of them in ethanol under 353K based on Gaussian 09. Item ∆fH (1Hartree) ∆fG (Hartree) Item ∆fH (Hartree) ∆fG (Hartree) ∆H (Hartree) ∆G (Hartree) B1 −366.48955 −366.545969 Z1 −848.007794 −848.103322 0.015431 0.019032 B2 −990.24633 −990.315512 Z2 −1471.772855 −1471.88174 0.00715 0.010157 B3 −990.238486 −990.308045 Z3 −1471.762939 −1471.871835 0.009222 0.012595 B4 −990.249985 −990.320103 Z4 −1471.761824 −1471.871864 0.021836 0.024624 B5 −555.034403 −555.097776 Z5 −1036.548901 −1036.654302 0.019177 0.019859 B6 −555.02708 −555.092909 Z6 −1036.547848 −1036.651849 0.012907 0.017445 E −557.924048 −557.992916 F −76.390373 −76.416531 - - 1Hartree = 627.509 kcal mol−1 = 27.2116 eV. Table 7. The enthalpy and Gibbs free energy values of chemicals in second step and changes of them in ethanol under 353K based on Gaussian 09. Table 7. The enthalpy and Gibbs free energy values of chemicals in second step and changes of them in ethanol under 353K based on Gaussian 09. Int. J. Mol. Sci. 2019, 20, 5971 13 of 16 13 of 16 4. Conclusions After computational analyses of design, synthesis, ﬂuorescent properties, photo-stability, as well as estimation of aquatic environmental toxicity of the ﬂuorescent dyes, a dye molecule (Z2) was identiﬁed as possessing desired properties and potential environmental friendliness. Both dyes Z2 and Z5 were prepared and proven to possess ideal ﬂuorescent properties. The Z2 dyed fabrics showed improved light stability, while the Z5 dyed ones shown reduced light stability. The structures of the dye Z2 present features of containing a sulfonic acid group, which could reduce toxicity of the chemicals. Both carboxyl and sulfonic acid groups can improve dyeing properties/aﬃnity to ﬁve polymers. The diﬀerent substituent groups and positions have diﬀerent inﬂuences on the photo-stability of dyes. Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/20/23/ 5971/s1. Author Contributions: Conceptualization, S.T.; software, S.T.; supervision, G.C. and G.S.; writing—original draft, S.T. Author Contributions: Conceptualization, S.T.; software, S.T.; supervision, G.C. and G.S.; writing—original draft, S.T. Funding: This research was funded by Key R & D plan of Jiangsu Province, BE2019001-3. Acknowledgments: Songsong Tang is grateful for a scholarship from Soochow University for ﬁnancially sponsoring his research activities at University of California, Davis. Acknowledgments: Songsong Tang is grateful for a scholarship from Soochow University for ﬁnancially sponsoring his research activities at University of California, Davis. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. Conﬂicts of Interest: The authors declare no conﬂict of interest. References Kolmakov, K.; Hebisch, E.; Wolfram, T.; Lars, A.; Nordwig, L.A.; Wurm, C.A.; Ta, H.; Westphal, V.; Belov, V.S.W. Far-Red Emitting Fluorescent Dyes for Optical Nanoscopy: Fluorinated Silicon–Rhodamines (SiRF Dyes) 7. Kolmakov, K.; Hebisch, E.; Wolfram, T.; Lars, A.; Nordwig, L.A.; Wurm, C.A.; Ta, H.; Westphal, V.; Belov, V.S.W. Far-Red Emitting Fluorescent Dyes for Optical Nanoscopy: Fluorinated Silicon–Rhodamines (SiRF Dyes) and Phosphorylated Oxazines. Chem. Eur. J. 2015, 21, 13344–13356. [CrossRef] 8. Sahl, S.J.; Hell, S.W.; Jakobs, S. Fluorescence nanoscopy in cell biology. Nat. Rev. Mol. Cell. Biol. 2017, 18, 685–701. [CrossRef] 9. Ahlström, L.H.; Eskilsson, C.S.; Björklund, E. Determination of banned azo dyes in consumer goods. TrAC Trends Anal. Chem. 2005, 24, 49–56. [CrossRef] 10. Brüschweiler, B.J.; Küng, S.; Bürgi, D.; Muralt, L.; Erich Nyfeler, E. Identiﬁcation of non-regulated aromatic amines of toxicological concern which can be cleaved from azo dyes used in clothing textiles. Regul. Toxicol. Pharmacol. 2014, 69, 263–272. [CrossRef] 11. Fadadu, K.B.; Soni, S.S. Spectral sensitization of TiO2 by new hemicyanine dyes in dye solar cell yielding enhanced photovoltage: Probing chain length eﬀect on performance. Electrochim. Acta 2013, 88, 270–277. [CrossRef] 12. J˛edrzejewska, B.; Kabatc, J.; P ˛aczkowski, J. Dichromophoric hemicyanine dyes. Synthesis and spectroscopic investigation. Dye. Pigment. 2007, 74, 262–268. [CrossRef] 13. Zhang, C.; Liu, Z.; Chen, Y.; Ma, J.; Chen, H.S.; Zhang, M.L. Density Functional Theory Study on Organic Dye Sensitizers Containing Bis-dimethylﬂuorenyl Amino Benzofuran. Chin. J. Chem. Phys. 2009, 22, 489–496. [CrossRef] 14. Wang, Z.S.; Li, F.Y.; Huang, C.H. Photocurrent enhancement of hemicyanine dyes containing RSO3-group through treating TiO2 ﬁlms with hydrochloric acid. J. Phys. Chem. B 2001, 105, 9210–9217. [CrossRef] through treating TiO2 ﬁlms with hydrochloric acid. J. Phys. Chem. B 2001, 105, 9210–9217. [CrossRef] 15. Das, P.J.; Das, J.; Das, D. An Eﬃcient Conversion of Alcohols to Alkyl Bromides Using Pyridinium Based Ionic Liquids: A Green Alternative to Appel Reaction Asian J Chem 2018 30 651–654 [CrossRef] 15. Das, P.J.; Das, J.; Das, D. An Eﬃcient Conversion of Alcohols to Alkyl Bromides Using Pyridinium Based Ionic Liquids: A Green Alternative to Appel Reaction. Asian J. Chem. 2018, 30, 651–654. [CrossRef] 16. Jin, F.; Cai, Z.B.; Huang, J.Q.; Sheng, L.L.; Yu, P.T. Investigation of two-photon absorption properties in new A–D–A compounds emitting blue and yellow ﬂuorescence. J. Mol. Struct. 2015, 1093, 33–38. [CrossRef] 17. Krieg, R.; Eitner, A.; Günther, W.; Halbhuber, K.J. References 1. Das, P.P.; Roy, A.; Agarkar, S.; Devi, P.S. Hydrothermally synthesized ﬂuorescent Zn2SnO4 nanoparticles for dye sensitized solar cells. Dye. Pigment. 2018, 154, 303–313. [CrossRef] 1. Das, P.P.; Roy, A.; Agarkar, S.; Devi, P.S. Hydrothermally synthesized ﬂuorescent Zn2SnO4 nanoparticles for dye sensitized solar cells. Dye. Pigment. 2018, 154, 303–313. [CrossRef] y y g 2. Ooyama, Y.; Nagano, T.; Inoue, S.; Imae, I.; Komaguchi, K.; Ohshita, J.; Harima, Y. Dye-Sensitized Solar Cells Based on Donor-π-Acceptor Fluorescent Dyes with a Pyridine Ring as an Electron-Withdrawing-Injecting Anchoring Group. Chem. Eur. J. 2011, 17, 14837–14843. [CrossRef] [PubMed] 2. Ooyama, Y.; Nagano, T.; Inoue, S.; Imae, I.; Komaguchi, K.; Ohshita, J.; Harima, Y. Dye-Sensitized Solar Cells Based on Donor-π-Acceptor Fluorescent Dyes with a Pyridine Ring as an Electron-Withdrawing-Injecting Anchoring Group. Chem. Eur. J. 2011, 17, 14837–14843. [CrossRef] [PubMed] g p 3. Cao, Y.; Wang, X.; Shi, X.; Clee, S.M.; McGeer, P.L.; Wolf, M.O.; Orvig, C. Biological Imaging with Medium-Sensitive Bichromatic Flexible Fluorescent Dyes. Angew. Chem. Int. Ed. 2017, 56, 15603–15606. [CrossRef] [PubMed] 3. Cao, Y.; Wang, X.; Shi, X.; Clee, S.M.; McGeer, P.L.; Wolf, M.O.; Orvig, C. Biological Imaging with Medium-Sensitive Bichromatic Flexible Fluorescent Dyes. Angew. Chem. Int. Ed. 2017, 56, 15603–15606. [CrossRef] [PubMed] 4. Li, Q.; Liu, W.; Wu, J.; Zhou, B.; Niu, G.; Zhang, H.; Ge, J.; Wang, P. Deep-red to near-infrared ﬂuorescent dyes: Synthesis, photophysical properties, and application in cell imaging. Spectrochim. Acta A 2016, 164, 8–14. [CrossRef] [PubMed] 4. Li, Q.; Liu, W.; Wu, J.; Zhou, B.; Niu, G.; Zhang, H.; Ge, J.; Wang, P. Deep-red to near-infrared ﬂuorescent dyes: Synthesis, photophysical properties, and application in cell imaging. Spectrochim. Acta A 2016, 164, 8–14. [CrossRef] [PubMed] 5. Tang, Y.; Lee, D.; Wang, J.; Li, G.; Yu, J.; Lin, W.; Yoon, J. Development of ﬂuorescent probes based on protection–deprotection of the key functional groups for biological imaging. Chem. Soc. Rev. 2015, 44, 5003–5015. [CrossRef] 5. Tang, Y.; Lee, D.; Wang, J.; Li, G.; Yu, J.; Lin, W.; Yoon, J. Development of ﬂuorescent probes based on protection–deprotection of the key functional groups for biological imaging. Chem. Soc. Rev. 2015, 44, 5003–5015. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5971 14 of 16 14 of 16 6. Cotruvo, J.A., Jr.; Aron, A.T.; Ramos-Torres, K.M.; Chang, C.J. Synthetic ﬂuorescent probes for studying copper in biological systems. Chem. Soc. Rev. 2015, 44, 4400–4414. [CrossRef] [PubMed] 7. References Optimization of heterocyclic 4-hydroxystyryl derivatives for histological localization of endogenous and immunobound peroxidase activity. Biotech. Histochem. 2007, 82, 235–262. [CrossRef] 18. Ikeda, T.; Tadaki, Y.; Funajima, R.; Tatewaki, Y.; Okad, S. Aminostilbazolium Derivatives Substituted by Hydroxyethyl Groups for Second-Order Nonlinear Optics. Mol. Cryst. Liq. Cryst. 2011, 539, 142–147. [CrossRef] 19. Wei, B.X.; Zhao, L.; Wang, T.J.; Gao, H.; Wu, H.X.; Jin, Y. Photo-stability of TiO2 particles coated with several transition metal oxides and its measurement by rhodamine-B degradation. Adv. Powder Technol. 2013, 24, 708–713. [CrossRef] 20. Yang, S.; Tian, H.; Xiao, H.; Shang, H.; Gong, X.D.; Yao, S.; Chen, K. Photodegradation of cyanine and merocyanine dyes. Dye. Pigment. 2001, 49, 93–101. [CrossRef] 21. Chang, X.P.; Gao, Y.J.; Fang, W.H.; Cui, G.; Thiel, W. Quantum Mechanics/Molecular Mechanics Study on the Photoreactions of Dark-and Light-Adapted States of a Blue-Light YtvA LOV Photoreceptor. Angew. Chem. 2017, 129, 9469–9473. [CrossRef] 2. Zhan, Y.Z.; Zhao, X.; Wang, W. Theoretical study of the interaction energy of benzodifuranone dye mole rings. Color. Technol. 2017, 133, 50–56. [CrossRef] 23. Zhan, Y.; Zhao, X.; Wang, W. Synthesis of phthalimide disperse dyes and study on the interaction energy. Dye. Pigment. 2017, 146, 240–250. [CrossRef] 24. Sokołowska, J.; Czajkowski, W.; Podsiadły, R. The photostability of some ﬂuorescent disperse dyes derivatives of coumarin. Dye. Pigment. 2001, 49, 187–191. [CrossRef] 25. Patel, D.R.; Bilimoriya, J.T.; Patel, B.M.; Paresh, S.; Patel, S.H.; Mehta, K.M.; Patel, S.A.; Patel, K.C.; Patel, B.S.; Sahoo, S.K. Monoazo Styryl Quinazolinone Reactive Dyes: Their Synthesis, Application and Density Function Theory (DFT) Calculation. Proc. Nat. Acad. Sci. India Sect. A 2017, 87, 339–348. [CrossRef] 26. Cash, G.G. Prediction of chemical toxicity to aquatic organisms: ECOSAR vs. Microtox® Assay. Environ. Toxicol. Water Qual. 1998, 13, 211–216. [CrossRef] 27. Tang, S.; Zhao, C.; Chen, G.; Sun, G. A study on computerized selection of ﬂuorescent dyes for environmentally friendly textile applications. Dye. Pigment. 2019, 165, 256–263. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5971 15 of 16 15 of 16 28. Karst, D.; Yang, Y. Using the solubility parameter to explain disperse dye sorption on polylactide. J. Appl. Polym. Sci. 2005, 96, 416–422. [CrossRef] 29. Tamizifar, M.; Sun, G. Control of surface radical graft polymerization on polyester ﬁbers by using Hansen solubility parameters as a measurement of the aﬃnity of chemicals to materials. RSC Adv. 2017, 7, 13299–13303. [CrossRef] 30. Knox, B.H.; Weigmann, H.D.; Scott, M.G. References Interactions of nonaqueous solvents with textile ﬁbers: Part V: Application of the solubility parameter concept to polyester ﬁber-solvent interactions. Text. Res. J. 1975, 45, 203–217. [CrossRef] 31. Montgomery, J.A., Jr.; Frisch, M.J.; Ochterski, J.W.; Petersson, G. A complete basis set model chemistry. VI. Use of density functional geometries and frequencies. J. Chem. Phys. 1999, 110, 2822–2827. [CrossRef] 32. Baboul, A.G.; Curtiss, L.A.; Redfern, P.C.; Redfern, P.C.; Raghavachari, K. Gaussian-3 theory using density functional geometries and zero-point energies. J. Chem. Phys. 1999, 110, 7650–7657. [CrossRef] 33. Zhao, Y.; Truhlar, D.G. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: Two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor. Chem. Acc. 2008, 120, 215–241. 34. Curtiss, L.A.; Redfern, P.C.; Raghavachari, K.; Raghavachari, K.; Pople, J.A. Gaussian-3X (G3X) theory: Use of improved geometries, zero-point energies, and Hartree–Fock basis sets. J. Chem. Phys. 2001, 114, 108–117. [CrossRef] 35. Cai, Z.L.; Crossley, M.J.; Reimers, J.R.; Kobayashi, R.; Amos, R.D. Density functional theory for charge transfer: The nature of the N-bands of porphyrins and chlorophylls revealed through CAM-B3LYP, CASPT2, and SAC-CI calculations. J. Phys. Chem. B 2006, 110, 15624–15632. [CrossRef] 36. Civalleri, B.; Zicovich-Wilson, C.M.; Valenzano, L.; Ugliengo, P. B3LYP augmented with an empirical dispersion term (B3LYP-D) as applied to molecular crystals. Cryst Eng Comm 2008, 10, 405–410. [CrossRef] è d f h b f f d l 36. Civalleri, B.; Zicovich-Wilson, C.M.; Valenzano, L.; Ugliengo, P. B3LYP augmented with an empirical dispersion term (B3LYP-D) as applied to molecular crystals. Cryst Eng Comm 2008, 10, 405–410. [CrossRef] 37. Jacquemin, D.; Preat, J.; Perpète, E.A. A TD-DFT study of the absorption spectra of fast dye salts. Ch Ph L tt 2005 410 254 259 [C R f] 37. Jacquemin, D.; Preat, J.; Perpète, E.A. A TD-DFT study of the absorption spectra of fast dye salts. Chem. Phys. Lett. 2005, 410, 254–259. [CrossRef] 38. Guillaumont, D.; Nakamura, S. Calculation of the absorption wavelength of dyes using time-dependent density-functional theory (TD-DFT). Dye. Pigment. 2000, 46, 85–92. [CrossRef] 39. Xie, L.; Chen, Y.; Wu, W.; Guo, H.; Zhao, J.; Yu, X. Fluorescent coumarin derivatives with large stokes shift, dual emission and solid state luminescent properties: An experimental and theoretical study. Dye. Pigment. 2012, 92, 1361–1369. [CrossRef] 40. Shao, J.; Sun, H.; Guo, H.; Ji, S.; Zhao, J.; Wu, W.; James, T.D. References A highly selective red-emitting FRET ﬂuorescent molecular probe derived from BODIPY for the detection of cysteine and homocysteine: An experimental and theoretical study. Chem. Sci. 2012, 3, 1049–1061. [CrossRef] 41. Huynh, H.V.; He, X.; Baumgartner, T. Halochromic generation of white light emission using a single dithienophosphole luminophore. Chem. Commun. 2013, 49, 4899–4901. [CrossRef] 42. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Petersson, G.A. Gaussian 09, revision A. 1. Gaussian Inc. Wallingford CT 2009, 27, 34. 43. Netzeva, T.I.; Pavan, M.; Worth, A.P. Review of (quantitative) structure–activity relationships for acute aquatic toxicity. QSAR Comb. Sci. 2008, 27, 77–90. [CrossRef] 44. Cronin, M.T.D. Predicting Chemical Toxicity and Fate. In The Use by Governmental Regulatory Agencies of quantitative Structure-Activity Relationships and Expert Systems to Predict Toxicity; CRC Press: Boca Raton, FL, USA, 2004; pp. 414–425. 45. Nickell, J.R.; Culver, J.P.; Janganati, V.; Zheng, G.; Dwoskin, L.P.; Crooks, P.A. 1, 4-Diphenalkylpiperidines: A new scaﬀold for the design of potent inhibitors of the vesicular monoamine transporter-2. Bioorganic Med. Chem. Lett. 2016, 26, 2997–3000. [CrossRef] 46. Hauser, S.A.; Korinth, V.; Herdtweck, E.; Cokoja, M.; Herrmann, W.A.; Kühn, F.E. Chromophoric Lewis Base Adducts of Methyltrioxorhenium: Synthesis, Catalysis and Photochemistry. Eur. J. Inorg. Chem. 2010, 26, 4083–4090. [CrossRef] 47. Lacroix, P.G.; Munoz, M.C.; Gaspar, A.B.; Gaspar, A.B.; Real, J.A.; Bonhommeau, S.; Rodriguez, V.; Nakatani, K. Synthesis, crystal structures, and solid state quadratic nonlinear optical properties of a series of stilbazolium cations combined with gold cyanide counter-ion. J. Mater. Chem. 2011, 21, 15940–15949. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5971 16 of 16 16 of 16 48. Agustí, G.; Gaspar, A.B.; Muñoz, M.C.; Lacroix, P.G.; Real, J.A. Spin Crossover and Paramagnetic Behaviour in Two-Dimensional Iron (ii) Coordination Polymers with Stilbazole Push–Pull Ligands. Aust. J. Chem. 2009, 62, 1155–1165. [CrossRef] 9. Cicchi, S.; Fabbrizzi, P.; Ghini, G.; Brandi, A.; Foggi, P.; Marcelli, A.; Botta, C. Pyrene-Excimers-Based Ante Systems. Chem. Eur. J. 2009, 15, 754–764. [CrossRef] 50. Li, D.; Sun, X.; Shao, N.; Wu, J.; Tian, Y. Self-Assembly of Organic Chromosphere Cations with Inorganic Lanthanide (III) Complex Counterions. Zeitschrift Für Anorganische Und Allgemeine Chemie 2014, 640, 2283–2286. [CrossRef] 51. Hao, F.; Zhang, X.; Tian, Y.; Zhou, H.; Li, L.; Wu, J.; Zhou, G. Design, crystal structures and enhanced frequency-upconverted lasing eﬃciencies of a new series of dyes from hybrid of inorganic polymers and organic chromophores. J. Mater. Chem. 2009, 19, 9163–9169. [CrossRef] 52. Guggenheim, E.A. References Thermodynamics-an advanced treatment for chemists and physicists. In Thermodynamics—An Advanced Treatment for Chemists and Physicists, 7th ed.; Amsterdam, North-Holland: New York, NY, USA, 1985; p. 414. 53. Zumdahl, S.S. Thermochemistry. In Chemistry; Cengage Learning: Boston, MA, USA, 2008; p. 243. 54. Strickler, S.J.; Berg, R.A. Relationship between absorption intensity and ﬂuorescence lifetime of molecules. J. Chem. Phys. 1962, 37, 814–822. [CrossRef] 55. Mohammad, M.A.; Alhalaweh, A.; Velaga, S.P. Hansen solubility parameter as a tool to predict cocrystal formation. Int. J. Pharm. 2011, 407, 63–71. [CrossRef] 56. Hansen, C.M. Aspects of solubility, surfaces and diﬀusion in polymers. Prog. Org. Coat. 2004, 51, 55–66. [CrossRef] 57. Qi, H.; Zhao, C.; Qing, F.L.; Yan, K.; Sun, G. Antiwrinkle ﬁnishing of cotton fabrics with 5-(Carbonyloxy succinic)-benzene-1, 2, 4-tricarboxylic acid: Comparison with other acids. Ind. Eng. Chem. Res. 2016, 55, 11850–11856. [CrossRef] 58. Launay, H.; Hansen, C.M.; Almdal, K. Hansen solubility parameters for a carbon ﬁber/epoxy composite. Carbon 2007, 45, 2859–2865. [CrossRef] 59. Ji, B.; Zhao, C.; Yan, K.; Sun, G. Eﬀects of acid diﬀusibility and aﬃnity to cellulose on strength loss of polycarboxylic acid crosslinked fabrics. Carbohydr. Polym. 2016, 144, 282–288. [CrossRef] 60. Qin, C.; Qin, Z. Research on the ﬂuorescent whiteners for textiles. Text. Aux. 2005, 9, 1–3. 61. Christie, R.M. Review of Progress in Coloration and Related Topics. Fluoresc. Dye. 1993, 23, 1–18. 62. Bi, L.; Zhang, W.; QIN, C. Dyeing behaviors of three phthalocyanine dyes on acrylic fabric. China Dye. Finish. 2014, 8, 1–4. 63. Tang, S.; Qin, C.; Chen, G.; Bao, Z.M.; Hou, X.N. Synthesis and dyeing properties of three hemicyanine-based ﬂuorescent polymeric dyes. AATCC J. Res. 2018, 5, 19–25. [CrossRef] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W2602525120	https://www.epj-conferences.org/articles/epjconf/pdf/2017/04/epjconf_arena2017_06003.pdf	English	null	Development of an acoustic sensor for the future IceCube-Gen2 detector for neutrino detection and position calibration	EPJ web of conferences	2,017	cc-by	2,571	⋆e-mail: wickmann@physik.rwth-aachen.de © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). Development of an acoustic sensor for the future IceCube-Gen2 detector for neutrino detection and position calibration Stefan Wickmann1,⋆, Dmitry Eliseev1, Dirk Heinen1, Peter Linder1, Martin Rongen1, Franziska Scholz1, Lars Steffen Weinstock1, Christopher Wiebusch1, and Simon Zierke1 1III. Physikalisches Institut B, RWTH Aachen University, D-52056 Aachen, Germany Abstract. For the planned high-energy extension of the IceCube Neutrino Observatory in the glacial ice at the South Pole the spacing of detector modules will be increased with respect to IceCube. Because of these larger distances the quality of the geometry calibration based on pulsed light sources is expected to deteriorate. To counter this an independent acoustic geometry calibration system based on trilateration is introduced. Such an acoustic positioning system (APS) has already been developed for the Enceladus Explorer Project (EnEx), initiated by the DLR Space Administration. In order to inte- grate such APS-sensors into the IceCube detector the power consumption needs to be minimized. In addition, the frequency response of the front end electronics is optimized for positioning as well as the acoustic detection of neutrinos. The new design of the acoustic sensor and results of test measurements with an IceCube detector module will be presented. , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 DOI: 10.1051/ 7135 epjconf/201 06003 Development of an acoustic sensor for the future IceCube-Gen2 detector for neutrino detection and position calibration Stefan Wickmann1,⋆, Dmitry Eliseev1, Dirk Heinen1, Peter Linder1, Martin Rongen1, Franziska Scholz1, Lars Steffen Weinstock1, Christopher Wiebusch1, and Simon Zierke1 1III. Physikalisches Institut B, RWTH Aachen University, D-52056 Aachen, Germany 1 Introduction The IceCube Neutrino Observatory [1], located at the geographical South Pole, instruments about one cubic kilometer of Antarctic ice to detect neutrinos. The detector consists of 86 strings, each with 60 digital optical modules (DOM) forming a photomultiplier (PMT) array with a spacing of 125 m between the strings. Most of the DOMs are equally spaced (17 m) among the strings at a depth between 1450 m and 2450 m beneath the ice surface. Geometry calibration of the detector is performed with deployment data of the ice drilling and an optical calibration system, with which an accuracy of less than 1m can be achieved. IceCube-Gen2 is a planned extension of IceCube [2]. It will allow for the detection of high- energetic neutrinos with an increased detection rate and precision. To accomplish this an extension of the detector volume is planned by adding more strings to the existing IceCube array. The spacing be- tween these additional strings will therefore be increased to 240 m - 300 m, depending on the planned conﬁgurations of IceCube-Gen2. Because of the larger spacing between the strings the quality of the optical geometry calibration is expected to deteriorate due to the extinction length (20 m - 200 m) [3] of optical signals in the Antarctic ice. A possible solution could be the use of an independent geometry calibration system based on acoustic signals due to their larger attenuation length (~300 m, [4]). , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 DOI: 10.1051/ 7135 epjconf/201 06003 DOI: 10.1051/ 7135 epjconf/201 06003 The idea is to integrate acoustic sensors into the IceCube-Gen2 DOMs to use them for acoustic positioning independent of optical properties. In addition, such sensors could also be used for acoustic neutrino detection based on the thermo-acoustic eﬀect [5] [4]. The required steps to integrate such a system into IceCube-Gen2 would be to test the concept of acoustic positioning with a DOM, to develop a suitable acoustic sensor for the required frequency ranges of the acoustic positioning and the acoustic neutrino detection with the desired low power consumption and ﬁnally to integrate the sensor into the IceCube-Gen2 DOMs. Acoustic emitters could be deployed as separate modules and do not need to be integrated into the DOMs. 1 Introduction Such a sensor for in-ice acoustic positioning has already been developed by the Enceladus Ex- plorer Project (EnEx) [6]. Within this project navigation technology for a future space mission has been developed. A main aspect of the project has been the development of an in-ice acoustic position- ing system for a maneuverable melting probe. The system measures the propagation time of acoustic signals between six ultrasonic emitters at the ice surface and four acoustic sensors located in the melt- ing head of the probe. The probes position is then determined by trilateration with an accuracy of better than 1 m which is comparable to the accuracy required for IceCube’s geometry calibration with optical signals. The acoustic sensor developed in the EnEx-project is a piezo-based sensor, optimized for ultra- sonic signals (18 kHz) [7]. It consists of a piezo disc (Ø16 mm × 3 mm) and two stacked sensor front end boards. One board ampliﬁes, ﬁlters and digitizes the received signals and the second board con- verts the single-ended digital signals to LVDS signals. The two front end boards and the piezo disc are encapsulated in a brass housing (Ø20 mm). The sensor front end contains two channels. After passing diﬀerent ﬁlters the signal is digitized for each channel with a sampling rate of up to 3 MS/s at a resolution of 12 bits. The power consumption of one sensor front end with two channels is about 150 mW (30 mA at 5 V). 2 Prototype Development To meet the requirements of IceCube-Gen2 the EnEx acoustic sensor front end was modiﬁed in respect of the desired frequency ranges and power consumption [8]. The new prototype front end utilizes two channels as well: channel one with a frequency range from 16 kHz to 20 kHz, optimized with regard to noise suppression of the acoustic positioning and channel two with a frequency range from 10 kHz to 100 kHz, optimized with regard to sensitivity of the acoustic neutrino detection. To substantially reduce the power consumption the LVDS conversion was removed which also leads to a reduced pin count of the sensor front end. The power consumption of the new prototype front end was reduced to typically 50 mW (max. 53 mW) in idle mode (no digitization), which is about 33% compared to the original EnEx front end, and typically 75 mW (max. 78 mW) when digitizing both channels. Further improvement could be achieved by implementing a one channel operating mode, where only one channel is digitized and the other is in idle mode which will lead to an estimated power consumption of about 35 mW. By using enhanced electronic components optimized for low power application further reduction is possible. A basic FPGA-ﬁrmware module was developed which controls and reads-out the prototype board. The Firmware was designed as a QSys component for use with the Altera QSys build system which will likely be used in IceCube-Gen2. 3 Test Measurements The acoustic positioning concept for IceCube-Gen2 and its performance was evaluated in water (swimming pool) [8]. For this purpose, three EnEx acoustic sensors were integrated into an Ice- 2 2 , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 DOI: 10.1051/ 7135 epjconf/201 06003 Cube DOM1 (Acoustic DOM) along with a data acquisition system (see ﬁg. 1). The sensors were mounted into the Acoustic DOM using an aluminum construction that was glued to the glass sphere with vacuum cups. The sensor itself was pressed towards the glass sphere with metal springs (see ﬁg. 1). The Acoustic DOM was mounted under a ﬂoating platform and held under water with weights. On top of the ﬂoating platform two targets were mounted for reference measurements with a laser range-ﬁnder (LRF): one to measure the position of the Acoustic DOM and another to determine its heading. In addition six EnEx acoustic emitters were distributed in the swimming pool at diﬀerent depths to reconstruct the position of the Acoustic DOM via acoustic measurements. Positions (laser range-fnder) x [m] y [m] E1 E2 E3 E4 E5 E6 P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 EnEx sensors cable connector spring vacuum cup EnEx sensor double-sided adhesive tape tefon ring IceCube DOM Figure 1. left: positions of the six acoustic emitters (E1 - E6, circles) and 15 measurement positions of the Acoustic DOM (P1 - P15, rectangles) in the swimming pool, measured with the laser range-ﬁnder (LRF); cen- ter: Acoustic DOM: IceCube DOM1 with three EnEx sensors, read-out electronics and a pressure-tight cable connector; right: mounting of the acoustic sensor inside the IceCube DOM. EnEx sensors cable connector Positions (laser range-fnder) x [m] y [m] E1 E2 E3 E4 E5 E6 P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 spring vacuum cup EnEx sensor double-sided adhesive tape tefon ring IceCube DOM Figure 1. left: positions of the six acoustic emitters (E1 - E6, circles) and 15 measurement positions of the Acoustic DOM (P1 - P15, rectangles) in the swimming pool, measured with the laser range-ﬁnder (LRF); cen- ter: Acoustic DOM: IceCube DOM1 with three EnEx sensors, read-out electronics and a pressure-tight cable connector; right: mounting of the acoustic sensor inside the IceCube DOM. 3 Test Measurements After measuring the positions of the six acoustic emitters with the LRF, the measurement pro- cedure continues as follows: First, the reference position and heading of the Acoustic DOM at the respective measurement position is measured with the LRF. Then a sine-burst signal is sent consec- utively from each emitter and data is taken synchronously with the Acoustic DOM for each emitter. After changing the position of the Acoustic DOM the above steps are repeated (see ﬁg. 1). The po- sition was reconstructed by determining the propagation time between the acoustic emitters and the acoustic sensors by detecting the start of the sine-burst as the point, where the signal exceeds 5σ of the preceding noise signal. After that, a least square ﬁt was applied to the measured data for each channel of the sensors to reconstruct the position of the sensor. 1kindly provided by Rolf Nahnhauer, DESY 4 Performance The reconstructed positions are in good agreement with the reference positions measured with the LRF and the heading of the Acoustic DOM can be estimated (see ﬁg. 2). However, there are some systematic errors that are not yet fully understood. To quantify the performance of the system, the distance of the measured position from the refer- ence position (ﬁg. 2), both in full space and in the x-y-plane, were histogrammed. The accuracy in the x-y-plane (median: 6.31 cm) is better than the overall 3D accuracy (median: 26.48 cm) which results from smaller variation of the acoustic emitters’ positions in z-direction compared to the x-y-plane, due 3 3 DOI: 10.1051/ 7135 epjconf/201 06003 , (2017) 135 EPJ Web of Conferences ARENA 2016 06003 , (2017) 135 EPJ Web of Conferences 06003 Position 4 x [m] y [m] 6.9 6.8 6.7 6.6 6.5 2.7 2.8 2.9 3.0 s1 s2 s3 c1(s1) c0(s1) c1(s2) c0(s2) c1(s3) c0(s3) Δ = 0.0631 m (median) Performance (x-y-projection) 0 1 2 3 4 5 6 0.0 0.2 0.4 0.6 0.8 1.0 ∆ [m] 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 Performance (3D) ∆ [m] Δ = 0.2648 m (median) Figure 2. left: reconstructed positions of the acoustic sensors at Acoustic DOM pos. 4, cX(Y) labels channel X of sensor Y, s1 - s3 are the reference positions of the acoustic sensors deduced from the measurements with the LRF, the circle shows the diameter of the Acoustic DOM; center, right: Histogram of the distances of the acoustically measured positions from the true positions in full space and in the x-y-plane, respectively. Δ = 0.0631 m (median) Performance (x-y-projection) 0 1 2 3 4 5 6 0.0 0.2 0.4 0.6 0.8 1.0 ∆ [m] Position 4 x [m] y [m] 6.9 6.8 6.7 6.6 6.5 2.7 2.8 2.9 3.0 s1 s2 s3 c1(s1) c0(s1) c1(s2) c0(s2) c1(s3) c0(s3) 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 Performance (3D) ∆ [m] Δ = 0.2648 m (median) Figure 2. left: reconstructed positions of the acoustic sensors at Acoustic DOM pos. 4 Performance 4, cX(Y) labels channel X of sensor Y, s1 - s3 are the reference positions of the acoustic sensors deduced from the measurements with the LRF, the circle shows the diameter of the Acoustic DOM; center, right: Histogram of the distances of the acoustically measured positions from the true positions in full space and in the x-y-plane, respectively. to the limited depth of the swimming pool. It is expected that a more sophisticated data selection for noise and outliers, as well as orientation dependent corrections would further improve the accuracy. For use in ice, the media properties involve larger uncertainties which, however, do not vary with time. A similar performance is expected if the system provides suﬃcient redundancy for determining and correcting these uncertainties. to the limited depth of the swimming pool. It is expected that a more sophisticated data selection for noise and outliers, as well as orientation dependent corrections would further improve the accuracy. For use in ice, the media properties involve larger uncertainties which, however, do not vary with ti A i il f i t d if th t id ﬃi t d d f d t i i to the limited depth of the swimming pool. It is expected that a more sophisticated data selection for noise and outliers, as well as orientation dependent corrections would further improve the accuracy. noise and outliers, as well as orientation dependent corrections would further improve the accuracy. For use in ice, the media properties involve larger uncertainties which, however, do not vary with time. A similar performance is expected if the system provides suﬃcient redundancy for determining and correcting these uncertainties. 5 Conclusion A concept of an acoustic positioning system for IceCube-Gen2 based on the EnEx positioning system has been evolved by developing a prototype sensor front end based on the EnEx sensor front end and a ﬁrmware module for IceCube-Gen2. The frequency ranges of the front end have been adapted to meet the requirements of IceCube-Gen2. Also the power consumption of the front end was reduced to a third of the power consumption of the EnEx sensor front end. The EnEx acoustic sensor has been integrated into an IceCube DOM and the acoustic positioning of an IceCube detector module has been demonstrated in water with a measured accuracy better than 7 cm on average. The next steps to further evolving the concept of acoustic positioning for IceCube-Gen2 would be to improve the prototype front end by means of power consumption and actually integrate it into an IceCube-Gen2 DOM. Furthermore, a concept for design and distribution of acoustic emitters for IceCube-Gen2 needs to be developed. References [1] A. Achterberg et al., Astroparticle Physics 26, 155 (2006) [2] M.G. Aartsen et al. (2014), arXiv:1412.5106 [3] M. Aartsen et al., Nuclear Instruments and Methods in Physics Research Section A 711, 73 (2013) [4] R. Abbasi et al., Astroparticle Physics 34, 382 (2011) [5] G. Askariyan et al., Nuclear Instruments and Methods 164, 267 (1979) [6] J. Kowalski et al., Cold Regions Science and Technology 123, 53 (2016) [7] D. Eliseev et al., Annals of Glaciology 55, 253 (2014) [8] S. Wickmann, master’s thesis, RWTH Aachen University (2016) 4
https://openalex.org/W4389298776	https://www.qeios.com/read/OC5VHB/pdf	English	null	Review of: "Redefining borders in the contested territory between San Pedro and San Andres Cholula"	null	2,023	cc-by	359	Qeios, CC-BY 4.0 · Review, December 4, 2023 Qeios ID: OC5VHB · https://doi.org/10.32388/OC5VHB Review of: "Redefining borders in the contested territory between San Pedro and San Andres Cholula" Meehwa Cho1 1 Universidad Politécnica de Cataluna Meehwa Cho1 1 Universidad Politécnica de Cataluna Potential competing interests: No potential competing interests to declare. Potential competing interests: No potential competing interests to declare. There is an effort to visualize different urban situations throughout the text and the provided information may be valid at some points. But various questions arise from the beginning, from the premise of using "boundaries" as a tool to define social identity and as a strategic planning tool. Even if there were clearly two different identities since pre-hispanic, it is unlikely that they have been maintained to the point where they can be clearly distinguished by a boundary, six centuries after they were transferred into another logic. And a series of events that occurred thereafter doesn't seem to match completely to the same boundary neither. Moreover, clear political-administrative division with single lines is only possible after modern land surveying, and before that, boundaries used to include physical surface or volumes. Moreover, clear political-administrative division with single lines is only possible after modern land surveying, and before that, boundaries used to include physical surface or volumes. On the map, it seems that there are no morphological differences between San Pedro and San Andres. And the author himself mentions that there is no "difference or separation in the perception of their everyday inhabitants and visitors". If so, the question arises again whether there is a need to reinforce dichotomous thinking by identifying virtual boundaries that do not correspond to physical elements. If differences in historical, social and cultural identity are real, it would seem more appropriate to visualize them as urban tissues with land use, social organizations, specific streets, symbolic buildings, etc. instead of boundary, and to find “border zone” where different identities meet or mix. Furthermore, if a discrepancy between “border zone” and administrative “boundary” exists, it can be understood as a conflict, and strategic plans or projects are more necessary at that point. Qeios ID: OC5VHB · https://doi.org/10.32388/OC5VHB 1/1
https://openalex.org/W4292057494	https://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/144/10/101004/6912465/gtp_144_10_101004.pdf	English	null	Off-Design Modeling and Operational Optimization of Trans-Critical Carbon Dioxide Heat Pumps	Journal of engineering for gas turbines and power	2,022	cc-by	9,286	1 Introduction operating CO2 heat pumps through a trans-critical cycle, where the maximum pressure is signiﬁcantly higher than the critical pressure of CO2 (71.1 bar). This avoids being limited in the heat delivery temperatures, hence widening the operation range and improving the performance of the thermodynamic process. With a contribution of more than 51% of the total worldwide energy consumption, the heating and cooling sector plays a major role in the transition toward more sustainable energy systems and in the mitigation of climate change [1]. In this context, the installed capacity of heat pumps increased signiﬁcantly in recent years and is projected to grow even more in the future, as shown in Fig. 1 for the two major heat pump applications in Europe, namely, district heating and industrial heating [2]. The former refers to heat pumps for space and water heating; the latter refers to heat pumps used for heat recovery and energy efﬁciency improvements in different industrial sectors. p g p y p Several studies dealt with the modeling and analysis of trans- critical CO2 heat pumps in the past. For example, Sarkar et al. provided a thorough assessment of different system designs and carried out an extensive experimental campaign showing that trans-critical CO2 heat pumps can be used for a wide range of applications ranging from residential to commercial heating to agriculture [6]. Dai et al. evaluated heat pumps integrated with mechanical subcooling in terms of energy, exergy, and economic performance [7], whereas Wang et al. investigated the feasibility and performance of heat pumps integrated with thermal energy storage for space heating [8]. Recently, He et al. presented a modiﬁed trans-critical CO2 heat pump system with new water ﬂow conﬁguration for residential space heating [9], whereas Wang et al. proposed a novel concept for simultaneous space heat- ing and cooling [10]. Cao et al. reviewed the developments of air source trans-critical CO2 heat pumps using direct-heated type and recirculating-heated type, whereas Yao et al. performed a compar- ative study of upgraded heat pump systems with different heat sinks [11]. Lo Basso et al. studied the potential role of trans- critical CO2 heat pumps within a solar cooling system for building services [12]. Despite the latest advancements, a gap exists con- cerning the modeling and optimization of trans-critical CO2 heat pumps operating under different operating conditions throughout the year [13]. Paolo Gabrielli1 Institute of Energy and Process Engineering, ETH Zurich, Zurich 8092, Switzerland e-mail: gapaolo@ethz.ch Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/144/10/101004/6912465/gtp_144_10_101004.pdf by guest on 24 October 2024 Industrial heat pumps, and speciﬁcally those using carbon dioxide (CO2) as a refrigerant, can play a key role in the decarbonization of the heating and cooling sector, due to their low global warming potential, toxicity and ﬂammability. However, challenges arise when dealing with the modeling and optimization of CO2 heat pumps under different operating conditions. We address this challenge by presenting a modeling and optimization tool to predict and optimize the operation of heat pumps in off-design conditions. The tool improves on the current state-of-the-art in several ways. First, it describes a novel ther- modynamic cycle, which features higher performance than conventional heat pumps. Also, it is based on a mathematical model that describes accurately the behavior of CO2 across a wide range of thermodynamic conditions, especially near its critical region, and takes into account effects of motor-cooling, leakages and performance limits. Further- more, it maximizes the coefﬁcient of performance (COP) of the heat pump via an accurate and computationally efﬁcient optimization problem. The capabilities of the model are illustrated by looking at different typical heat pump applications based on real-world projects within the heating and cooling sector. Different case studies are considered, showing how the heat pump is optimally operated during the year to maximize its COP while meeting the varying boundary conditions. [DOI: 10.1115/1.4055233] Luis Sanz Garcia1 MAN Energy Solutions Schweiz AG, Hardstrasse 319, Zurich 8005, Switzerland e-mail: luis.sanz-garcia@man-es.com Emmanuel Jacquemoud1 MAN Energy Solutions Schweiz AG, Hardstrasse 319, Zurich 8005, Switzerland e-mail: emmanuel.jacquemoud@man-es.com Emmanuel Jacquemoud1 MAN Energy Solutions Schweiz AG, Hardstrasse 319, Zurich 8005, Switzerland e-mail: emmanuel.jacquemoud@man-es.com Philipp Jenny MAN Energy Solutions Schweiz AG, Hardstrasse 319, Zurich 8005, Switzerland Philipp Jenny MAN Energy Solutions Schweiz AG, Hardstrasse 319, Zurich 8005, Switzerland 1Corresponding authors. Manuscript received July 17, 2022; ﬁnal manuscript received July 24, 2022; published online August 25, 2022. Editor: Jerzy T. Sawicki. Journal of Engineering for Gas Turbines and Power OCTOBER 2022, Vol. 144 / 101004-1 Copyright V C 2022 by ASME; reuse license CC-BY 4.0 Off-Design Modeling and Operational Optimization of Trans-Critical Carbon Dioxide Heat Pumps Siddhant Singh Institute of Energy and Process Engineering, ETH Zurich, Zurich 8092, Switzerland Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/144/10/101004/6912465/gtp_144_10_101004.pdf by guest on 24 October 2024 1 Introduction Different heat pump conﬁgurations and refrigerants are cur- rently used depending on the end applications. After the seminal work of Lorentzen [3] and Neksa˚ [4], natural refrigerants such as carbon dioxide (CO2) have been gaining traction in recent years due to their low global warming potential, toxicity, and ﬂamma- bility compared to traditional refrigerants, such as ammonia or R134a amongst many others [5]. However, having a considerably lower critical temperature (31 C) than conventional refrigerants, CO2 is subject to a number of challenges when applying it in a subcritical cycle: (i) limited operational temperature range at the sink side of the cycle; (ii) low enthalpy of vaporization; (iii) heat rejection temperature similar to the refrigerant’s critical tempera- ture; and (iv) ambient temperature in summer at similar level as the refrigerant’s critical temperature. All these factors translate into limitations and poor performance under certain operating conditions [4]. These challenges are tackled by designing and While heat pumps are typically designed for speciﬁc nominal operating points, their performance is affected by variable bound- ary conditions, such as the heat duty and the water supply temper- ature required by the ﬁnal end user and the ambient temperature g for Gas Turbines and Power OCTOBER 2022, Vol. 144 / 101004-1 Copyright V C 2022 by ASME; reuse license CC-BY 4.0 Fig. 2 Schematic of CO2 heat pump cycle used in this study Fig. 1 Heat generation via heat pump from 2015 to 2050 in the European heating sector; adapted from Ram et al. [2] Fig. 2 Schematic of CO2 heat pump cycle used in this study Fig. 1 Heat generation via heat pump from 2015 to 2050 in the European heating sector; adapted from Ram et al. [2] is only needed when the end user requires high-return tem- peratures on the gas cooler water return side. For low return temperatures, the recuperator does not improve the heat pump COP and simply increases the investment cost of the system [6]. Compared to the gas cooler, the recuperator has a much smaller heat exchange area, because of the much higher temperature difference required across the HEX, and the lower heat duty. [14]. We address this challenge by presenting a modeling and optimization tool to predict and optimize the operation of trans- critical CO2 heat pumps in off-design conditions. 1 Introduction Whereas the model developed is general, it is applied for a custom-made prod- uct developed at MAN Energy Solutions Schweiz AG, which fea- tures higher performance with respect to conventional heat pumps [15]. We present the mathematical model of such a novel thermo- dynamic cycle, which describes the behavior of CO2 across a wide range of thermodynamic conditions. A high level of accu- racy is reached by the model thanks to the detailed modeling of the heat exchangers (HEXs) and of the compressor, which takes into account effects of motor-cooling, leakages and performance limits. Furthermore, the model maximizes the coefﬁcient of per- formance (COP) of the heat pump via an accurate and computa- tionally efﬁcient heuristic optimization algorithm. This allows (i) to improve the overall heat pump performance during the year, hence reducing the operational costs, and (ii) to investigate the tradeoff between the optimal design under nominal and under actual operating conditions. An additional heat exchanger, the rejection HEX, may be required upstream the expander to meet the required volume ﬂow conditions. Heat is rejected to a source ﬂow prior to entering the evaporator, and the trans-critical CO2 is cooled down to suitable conditions at the expander suction. An expander stage is deployed between the rejection HEX (if required) and the expansion valve to mechanically exploit the pressure difference between the high- and low-pressure sides of the thermodynamic cycle. The expander stage is mounted on the same shaft as the compressor, thus reducing the overall mechanical shaft power of the compressor. The expander stage rotational speed is therefore imposed by the compressor motor. This paper is structured as follows. Section 2 describes the CO2 heat pump conﬁguration considered in this work. Section 3 presents the mathematical model and the optimization algorithm, and Sec. 4 presents the validation of the mathematical model. Sec- tion 5 discusses the results by means of two case studies. Finally, Sec. 6 summarizes the study and draws conclusions. 3 Mathematical Model This section presents an overview of the mathematical models of all relevant heat pump components (see Sec. 2), as well as the optimization algorithm used to maximize the heat pump COP under varying boundary conditions. 2 System Description 3.1 Model Overview. The major model inputs can be grouped into three categories: The trans-critical heat pump considered in this work consists of a compressor, a gas cooler HEX, an expansion manifold (i.e., expander stage followed by an expansion valve), and an evapora- tor HEX. The main difference with respect to a conventional sub- critical cycle is the fact that the working ﬂuid (also called refrigerant) is compressed to a pressure above the critical point, hence to its supercritical thermodynamic conditions. This implies that the conventional condensation at the sink side of the process is replaced by the sensible cooling of the refrigerant. Load boundary conditions, which characterize the end-use application of the heat pump, independently of its speciﬁc components. These include the overall heat load that has to be met during the year, the temperatures of the heat recovery ﬂuid (i.e., the ﬂuid on the sink side, water here) required at the inlet and the outlet of the heat pump, and the temperature and pressure of the cold source of the heat pump (water here). p p p Turbomachinery operating maps, which include the nondi- mensional maps of both the compressor and the expander. Figure 2 shows a process ﬂow schematic of the considered CO2 heat pump, which is also a custom-made product developed at MAN Energy Solutions Schweiz AG [15]. This system improves on conventional trans-critical CO2 heat pumps as follows: p p p HEX design, which describes the exchange area of the HEXs, as well as their operating conditions at the design point. The overall modeling and optimization procedure is shown in Fig. 3. First, the aforementioned input data are deﬁned, alongside an initial guess of the decision variables of the optimization prob- lem, i.e., the operation variables of the heat pump. Based on this, the model determines the conditions of CO2 along the heat pump The purpose of the recuperator HEX is to pre-heat the refrig- erant after the evaporator, on the low-pressure side of the cycle, by using the excess heat carried after the gas cooler, on the high-pressure side of the cycle. In fact, the recuperator The purpose of the recuperator HEX is to pre-heat the refrig- erant after the evaporator, on the low-pressure side of the cycle, by using the excess heat carried after the gas cooler, on the high-pressure side of the cycle. 2 System Description The modeling steps are embedded within an optimization procedure, which deter- mines the optimal combination of operation variables that maxi- mizes the COP of the system while minimizing the penalties of the individual components. The optimization procedure is based on a genetic algorithm, which solves the model iteratively and stops when predeﬁned convergence criteria are met (see Sec. 3.3). Table 1 Constant parameters used to compute the heat trans- fer coefﬁcients of the hot (a and b) and cold ﬂuids (c and d) HEX Hot fluid Cold fluid aH bH aC bC N Gas cooler CO2 Water 0.721 0.33 0.8 0.4 10 Recuperator CO2 CO2 0.55 0.23 0.55 0.23 5 Rejection HEX CO2 Water 0.55 0.23 0.8 0.4 5 Evaporator Water CO2 0.8 0.4 0.721 0.33 5 OCTOBER 2022, Vol. 144 / 101004-3 Table 1 Constant parameters used to compute the heat trans- fer coefﬁcients of the hot (a and b) and cold ﬂuids (c and d) 3.2 Heat Pump Components 3.2.1 Heat Exchangers. The considered HEXs include the gas cooler, the recuperator, and the rejection HEX (see Fig. 2), which are all modeled through the same approach. Various approaches 2 System Description The heat transfer within all seg- ments i 2 f1; …; Ng is modeled through following equations: Qi ¼ UiAiLMTDi (1) Qi ¼ mHDhH;i ¼ mCDhC;i (2) 1 Ui ¼ 1 aH;i þ 1 aC;i (3) LMTDi ¼ DTi DTi1 ln DTi DTi1 (4) Qi ¼ UiAiLMTDi (1) Qi ¼ mHDhH;i ¼ mCDhC;i (2) 1 Ui ¼ 1 aH;i þ 1 aC;i (3) LMTDi ¼ DTi DTi1 ln DTi DTi1 (4) (1) (3) LMTDi ¼ DTi DTi1 ln DTi DTi1 (4) (4) where the ﬂuid heat transfer coefﬁcients are expressed, for both the cold and the hot ﬂuid, through: where the ﬂuid heat transfer coefﬁcients are expressed, for both the cold and the hot ﬂuid, through: ai ¼ B maPrb i ki la i (5) (5) Here, Qi is the heat load, Ui is the overall heat transfer coefﬁcient, Ai is the heat exchange area, and LMTD is the logarithmic mean temperature difference across the ith segment of the HEX, which is computed by using DTi ¼ TH;i TC;i. mH and mC are the mass ﬂow rates of the hot and cold ﬂuids, respectively, which are the same through all segments; DhH;i and DhC;i are the enthalpy varia- tions of the hot and cold ﬂuids across the ith segment of the HEX, respectively. aH;i and aC;i are heat transfer coefﬁcients of the hot and cold ﬂuids, respectively, across the ith segment of the HEX; they are calculated by using the values of the ﬂuid thermal con- ductivity, k, viscosity, l, and Prandtl number, Pr. The coefﬁcient B is the base heat transfer coefﬁcient, which only depends on the Fig. 3 Optimization procedure ﬂowchart cycle, the COP of the system, as well as the penalties associated with all individual components (see Sec. 3.3). The modeling steps are embedded within an optimization procedure, which deter- mines the optimal combination of operation variables that maxi- mizes the COP of the system while minimizing the penalties of the individual components. The optimization procedure is based on a genetic algorithm, which solves the model iteratively and stops when predeﬁned convergence criteria are met (see Sec. 3.3). cycle, the COP of the system, as well as the penalties associated with all individual components (see Sec. 3.3). 2 System Description In fact, the recuperator 101004-2 / Vol. 144, OCTOBER 2022 Transactions of the ASME Fig. 3 Optimization procedure ﬂowchart can be found in the literature for modeling HEXs in CO2 heat pumps. For example, Sarkar et al. [16], Li and Wang [17], and Ye et al. [18] modeled tube HEX based on their detailed geometrical characteristics. However, HEXs currently used in CO2 heat pumps are often characterized by complex geometries, which results in models with high computational complexity. To reduce the com- putational complexity, a novel model is developed that enables the calculation of the heat transfer coefﬁcients in off-design con- ditions based on the behavior of the HEXs at the design point. Such a novel approach is detailed in the following for a generic HEX. It requires the following input data: the type of HEX and its exchange area, A the inlet temperatures, T, and pressures, p, of both the hot (H) and the cold (C) ﬂuids within the HEX the inlet temperatures, T, and pressures, p the inlet temperatures, T, and pressures, p, of both the hot (H) and the cold (C) ﬂuids within the HEX (H) and the cold (C) ﬂuids within the HEX performance of the HEX at the design point, which includes the pressure drops, Dp, and the inlet and outlet temperature, pressure, and mass ﬂow rates, m, of both the hot and cold ﬂu- ids, hence the total heat load of the unit, Q The HEX is divided into N segments (where N is speciﬁc for the different HEXs, see Table 1) to describe the variation of the ﬂuid properties and conditions between inlet and outlet sections. A ﬁne enough discretization is key as the properties of CO2 change abruptly near the critical point. All segments have equal heat load, and the sum of all segment contributions is the total heat load of the system. Also, a linear pressure proﬁle is assumed from inlet to outlet conditions. OCTOBER 2022, Vol. 144 / 101004-3 Journal of Engineering for Gas Turbines and Power 4), the polytropic head is calculated for all segments i 2 f1; …; Ng by ﬁrst calculating the entropy at State 3, s3;i, as a function of the inlet pressure, p1;i, and the outlet enthalpy, h2;i. The outlet entropy s2;i is calculated through Eqs. (10) and (11) for compressor [20] and expander [21], respectively By using the values of the base heat transfer coefﬁcients, of the off-design heat load, and of the HEX exchange area, Eqs. (1) to (5) are used to determine the mass ﬂow rate, pres- sure, and temperature proﬁles along the HEX in off-design conditions. For both the cold and hot ﬂuids, the pressure drops along the HEX are computed as By using the values of the base heat transfer coefﬁcients, of the off-design heat load, and of the HEX exchange area, Eqs. (1) to (5) are used to determine the mass ﬂow rate, pres- sure, and temperature proﬁles along the HEX in off-design conditions. For both the cold and hot ﬂuids, the pressure drops along the HEX are computed as s2;i ¼ s1;i þ ð1 gpÞðs3;i s1;iÞ (10) s2;i ¼ s3;i þ s1;i s3;i gp (11) (11) The inlet conditions of segment i are set equal to the outlet condi- tions of segment i – 1, and Eqs. (10) and (11) are applied for all segments i 2 f1; …; Ng. The outlet conditions of segment N deﬁne the outlet conditions of the turbomachinery unit. Dp ¼ Dpdes m mdes 2 (6) Dp ¼ Dpdes m mdes 2 (6) where the subscript “des” denotes design conditions. This iterative procedure proves to model the heat transfer across all HEX at the same time with a high enough accuracy and a low computational complexity. The accuracy of the model is assessed by comparing the total heat load of the system and the HEX pres- sure drops against manufacturer data. The model validation is shown in Sec. 4. Fig. 4 Polytropic path of a compressor (blue line) and multi- step process to determine the stage outlet thermodynamic states 3.2.2 Turbomachinery. Here, turbomachinery include the compressor and the expander of the trans-critical CO2 heat pump (see Fig. 2), which use CO2 as the working ﬂuid. The turbomachi- nery performance is often modeled through nondimensional maps of work versus ﬂow characteristics, which are typically provided by manufacturers. Journal of Engineering for Gas Turbines and Power The total enthalpy difference across the turbomachi- nery is then calculated as The work coefﬁcient l0 and the polytropic efﬁciency gp are determined through the turbomachinery nondimensional map by knowing the ﬂow coefﬁcient and the inlet stage Mach number. The total enthalpy difference across the turbomachi- nery is then calculated as With known pressure and temperature (or enthalpy) proﬁles, the thermodynamic properties of the HEX ﬂuids, namely, k, l, and Pr, are computed via the NIST standard reference database [19] for both ﬂuids. Furthermore, a ﬁrst guess for the value of the base heat transfer coefﬁcients, BH and BC, is chosen based on the type of ﬂuids and on the type and geom- etry of the HEX. These quantities are used in Eq. (5) to deter- mine the heat transfer coefﬁcients, aH;i and aC;i, of both the hot and the cold ﬂuids. The ﬂuid heat transfer coefﬁcients are then used to compute the overall heat transfer coefﬁcient of the HEX, Ui. Dhtot ¼ l0u2 (9) (9) The remaining outlet conditions of the turbomachinery are calculated by following the polytropic head corresponding to gp and illustrated in Fig. 4. The polytropic head is divided into N segments of equal enthalpy difference Dhi deﬁning the polytropic path. Here, N ¼ 30 proves to be the optimal tradeoff between accuracy and computational complexity. The remaining outlet conditions of the turbomachinery are calculated by following the polytropic head corresponding to gp and illustrated in Fig. 4. The polytropic head is divided into N segments of equal enthalpy difference Dhi deﬁning the polytropic path. Here, N ¼ 30 proves to be the optimal tradeoff between accuracy and computational complexity. The heat load, the LMTD, and the overall heat transfer coef- ﬁcients are used to determine the area of the ith segment, Ai. The total area calculated from the sum of the segment areas is compared against the actual area of the HEX, A. The val- ues of the base heat transfer coefﬁcients of bot ﬂuids are changed iteratively until the calculated HEX area matches the actual one. Starting from the inlet conditions (h1 and p1 in Fig. 4) and from the outlet enthalpy (h2 in Fig. Journal of Engineering for Gas Turbines and Power The following calculation steps are performed to estimate the outlet conditions and the performance for both compressor [20] and the expander units [21]: type of ﬂuid and on the type and geometry of the HEX; a and b are constant parameters that depend on the type of ﬂuid and on the type of HEX, and are reported in Table 1. The following calculation steps are performed for all segments i 2 f1; …; Ng. The rotational speed (in revolutions per minute, rpm) and the diameter of the turbomachinery, D, are used to determine the tip wheel velocity, u The rotational speed (in revolutions per minute, rpm) and the diameter of the turbomachinery, D, are used to determine the tip wheel velocity, u The heat load, Qi ¼ Q=N, is equal for all segments and is known from the total heat load of the system. u ¼ px 60D (7) The known values of mass ﬂow rates at design conditions, mH and mC, are used to compute the design enthalpy changes for both ﬂuids, DhH;i and DhC;i, via Eq. (2). The known inlet temperatures and pressures are used to determine the enthalpy changes, hence the enthalpy proﬁles across the HEX. Such enthalpy proﬁles are combined with the linear pressure proﬁles assumed between the inlet and outlet ﬂuid pressures at design conditions to determine the temperature proﬁles, DTi, via the NIST standard reference database [19]. The temperature proﬁles are then used to determine LMTDi via Eq. (4). (7) The tip wheel velocity and the known volume ﬂow rate of the working ﬂuid at inlet conditions, V, are used to compute the ﬂow coefﬁcient The tip wheel velocity and the known volume ﬂow rate of the working ﬂuid at inlet conditions, V, are used to compute the ﬂow coefﬁcient / ¼ V uD2 (8) (8) The work coefﬁcient l0 and the polytropic efﬁciency gp are determined through the turbomachinery nondimensional map by knowing the ﬂow coefﬁcient and the inlet stage Mach number. The total enthalpy difference across the turbomachi- nery is then calculated as The work coefﬁcient l0 and the polytropic efﬁciency gp are determined through the turbomachinery nondimensional map by knowing the ﬂow coefﬁcient and the inlet stage Mach number. 4 Model Validation p y The decision variables of the optimization problem are: Higher relative errors are observed when considering the pres- sure drops along the HEXs. MAPEs of 8.8% and 13% are obtained for the water and the CO2 side of the gas cooler, respec- tively, and MAPES of 34% and 2.1% are obtained for the water and the CO2 side of the evaporator, respectively. On the one hand, such larger relative discrepancies are due to the simple model adopted for pressure losses, given by Eq. (6). On the other hand, they correspond to absolute discrepancies smaller than about 100 kPa, which are deemed acceptable for the applications of inter- est, especially when comparing them to the operating pressures of the trans-critical CO2 heat pump (in the range of 115–140 bar). the mass ﬂow rate of CO2 in the loop (kg/s) the speed of the compressor (rpm) the outlet temperature of the recuperator HEX (if present) (C) the temperature difference across the rejection HEX (if pres- ent) (C) the mass ﬂow rate of cold source (kg/s) the evaporator HEX pressure (bar). The penalty functions are deﬁned by multiplying a penalty coefﬁcient, which is scaled according to the magnitude of the objective function, by the following discrepancies: (P1) Compressor feasibility: Difference between the work coef- ﬁcient of the compressor obtained by the model and its lower and upper values; it quantiﬁes the violation of the operation range of the compressor. 4.2 Turbomachinery. The model of the compressor is assessed by comparing the discharge temperature, the discharge pressure, and the shaft power against manufacturer data. The vali- dation results are presented in Fig. 5(b), which shows the relative error for two technology designs and the aforementioned load points. These load points are selected as they cover the majority of the compressor operation, with the discharge temperature rang- ing between 390 and 450 K, the discharge pressure between 115 and 145 bar, and the shaft power between 6 and 10 MW electrical power across all considered ambient temperatures and load points. p g p (P2) System target load: Difference between the heat load of the gas cooler obtained by the model with that of the required heat load demand from the system. q y (P3) Rejection HEX penalty: Difference between the area of rejection HEX obtained by the model and its minimum and maximum values. Journal of Engineering for Gas Turbines and Power Here, the input data required by the turboma- chinery models are: the type and geometry of the turbomachinery yp g y y the nondimensional performance maps of the turbo- machinery in terms of work coefﬁcient, l0, and ﬂow coefﬁ- cient, / / the inlet conditions of the working ﬂuid, namely temperature, pressure, and mass ﬂow rate / the inlet conditions of the working ﬂuid, namely temperature, d ﬂ t Fig. 4 Polytropic path of a compressor (blue line) and multi- step process to determine the stage outlet thermodynamic states pressure, and mass ﬂow rate the rotational speed, x, of the turbomachinery. Transactions of the ASME 101004-4 / Vol. 144, OCTOBER 2022 Once the outlet entropy of segment i is known, the remaining properties, i.e. the temperature and pressure proﬁles, are cal- culated as functions of entropy and enthalpy. which include the recuperator HEX, the average time per iteration is close to 3 s, and requires approximately 1,100 iterations to meet the convergence criteria. This results in a typical run time shorter than 1 h. 3.3 Optimization. The optimization algorithm considers all components of the trans-critical CO2 heat pump and maximizes its COP while complying with the heat load and water tempera- tures required by the end user. In its general form, the optimiza- tion problem is written as follows: 4 Model Validation (P4) Evaporator heat load penalty: Difference between the heat load of the evaporator obtained by the model and its maxi- mum and minimum values. Mean absolute percentage errors of 0.02%, 0.23%, and 0.09% are obtained for the discharge temperature, the discharge pressure and the shaft power, respectively, highlighting the excellent pre- dicting performance of the compressor model. (P5) Motor torque feasibility: Difference between the torque required by the compressor and maximum allowed torque that the motor is able to deliver. 4 Model Validation The validation of the model is presented in Fig. 5 for the sys- tem’s most relevant components, namely, the gas cooler, the evap- orator, and the compressor. No validation is shown for the expander due the difﬁculty in measuring its performance under real operating conditions. The model results are assessed in terms of relative errors with respect to manufacturer data for different technology designs and for a variety of operating conditions. The overall model performance is assessed via the mean absolute per- centage error (MAPE) [23] across all considered designs and condi- tions (i.e., the average of the absolute values of the relative errors). min x fðxÞ ¼ ðCOP þ X M i PiðxÞÞ subject to gjðxÞ 0; j ¼ 1; …; J hkðxÞ ¼ 0; k ¼ 1; …; K 4.1 Heat Exchangers. The models of the gas cooler and the recuperator HEXs are assessed by comparing the total heat pro- vided, Q, against manufacturer data. The validation results are presented in Fig. 5(a), which shows the relative error for two tech- nology designs, simply denoted as Design 1 and Design 2, and four load points (LP), which are typical for the different seasons. These load points are selected as they cover the majority of the HEXs operation, with the heat load ranging from about 15 MWth to about 30 MWth of heat across case studies and load points. Here, x 2 RX is the vector deﬁning the decision variables of the optimization problem, with X being the dimensionality of x; Pi 2 R are the penalties of the individual components, and M the number of penalties considered, namely: (1) the compressor feasi- bility, (2) the system target load, (3) the area of the rejection HEX, (4) the area of the evaporator, and (5) the motor torque fea- sibility (see Fig. 3). The inequality constraints, gj, include the lower- and upper-bound constraints for all optimization variables, which can range between a minimum and a maximum value. The equality constraints, hk, include the equations describing the behavior of HEXs and turbomachinery presented in Secs. 3.2.1 and 3.2.2, respectively. Smaller errors are generally obtained for the gas cooler than for the evaporator, resulting in a MAPE of 0.29% for the former and 1.31% for the latter. Both values are deemed satisfactory for the applications of interest and improve the performance of earlier work presented for the gas cooler HEX [24]. 5 Results and Discussion The optimization solver adopted is a genetic algorithm based on a covariance matrix adaptation evolution strategy [22]. The tol- erance of the solver is the difference between the values of the decision variables in two subsequent iterations and is set to 104. The computational time required to meet the convergence criteria highly depends on the complexity of the system, i.e., presence or not of the rejection and recuperator HEXs. This increases both the number of optimization variables, hence of iterations required, and the time required per iteration. For the cases considered here, In this section, the off-design model presented in Sec. 3 is applied to two real-world case studies to determine the optimal operational strategy that maximizes the heat pump performance, as well as its sensitivity to the most relevant decision variables, hereafter denoted as control variables. 5.1 Control Variables for Optimal Heat Pump Operation. The most relevant control variables are the compressor rotational Journal of Engineering for Gas Turbines and Power Journal of Engineering for Gas Turbines and Power OCTOBER 2022, Vol. 144 / 101004-5 Fig. 5 Validation of the mathematical models for the system’s most relevant components: (a) gas cooler and evaporator HEXs and (b) compressor. Both for the HEXs and the compressor, two technology designs are considered for four LP typical for the different seasons: winter (LP1), spring (LP2), summer (LP3), and autumn (LP4). Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/144/10/101004/6912465/gtp_144_10_101004.pdf by guest on 24 October 2024 Fig. 5 Validation of the mathematical models for the system’s most relevant components: (a) gas cooler and evaporator HEXs and (b) compressor. Both for the HEXs and the compressor, two technology designs are considered for four LP typical for the different seasons: winter (LP1), spring (LP2), summer (LP3), and autumn (LP4). 5.2.1 Case Study 1: Industrial Heat Generation Using a Large Body of Water as Heat Source. Large bodies of water such as rivers, lakes, or seawater has been used for many years in the power industry to condense steam in Rankine cycles. The know- how gathered from this industry can be tapped when dealing with the heat source of heat pumps. This allows to use large ﬂow rates of water, which result in lower temperature differences in the evaporator HEX, thus in higher evaporation temperatures hence higher COP. Moreover, the small amplitudes and slow variation in source temperature throughout the year facilitates the system operation, as it allows to gradually change the compressor rota- tional speed and the valve position as load conditions change. speed and the loop resistance, which deﬁne the pressure rise across the compressor and ultimately its operating point. Whereas the rotational speed determines the power consumption of the compressor motor, and the resistance created by the expander manifold (see Fig. 2) determines the mass ﬂowrate and the pres- sure ratio of the loop. Since the pressure ratio of the expander is given by its design characteristics and the rotational speed, the only way to control the overall resistance of the system is by adjusting the position of the expansion valve (hereafter referred as “valve”) downstream the expander stage. For a given rotational speed, closing the valve results in increasing the loop resistance (i.e., higher pressure ratio) and lowering the mass ﬂowrate; this can be done until the surge line of the compressor is reached. Journal of Engineering for Gas Turbines and Power In contrast, opening the valve results in a lower pressure ratio and a higher mass ﬂowrate in the loop; this can be done until the com- pressor reaches choke conditions. p p g The ﬁrst considered case study describes an industry located near a large body of water, which installs a trans-critical CO2 heat pump to heat up water from 40 C to 95 C for process heating application. Due to regulatory constraints based on realistic enqui- ries, it is allowed to extract a maximum water ﬂowrate of 1,700 kg/s, which is fed via a circulation pump to the evaporator HEX. Given the ﬂuctuations in the heating demand of the indus- trial end user during the year, the capability of the heat pump to meet different heat loads is a key feature of the system installed. Moreover, it is necessary not only to estimate the expected per- formance and operation of the system but also to predict its opera- tional limitations in terms of heat output, both on the upper and lower bounds (i.e., part-load), throughout the year. The operation range of the heat pump (also called envelope), as well as its per- formance and control variables, for all possible operating condi- tions is shown in Fig. 6; the normalized compressor map selected and optimized for this case study is shown in Fig. 7, whereas the parameters of the HEXs are reported in Table 2. Different operat- ing conditions, namely, winter, summer, and minimum load oper- ations, are explicitly reported. The optimal value of the valve position is thus a compromise, which takes into account the performance characteristics of all the components of the system, namely, the turbomachinery efﬁcien- cies and the performance of the HEXs. Although the compressor performance can be easily predicted through its characteristic map, the variation of the heat transfer coefﬁcients of the HEXs due to changes in the thermal properties of supercritical CO2, mostly in the gas cooler, adds another degree of complexity when operating trans-critical CO2 heat pumps compared to conventional subcritical heat pumps. 5.2 Case Studies. The considered case studies include (1) industrial heat generation using a large body of water as heat source and (2) high-temperature district heating using the waste heat from data center (DC) as heat source. Transactions of the ASME Transactions of the ASME 101004-6 / Vol. 144, OCTOBER 2022 Fig. Journal of Engineering for Gas Turbines and Power Table 2 HEX parameters for case study 1 HEX A (m2) BH (m–2) BC (m–2) Evaporator 2,580 2.7410–2 1.75105 Gas cooler 2,327 0.941 1.034 Recuperator 412 58.96 14.60 Table 2 HEX parameters for case study 1 source water. For warmer source temperatures, the compressor remains the limiting component. Operation strategy. Figure 6 shows the interdependence of expansion valve position and compressor speed. The higher the heat load, and the higher the required temperature lift, the higher the compressor speed to increase the necessary pressure ratio. This also leads to a wider valve opening to increase the CO2 mass ﬂowrate. Performance sensitivity. Two major trends are observed when looking at the heat pump operation: Signiﬁcant increase in COP for warmer source temperature. For a given heat load, the lower temperature lift that the com- pressor carries out results in a lower electrical consumption hence higher COP. While this is expected, the results allow Fig. 6 Normalized nondimensional compressor map selected and optimized for case study 1. The isolines represent relative rotational speed. The winter (cross), summer (triangle), and minimum load operations (diamond) are shown. Fig. 8 Normalized COP variation of the trans-critical CO2 heat pump for different source temperatures OCTOBER 2022 Vol 144 / 101004 7 Operation range. The operation range of the heat pump is delimited on the lower heat load end by the compressor design, as instabilities arise when operating the compressor at low mass ﬂow rates near surge conditions. On the upper end, the operation is lim- ited by the maximum allowed torque of the compressor motor, which limits the electrical power input to the compressor hence the heat load of the heat pump. Similarly, on the high end of the operation range, the maximum attainable ﬂowrate of the compres- sor may be limited by choke conditions. Therefore, the size and shape of the heat pump envelope depend on the design criteria of the compressor. The maximum heat produced by the heat pump can also be limited by the maximum heat that can be extracted from the water source, in this case a lake or river. As shown in Fig. 6, for source temperatures lower than 3 C, the maximum heat output is limited by the potential risk of freezing of the Fig. 8 Normalized COP variation of the trans-critical CO2 heat pump for different source temperatures Journal of Engineering for Gas Turbines and Power 6 Normalized nondimensional compressor map selected and optimized for case study 1. The isolines represent relative rotational speed. The winter (cross), summer (triangle), and minimum load operations (diamond) are shown. source water. For warmer source tempera remains the limiting component. Operation strategy. Figure 6 shows the expansion valve position and compressor s heat load, and the higher the required temp the compressor speed to increase the nec This also leads to a wider valve opening to i ﬂowrate. Performance sensitivity. Two major tren looking at the heat pump operation: Signiﬁcant increase in COP for warme For a given heat load, the lower temper pressor carries out results in a lower e hence higher COP. While this is expec Fig. 6 Normalized nondimensional compressor map selected and optimized for case study 1. The isolines represent relative rotational speed. The winter (cross), summer (triangle), and minimum load operations (diamond) are shown. Fig. 7 Operation range of the heat pump ing a constant maximum ﬂowrate of 1,700 k ture demands of 40 C/95 C. Heat pump compressor speed (middle) and valve op winter (cross), summer (triangle) and mini (diamond) are shown. Table 2 HEX parameters for ca HEX A (m2) BH (m Evaporator 2,580 2.741 Gas cooler 2,327 0.941 Recuperator 412 58.96 Fig. 7 Operation range of the heat pump (or envelope) assum- ing a constant maximum ﬂowrate of 1,700 kg/s and hot tempera- ture demands of 40 C/95 C. Heat pump COP (top), relative compressor speed (middle) and valve opening (bottom). The winter (cross), summer (triangle) and minimum load operations (diamond) are shown. Fig. 7 Operation range of the heat pump (or envelope) assum- ing a constant maximum ﬂowrate of 1,700 kg/s and hot tempera- ture demands of 40 C/95 C. Heat pump COP (top), relative compressor speed (middle) and valve opening (bottom). The winter (cross), summer (triangle) and minimum load operations (diamond) are shown. Fig. 7 Operation range of the heat pump (or envelope) assum- ing a constant maximum ﬂowrate of 1,700 kg/s and hot tempera- ture demands of 40 C/95 C. Heat pump COP (top), relative compressor speed (middle) and valve opening (bottom). The winter (cross), summer (triangle) and minimum load operations (diamond) are shown. Journal of Engineering for Gas Turbines and Power Journal of Engineering for Gas Turbines and Power OCTOBER 2022, Vol. 144 / 101004-7 Fig. 9 Normalized nondimensional compressor map selected and optimized for case study 2. The isolines represent relative rotational speed. The winter (cross), summer (triangle), and spring operations (diamond) are shown. to quantify accurately the relative performance deviation depending on the seasonal conditions. This means that opera- tion cost variation throughout the year for varying operating conditions can be derived from the model results. For different source temperatures, the performance of the system is optimal at approximately 80% of the maximum heat load (see Fig. 8). Starting from the maximum heat load, the reduction in heat load results in an increase in COP until the 80% load capacity. This is due to lower exergetic losses in the HEXs while keeping the efﬁciency of the turbomachinery near their best efﬁciency at design point. When further decreasing the heat load, the differ- ence between operating point and design conditions results in lower turbomachinery efﬁciency, which offsets the lower exergy losses in the HEXs, and reduces the heat pump COP. The effect is accentuated by the reduction in turbulence in the HEX as a result of the mass ﬂow rates reduction, which leads to higher exergetic losses and further penalizes the COP. Downloaded from http://asmedigitalcollection.asme.org/gasturbinespower/article-pdf/144/10/101004/6912465/gtp_144_10_101004.pdf by guest on 24 October 2024 5.2.2 Case Study 2: High-Temperature District Heating Using Waste Heat From Data Centers as Heat Source. The second application presented in this study is the use of waste heat avail- able from a DC as the heat source for a local high-temperature dis- trict heating (DH) network [25]. This application enables sector- coupling and exploits both the high performance and operational ﬂexibility of trans-critical CO2 heat pumps by providing both heat and cold to two different consumers at the same time [26,27]. The optimal proﬁles of the supply and return temperatures and of the control variables during the year are presented in Fig. 9. The DC returns water at 15 C to the CO2 heat pump, delivering 17 MWth of chilled water at 7 C; this is used by the computer room air han- dler to cool down the servers room (also called white space). A constant cooling demand of the DC is assumed over the year according to the base load of a typical medium size DC. Journal of Engineering for Gas Turbines and Power Accord- ingly, a constant heat load is delivered to the local DH network of a middle size European city (þ200,000 inhabitants) at temperatures between 85 C and 120 C. Since the heat demand of DH networks varies seasonally, the coupling with the DC cooling demand is mostly possible for the base load of the DH heating demand. Fig. 9 Normalized nondimensional compressor map selected and optimized for case study 2. The isolines represent relative rotational speed. The winter (cross), summer (triangle), and spring operations (diamond) are shown. The normalized compressor map selected and optimized for this case study is shown in Fig. 10, whereas the parameters of the HEXs are reported in Table 3. Compared to case study 1, the nar- rower range of operation resulting from the constant demand allows to select a more efﬁcient compressor with a more limited operation range (see Fig. 10). Concerning the HEXs parameters, the major difference with respect to case study 1 is given by the evaporator (see Table 3). The higher temperature difference of the source side in the evaporator (8 K compared to approx. 3 K in case study 1) allows to increase the LMTD and to reduce the size of the component for equal heat load (see Eq. (1)). demand), the power input and the heat pump COP solely depend on the operating temperatures, which vary on the DH network side in a seasonal fashion. As shown in Fig. 9, the lower required temperature lift in summer allows to increase the COP of the sys- tem, ﬂuctuating between 2.87 and 3.06 depending on the season. Fig. 10 Yearly performance prediction and control of the trans- critical CO2 heat pump providing a constant cooling supply of 17 MWth at 7 C to a data center and heat to a local district heat- ing network Transactions of the ASME Operating strategy. The direct correlation between supply tem- perature and rotational speed of the compressor is directly linked to the discharge pressure that the compressor delivers in order to meet both the supply temperature to the DH (i.e., 85–120 C) and the cool- ing load to the DC (i.e., 17 MWth) at 7 C. As shown in Fig. 9, the higher supply temperatures required during the winter months result in higher rotational speeds. 6 Conclusions [11] Yao, L., Li, M., Hu, Y., Wang, Q., and Liu, X., 2021, “Comparative Study of Upgraded CO2 Transcritical Air Source Heat Pump Systems With Different Heat Sinks,” Appl. Therm. Eng., 184, p. 116289. This study presents the model of an optimization tool developed and used to predict the operation of trans-critical CO2 heat pumps in off-design conditions. As the operation of these systems strongly depend on the variable end user heat demand, it is key to predict the system performance for varying boundary conditions and to assess optimal operation strategies to maximize the heat pump performance, hence its proﬁtability, under realistic condi- tions. The tool allows at the same time high accuracy and fast computational performance and can in principle be used for any heat and cold industrial energy demand. Here, it is presented by referring to two case studies of interest, namely, (i) heat genera- tion using a large body of water as heat source and (ii) district heating using waste heat from data centers as heat source. [12] Lo Basso, G., de Santoli, L., Paiolo, R., and Losi, C., 2021, “The Potential Role of Trans-Critical CO2 Heat Pumps Within a Solar Cooling System for Building Services: The Hybridised System Energy Analysis by a Dynamic Simulation Model,” Renewable Energy, 164, pp. 472–490. [13] Cui, Q., Wang, C., Gao, E., and Zhang, X., 2022, “Pinch Point Characteristics and Performance Evaluation of CO2 Heat Pump Water Heater Under Variable Working Conditions,” Appl. Therm. Eng., 207, p. 118208. Working Conditions,” Appl. Therm. Eng., 207, p. 118208. [14] Rony, R., Yang, H., Krishnan, S., and Song, J., 2019, “Recent Advances in Transcritical CO2 (R744) Heat Pump System: A Review,” Energies, 12(3), p. 457. p [15] MAN Energy Solutions, 2021, “MAN Energy Solutions, Energy & Storage,” MAN Energy Solutions, Augsburg, Germany, accessed Dec. 2021, https:// www.man-es.com/energy-storage [16] Sarkar, J., Bhattacharyya, S., and Gopal, M. R., 2006, “Simulation of a Tran- scritical CO2 Heat Pump Cycle for Simultaneous Cooling and Heating Applications,” Int. J. Refrig., 29(5), pp. 735–743. The analysis of the optimal heat pump operation shows that the rotational speed of the compressor and the throttle valve opening are the most relevant control variables to maximize the heat pump performance over the year. 6 Conclusions The proposed model determines the set-point values of the main control variables to maximize the heat pump performance under different operating conditions, while complying with the operational limits imposed by the indi- vidual heat pump components. Such components include the com- pressor, the expander and the HEXs. As a result of the optimization procedure, the optimal values of the COP, of the compressor speed, and of the valve opening are shown for the complete operation envelopes of the system, i.e., for a wide range of source temperature and net heat load. It is worth noting that the maximum heat pump efﬁciency is predicted at about 80% of max heat load capacity, which suggests to maximize the operating time for these conditions whenever possible. [17] Li, Y.-M., and Wang, C.-C., 2021, “Investigation of the Performance of a Tran- scritical CO2 Heat Pump System Subject to Heated Water Conditions: Perspec- tive From the Second Law,” Appl. Therm. Eng., 193, p. 116999. [18] Ye, Z., Wang, Y., Zendehboudi, A., Hafner, A., and Cao, F., 2022, “Investigation on the Performance of Fluted Tube-in-Tube Gas Cooler in Transcritical CO2 Heat Pump Water Heater,” Int. J. Refrig., 135, pp. 208–220. [19] Lemmon, E. W., Bell, I., Huber, M. L., and McLinden, M. O., 2018, “NIST Standard Reference Database 23: Reference Fluid Thermodynamic and Trans- port Properties-REFPROP,” Version 10.0, National Institute of Standards and Technology, Gaithersburg, MD. [20] Casey, M. V., and Marty, F., 1986, Centrifugal Compressors-Performance at Design and Off-Design, The Institute of Refrigeration, London, UK. [21] Aungier, R. H., 2006, Turbine Aerodynamics: Axial-Flow and Radial-Flow Turbine Design and Analysis, ASME Press, New York. [22] Hansen, N., and Ostermeier, A., 2001, “Completely Derandomized Self- Adaptation in Evolution Strategies,” Evol. Comput., 9(2), pp. 159–195. [23] de Myttenaere, A., Golden, B., Le Grand, B., and Rossi, F., 2016, “Mean Abso- lute Percentage Error for Regression Models,” Neurocomputing, 192, pp. 38–48. Overall, the model can be used to assess and modify the opera- tional points set by different industrial end users during the year. This allows to (i) evaluate the effectiveness of a system design under realistic operating conditions and (ii) maximize the eco- nomic and environmental performance of an end user, while accounting for both design and off-design operations. [24] Garcia, L. S., and Emmanuel Jacquemoud, P. Table 3 HEX parameters for case study 2 Lappeenranta, Berlin,” Lappeenranta University of Technology, Lappeenranta, Finland, Research Reports No. 89. p [3] Lorentzen, G., 1994, “Revival of Carbon Dioxide as a Refrigerant,” Int. J. Refrig., 17(5), pp. 292–301. Neksa˚, P., 2002, “CO2 Heat Pump Systems,” Int. J. Refrig., 25(4), pp. 42 [5] ASHRAE, 2020, “Handbook - Refrigeration,” ASHRAE, Peachtree Corners, GA, accessed Dec. 2021, https://www.ashrae.org/technical-resources/ refrigeration [6] Sarkar, J., Bhattacharyya, S., and Ramgopal, M., 2009, “A Transcritical CO2 Heat Pump for Simultaneous Water Cooling and Heating: Test Results and Model Validation,” Int. J. Energy Res., 33(1), pp. 100–109. The sensitivity of heat pump performance on boundary condi- tions is key when design a heat pump, and especially the turboma- chinery. The developed model helps the decision-making process at design phase with the goal of optimizing the end user proﬁtabil- ity. As an example, the higher proﬁt expected from selling heat in winter leads to designs of the compressor and the expander that allows for the highest possible COP in winter, possibly reducing the COP during other periods of the year. This implies that the power consumption during summer operation could be further reduced if both compressor and expander designs were optimized speciﬁcally for these boundary conditions. [7] Dai, B., Qi, H., Liu, S., Ma, M., Zhong, Z., Li, H., Song, M., and Sun, Z., 2019, “Evaluation of Transcritical CO2 Heat Pump System Integrated With Mechani- cal Subcooling by Utilizing Energy, Exergy and Economic Methodologies for Residential Heating,” Energy Convers. Manage., 192, pp. 202–220. Residential Heating,” Energy Convers. Manage., 192, pp. 202–220. [8] Wang, Z., Wang, F., Ma, Z., Lin, W., and Ren, H., 2019, “Investigation on the Feasibility and Performance of Transcritical CO2 Heat Pump Integrated With Thermal Energy Storage for Space Heating,” Renewable Energy, 134, pp. 496–508. [9] He, Y.-J., Cheng, J.-H., Chang, M.-M., and Zhang, C.-L., 2021, “Modiﬁed Transcritical CO2 Heat Pump System With New Water Flow Conﬁguration for Residential Space Heating,” Energy Convers. Manage., 230, p. 113791. [10] Wang, J., Belusko, M., Liu, M., Semsarilar, H., Liddle, R., Alemu, A., Evans, M., Zhao, C., Hudson, J., and Bruno, F., 2021, “A Comprehensive Study on a Novel Transcritical CO2 Heat Pump for Simultaneous Space Heating and Cool- ing – Concepts and Initial Performance,” Energy Convers. Manage., 243, p. 114397. 6 Conclusions J., 2022, “Large Scale Tri-Gener- ation Energy Storage System for Heat, Cold and Electricity Based on Transcrit- ical CO2 Cycles,” 7th International Supercritical CO2 Power Cycles Symposium, San Antonio, TX, Feb. 21–24. [25] Li, J., Yang, Z., Li, H., Hu, S., Duan, Y., and Yan, J., 2021, “Optimal Schemes and Beneﬁts of Recovering Waste Heat From Data Center for District Heating by CO2 Transcritical Heat Pumps,” Energy Convers. Manage., 245, p. 114591. [2] Ram, M., Bogdanov, D., Aghahosseini, A., Gulagi, A., Oyewo, A. S., Child, M., and Caldera, I., 2018, “Global Energy System Based on 100% Renewable Energy – Energy Transition in Europe Across Power, Heat, Transport and Desalination Sectors, Study by LUT University and Energy Watch Group, Journal of Engineering for Gas Turbines and Power Moreover, the relatively small variation in speed throughout the year due to constant cooling supply allows to design the turbomachinery with a more restricted operating range, hence optimizing the efﬁciency over a narrower operating envelope The variation in valve position between summer and winter operation is explained by the lower loop resistance (i.e., lower compressor discharge pressure) required to satisfy the thermal demands during summer, and therefore the higher CO2 mass ﬂow- rate is necessary to deliver the cooling demand, maximizing both heat pump performance and proﬁtability in operation. Fig. 10 Yearly performance prediction and control of the trans- critical CO2 heat pump providing a constant cooling supply of 17 MWth at 7 C to a data center and heat to a local district heat- ing network Performance sensitivity. Since the heat pump system delivers a constant cooling load of 17 MWth (i.e., dictated by the DC Transactions of the ASME 101004-8 / Vol. 144, OCTOBER 2022 Transactions of the ASME Transactions of the ASME Transactions of the ASME Table 3 HEX parameters for case study 2 HEX Surface (m2) BH (m–2) BC (m–2) Evaporator 598 1.0210–1 1.69107 Gas cooler 2250 0.666 2.192 Recuperator 435 12.35 16.24 Table 3 HEX parameters for case study 2 [1] REN21, 2021, “Renewables 2021 Global Status Report,” REN21, Paris, France, accessed Dec. 2021, https://www.ren21.net/reports/global-status-report/ [2] Ram, M., Bogdanov, D., Aghahosseini, A., Gulagi, A., Oyewo, A. S., Child, M., and Caldera, I., 2018, “Global Energy System Based on 100% Renewable Energy – Energy Transition in Europe Across Power, Heat, Transport and Desalination Sectors, Study by LUT University and Energy Watch Group, References [26] Diaby, A. T., Byrne, P., and Mare, T., 2019, “Simulation of Heat Pumps for Simultaneous Heating and Cooling Using CO2,” Int. J. Refrig., 106, pp. 616–627. [1] REN21, 2021, “Renewables 2021 Global Status Report,” REN21, Paris, France, accessed Dec. 2021, https://www.ren21.net/reports/global-status-report/ [27] Gabrielli, P., Acquilino, A., Siri, S., Bracco, S., Sansavini, G., and Mazzotti, M., 2020, “Optimization of Low-Carbon Multi-Energy Systems With Seasonal Geothermal Energy Storage: The Anergy Grid of ETH Zurich,” Energy Convers. Manage.: X, 8, p. 100052. [2] Ram, M., Bogdanov, D., Aghahosseini, A., Gulagi, A., Oyewo, A. S., Child, M., and Caldera, I., 2018, “Global Energy System Based on 100% Renewable Energy – Energy Transition in Europe Across Power, Heat, Transport and Desalination Sectors, Study by LUT University and Energy Watch Group, OCTOBER 2022, Vol. 144 / 101004-9 Journal of Engineering for Gas Turbines and Power
https://openalex.org/W2953532584	https://europepmc.org/articles/pmc6544397?pdf=render	English	null	Mesenchymal stem cells confer chemoresistance in breast cancer via a CD9 dependent mechanism	Oncotarget	2,019	cc-by	10,449	1Interventional Regenerative Medicine and Imaging Laboratory, Stanford University School of Medicine, Department of Radiology, Palo Alto, CA 94304, USA 2Mid-Florida Research and Education Center, Department of Pathology, University of Florida, Apo 2Mid-Florida Research and Education Center, Department of Pathology, University of Florida, Apopka, FL 32703, USA Correspondence to: Mujib Ullah, email: ullah@stanford.edu Keywords: MSCs; CD9; chemoresistance; cytokine; xenograft tumors Received: March 28 2019 Accepted: May 05 2019 Published: May 28 2019 Correspondence to: Mujib Ullah, email: ullah@stanford.edu Keywords: MSCs; CD9; chemoresistance; cytokine; xenograft tumors Received: March 28, 2019 Accepted: May 05, 2019 Published: May 28, 2019 Copyright: Ullah et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: Ullah et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT The development of chemotherapy drug resistance remains a significant barrier for effective therapy in several cancers including breast cancer. Bone marrow- derived mesenchymal stem cells (BMMSCs) have previously been shown to influence tumor progression and the development of chemoresistance. In the present study, we showed that when GFP labelled BMMSCs and RFP labelled HCC1806 cells are injected together in vivo, they create tumors which contain a new hybrid cell that has characteristics of both BMMSCs and HCC1806 cells. By labelling these cells prior to their injection, we were then able to isolate new hybrid cell from harvested tumors using FACS (DP-HCC1806:BMMSCs). Interestingly, when DP-HCC1806:BMMSCs were then injected into the mammary fat pad of NOD/SCID mice, they produced xenograft tumors which were smaller in size, and exhibited resistance to chemotherapy drugs (i.e. doxorubicin and 5-fluorouracil), when compared tumors from HCC1806 cells alone. This chemoresistance was shown to associated with an increased expression of tetraspanins (CD9, CD81) and drug resistance proteins (BCRP, MDR1). Subsequent siRNA-mediated knockdown of BMMSC-CD9 in DP-HCC1806:BMMSCs resulted in an attenuation of doxorubicin and 5-fluorouracil chemoresistance associated with decreased BCRP and serum cytokine expression (CCL5, CCR5, CXCR12). Our findings suggest that within the tumor microenvironment, CD9 is responsible for the crosstalk between BMMSCs and HCC1806 breast cancer cells (via CCL5, CCR5, and CXCR12) which contributes to chemoresistance. Hence, BMMSC-CD9 may serve as an important therapeutic target for the treatment of breast cancer. Mujib Ullah1, Asma Akbar2, Nathan Norton Ng1, Waldo Concepcion1 and Avnesh S. Thakor1 1Interventional Regenerative Medicine and Imaging Laboratory, Stanford University School of Medicine, Department of Radiology, Palo Alto, CA 94304, USA www.oncotarget.com www.oncotarget.com www.oncotarget.com www.oncotarget.com INTRODUCTION remains a significant barrier for effective therapy [4]. Hence, uncovering the mechanisms that promote chemoresistance is important for developing therapies that can treat breast cancer by impeding tumor growth as well as preventing disease relapse. Breast cancer is the most diagnosed cancer and leading cause of cancer related death amongst women, with a global incidence of nearly 1.7 million new cases each year and over 520,000 deaths [1, 2]. It is categorized according to several subtypes and is highly heterogeneous in its disease progression, rate of metastasis, and prognosis, thereby making it a challenge to treat [3]. While chemotherapy is part of the standard of care for patients with breast cancer, the development of drug resistance The progression of breast cancer and its subsequent development of chemoresistance, is highly dependent on the paracrine and cell-cell interactions between the tumor and its surrounding microenvironment, which consists of fibroblasts, immune cells, endothelial cells, and mesenchymal stem cells (MSCs) [5]. MSCs are self- www.oncotarget.com Oncotarget 3435 tumors created by RFP-labeled HCC1806 cells (Figure 1E–1G). To further evaluate the interaction between RFP- labeled HCC1806 cells and GFP-labeled BMMSCs, both cells were injected in vivo with tumors isolated daily from animals over the next 4 days. The different cell populations were then sorted through a fluorescence activated cell sorter (FACS) (Figure 2A–2C). While there was no significant change in the percent expression of RFP- HCC1806 cells over 4 days, there was a steady decrease in the percent expression of GFP-BMMSCs and this was accompanied by an increase in the percent expression of a new population of HCC1806:BMMSCs (i.e. hybrid cells which were double positive (DP) for GFP-BMMSCs and RFP-HCC1806 cells) (Figure 2D, 2E). This new population of cell: DP-HCC1806:BMMSCs, was then specifically examined in all of our subsequent experiments. renewing multipotent cells, found in bone marrow, adipose tissue, umbilical cord blood, and placental tissue, that are capable of differentiating into cells of the mesodermal lineage such as adipocytes, chondrocytes, and osteocytes [6–8]. Although MSCs have an important well-defined therapeutic role in tissue repair and regenerative medicine [9–12], their role in cancer biology is less certain. While MSCs have been shown to be recruited to the site of tumors, via endocrine and paracrine signaling [13–15], there are multiple studies reporting their anti, as well as pro-, tumorigenic properties. Evaluating the in vivo effect of BMMSCs in xenograft breast cancer animal models y g y [ ] Hence, further clarification of the molecular interactions between different cancer subtypes and MSCs is warranted. In the present study, we investigated the effect of human bone marrow-derived MSCs (BMMSCs) on an ERα-, PR-, and HER2-negative (triple negative) breast cancer cell line (HCC1806). We found that BMMSCs and HCC1806 cells behave and interact differently depending on whether they are co-cultured in vitro or in vivo. Indeed, we found that when BMMSCs and HCC1806 cells are co-cultured in vivo, they create tumors which contain a new hybrid cell (DP-HCC1806:BMMSCs) that has characteristics of both BMMSCs and HCC1806 cells. Following isolation of DP-HCC1806:BMMSCs, we then showed in vivo that this hybrid cell could create tumors like conventional HCC1806 cells, but that these tumors were reduced in size and exhibited increased resistance to chemotherapeutic agents. Next, we determined that this effect was dependent on the expression of CD9 in BMMSCs. Taken together, our findings provide insight into the possible mechanism by which BMMSCs may influence breast cancer development and chemotherapy drug resistance. In NOD/SCID mice, the following cells were injected into the mammary fat pad: RFP-HCC1806 cells, GFP-BMMSCs, or DP-HCC1806:BMMSCs. Animals which received GFP-BMMSCs alone produced no tumors, however, animals which received either RFP- HCC1806 alone or DP-HCC1806:BMMSCs developed tumors. At week 8, although there was no difference in the body weight of animals (Figure 1B), the excised tumors from DP-HCC1806:BMMSCs xenograft animals had a decreased volume when compared to animals which received RFP-HCC1806 cells alone (Figure 2F). INTRODUCTION Indeed, some studies have shown that MSCs can promote tumor progression by stimulating tumor proliferation, angiogenesis, motility, metastasis, and tissue invasion [16–19]. In contrast, other studies have shown MSCs to have an inhibitory effect on cancer progression by inducing apoptosis, suppressing signaling pathways, initiating cell-cycle arrest, and increasing infiltration of monocyotes and granulocytes [20–23]. Evaluating the in vivo effects of RFP- HCC1806 cells and DP-HCC1806:BMMSCs to chemotherapeutic drugs In NOD/SCID mice, the following cells were injected into the mammary fat pad: RFP-HCC1806 cells or DP-HCC1806:BMMSCs. After 10 days, animals were subjected to 25 days of chemotherapy (i.e. either doxorubicin (Dox; 10 mg/kg), mithramycin A (MTR; 1 mg/kg), or 5-fluorouracil (5FU; 10 mg/kg). In xenograft animals created with RFP-HCC1806 cells, there was a reduction in both the rate and magnitude of tumor growth when animals were treated with Dox or 5FU compared to control animals which received no chemotherapy treatment. In contrast, in xenograft animals created with DP-HCC1806:BMMSCs, treatment with Dox and 5FU resulted in no change in the rate and magnitude of tumor growth compared to control animals, thereby demonstrating chemoresistance within these animals (Figure 3A–3C, 3F, 3G). The limited reduction in tumor volume in xenograft animals created with DP-HCC1806:BMMSCs was also accompanied by a reduction in caspase-3 activity following 5FU treatment (Figure 3H–3I). Regardless of the cells used to create Determining the interaction between BMMSCs and HCC1806 breast cancer cells To evaluate the effect and interaction between RFP-labeled HCC1806 cells and GFP-labeled BMMSCs, these cells were injected either alone or together into the mammary fat pad of NOD/SCID mice (Figure 1A– 1C). While BMMSCs did not grow into tumors in this environment, RFP-labeled HCC1806 cells did produce tumors which progressively increased in size over 8 weeks (Figure 1D). Interestingly, when RFP-labeled HCC1806 cells and GFP-labeled BMMSCs were injected together, the tumor size was markedly reduced in comparison to www.oncotarget.com Oncotarget 3436 cells demonstrating increased cell apoptosis (assessed qualitatively using Trypan blue exclusion (Figure 4A, 4B)); reduced cell viability (assessed quantitatively using an MTT assay (Figure 4C, 4D)); and increased cell proliferation (assessed quantitatively using RFP/ GFP fluorescence (Figure 4B)). In contrast, by day 6, DP-HCC1806:BMMSCs demonstrated no significant change in cell viability or proliferation following chemotherapy treatment. This was accompanied by DP-HCC1806:BMMSCs showing reduced expression of cytotoxic caspase-3 and thioredoxin reductase when compared to RFP-HCC1806 cells (Figure 4E). xenograft animals, there was no difference in either the rate or magnitude of tumor growth when animals were given MTR compared to animals which received no chemotherapy treatment (Figure 3D, 3E). Determining the mechanism of chemoresistance of DP-HCC1806:BMMSCs to Dox and 5FU Following 6 days of in vitro culture, RFP-HCC1806 cells or DP-HCC1806:BMMSCs were treated with either Dox (100 µM) or 5FU (300 µM). From day 2 to day 6, chemotherapy treatment resulted in RFP-HCC1806 Following 6 days of in vitro culture, RFP-HCC1806 cells or DP-HCC1806:BMMSCs were treated with either Dox (100 µM) or 5FU (300 µM). From day 2 to day 6, chemotherapy treatment resulted in RFP-HCC1806 chemotherapy treatment. This was accompanied by DP-HCC1806:BMMSCs showing reduced expression of cytotoxic caspase-3 and thioredoxin reductase when compared to RFP-HCC1806 cells (Figure 4E). Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimental design. Xenografts consisting of BMMSCs (1 × 106 cells), HCC1806 cells (1 × 106 cells), or a coculture of HCC1806:BMMSCs (1 × 106 cells per line) were injected into immunocompromised NOD/SCID mice. Tumor burden was assessed weekly in injected animals, for a total of 8 weeks. (B) Total animal body weight by week 8. (C) Representative images of bone marrow-derived BMMSCs, labeled with GFP, and HCC1806 breast cancer cells, labeled with RFP, that were cocultured prior to in vivo injection. (D) Weekly tumor volume in grafted animals over 8 weeks. (E) Week 8 tumor volume in excised tissue samples. (F) Week 8 tumor weight in excised tissue samples. (G) Representative images of tumor excised from sacrificed animals. Significance indicated by p < 0.05, p < 0.01 for comparison between HCC1806 and HCC1806:BMMSCs. ###p < 0.001 for comparison between BMMSCs and HCC1806:BMMSCs. Scale bar 100 µm. Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimental des Xenografts consisting of BMMSCs (1 × 106 cells), HCC1806 cells (1 × 106 cells), or a coculture of HCC1806:BMMSCs (1 × 106 c per line) were injected into immunocompromised NOD/SCID mice. Tumor burden was assessed weekly in injected animals, for a tota 8 weeks. (B) Total animal body weight by week 8. (C) Representative images of bone marrow-derived BMMSCs, labeled with GFP, HCC1806 breast cancer cells, labeled with RFP, that were cocultured prior to in vivo injection. (D) Weekly tumor volume in grafted anim 8 k (E) W k 8 t l i i d ti l (F) W k 8 t i ht i i d ti l (G) R t Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimental design. Determining the mechanism of chemoresistance of DP-HCC1806:BMMSCs to Dox and 5FU To further investigate the mechanisms underlying the chemoresistance of DP-HCC1806:BMMSCs, we evaluated the changes in the expression pattern of tetraspanin proteins (CD9 and CD81), drug resistance proteins (BCRP and MDR1), and common targets of cancer pathways (p-ERK, pMAPK, mTOR, PI3K, pAKT, and p53) (Figure 5A–5D). Western blot analysis demonstrated an increase in the protein expression of CD9, CD81, BCRP and MDR1 accompanied by a decrease in the protein expression of mTOR in DP- HCC1806:BMMSCs compared to both GFP-BMMSCs or RFP-HCC1806 cells. Flow cytometric distribution Determining the mechanism of chemoresistance of DP-HCC1806:BMMSCs to Dox and 5FU Xenografts consisting of BMMSCs (1 × 106 cells) HCC1806 cells (1 × 106 cells) or a coculture of HCC1806:BMMSCs (1 × 106 cells Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagr SCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimenta Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimental design. Xenografts consisting of BMMSCs (1 × 106 cells), HCC1806 cells (1 × 106 cells), or a coculture of HCC1806:BMMSCs (1 × 106 cells per line) were injected into immunocompromised NOD/SCID mice. Tumor burden was assessed weekly in injected animals, for a total of 8 weeks. (B) Total animal body weight by week 8. (C) Representative images of bone marrow-derived BMMSCs, labeled with GFP, and HCC1806 breast cancer cells, labeled with RFP, that were cocultured prior to in vivo injection. (D) Weekly tumor volume in grafted animals over 8 weeks. (E) Week 8 tumor volume in excised tissue samples. (F) Week 8 tumor weight in excised tissue samples. (G) Representative images of tumor excised from sacrificed animals. Significance indicated by p < 0.05, *p < 0.01 for comparison between HCC1806 and HCC1806:BMMSCs. ###p < 0.001 for comparison between BMMSCs and HCC1806:BMMSCs. Scale bar 100 µm. Figure 1: BMMSCs reduce tumor burden of HCC1806 xenografts in vivo. (A) Schematic diagram of our experimental design. Xenografts consisting of BMMSCs (1 × 106 cells), HCC1806 cells (1 × 106 cells), or a coculture of HCC1806:BMMSCs (1 × 106 cells per line) were injected into immunocompromised NOD/SCID mice. Tumor burden was assessed weekly in injected animals, for a total of 8 weeks. (B) Total animal body weight by week 8. (C) Representative images of bone marrow-derived BMMSCs, labeled with GFP, and HCC1806 breast cancer cells, labeled with RFP, that were cocultured prior to in vivo injection. (D) Weekly tumor volume in grafted animals over 8 weeks. (E) Week 8 tumor volume in excised tissue samples. (F) Week 8 tumor weight in excised tissue samples. (G) Representative images of tumor excised from sacrificed animals. Significance indicated by p < 0.05, p < 0.01 for comparison between HCC1806 and HCC1806:BMMSCs. ###p < 0.001 for comparison between BMMSCs and HCC1806:BMMSCs. Scale bar 100 µm. www.oncotarget.com Oncotarget 3437 analyses indicated that the greatest change was in the expression of CD9 and MDR1 (Figure 5B). The role of CD9 in mediating the chemoresistance of DP-HCC1806:BMMSC protein expression in DP-HCC1806:BMMSCs-siCD9 cells at day 6 (Figure 6D). to chemotherapy treatment with both Dox and 5FU showing a reduction in tumor volume at 4 weeks when compared to the same treatment given to xenograft animals created with DP-HCC1806:BMMSCs (Figure 6E). In vitro analysis of DP-HCC1806:BMMSCs-siCD9 also now demonstrated these cells to now be sensitive to chemotherapeutics with them exhibiting a decrease in cell viability (assessed quantitatively using an MTT assay) and increased apoptosis (assessed quantitatively using Trypan blue exclusion), compared to DP-HCC1806:BMMSCs after 6 days of treatment (Figure 6C, 6F, 6H). This reduced viability was also accompanied by a decrease in BCRP The role of CD9 in mediating the chemoresistance of DP-HCC1806:BMMSC Next, we performed siRNA-mediated knockdown of CD9 in BMMSCs and then co-cultured these cells in vivo with HCC1806 (Figure 6A, 6B). DP-HCC1806:BMMSCs- siCD9 were then isolated as previously described using FACS. When DP-HCC1806:BMMSCs-siCD9 were used to create a xenograft model, tumors were now responsive Figure 2: FACS-sorted GFP/RFP-double positive cells from HCC1806:BMMSC xenografts reduced tumor volume in vivo. (A) Animals injected with MSC-GFP, HCC1806-RFP, and HCC1806-RFP:MSC-GFP xenografts are monitored over a period of four days. On each day, xenografts are harvested and FACS-sorted into GFP-positive (BMMSCs), RFP-positive (HCC1806 cells), or GFP/RFP-double positive cells (DP-HCC1806:BMMSCs). (B) Schematic diagram of our experimental design: Sorted cells are reinjected into the animals to assess tumor burden. (C) FACS sorting of GFP-positive, RFP-positive, and GFP/RFP-double positive cells from xenografts isolated from vehicle, GFP-BMMSC, RFP-HCC1806, and DP-HCC1806:BMMSC xenografted animals (n = 6 animals). (D) Representative micrographs of FACS-sorted GFP-BMMSCs, RFP-HCC1806 cells, and GFP/RFP-double positive cells. (E) Percent total expression of cells sorted from harvested xenografts during a four day period. (F) Animals were reinjected with either FACS-sorted RFP-positive cells (HCC1806 cells), GFP/RFP-double positive cells (DP-HCC1806:BMMSCs), or GFP-positive cells (BMMSCs), and tumor volume was assessed at day 35 post-injection. Data are reported as mean ± s.e.m., with significance indicated by p < 0.001 and p < 0.01. Scale bar 100 µm. Figure 2: FACS-sorted GFP/RFP-double positive cells from HCC1806:BMMSC xenografts reduced tumor volume in vivo. (A) Animals injected with MSC-GFP, HCC1806-RFP, and HCC1806-RFP:MSC-GFP xenografts are monitored over a period of four days. On each day, xenografts are harvested and FACS-sorted into GFP-positive (BMMSCs), RFP-positive (HCC1806 cells), or GFP/RFP-double positive cells (DP-HCC1806:BMMSCs). (B) Schematic diagram of our experimental design: Sorted cells are reinjected into the animals to assess tumor burden. (C) FACS sorting of GFP-positive, RFP-positive, and GFP/RFP-double positive cells from xenografts isolated from vehicle, GFP-BMMSC, RFP-HCC1806, and DP-HCC1806:BMMSC xenografted animals (n = 6 animals). (D) Representative micrographs of FACS-sorted GFP-BMMSCs, RFP-HCC1806 cells, and GFP/RFP-double positive cells. (E) Percent total expression of cells sorted from harvested xenografts during a four day period. (F) Animals were reinjected with either FACS-sorted RFP-positive cells (HCC1806 cells), GFP/RFP-double positive cells (DP-HCC1806:BMMSCs), or GFP-positive cells (BMMSCs), and tumor volume was assessed at day 35 post-injection. Data are reported as mean ± s.e.m., with significance indicated by p < 0.001 and p < 0.01. Scale bar 100 µm. www.oncotarget.com Oncotarget 3438 protein expression in DP-HCC1806:BMMSCs-siCD9 cells at day 6 (Figure 6D). Analysis of serum and tissue cytokines in xenograft breast cancer animal models However, the level of all 3 of these proteins was reduced in the serum of animals which had been injected with DP-HCC1806:BMMSC-siCD9 cells (Figure 6I). We then examined the tumor specimens at the time of sacrifice and again showed that there was an increased expression of CXCL12 in DP-HCC1806:BMMSCs relative to DP- HCC1806:BMMSC-siCD9 was confirmed using real-time DP-HCC1806:BMMSCs and DP-HCC1806:BMMSC- siCD9 cells. At 4 weeks, animals which had been injected with DP-HCC1806:BMMSCs showed a higher expression of CCL5, CCR5 and CXCL12 compared to RFP-HCC1806 or GFP-BMMSC xenografts (Figure 6G). However, the level of all 3 of these proteins was reduced in the serum of animals which had been injected with DP-HCC1806:BMMSC-siCD9 cells (Figure 6I). We then examined the tumor specimens at the time of sacrifice and again showed that there was an increased expression of CXCL12 in DP-HCC1806:BMMSCs relative to DP- HCC1806:BMMSC-siCD9 was confirmed using real-time RT-PCR and an ELISA (Figure 6J) Of note, regardless of the cells which were injected, the serum expression of IL-6 was increased. Analysis of serum and tissue cytokines in xenograft breast cancer animal models In order to understand the molecular interactions and signaling pathways involved between BMMSCs and HCC1806 cells, we performed a screen for different cell surface proteins, growth factors, chemotactic factors, and inflammatory factors in the serum of mice which had been injected with RFP-HCC1806 cells, GFP-BMMSCs, Figure 3: DP-HCC1806:BMMSC xenografts mediate chemotherapeutic resistance. (A) Representative images showing the growth of sorted cells. (B) Tumor volume upon in vivo administration of doxorubicin (10 mg/kg) for 25 days. (C) Tumor volume at day 35 in doxorubicin-treated animals. (D) Tumor volume upon in vivo administration of mithramycin A (1 mg/kg) for 35 days. (E) Tumor volume at day 35 in mithramycin A-treated animals. (F) Tumor volume upon in vivo administration of 5-fluorouracil (5FU) (10 mg/kg) for 35 days. (G) Tumor volume at day 35 in 5FU-treated animals. (H, I) In vivo tumor reduction validated by flow cytometric confirmation of caspase-3 cell death assays in HCC1806 and DP-HC1806:BMMSCs xenografted animals, n = 6 animals). Data are reported as mean ± s.e.m., with significance indicated by *p < 0.001, p < 0.01 and p < 0.05. Scale bar 100 µm. Figure 3: DP-HCC1806:BMMSC xenografts mediate chemotherapeutic resistance. (A) Representative images showing the growth of sorted cells. (B) Tumor volume upon in vivo administration of doxorubicin (10 mg/kg) for 25 days. (C) Tumor volume at day 35 Figure 3: DP-HCC1806:BMMSC xenografts mediate chemotherapeutic resistance. (A) Representative images showing the growth of sorted cells. (B) Tumor volume upon in vivo administration of doxorubicin (10 mg/kg) for 25 days. (C) Tumor volume at day 35 in doxorubicin-treated animals. (D) Tumor volume upon in vivo administration of mithramycin A (1 mg/kg) for 35 days. (E) Tumor volume at day 35 in mithramycin A-treated animals. (F) Tumor volume upon in vivo administration of 5-fluorouracil (5FU) (10 mg/kg) for 35 days. (G) Tumor volume at day 35 in 5FU-treated animals. (H, I) In vivo tumor reduction validated by flow cytometric confirmation of caspase-3 cell death assays in HCC1806 and DP-HC1806:BMMSCs xenografted animals, n = 6 animals). Data are reported as mean ± s.e.m., with significance indicated by p < 0.001, p < 0.01 and p < 0.05. Scale bar 100 µm. www.oncotarget.com Oncotarget 3439 DP-HCC1806:BMMSCs and DP-HCC1806:BMMSC- siCD9 cells. At 4 weeks, animals which had been injected with DP-HCC1806:BMMSCs showed a higher expression of CCL5, CCR5 and CXCL12 compared to RFP-HCC1806 or GFP-BMMSC xenografts (Figure 6G). DISCUSSION In the present study, both RPF-HCC1806 cells and DP-HCC1806:BMMSCs were able to create xenograft tumor models when injected into the mammary fat pad of NOD/SCID mice. However, the tumors created by DP-HCC1806:BMMSCs were smaller as well as more resistant to chemotherapy. Further studies determined that Oncotarget 3440 www oncotarget com Figure 4: Increased cell viability and reduced cytotoxicity underlie chemoresistance to doxorubicin and 5FU. (A, B) In vitro cell death assessed qualitatively using trypan blue exclusion in HCC1806 and HCC1806:BMMSCs cocultures treated with either 300 µM 5FU or 100 µM doxorubicin for 6 days. (C) MTT assay for doxorubicin-treated cultures and percent viable cells by day 6 of treatment. (D) MTT assay for 5FU-treated cultures and percent viable cells by day 6 of treatment. (E) Degree of in vitro cytotoxicity to chemotherapeutic agents, assessed through caspase-3 and thioredoxin (TXN) reductase activity. Data are reported as mean ± s.e.m., with significance indicated by p < 0.01, p < 0.001. Scale bar 100 µm. Figure 4: Increased cell viability and reduced cytotoxicity underlie chemoresistance to doxorubicin and 5FU. (A, B) In vitro cell death assessed qualitatively using trypan blue exclusion in HCC1806 and HCC1806:BMMSCs cocultures treated with either 300 µM 5FU or 100 µM doxorubicin for 6 days. (C) MTT assay for doxorubicin-treated cultures and percent viable cells by day 6 of treatment. (D) MTT assay for 5FU-treated cultures and percent viable cells by day 6 of treatment. (E) Degree of in vitro cytotoxicity to chemotherapeutic agents, assessed through caspase-3 and thioredoxin (TXN) reductase activity. Data are reported as mean ± s.e.m., with significance indicated by p < 0.01, **p < 0.001. Scale bar 100 µm. www.oncotarget.com Oncotarget 3440 our data has shown that BMMSCs are not inherently tumorigenic, especially given that they were not able to induce neoplasms when injected into the mammary fat pad. However, BMMSCs have been shown to facilitate tumorigenic behavior when cultured with breast cancer cells via direct (cell-cell) and indirect (endocrine and paracrine signaling) interactions [16–19, 28, 29]. In support of this, Park et al. have reported that while MSCs were not capable of inducing neoplastic transformation, they did however create an inflammatory microenvironment conducive towards tumor growth and angiogenesis [27]. In our studies, we have shown that BMMSCs interact closely with breast cancer cells (i.e. DISCUSSION Although the exact mechanism by which BMMSCs might be inducing these changes in HC1806 CD9 is an integral membrane protein which contains four-membrane spanning domains and is found on the cell surface [30, 31], in exosomes [32, 33], and in nuclei [34]. Studies have shown that CD9 plays a diverse role in both cancer and stem cell biology by regulating numerous cellular processes, such as cell adhesion, proliferation, apoptosis, motility, mitosis, and even extracellular vesicle (EV) secretion [30–38]. In Figure 6: BMMSCs-CD9 siRNA knockdown in DP-HCC1806:BMMSC xenografts reduces chemotherapeutic resistance to 5FU and doxorubicin in vivo. (A, B) Western blot and real time PCR analyses confirming siRNA-mediated knockdown of CD9 in BMMSCs. (C) Percent viability of 5FU-treated DP-HCC1806:BMMSCs and DP-HCC1806:BMMSCs-siCD9 cultures. (D) Western blot analyses of BCRP expression. (E) In vivo tumor volume in animals co-injected with DP-HCC1806:BMMSCs or DP- HCC1806:BMMSCs-siCD9 cells, evaluated at week 4 post-injection. (F) MTT assay for BMMSC, HCC1806, DP-HCC1806:BMMSCs, Figure 6: BMMSCs-CD9 siRNA knockdown in DP-HCC1806:BMMSC xenografts reduces chemotherapeutic resistance to 5FU and doxorubicin in vivo. (A, B) Western blot and real time PCR analyses confirming siRNA-mediated knockdown of CD9 in BMMSCs. (C) Percent viability of 5FU-treated DP-HCC1806:BMMSCs and DP-HCC1806:BMMSCs-siCD9 cultures. (D) Western blot analyses of BCRP expression. (E) In vivo tumor volume in animals co-injected with DP-HCC1806:BMMSCs or DP- HCC1806:BMMSCs-siCD9 cells, evaluated at week 4 post-injection. (F) MTT assay for BMMSC, HCC1806, DP-HCC1806:BMMSCs, and DP-HCC1806:BMMSCs-siCD9 cultures. (G) Serum levels of various cytokines, and inflammatory factors from mice injected with HCC1806, BMMSCs, and DP-HCC1806:BMMSC xenografts. (H) In vitro cell death assessed qualitatively using trypan blue exclusion. (I) Cytokine profile for BMMSCs, HCC1806 cells, and DP-HCC1806:BMMSCs-siCD9 cells. (J, right) Cytokine ELISA assay and real- time RT-PCR analyses of CXCL12 protein and mRNA levels in excised tumor tissue. Mean ± s.e.m, n = 6. Significant difference indicated by p < 0.05, p < 0.01, p < 0.001 for comparison between HCC1806 and DP-HCC1806:BMMSCs. #p < 0.05, ##p < 0.01, ###p < 0.001 for comparison between BMMSCs and DP-HCC1806:BMMSCs. Scale bar 100 µm. Figure 6: BMMSCs-CD9 siRNA knockdown in DP-HCC1806:BMMSC xenografts reduces chemotherapeutic resistance to 5FU and doxorubicin in vivo. (A, B) Western blot and real time PCR analyses confirming siRNA-mediated knockdown of CD9 in BMMSCs. (C) Percent viability of 5FU-treated DP-HCC1806:BMMSCs and DP-HCC1806:BMMSCs-siCD9 cultures. (D) Western blot analyses of BCRP expression. DISCUSSION HCC1806), and when co-cultured together in vivo, they actually produce a new this chemoresistance was due to an increase in expression of CD9, CD81, BCRP and MDR1 proteins. The effects of CD9, which demonstrated the strongest expression in DP-HCC1806:BMMSCs, appeared to be mediated by the CXCL12 protein. When CD9 was silenced in BMMSCs, xenograft animals created with DP-HCC1806:BMMSC- siCD9 now exhibited reduced CXCL12 protein expression with a resulting increased sensitivity to chemotherapy (Figure 7). For BMMSCs to be considered as a “safe” cell to be used in regenerative medicine, tissue engineering and stem cell therapy applications, it is important to carefully assess whether the cells have any tumorigenic potential. Consistent with previous studies [24–27], t t Figure 5: Differential expression of tetraspanins and drug resistance proteins in DP-HCC1806:BMMSCs. (A) Western blot analyses of sorted cells for CD9, BCRP, MDR1 and CD81 expression. (B) Flow cytometry analysis of CD9, BCRP, MDR1, and CD81 protein expression distribution. (C) Western blot analyses of sorted cells for p-ERK, pMAPK, mTOR, PI3K, p53, and pAKT expression. (D) Quantification of western blot data proteins confirming increased expression of proteins CD9, BCRP, MDR1 and CD81 protein. Data are reported as mean ± s.e.m., with significance indicated by p < 0.05, p < 0.01, p < 0.001. Figure 5: Differential expression of tetraspanins and drug resistance proteins in DP-HCC1806:BMMSCs. (A) Western blot analyses of sorted cells for CD9, BCRP, MDR1 and CD81 expression. (B) Flow cytometry analysis of CD9, BCRP, MDR1, and CD81 protein expression distribution. (C) Western blot analyses of sorted cells for p-ERK, pMAPK, mTOR, PI3K, p53, and pAKT expression. (D) Quantification of western blot data proteins confirming increased expression of proteins CD9, BCRP, MDR1 and CD81 protein. Data are reported as mean ± s.e.m., with significance indicated by p < 0.05, p < 0.01, p < 0.001. www.oncotarget.com Oncotarget 3441 cells is unclear, our work has shown that this may be mediated, in part, by CD9. hybrid cell that has molecular markers of both BMMSCs and HCC1806 cells (DP-HCC1806:BMMSCs); this is likely due to either the BMMSCs being internalized by HCC1806 cells or membrane fragments of BMMSCs being attached to HC1806 cells. Interestingly, this changed the phenotype of HCC1806 cells with the new hybrid DP- HCC1806:BMMSCs producing smaller tumors that were more chemoresistant when injected into the mammary fat pad of animals. DISCUSSION Interestingly, our results showed that DP-HCC1806:BMMSCs produced smaller tumors; however, these tumors were more chemoresistant to Dox and 5FU which we found was dependent on their increased CD9 expression. In keeping with this, studies in small cell lung cancer (SCLC) cells have shown that ectopic overexpression of CD9 enhances β1 integrin-mediated cell adhesion to extracellular matrix (ECM) fibronectin which has been implicated in cell adhesion mediated drug resistance (CAM-DR) [36, 39]. In addition, there was also a CD9 dependent increase in MDR1 and BCRP; both of which are ABC transporters that have been shown in other cancer models to confer multidrug resistance to chemotherapies such as Dox and 5FU [40, 41]. Although MTR (1 mg/kg) has previously been shown to have anti- tumor effects [42, 43], it had no effect in the present study on xenograft tumors from either RFP-HCC1806 cells or DP-HCC1806:BMMSCs. One explanation for this is that MTR is not able to be transported through MDR1 or BCRP, and has even been shown to downregulate MDR1/BCRP expression [40, 44, 45]. This data therefore suggests that the resistance of RFP-HCC1806 to MTR is independent of CD9-associated BCRP and MDR1 overexpression. Given that the inhibition of BMMSC-CD9 expression failed to fully restore chemotherapeutic sensitivity in tumors created from DP-HCC1806:BMMSCs-siCD9 cells, this suggests that other molecular mechanisms Figure 7: Flow diagram illustrating doxorubicin and 5FU chemoresistance in breast cancer tumors. Schematic diagram showing BMMSCs and HCC1806 coculture. BMMSCs are green due to GFP expression and HCC1806 are red due to RFP expression. In vivo co-culture of HCC1806 and BMMSCs cells produced a hybrid cell (DP-HCC1806:BMMSC) whose interaction is dependent on the CD9 expression in BMMSCs. DP-HCC1806:BMMSCs were then used to create xenograft tumors which were smaller but more chemoresistant to agents such as doxorubicin and 5FU. Next, CD9 was inhibited by siRNA in BMMSCs and then co-cultured with HCC1806 cells. These DP-HCC1806:BMMSCs-siCD9 cells were then injected into mice to create xenograft tumors which were now no longer resistant to chemotherapeutic agents. CD9 inhibition also reversed the expression of tumor resistant proteins such as BCRP, CCL5, CCR5, and CXCL12. Figure 7: Flow diagram illustrating doxorubicin and 5FU chemoresistance in breast cancer tumors. Schematic diagram showing BMMSCs and HCC1806 coculture. BMMSCs are green due to GFP expression and HCC1806 are red due to RFP expression. DISCUSSION (E) In vivo tumor volume in animals co-injected with DP-HCC1806:BMMSCs or DP- HCC1806:BMMSCs-siCD9 cells, evaluated at week 4 post-injection. (F) MTT assay for BMMSC, HCC1806, DP-HCC1806:BMMSCs, and DP-HCC1806:BMMSCs-siCD9 cultures. (G) Serum levels of various cytokines, and inflammatory factors from mice injected with HCC1806, BMMSCs, and DP-HCC1806:BMMSC xenografts. (H) In vitro cell death assessed qualitatively using trypan blue exclusion. (I) Cytokine profile for BMMSCs, HCC1806 cells, and DP-HCC1806:BMMSCs-siCD9 cells. (J, right) Cytokine ELISA assay and real- time RT-PCR analyses of CXCL12 protein and mRNA levels in excised tumor tissue. Mean ± s.e.m, n = 6. Significant difference indicated by p < 0.05, p < 0.01, *p < 0.001 for comparison between HCC1806 and DP-HCC1806:BMMSCs. #p < 0.05, ##p < 0.01, ###p < 0.001 for comparison between BMMSCs and DP-HCC1806:BMMSCs. Scale bar 100 µm. www.oncotarget.com Oncotarget 3442 Alternatively, there is also growing evidence to suggest that the chemoresistance of DP- HCC1806:BMMSCs may be due to exosomes given the increased expression of both CD9 and CD81 (Figure 5D), which are two tetraspanins that also happen to be enriched in exosome membranes and serve as exosomal biomarkers [46]. Furthermore, CD9 overexpression been implicated in the increased secretion of extracellular vesicles (EVs), including exosomes [32, 33], and there are studies which have documented acquired chemoresistance of cells through exosome-mediated mechanisms [38, 47]. Koch et al. discovered that (i) chemoresistance occurred when B-cell lymphomas sequestered Dox within CD9-positive exosomes which were then exported out of the cell, and (ii) that inhibition of ABC/A3-supported exosome biosynthesis resulted in greater Dox retention within tumor cells [38]. Similarly, Ji et al. reported that MSC- derived exosomes conferred drug resistance to 5FU in gastric cancer cells by activating a calcium/calmodulin- dependent protein kinase Raf/MEK/ERK pathway [47]. Based on this literature, it is plausible that the CD9- mediated chemoresistance of DP-HCC1806:BMMSCs to both Dox and 5FU may be due to an increase in BMMSCs- exosome-associated signaling. Hence, future studies will aim to define the exact cellular localization of CD9 within DP-HCC1806:BMMSCs cells. regenerative medicine, activation of BMMSCs increases their CD9 expression which, in turn, has been shown to stimulate their proliferation and regenerative potential [35]. In oncology, CD9 overexpression in tumors has been associated increased risk of invasion and development of metastasis, especially when it forms a complex with its molecular partner CD81, where it then facilitates long-term tumor growth [30, 34]. HCC1806 cells A triple negative human breast cancer cell line (HCC1806) was purchased from ATCC (Manassas, VA, USA; CRL2335) and cultured in an α-minimum essential medium (α-MEM; Gibco), containing 10% FBS, 2 mM glutamine, 1 mM sodium pyruvate, and 1 mM nonessential amino acids at 37° C in 5% CO2. Cells were passaged every 3 days by incubating for 5 minutes at 37° C first in 0.5 mM EDTA dissolved in PBS followed by 0.5% of trypsin. All other cancer cells were expanded in T-175 culture flasks with filter tops (Corning) using cancer growth medium consisting of αMEM, 10% FBS, 100 units/ml penicillin, and 100 µg/ml streptomycin. The medium was changed every 2–3 days. For all experiments, cancer cells were used when they reached 70–80% confluence. In animals injected with DP-HCC1806:BMMSCs, we also detected a several fold increase in the serum levels of IL6, CCL5, and CCR5. Interestingly, IL6 is a cytokine that has been shown to be secreted from both tumors and BMMSCs [52, 53] in response to β1 integrin adhesion and has been implicated in maintaining CAM-DR via JAK/ STAT3 signaling [54–58]. Likewise, the inflammatory chemokine CCL5, interacting primarily with G protein- coupled receptor CCR5, is highly expressed in several cancers, including breast cancer, and has been shown to play an important pro-oncogenic role via immune cell recruitment, tumor growth, chemotaxis, and apoptosis [59–61]. In addition, CCL5 has been shown to increase αvβ3 integrin expression in cancer cells and increase NF-κB-mediated resistance to drug-induced apoptosis, thus facilitating enhanced integrin-mediated tumor invasion and an immunosuppressive, anti-apoptotic tumor microenvironment [60–63]. Taken together, in this study, the acquisition of 5FU and Dox chemoresistance in DP-HCC1806:BMMSCs is likely due to a CAM- DR mechanism, in which CD9, IL6, and CCL5/CCR5 collectively mediate tumor-growth, adhesion and anti- apoptotic signals, via integrin-dependent mechanisms. Interaction between BMMSCs and HCC1806 cells BMMSCs and HCC1806 cells were labeled with cell fluorescent green and red florescent red proteins, respectively, as previously described [65]. Phase contrast GFP and RFP images were acquired using a Nikon inverted microscope with an epifluorescence attachment. GFP-BMMSCs and RPF-HCC1806 cells were harvested using trypsin/EDTA and collected by centrifugation at 400 × g for 7 min. The cells were then co-cultured together at 37° C for 3 days in a humidified atmosphere with 5% CO2. Unless otherwise indicated, co-cultures of GFP- BMMSCs and RFP-HCC1806 cells were prepared by mixing cell suspensions at a 1:1 ratio. The co-cultures were then expanded by plating at 1000 cells/cm2 in complete culture medium, except that FBS was reduced to 10%. In some experiments, co-cultures were initiated in the presence of the CD9 siRNA. Hence, to fully adopt BMMSCs for clinical applications it will be important for future studies to further define the cellular roles through which BMMSCs interact within a tumor microenvironment. DISCUSSION In vivo co-culture of HCC1806 and BMMSCs cells produced a hybrid cell (DP-HCC1806:BMMSC) whose interaction is dependent on the CD9 expression in BMMSCs. DP-HCC1806:BMMSCs were then used to create xenograft tumors which were smaller but more chemoresistant to agents such as doxorubicin and 5FU. Next, CD9 was inhibited by siRNA in BMMSCs and then co-cultured with HCC1806 cells. These DP-HCC1806:BMMSCs-siCD9 cells were then injected into mice to create xenograft tumors which were now no longer resistant to chemotherapeutic agents. CD9 inhibition also reversed the expression of tumor resistant proteins such as BCRP, CCL5, CCR5, and CXCL12. www.oncotarget.com Oncotarget 3443 may be also contributing towards drug resistance. Although Western blot analyses revealed reduced mTOR protein levels in DP-HCC1806:BMMSCs, there was no significant differences in protein expression of pERK, pMAPK, PI3K, p53, and pAKT, suggesting that these oncogenic pathways (i.e. PI3K/AKT/mTOR and pMAPK/ pERK [48, 49]) are not directly involved in mediating the resistance of DP-HCC1806:BMMSCs to Dox and 5FU. However, serum cytokine analyses identified a several fold increase in CXCL12 levels in DP-HCC1806:BMMSCs, which was reduced upon CD9 knockdown, suggesting that CXCL12 expression is correlated with CD9 levels. Interestingly, studies have reported that CD9 and CXCL12 are associated in a CD9/CXCL12/CXCR4 signaling pathway, and activation of this pathway is linked to increased tumor invasion, metastasis, and chemoresistance [50, 51]. Furthermore, in both in vivo and in vitro models of colorectal cancer, Yu et al. reported that activation of CXCL12/CXCR4 conferred miR-125b-mediated resistance to 5FU, causing reduced chemotherapy-induced apoptosis and enhanced autophagy [51]. medium consisting of α-minimum essential medium (α-MEM; Gibco), 20% FBS (Atlanta Biologicals), 100 U/ mL penicillin (Gibco), 100 µg/mL streptomycin (Gibco), and 2 mM l-glutamine (Gibco) on a 152-cm2 culture dish (Corning). After 24 h, cells were washed with PBS, and the adherent viable cells were harvested by using 0.25% trypsin and 1 mM EDTA (Gibco) for 5 min at 37° C. The harvested cells were plated at 1000 cells/cm2 in culture dishes and expanded for 7 days until 70–80% confluence. The culture medium was changed every 2–3 days and were cells were kept in incubators at 37° C with 5% CO2. Preparation of single cells from in vivo tumors To collect single cells, tumors were harvested from the mammary fat pad, mechanically processed into small pieces, washed with PBS buffer and filtered using a cellular strainer to remove undigested cellular derbies. The filtered cells were then transferred to 15 ml conical Western immunoblotting Tissues or cells were lysed in a lysis buffer containing 25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 1% sodium deoxycholate, and 0.1% SDS supplemented with protease and phosphatase inhibitor tablets (Roche). Lysates were clarified by centrifugation at 10,000 × g for 10 min and the supernatant containing equal amounts of protein were separated by SDS-PAGE, transferred to PVDF membranes, and probed with primary antibodies followed by peroxidase-conjugated secondary antibodies. Protein levels in the samples were determined by BCA assay kit (Thermo Fisher Scientific, Inc). Approximately 20 µg protein was mixed with SDS buffer (Life Technologies) containing mercaptoethanol (Sigma) and heated at 95° C for 5 minutes. Denatured proteins were separated by electrophoresis on polyacrylamide gels (Biorad Gels; Biorad) and transferred to nitrocellulose membranes using the Biorad turbo transfer System (Biorad). Membranes were blocked for 1 h at room temperature with 5% BSA (Theromofisher) in PBS containing 0.1% Tween-20 (PBST, Biorad). After blocking, membranes were incubated overnight at 4° C with primary antibodies diluted 1:1000 in blocking buffer. Membranes were washed 3 times in PBST and incubated with HRP-conjugated secondary antibodies (Cell Signaling Technologies) diluted 1:2000 in PBST for 2 hours at room temperature. Membranes were developed RNA interference (siRNA) CD9 double-stranded synthetic 21-mer RNA oligonucleotides were used at a final concentration of 200 nM with the Lipofectamine transfection reagent. To knock down CD9 expression in BMMSCs, 200 nM CD9-specific small interfering RNA (siRNA) (5′-GACGUACUCGAAACCUUCA-3′) was transfected. Scrambled siRNA was used as a negative control. 3 days after transfection, cells were analyzed for knockdown efficiency by western blot under non-reducing condition. CD9 siRNA oligomer duplexes targeting CD9 were prepared according to the manufacturer’s instructions. For BMMSCs-siRNA transfection, the cells were transfected with an siRNA mixture against human CD9 or control random siRNAs (life technologies) using Lipofectamine RNAiMAX (Invitrogen). The cells were cultured for 3 days, and the gene-silencing effect of the siRNAs was assessed by immunoblotting with anti-CD9 monoclonal antibody (Santa Cruz). g y y In some experiments, cells obtained from mice after 4 days and small sections of the tissue were suspended in PBS containing 2%FBS and 1mM EDTA at ~5,000 c/µL and incubated with antibodies (BD Biosciences) for 25 min on ice. Samples were then washed twice in PBS/2%FBS/1 mM EDTA and suspended at a concentration of 2 million cells per mL for FACS. The viable RFP-HCC1806 cell population, GFP-BMMSC cell population, and double positive HCC1806/BMMSCs cell population (i.e. HCC1806 cells with BMMSCs fragments, DP-HCC1806:BMMSCs) were then gated and sorted using a cell sorter (Beckman Coulter). The cells collected were centrifuged at 400 × g for 7 min and washed in PBS. In some experiments, combinations of fluorochrome- conjugated monoclonal antibodies or their respective isotype controls were added to the cell suspension at concentrations recommended by the manufacturer (BD Biosciences) and incubated at 4° C in the dark for 30 min. The labeled cells were washed in PBS and then analyzed using a FACS (BD Biosciences). Gating was set to relevant isotype controls (GFP-FL1 and RFP-FL4) and labeled cells for each cell line. A fraction of the sorted cells was also analyzed on the flow cytometer to ensure viability, complete elimination of the unlabeled cells, and for verification of bright BMMSCs or HCC1806 cells populations. The purity of sorted cells was more than 95% by additional flow cytometric analysis. Sorted populations were injected into mice for tumor evaluation or plated for proteomic assay or processed for genomic assays. Human BMMSCs A frozen vial of 1 × 106 passage-2 human BMMSCs was thawed at 37° C and cultured as previously described [64]. Briefly, BMMSCs were plated in complete culture www.oncotarget.com Oncotarget 3444 followed by 40 cycles of 95° C for 1 s and 60° C for 20 s. Data was analyzed with Sequence Detection Software V2.3 (Applied Biosystems) and relative quantities (RQs) were calculated with comparative CT method using RQ Manager V1.2 (Applied Biosystems). If no amplification occurred, a CT value of ~ 40 was used in calculating the RQs. tubes (Falcon), washed with PBS, and centrifuged at 400 × g for 7 min. To obtain a single-cell suspension, tumors were incubated with trypsin/EDTA at 37° C for 10 min followed by 5 min with collagenase. Every 2 minutes, cells were mechanically disrupted by pipetting 5–10 times. When most aggregates were no longer visible, cells were collected by centrifugation at 400 × g for 7 min. Subsequently, cells were passed through a 40–70 µm cell strainer (Falcon) to remove any remaining cell clusters before staining for flow cytometry. Breast cancer xenograft model Female NOD/SCID mice were supplied by Jackson Laboratory (Bar Harbor, ME) and used under a protocol approved by the Institutional Animal Care and Use Committee. RFP-HCC1806 cells (1 × 106 in 100 µL HBSS) were injected into the left mammary fat pad of mice at 12 weeks of age. The RFP-HCC1806 cells were obtained from standard monolayer cultures (experiment 1), from pre-incubated in mice for 4 days cultures (experiment 2), and from BMMSC-CD9 knockout co- cultured BMMSCs:HCC1806 (experiment 6). In addition, animals were injected with 1 × 106 BMMSCs or HCC1806 as a control group. Mice were observed weekly for 8 weeks and were sacrificed after 56 days of tumor cell inoculation. Tumorigenesis was determined via palpation during animal observation and was confirmed by visual assessment of the tumors upon excision. On day 56, animals were euthanized by intraperitoneal injection of ketamine/xylazine. Tumors were excised, photographed, and weighed. HCC1806 growth assays Cell growth was determined on RFP-HCC1806 cells, GFP-BMMSCs, and DP-HCC1806:BMMSCs. Cells were seeded into 6-well plates (Corning) at 25,000 cells per well and cultured up to 7 days. On day 7, images were acquired and the cells harvested with trypsin/EDTA for cell counts. For MTT (3-(4, 5-dimethylthiazolyl- 2)-2,5-diphenyltetrazolium bromide) (ThermoFisher Scientific), RFP-HCC1806 cells, GFP-BMMSCs, and ELISA Blood serum was collected from mice after injection RFP-HCC1806 cells, GFP-BMMSCs, DP- HCC1806:MSCs, and DP-HCC1806:MSCs-siCD9. Similarly, the supernatant was collected from co-cultures, with or without CD9 knockdown in BMMSCs cells, and clarified by centrifugation, first at 500xg for 5 minutes then at 10,000 × g for 10 minutes. The medium was then aliquoted and processed using an ELISA. Levels of inflammatory factors (CXCL12, CCL5) were determined from the serum using commercially available ELISA kits (R&D systems). Prior to use, a frozen aliquot of conditioned medium was thawed on ice and appropriately diluted with buffers recommended by the manufacturer. Optical density (OD) was measured on a plate reader (FLUOstar Omega; BMG Labtech) at an absorbance of 450 nm. Protein concentration was determined after correcting for optical imperfections in the plate by subtracting OD values at 540/570 nm from those obtained at 450 nm. HCC1806 growth and drug treatment and preparation The cell lines HCC1806 were cultured in cultured medium as described above, containing 10% fetal bovine serum (FBS), 1.5 g/L sodium bicarbonate, and 1 mM sodium pyruvate. These cells were maintained in a humidified atmosphere with 5% CO2 at 37° C. Mithramycin A was purchased from Sigma-Aldrich (St. Louis, USA) and reconstituted in DMSO to a final concentration of 100 mM while Doxycycline was purchased from Sigma Aldrich at a stock concentration of 100 mg/mL in DMSO and stored at −20° C. 5-Fluorouracil (Sigma-Aldrich, USA) stock solution was prepared at concentration 50 mg/mL in DMSO, and then diluted 5 times in PBS and stored at −4° C. For CD9 knockout experiment 12 6-week-old female NOD/SCID mice were purchased from Jackson Laboraties USA). 1 × 106 DP-HCC1806:BMMSCs-siCD9 cells were implanted into mammary fat pads of the mice. Tumor sizes were measured using Vernier calipers once tumors became palpable. Tumor volumes were calculated using the following equation: tumor volume (cm3) = л(length × width2)/6. Tumor size was monitored weekly. All mice were sacrificed and tumors were collected for analysis. Real-Time RT-PCR Isolation of total RNA was performed using the RNeasy mini kit (Qiagen). Total RNA concentration was measured using a NanoDrop ND1000 spectrophotometer (BioRad). RNA (1 μg) was reverse transcribed using oligo dT and transcript or first strand cDNA synthesis kit (Roche). RT-qPCR reactions were carried out in a 20- μL reaction volume containing 3 nM of each primer and SYBR Green PCR Mastermix (Applied Biosystems). Real-time RT-PCR was performed for CXCL12, TWIST1, CD9 and GAPDH using Taqman® Gene Expression Assays (Applied Biosystems). A total of 20–40 ng of cDNA was used for each 20 μl reaction. Thermal cycling was performed with a 7900HT System (Applied Biosystems) by incubating the reactions at 95° C for 20 s www.oncotarget.com www.oncotarget.com Oncotarget 3445 DP-HCC1806:BMMSCs were seeded into 96-well plates (Corning) at 2,000 cells per well. After 7 days, 20 µL of the MTT solution was added to each well. Plates were incubated at 37° C for 2 h in a humidified atmosphere and 5% CO2. Absorbance was recorded at 490 nm using a plate reader. The number of viable cells was determined using a Cell Counting Kit (BioRad). Cells (3.0 × 104), transfected with siRNA against human CD9 or control RNAs, were seeded on 96-well plates and incubated overnight. After further incubation for 3 days, the kit reagent was added to the medium, and the cells were incubated for 1 h. The absorbance of samples (450 nm) was determined using a scanning multi-well spectrophotometer. For the apoptosis assay, 3.0 × 104cells were seeded on 96-well plates and incubated overnight. After further incubation for 3 days, quantitative viability/ cell death in cultured cells were measured using an MTT assay (Roche). in a 100 mM Tris base solution (pH 8.2) containing hydrogen peroxide, para-coumaric acid, and luminol (all from Biorad). Images were captured on a VersaDocTM MP4000 Molecular Imager (Biorad Laboratories). Primary antibodies used included CD9, BCRT, MDR1, CD81, mTOR, ERK, pAKT, pMAPK, p13AKT, and GAPDH (all from Cell Signaling Technologies). org/10.1161/JAHA.115.002856. [PubMed] 12. Tao SC, Guo SC, Li M, Ke QF, Guo YP, Zhang CQ. Chitosan Wound Dressings Incorporating Exosomes Derived from MicroRNA-126-Overexpressing Synovium Mesenchymal Stem Cells Provide Sustained Release of Exosomes and Heal Full-Thickness Skin Defects in a Diabetic Rat Model. Stem Cells Transl Med. 2017; 6:736– 47. https://doi.org/10.5966/sctm.2016-0275. [PubMed] 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015; 65:87–108. https://doi.org/10.3322/caac.21262. [PubMed] 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin. 2015; 65:87–108. https://doi.org/10.3322/caac.21262. [PubMed] 2. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 2018; 68:7–30. https://doi.org/10.3322/ caac.21442. [PubMed] 3. Verma R, Bowen RL, Slater SE, Mihaimeed F, Jones JL. Pathological and epidemiological factors associated with advanced stage at diagnosis of breast cancer. Br Med Bull. 2012; 103:129–45. https://doi.org/10.1093/bmb/lds018. [PubMed] 3. Verma R, Bowen RL, Slater SE, Mihaimeed F, Jones JL. Pathological and epidemiological factors associated with advanced stage at diagnosis of breast cancer. Br Med Bull. 2012; 103:129–45. https://doi.org/10.1093/bmb/lds018. [PubMed] 13. Dwyer RM, Potter-Beirne SM, Harrington KA, Lowery AJ, Hennessy E, Murphy JM, Barry FP, O’Brien T, Kerin MJ. Monocyte chemotactic protein-1 secreted by primary breast tumors stimulates migration of mesenchymal stem cells. Clin Cancer Res. 2007; 13:5020–27. https://doi. org/10.1158/1078-0432.CCR-07-0731. [PubMed] 4. Nounou MI, ElAmrawy F, Ahmed N, Abdelraouf K, Goda S, Syed-Sha-Qhattal H. Breast Cancer: Conventional Diagnosis and Treatment Modalities and Recent Patents and Technologies. Breast Cancer (Auckl). 2015 (Suppl 2); 9:17– 34. https://doi.org/10.4137/BCBCR.S29420. [PubMed] 14. Rattigan Y, Hsu JM, Mishra PJ, Glod J, Banerjee D. Interleukin 6 mediated recruitment of mesenchymal stem cells to the hypoxic tumor milieu. Exp Cell Res. 2010; 316:3417–24. https://doi.org/10.1016/j.yexcr.2010.07.002. [PubMed] 5. Ridge SM, Sullivan FJ, Glynn SA. Mesenchymal stem cells: key players in cancer progression. Mol Cancer. 2017; 16:31. https://doi.org/10.1186/s12943-017-0597-8. [PubMed] 15. Bhoopathi P, Chetty C, Gogineni VR, Gujrati M, Dinh DH, Rao JS, Lakka SS. MMP-2 mediates mesenchymal stem cell tropism towards medulloblastoma tumors. Gene Ther. 2011; 18:692–701. https://doi.org/10.1038/gt.2011.14. [PubMed] 6. Ullah M, Stich S, Notter M, Eucker J, Sittinger M, Ringe J. Transdifferentiation of mesenchymal stem cells- derived adipogenic-differentiated cells into osteogenic- or chondrogenic-differentiated cells proceeds via dedifferentiation and have a correlation with cell cycle arresting and driving genes. Differentiation. 2013; 85:78– 90. https://doi.org/10.1016/j.diff.2013.02.001. [PubMed] 16. Author contributions MU: designed the study, performed the experiments, analyzed the data and prepared the manuscript. AAK: designed the study, assisted in animal experiments, www.oncotarget.com Oncotarget 3446 10. Qin Y, Wang L, Gao Z, Chen G, Zhang C. Bone marrow stromal/stem cell-derived extracellular vesicles regulate osteoblast activity and differentiation in vitro and promote bone regeneration in vivo. Sci Rep. 2016; 6:21961. https:// doi.org/10.1038/srep21961. [PubMed] Western blots and real-time RT-PCR. NNN: helped with data analysis and manuscript preparation; WC: helped with data analysis and manuscript preparation; AST: Directed the research and prepared the manuscript. CONFLICTS OF INTEREST 11. Zhang Z, Yang J, Yan W, Li Y, Shen Z, Asahara T. Pretreatment of Cardiac Stem Cells With Exosomes Derived From Mesenchymal Stem Cells Enhances Myocardial Repair. J Am Heart Assoc. 2016; 5:e002856. https://doi. org/10.1161/JAHA.115.002856. [PubMed] The authors have no competing financial interests. org/10.1161/JAHA.115.002856. [PubMed] 2010; 89:843–52. https://doi.org/10.1016/j. ejcb.2010.07.004. [PubMed] 21. Khakoo AY, Pati S, Anderson SA, Reid W, Elshal MF, Rovira II, Nguyen AT, Malide D, Combs CA, Hall G, Zhang J, Raffeld M, Rogers TB, et al. Human mesenchymal stem cells exert potent antitumorigenic effects in a model of Kaposi’s sarcoma. J Exp Med. 2006; 203:1235–47. https:// doi.org/10.1084/jem.20051921. [PubMed] 32. Schiller LT, Lemus-Diaz N, Rinaldi Ferreira R, Böker KO, Gruber J. Enhanced Production of Exosome-Associated AAV by Overexpression of the Tetraspanin CD9. Mol Ther Methods Clin Dev. 2018; 9:278–87. https://doi. org/10.1016/j.omtm.2018.03.008. [PubMed] 22. Qiao L, Xu ZL, Zhao TJ, Ye LH, Zhang XD. Dkk-1 secreted by mesenchymal stem cells inhibits growth of breast cancer cells via depression of Wnt signalling. Cancer Lett. 2008; 269:67–77. https://doi.org/10.1016/j.canlet.2008.04.032. [PubMed] 33. Böker KO, Lemus-Diaz N, Rinaldi Ferreira R, Schiller L, Schneider S, Gruber J. The Impact of the CD9 Tetraspanin on Lentivirus Infectivity and Exosome Secretion. Mol Ther. 2018; 26:634–47. https://doi.org/10.1016/j. ymthe.2017.11.008. [PubMed] 23. Ohlsson LB, Varas L, Kjellman C, Edvardsen K, Lindvall M. Mesenchymal progenitor cell-mediated inhibition of tumor growth in vivo and in vitro in gelatin matrix. Exp Mol Pathol. 2003; 75:248–55. https://doi.org/10.1016/j. yexmp.2003.06.001. [PubMed] 34. Rappa G, Green TM, Lorico A. The nuclear pool of tetraspanin CD9 contributes to mitotic processes in human breast carcinoma. Mol Cancer Res. 2014; 12:1840–50. https://doi.org/10.1158/1541-7786.MCR-14-0242. [PubMed] 24. Toupet K, Maumus M, Peyrafitte JA, Bourin P, van Lent PL, Ferreira R, Orsetti B, Pirot N, Casteilla L, Jorgensen C, Noël D. Long-term detection of human adipose-derived mesenchymal stem cells after intraarticular injection in SCID mice. Arthritis Rheum. 2013; 65:1786–94. https:// doi.org/10.1002/art.37960. [PubMed] 35. Kim YJ, Yu JM, Joo HJ, Kim HK, Cho HH, Bae YC, Jung JS. Role of CD9 in proliferation and proangiogenic action of human adipose-derived mesenchymal stem cells. Pflugers Arch. 2007; 455:283–96. https://doi.org/10.1007/ s00424-007-0285-4. [PubMed] 25. Ra JC, Shin IS, Kim SH, Kang SK, Kang BC, Lee HY, Kim YJ, Jo JY, Yoon EJ, Choi HJ, Kwon E. Safety of intravenous infusion of human adipose tissue-derived mesenchymal stem cells in animals and humans. Stem Cells Dev. 2011; 20:1297– 308. https://doi.org/10.1089/scd.2010.0466. [PubMed] 36. Kohmo S, Kijima T, Otani Y, Mori M, Minami T, Takahashi R, Nagatomo I, Takeda Y, Kida H, Goya S, Yoshida M, Kumagai T, Tachibana I, et al. Cell surface tetraspanin CD9 mediates chemoresistance in small cell lung cancer. Cancer Res. 2010; 70:8025–35. https://doi.org/10.1158/0008-5472. CAN-10-0996. [PubMed] 26. org/10.1161/JAHA.115.002856. [PubMed] Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW, Richardson AL, Polyak K, Tubo R, Weinberg RA. Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature. 2007; 449:557–63. https:// doi.org/10.1038/nature06188. [PubMed] 7. Ullah M, Eucker J, Sittinger M, Ringe J. Mesenchymal stem cells and their chondrogenic differentiated and dedifferentiated progeny express chemokine receptor CCR9 and chemotactically migrate toward CCL25 or serum. Stem Cell Res Ther. 2013; 4:99. https://doi.org/10.1186/scrt310. [PubMed] 17. Suzuki K, Sun R, Origuchi M, Kanehira M, Takahata T, Itoh J, Umezawa A, Kijima H, Fukuda S, Saijo Y. Mesenchymal stromal cells promote tumor growth through the enhancement of neovascularization. Mol Med. 2011; 17:579–87. https://doi.org/10.2119/molmed.2010.00157. [PubMed] 8. Shabbir A, Cox A, Rodriguez-Menocal L, Salgado M, Van Badiavas E. Mesenchymal Stem Cell Exosomes Induce Proliferation and Migration of Normal and Chronic Wound Fibroblasts, and Enhance Angiogenesis In Vitro. Stem Cells Dev. 2015; 24:1635–47. https://doi.org/10.1089/ scd.2014.0316. [PubMed] 18. Corcoran KE, Trzaska KA, Fernandes H, Bryan M, Taborga M, Srinivas V, Packman K, Patel PS, Rameshwar P. Mesenchymal stem cells in early entry of breast cancer into bone marrow. PLoS One. 2008; 3:e2563. https://doi. org/10.1371/journal.pone.0002563. [PubMed] 9. Spees JL, Olson SD, Ylostalo J, Lynch PJ, Smith J, Perry A, Peister A, Wang MY, Prockop DJ. Differentiation, cell fusion, and nuclear fusion during ex vivo repair of epithelium by human adult stem cells from bone marrow stroma. Proc Natl Acad Sci U S A. 2003; 100:2397–402. https://doi.org/10.1073/pnas.0437997100. [PubMed] 19. Djouad F, Plence P, Bony C, Tropel P, Apparailly F, Sany J, Noël D, Jorgensen C. Immunosuppressive effect of mesenchymal stem cells favors tumor growth in allogeneic animals. Blood. 2003; 102:3837–44. https://doi. org/10.1182/blood-2003-04-1193. [PubMed] www.oncotarget.com Oncotarget 3447 20. Cousin B, Ravet E, Poglio S, De Toni F, Bertuzzi M, Lulka H, Touil I, André M, Grolleau JL, Péron JM, Chavoin JP, Bourin P, Pénicaud L, et al. Adult stromal cells derived from human adipose tissue provoke pancreatic cancer cell death both in vitro and in vivo. PLoS One. 2009; 4:e6278. https:// doi.org/10.1371/journal.pone.0006278. [PubMed] 30. Gustafson-Wagner E, Stipp CS. The CD9/CD81 tetraspanin complex and tetraspanin CD151 regulate α3β1 integrin- dependent tumor cell behaviors by overlapping but distinct mechanisms. PLoS One. 2013; 8:e61834. https://doi. org/10.1371/journal.pone.0061834. [PubMed] 31. Castro-Sanchez L, Soto-Guzman A, Navarro-Tito N, Martinez-Orozco R, Salazar EP. Native type IV collagen induces cell migration through a CD9 and DDR1- dependent pathway in MDA-MB-231 breast cancer cells. Eur J Cell Biol. org/10.1161/JAHA.115.002856. [PubMed] Eterno V, Zambelli A, Pavesi L, Villani L, Zanini V, Petrolo G, Manera S, Tuscano A, Amato A. Adipose-derived Mesenchymal Stem Cells (ASCs) may favour breast cancer recurrence via HGF/c-Met signaling. Oncotarget. 2014; 5:613–33. https://doi.org/10.18632/oncotarget.1359. [PubMed] 37. Rappa G, Green TM, Karbanová J, Corbeil D, Lorico A. Tetraspanin CD9 determines invasiveness and tumorigenicity of human breast cancer cells. Oncotarget. 2015; 6:7970–91. https://doi.org/10.18632/oncotarget.3419. [PubMed] 27. Park SJ, Kim HJ, Kim W, Kim OS, Lee S, Han SY, Jeong EJ, Park HS, Kim HW, Moon KS. Tumorigenicity Evaluation of Umbilical Cord Blood-derived Mesenchymal Stem Cells. Toxicol Res. 2016; 32:251–58. https://doi. org/10.5487/TR.2016.32.3.251. [PubMed] 38. Koch R, Aung T, Vogel D, Chapuy B, Wenzel D, Becker S, Sinzig U, Venkataramani V, von Mach T, Jacob R, Truemper L, Wulf GG. Nuclear Trapping through Inhibition of Exosomal Export by Indomethacin Increases Cytostatic Efficacy of Doxorubicin and Pixantrone. Clin Cancer Res. 2016; 22:395–404. https://doi.org/10.1158/1078-0432.CCR- 15-0577. [PubMed] 28. Kucerova L, Matuskova M, Hlubinova K, Altanerova V, Altaner C. Tumor cell behaviour modulation by mesenchymal stromal cells. Mol Cancer. 2010; 9:129. https://doi.org/10.1186/1476-4598-9-129. [PubMed] 29. Maffey A, Storini C, Diceglie C, Martelli C, Sironi L, Calzarossa C, Tonna N, Lovchik R, Delamarche E, Ottobrini L, Bianco F. Mesenchymal stem cells from tumor microenvironment favour breast cancer stem cell proliferation, cancerogenic and metastatic potential, via ionotropic purinergic signalling. Sci Rep. 2017; 7:13162. https://doi.org/10.1038/s41598-017-13460-7. [PubMed] 39. Sethi T, Rintoul RC, Moore SM, MacKinnon AC, Salter D, Choo C, Chilvers ER, Dransfield I, Donnelly SC, Strieter R, Haslett C. Extracellular matrix proteins protect small cell lung cancer cells against apoptosis: a mechanism for small cell lung cancer growth and drug resistance in vivo. www.oncotarget.com Oncotarget 3448 Nat Med. 1999; 5:662–68. https://doi.org/10.1038/9511. [PubMed] 50. Iachelli V, Andriano N, Bonaccorso P, La Rosa M, Cannata E, Samperi P, La Spina M, Di Cataldo A, Russo G, Lo Nigro L. Characterization of CD9-CXCL12-CXCR4 Expression in Biological Subtypes of Childhood ACUTE Lymphoblastic Leukemia: Preliminary Findings and Future Perspectives. Blood. 2016; 128:5278. 40. Tornin J, Martinez-Cruzado L, Santos L, Rodriguez A, Núñez LE, Oro P, Hermosilla MA, Allonca E, Fernández- García MT, Astudillo A, Suarez C, Morís F, Rodriguez R. Inhibition of SP1 by the mithramycin analog EC-8042 efficiently targets tumor initiating cells in sarcoma. Oncotarget. 2016; 7:30935–50. https://doi.org/10.18632/ oncotarget.8817. [PubMed] 51. Yu X, Shi W, Zhang Y, Wang X, Sun S, Song Z, Liu M, Zeng Q, Cui S, Qu X. org/10.1161/JAHA.115.002856. [PubMed] CXCL12/CXCR4 axis induced miR-125b promotes invasion and confers 5-fluorouracil resistance through enhancing autophagy in colorectal cancer. Sci Rep. 2017; 7:42226. https://doi.org/10.1038/ srep42226. [PubMed] 41. Holohan C, Van Schaeybroeck S, Longley DB, Johnston PG. Cancer drug resistance: an evolving paradigm. Nat Rev Cancer. 2013; 13:714–26. https://doi.org/10.1038/nrc3599. [PubMed] 52. Damiano JS, Cress AE, Hazlehurst LA, Shtil AA, Dalton WS. Cell adhesion mediated drug resistance (CAM-DR): role of integrins and resistance to apoptosis in human myeloma cell lines. Blood. 1999; 93:1658–67. [PubMed] 42. Choi ES, Nam JS, Jung JY, Cho NP, Cho SD. Modulation of specificity protein 1 by mithramycin A as a novel therapeutic strategy for cervical cancer. Sci Rep. 2014; 4:7162. https://doi.org/10.1038/srep07162. [PubMed] 53. Mi F, Gong L. Secretion of interleukin-6 by bone marrow mesenchymal stem cells promotes metastasis in hepatocellular carcinoma. Biosci Rep. 2017; 37:37. https:// doi.org/10.1042/BSR20170181. [PubMed] 43. Saha S, Mukherjee S, Mazumdar M, Manna A, Khan P, Adhikary A, Kajal K, Jana D, Sa G, Mukherjee S, Sarkar DK, Das T. Mithramycin A sensitizes therapy-resistant breast cancer stem cells toward genotoxic drug doxorubicin. Transl Res. 2015; 165:558–77. https://doi.org/10.1016/j. trsl.2014.10.011. [PubMed] 54. Uchiyama H, Barut BA, Mohrbacher AF, Chauhan D, Anderson KC. Adhesion of human myeloma-derived cell lines to bone marrow stromal cells stimulates interleukin-6 secretion. Blood. 1993; 82:3712–20. [PubMed] 44. Tagashira M, Kitagawa T, Isonishi S, Okamoto A, Ochiai K, Ohtake Y. Mithramycin represses MDR1 gene expression in vitro, modulating multidrug resistance. Biol Pharm Bull. 2000; 23:926–29. https://doi.org/10.1248/bpb.23.926. [PubMed] 55. Shain KH, Yarde DN, Meads MB, Huang M, Jove R, Hazlehurst LA, Dalton WS. Beta1 integrin adhesion enhances IL-6-mediated STAT3 signaling in myeloma cells: implications for microenvironment influence on tumor survival and proliferation. Cancer Res. 2009; 69:1009–15. https://doi.org/10.1158/0008-5472.CAN-08-2419. [PubMed] 45. Zhang M, Mathur A, Zhang Y, Xi S, Atay S, Hong JA, Datrice N, Upham T, Kemp CD, Ripley RT, Wiegand G, Avital I, Fetsch P, et al. Mithramycin represses basal and cigarette smoke-induced expression of ABCG2 and inhibits stem cell signaling in lung and esophageal cancer cells. Cancer Res. 2012; 72:4178–92. https://doi. org/10.1158/0008-5472.CAN-11-3983. [PubMed] 56. Hu X, Xuan H, Du H, Jiang H, Huang J. Down-regulation of CD9 by methylation decreased bortezomib sensitivity in multiple myeloma. PLoS One. 2014; 9:e95765. https://doi. org/10.1371/journal.pone.0095765. [PubMed] 57. Furukawa Y, Kikuchi J. Epigenetic mechanisms of cell adhesion-mediated drug resistance in multiple myeloma. Int J Hematol. 2016; 104:281–92. https://doi.org/10.1007/ s12185-016-2048-5. [PubMed] 46. Andreu Z, Yanez-Mo M. org/10.1161/JAHA.115.002856. [PubMed] Tetraspanins in extracellular vesicle formation and function. Front Immunol. 2014; 5:442. https://doi.org/10.3389/fimmu.2014.00442. [PubMed] 58. Johnson DE, O’Keefe RA, Grandis JR. Targeting the IL-6/ JAK/STAT3 signalling axis in cancer. Nat Rev Clin Oncol. 2018; 15:234–48. https://doi.org/10.1038/nrclinonc.2018.8. [PubMed] 47. Ji R, Zhang B, Zhang X, Xue J, Yuan X, Yan Y, Wang M, Zhu W, Qian H, Xu W. Exosomes derived from human mesenchymal stem cells confer drug resistance in gastric cancer. Cell Cycle. 2015; 14:2473–83. https://doi.org/10.10 80/15384101.2015.1005530. [PubMed] 59. Luboshits G, Shina S, Kaplan O, Engelberg S, Nass D, Lifshitz-Mercer B, Chaitchik S, Keydar I, Ben-Baruch A. Elevated expression of the CC chemokine regulated on activation, normal T cell expressed and secreted (RANTES) in advanced breast carcinoma. Cancer Res. 1999; 59:4681– 87. [PubMed] 48. Saini KS, Loi S, de Azambuja E, Metzger-Filho O, Saini ML, Ignatiadis M, Dancey JE, Piccart-Gebhart MJ. Targeting the PI3K/AKT/mTOR and Raf/MEK/ ERK pathways in the treatment of breast cancer. Cancer Treat Rev. 2013; 39:935–46. https://doi.org/10.1016/j. ctrv.2013.03.009. [PubMed] 60. Aldinucci D, Casagrande N. Inhibition of the CCL5/CCR5 Axis against the Progression of Gastric Cancer. Int J Mol Sci. 2018; 19:E1477. https://doi.org/10.3390/ijms19051477. [PubMed] 49. Du J, Tong A, Wang F, Cui Y, Li C, Zhang Y, Yan Z. The Roles of PI3K/AKT/mTOR and MAPK/ERK Signaling Pathways in Human Pheochromocytomas. Int J Endocrinol. 2016; 2016:5286972. https://doi. org/10.1155/2016/5286972. [PubMed] 61. Aldinucci D, Colombatti A. The inflammatory chemokine CCL5 and cancer progression. Mediators Inflamm. 2014; www.oncotarget.com Oncotarget 3449 2014:292376. https://doi.org/10.1155/2014/292376. [PubMed] 64. Bartosh TJ, Ullah M, Zeitouni S, Beaver J, Prockop DJ. Cancer cells enter dormancy after cannibalizing mesenchymal stem/stromal cells (MSCs). Proc Natl Acad Sci U S A. 2016; 113:E6447–56. https://doi.org/10.1073/ pnas.1612290113. [PubMed] 62. Lv D, Zhang Y, Kim HJ, Zhang L, Ma X. CCL5 as a potential immunotherapeutic target in triple-negative breast cancer. Cell Mol Immunol. 2013; 10:303–10. https://doi. org/10.1038/cmi.2012.69. [PubMed] 65. Ullah M, Kuroda Y, Bartosh TJ, Liu F, Zhao Q, Gregory C, Reger R, Xu J, Lee RH, Prockop DJ. iPS-derived MSCs from an expandable bank to deliver a prodrug-converting enzyme that limits growth and metastases of human breast cancers. Cell Death Discov. 2017; 3:16064. https://doi. org/10.1038/cddiscovery.2016.64. [PubMed] 63. Wang SW, Wu HH, Liu SC, Wang PC, Ou WC, Chou WY, Shen YS, Tang CH. CCL5 and CCR5 interaction promotes cell motility in human osteosarcoma. PLoS One. 2012; 7:e35101. https://doi.org/10.1371/journal.pone.0035101. [PubMed] www.oncotarget.com Oncotarget 3450
https://openalex.org/W3203343365	https://abdira.org/index.php/abdira/article/download/6/pdf	Indonesian	null	SOSIALISASI POLA HIDUP BERSIH DAN SEHAT UNTUK MENINGKATKAN KUALITAS KESEHATAN MASYARAKAT DI DESA RIDAN PERMAI	Jurnal Pengabdian Masyarakat	2,021	cc-by-sa	1,732	A A ABDIRA Volume 1 Nomor 1 Tahun 2021 Halaman 33–37 JURNAL PENGABDIAN MASYARAKAT Research & Learning in Faculty of Education ABDIRA Volume 1 Nomor 1 Tahun 2021 Halaman 33–37 JURNAL PENGABDIAN MASYARAKAT Research & Learning in Faculty of Education ABDIRA Volume 1 Nomor 1 Tahun 2021 Halaman 33–3 JURNAL PENGABDIAN MASYARAKAT Research & Learning in Faculty of Education SOSIALISASI POLA HIDUP BERSIH DAN SEHAT UNTUK MENINGKATKAN KUALITAS KESEHATAN MASYARAKAT DI DESA RIDAN PERMAI Dedi Ahmadi Program Studi Penjaskesrek Fakultas Ilmu Pendidikan Universitas Pahlawan Tuanku Tambusai e-mail: ammardzoky@gmail.com Abstrak Banyak dari masyarakat yang membuang sampah sembarangan, yang tidak mengubur kaleng-kaleng bekas, yang tidak menguras bak mandi mereka, yang tidak memperhatikan kebersihan kandang hewan peliharaan mereka. Kegiatan Pengabdian Kepada Masyarakat ini bertujuan untuk memberi sosialisasi pola hidup bersih demi kesejahteraan masyarakat dan menciptakan generasi masyarakat yang peduli akan kesehatan. Hal itu kami lakukan agar warga masyarakat lebih peduli terhadap pola hidupnya, dan agar mengurangi berbagai panyakit yang mungkin saja dapat ditimbulkan akibat pola hidup yang tidak sehat. pemberian pemahaman terhadap masyarakat tentang pentingnya pola hidup bersih dan sehat guna menjaga kesehatan masyarakat sangat diperlukan untuk memotivasi pola hidup sehat. Pemberian pemahaman ini diberikan dengan metode ceramah di kantor desa Ridan Permai; pola hidup bersih dan sehat yang harus diterapkan oleh masyarakat di Desa Ridan Permai adalah dengan menjaga kebersihan diri, lingkungan, dan berolah raga. Kata Kunci: sosialisasi hidup sehat, kualitas kesehatan masyarakat Keyword: socialization of healthy life, quality of public health Abstract Many people don't litter, who don't bury used cans, who don't drain their bathtubs, who don't pay attention to the cleanliness of their pet cages. This Community Service activity aims to provide socialization of a clean lifestyle for the welfare of the community and create a generation of people who care about health. We do this so that the community members are more concerned about their lifestyle, and to reduce various diseases that may be caused by an unhealthy lifestyle. Providing understanding to the community about the importance of a clean and healthy lifestyle in order to maintain public health is very necessary to motivate a healthy lifestyle. This understanding was given using the lecture method at the Ridan Permai village office; A clean and healthy lifestyle that must be applied by the community in Ridan Permai Village is to maintain personal hygiene, the environment, and exercise. Keyword: socialization of healthy life, quality of public health Jurnal Pengabdian Masyarakat (Abdira) Vol.1, No.1, 2021 METODE Metode pengembangan yang akan dilaksanakan merupakan sebuah rangkaiantahapan yang disusun secara sistematis, diantaranya : persiapan (membentuk kerja sama dengan aparat desa setempat, membentuk kerjasama dengan pengisi materi (dari Dinas Kesehatan/dokter)), dan menyiapkan materi. Pelaksanaan (sosialisasi kepada masyarakat Ridan Permai, Bangkinang, Kampar, tentang pohiber berupa penyuluhan mengenai penggunaan sarung tangan dan sepatu boot ketika membersihkan kandang, sosialisasi mengenai pentingnya menjaga kesehatan sebagai contohnya mencuci tangan menggunakan sabun, sosialisasi mengenai pentingnya menjaga kebersihan lingkungan dengan mengumpulkan sampah dan memilah sesuai jenisnya, sosialisasi mengenai pentingnya penerapan pola hidup bersih di masyarakat), evaluasi dan laporan akhir. PENDAHULUAN Banyak permasalahan yang timbul di lingkungan sekitar masyarakat sekarang ini, seperti contohnya kenakalan remaja, tingginya tingkat kriminalitas, dan juga pola hidup bersih dan sehat di lingkungan masyarakat. Banyaknya warga masyarakat yang kurang peduli terhadap kebersihan lingkungan di sekitar mereka dapat membahayakan bagi kesehatan lingkungan masyarakat sekitar. Pola hidup sehat masyarakat yang kurang terjaga bisa menimbulkan berbagai macam penyakit yang ada, seperti penyakit kulit, diare, gangguan pernapasan, bahkan yang sedang marak terjadi akhir-akhir ini adalah demam berdarah. Banyak sekali penyakit yang muncul akibat pola hidup masyarakat yang tidak bersih, serta tidak memperdulikan lingkungan tempat tinggal mereka. Munculnya berbagai penyakit ini banyak terjadi pada masyarakat pedesaan yang kurang memperdulikan kebersihan lingkungan sekitar dan pola hidup mereka Hal inilah yang mengakibatkan mereka mudah terserang berbagai penyakit. Banyak dari masyarakat yang membuang sampah sembarangan, yang tidak mengubur kaleng-kaleng bekas, yang tidak menguras bak mandi mereka, yang tidak memperhatikan kebersihan kandang hewan peliharaan mereka. Kehidupan lingkungan masyarakat di desa yang seperti itu dan sudah menjadi kebiasaan setiap harinya seringkali melupakan hal sepele yang dapat menimbulkan banyak masalah seperti halnya, mencuci tangan sebelum makan. Hal itu bisa menyebabkan munculnya berbagai penyakit yang diderita warga masyarakat itu sendiri. Berdasar pernyataan Kepala Desa Ridan Permai Bapak Heru Pramana, permasalahan yang paling banyak ditemui di desa Ridan Permai adalah permasalahan yang berkaitan dengan sampah. Lahan pekarangan yang sempit dan keterbatasan biaya untuk membeli bak sampah di depan rumah warga yang menyebabkan permasalahan ini menjadi permasalahan pokok di desa Ridan Permai. Melihat latar belakang masalah tersebut, muncul kepedulian kami terhadap warga masyarakat untuk mengadakan sosialisasi pola hidup bersih demi kesejahteraan masyarakat dan menciptakan generasi masyarakat yang peduli akan kesehatan. Hal itu kami lakukan agar warga masyarakat lebih peduli terhadap pola hidupnya, dan agar mengurangi berbagai panyakit yang mungkin saja dapat ditimbulkan akibat pola hidup yang tidak sehat. Khususnya mengenai masalah penanganan sampah, kami berencana mensosialisasikan serta mengadakan pembelian bak sampah, agar warga dapat mengatasi masalah sampah yang ada, serta penggolongan sampah sesuai dengan jenisnya. Agar nantinya sampah-sampah itu bias dimanfaatkan, diolah kembali dan dijual sehingga meningkatkan kesejahteraan masyarakat dan menjadikan pola hidup bersih bagi masyarakat. 34 \| Jurnal Pengabdian Masyarakat (Abdira) Vol.1, No.1, 2021 HASIL DAN PEMBAHASAN Jogging adalah salah satu bentuk olahraga yang dilakukan dengan cara berjalan atau berlari kecil-kecil. Seorang ahli dalam bidang Jogging, Dr. George Sheehan, dalam bukunya mendefinisikan bahwa jogging adalah aktivitas berlari dengan kecepatan di bawah 6 mil/jam, atau sama dengan 9,7 km/jam. Aktivitas tersebut sama saja berlari sejauh 1 km yang ditempuh dalam waktu 6,2 menit, sehingga kecepatan berlari di atas 9,7 km/jam disebut dengan lari/running (Daniel Hartono, 2010). Jogging sama dengan continous slower running yaitu cara lari secara terus menerus dengan intensitas rendah sampai dengan sedang menempuh jarak yang jauh atau dalam jangka waktu antara 30 menit sampai dengan 1 jam yang dapat memacu timbulnya kadar asam laktat. Jogging adalah latihan yang paling efektif untuk meningkatkan kemampuan menghirup udara semaksimal mungkin (VO2max) yang berhubungan erat dengan tingkat kebugaran seseorang serta latihan yang paling efektif untuk menurunkan berat badan (Daniel Hartono, 2010). Jogging merupakan salah satu olahraga yang dapat dilakukan untuk menjaga kesehatan, khususnya organ otak dan jantung. Olahraga jogging tidak Perlu punya keahlian khusus agar dapat melakukannya. Semua orang dari segala usia dapat melakukan jogging. Olahraga jogging tersebut oleh sebab itu termasuk salah satu olahraga yang paling banyak dilakukan. Jogging yang dilakukan secara teratur akan memberikan banyak manfaat bagi kondisi fisik dan kesehatan diantaranya : a) membuat jantung kuat, dimana semakin memperlancar peredaran darah dan pernafasan, b) mempercepat sistem pencernaan, membantu mengatasi masalah pencernaan. c) menetralkan depresi, d) meningkatkan kapasitas untuk bekerja dan mengarahkan pada kehidupan yang aktif, e) membantu membakar lemak dan mengatasi kegemukan, f) memperbaiki pola makan, g) mengencangkan otot kaki, paha dan punggung, h) 35 \| Jurnal Pengabdian Masyarakat (Abdira) Vol.1, No.1, 2021 membuat tidur lebih nyenyak. Jogging juga dapat memberikan kesenangan secara fisik maupun mental (Ari, 2010). membuat tidur lebih nyenyak. Jogging juga dapat memberikan kesenangan secara fisik maupun mental (Ari, 2010). Saat melakukan jogging, irama dan kedalaman napas cenderung meningkat. Jogging meningkatkan fungsi saluran pernapasan. Jogging menyebabkan tubuh bergerak terus menerus tanpa henti, menghasilkan keringat menjadi bercucuran. Keringat tersebut keluar zat-zat yang tidak berguna, diantaranya sel-sel yang dapat menyebabkan kanker. Sel-sel tersebut dalam waktu singkat diganti dengan yang baru oleh tubuh. Olahraga jogging meningkatkan produksi sel darah putih yang berguna untuk memerangi sel kanker. Melakukan aktivitas jogging, sirkulasi darah menjadi lancar. Sirkulasi darah di saat lancar tersebut yang menyebabkan sel kanker tidak dapat bertahan atau berkembang biak. Penelitian telah mengungkapkan bahwa penderita kanker mengalami gangguan emosi sebelum terkena kanker (dr. Triangto, 2009:48). UCAPAN TERIMA KASIH Penulis mengucapkan terima kasih kepada masyarakat desa ridan permai yang telah memberikan waktu kepada peneliti untuk melaksanakan pengabdian mengenai sosialisasi pola hidup sehat (lari pagi). Terimakasih juga kepada semua pihak yang telah memberikan banyak saran berharga dan turut mensukseskan kegiatan program pengabdian masyarakat ini. HASIL DAN PEMBAHASAN Olahraga seperti jogging bermanfaat untuk memperbaiki suasana hati, jogging juga dapat memberikan kesenangan secara fisik maupun mental. Olahraga jogging sendiri apabila dilakukan dengan benar, kelelahan tidak akan terasa meskipun telah menyelesaikan satu tur lebih dari yang dilakukan sebelumnya. Manfaat yang dirasakan ialah merasa nyaman di otot selama jogging dan setelahnya (Ari, 2010). Sebelum melakukan jogging, ada faktor-faktor penting yang harus diperhatikan sebelum melakukan jogging, diantaranya : Bagaimana memulainya. Sebelum memulai jogging, cek kondisi kesehatan dengan berkonsultasi dengan dokter sebagai pertimbangan untuk melakukan olahraga jogging. Pertama-tama, berlari di tempat selama 10 menit. Dapat menambah waktu, jarak dan kecepatan setelah terbiasa. Pakaian dan sepatu. Pakaian yang dikenakan harus sesuai dengan udara saat itu. Saat udara hangat, celana pendek dan t-shirt cukup nyaman untuk dikenakan. Pemilihan pakaian yang dapat memberikan ventilasi bagus, hindari yang penuh jahitan, bertepi tajam atau yang membungkus dengan ketat. Jenis sepatu, kenakan yang lembut dan nyaman, tapi dengan bentuk yang pas dan cocok di kaki. Pemilihan alas kaki yang alasnya dapat ditekuk dengan lentur dalam pergerakan kaki tapi cukup mendukung saat terhentak dengan tanah. Rute dan sesi Jogging. Jogging dapat ditempuh dalam berbagai cara: Jarak yang panjang antara 2-20 km dalam kecepatan biasa, jarak 3-6 km dalam kecepatan tinggi, jogging dengan kecepatan sedang ditempuh dalam 4-8 km. Peregangan tubuh dan setelah beberapa sesi. Sangat disarankan untuk melakukan peregangan sebelum melakukan sesi jogging, dan bukan hanya pada otot kaki, tetapi juga keseluruhan tubuh, lakukan selama 2 menit sebelumnya dan 3-4 menit setelahnya. 36 \| Jurnal Pengabdian Masyarakat (Abdira) Vol.1, No.1, 2021 SIMPULAN Berdasarkan hasil penelitian dan pembahasan diperoleh simpulan bahwa: pemberian pemahaman terhadap masyarakat tentang pentingnya pola hidup bersih dan sehat guna menjaga kesehatan masyarakat sangat diperlukan untuk memotivasi pola hidup sehat. Pemberian pemahaman ini diberikan dengan metode ceramah di kantor desa Ridan Permai; pola hidup bersih dan sehat yang harus diterapkan oleh masyarakat di Desa Ridan Permai adalah dengan menjaga kebersihan diri, lingkungan, dan berolah raga. Salah satu olah raga ringan dan biaya yang minimal yaitu lari pagi; cara membujuk warga masyarakat agar mau melaksanakan dan menerapkan pola hidup bersih dan sehat adalah dengan memberikan pemahaman bahwa hidup bersih dan sehat itu diperlukan untuk mengurangi resiko penyakit pada diri dan keluarga. Ada pengaruh jogging pagi hari terhadap kadar asam laktat pada masyarakat Desa Ridan Permai Daniel Hartono. (2010). Pengaruh Olahraga Jogging Terhadap Kesehatan Fisik dan Mental. Bandung: Remaja Rosdakarya. DAFTAR PUSTAKA Ari, R. (2010). Manfaat Jogging Bagi Kesehatan Manusia. Yogyakarta: Graha Ilmu. Ari, R. (2010). Manfaat Jogging Bagi Kesehatan Manusia. Yogyakarta: Graha Ilmu. Daniel Hartono. (2010). Pengaruh Olahraga Jogging Terhadap Kesehatan Fisik dan Mental. Bandung: Remaja Rosdakarya. 37 \|
W2601052105.txt	https://zenodo.org/record/2384953/files/article.pdf	de		XVII. Bemerkungen zu den Vorsokratikern und Sophisten	Philologus	1,908	public-domain	19,546	XVII. Bemerkungen zu den Vorsokratikern und Sophisten. Nachdem im Anfang des Jahres meine deutsche Auswahl aus den vorsokratischen Philosophen erschienen ist (bei E. Diederichs. Jena. 1908), deren auf weitere Kreise der Gebildeten berechnete Anlage eine wissenschaftliche Begründung der gegebenen Uebersetzung sowie der in der Einleitung niedergelegten Auffassung der Ideen der verschiedenen Denker ausschloß, möge es mir gestattet sein, hier einige Bemerkungen dazu zu machen, teils um einige wichtigere Abweichungen von Diels zu begründen, teils um einige Beobachtungen darzulegen, die sich mir bei der Durcharbeitung der Texte aufgedrängt haben. Ich zähle die Bruchstücke nach Diels, Die Fragmente der Vorsokratiker, 2. Auflage (I. II. 1) 1907 und füge meine Numerierung, wo sie abweicht, in Klammern bei. Xenoplianes. Fr. 1, 16. ταϋτα γάρ ών έστι προχεφότερον. „Denn das ist doch das bessere Teil", Diels. „Denn dies liegt zu erfleh'n uns zunächst". N. „Hoc enim magis expeditum et facile est", übersetzt J. Dalechamp in der Ausgabe des Athenaeus von Casaubonus (1611) und ergänzt laut Randbemerkung: „quam sceleribus vitam suam contaminare" (p. 462c). Also wäre der Sinn: es ist leichter gut als böse sein ; wahrhaftig eine sehr gewagte Behauptung, zumal im Munde eines Philosophen. Was Diels mit seiner Uebersetzung meint, ist nicht ganz deutlich. Wenn ich ihn recht verstehe, so bezieht er ταϋτα auf die Handlungen des υμνείν, σπένδειν, εϋχεσθαι, welche zusammen „das bessere Teil" der Zusammenkunft, im Gegensatz zu dem nun folgenden πίνειν, ausmachen. Aber warum schreibt 34* Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 532 W. N e s t l e , er dann nicht „ d e r Teil"? Soll vielleicht der Gebrauch des Neutrums doch eine andere Auffassung andeuten, derart daß ταύτα auf τα δίκαια gienge? Das Gerechte ist doch das bessere Teil sc. im Leben, im Gegensatz zum Unrecht. Πρόχειρος ist was προ χειρός ist, was vor der Hand, was nahe liegt. Dies kann etwas Fertiges sein, wie Waffen, Geräte u. dgl. ζ. Β. Ψέλια Aisch. Proni. 5 4 : dann heißt πρόχειρου είναι ,bei der Hand sein'. Es kann aber ebenso etwas bezeichnen, woran erst Hand angelegt werden soll, wonach man erst mit der Hand greifen w i l l , auch in übertragenem Sinne: Demostb. Contr. Tim. (24) 1 : εν δ', δ μέγιστον εχω και προχειρότατου προς υμάς ειπείν : das Wichtigste und Nächstliegende, was ich euch sagen will ; ib. 76 εύροι άν τις τοΰτο προχειρότατον die nächstliegende Antwort, und 163 τοϋ τ' αν εύροι τε προχειρότατον den nächstliegenden Grund. Demgemäß wäre auch bei Xenoplianes zunächst wörtlich zu übersetzen: „denn dies ist nun einmal das Näherliegende". Wiederum kann man zweifeln, ob damit sachlich die religiöse Einleitung des Gelages gemeint ist oder τα δίκαια, und ganz sicher läßt sich dies m. E. nicht entscheiden. Im ersteren Falle wäre zu ergänzen τοϋ συμποσίου, im zweiten scheint sich mir der Comparativ durch ein hinzuzudenkendes των άλλων zu erklären. Und diese letztere Deutung gibt jedenfalls einen tieferen Sinn: ,wir wollen beten um die Fähigkeit, recht zu handeln; denn dies — eben das recht handeln — liegt näher, ist wichtiger als alles andere, als die äußeren Glücksgüter nämlich, die für gewöhnlich den Gegenstand des Gebets bilden'. Auch so sind immer noch zwei verschiedene Nuancierungen des Gedankens möglich: entweder das δίκαια πρήσσειν δύνασθ-αι ist der nächstliegende Gegenstand des Gebets oder die nächstliegende Aufgabe, Forderung des Lebens und eben darum auch in erster Linie zu erbitten. Doch läuft dies beides schließlich auf dasselbe hinaus, während grammatisch das erstere sich einfacher ergibt: ταϋτα γαρ ών ευξασθ·αι προχειρότεράν έστι sc. των άλλων oder ausführlicher τοϋ τά άλλα ευξασθ·αι. — Vgl. auch v. Wilamowitz zu Eur. Herakles 161. Fr. 8 (5), 2. βληστρίζοντες έμήν φροντίδα. Was bedeutet hier, wo Xenoplianes von seinem Wanderleben spricht, ,φρονBrought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zn den Vorsokratikern und Sophisten. 533 τίς'? Diels 1 : Geist; Diels 2 : Kummer, φροντίς heißt bekanntlich häufig ,Sorge'. Aber obwohl Xenophanes nicht reich war (Diels 2 S. 35 Nr. 11), so erwartet man hier doch, daß mit φροντίς etwas bezeichnet werde, was sein langes Leben ausfüllte, und das war doch nicht bloß Kummer und Sorge. ,Geist' scheint mir zu allgemein, φροντίς bedeutet aber auch wie φρονεΐν das Nachdenken: Soph. Philokt. 863 το δ' άλώσιμον άμα φροντίδι. Im Oid. Col. 132 heißt εϋφαμος φροντίς andächtiges Nachdenken. So dürfte auch bei Xenophanes in dem Wort die Tätigkeit des Verstandes und die Empfindung des Gemüts zugleich liegen, was ich mit „mein sinnend Gemüt" auszudrücken suchte. Fr. 38 (14) : „Wenn Gott nicht den gelblichen Honig erschaffen hätte, so würde man meinen, die Feigen seien viel süßer", wozu Diels 2 hinzufügt: „als alles andere". Diesen Zusatz halte ich für falsch. Warum sollte man denn, wenn es keinen Honig gäbe, ζ. B. meinen, die Feigen wären viel süßer als die Trauben ? Die Ergänzung kann nur lauten: ,als sie uns jetzt erscheinen, da wir den Honig kennen'. Das Beispiel wurde auch von Protagoras und Demokrit zur Veranschaulichung der Relativität der Sinneswahrnehmung verwendet, von dem ersteren nur der allgemeine Gegensatz von Süß und Bitter (Plato, Theaet. 166 E bei Diels 1 S. 519), von dem letzteren auch der Honig selbst (Diels 2 S. 373 Nr. 134; vgl. S. 375 Theophr. de sens. 63). Der Unterschied ist nur, daß Protagoras und Demokrit mit der verschiedenen Empfindung mehrerer Subjekte bei einem und demselben Gegenstand operieren, während Xenophanes die Empfindungen eines und desselben Subjekts auf Grund einer Vergleichung verschiedener Sinnesreize untersucht. Herakleitos. Fr. 21 (101). θάνατος έστιν όκόσα έγερθ-έντες όρέομεν, όκόσα δε ευδοντες ύπνος. Dazu bemerkt Diels in der Separatausgabe des Herakleitos (1901) S. 7: „Salzlos, wenn nicht folgte όκόσα δε τεθνηκότες ζωή. Leben, Schlaf, Tod ist die dreifache Leiter wie in der Physik Feuer, Wasser, Erde. Vgl. fr. 26. Daher ύπνος, nicht ένύπνιον". Kürzer, aber im gleiBrought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 534 W. N e s t l e , chen Sinn: Vorsokr. 2 S. 662 mit dem Zusatz: „όρέομεν": vielleicht besser ,erleben'". Ich kann dem Satze: „was wir schlafend sehen (oder erleben) ist Schlaf" keinen Sinn abgewinnen außer den einer vollkommenen Selbstverständlichkeit. Man erwartet statt dessen vielmehr ein Paradoxon wie in der ersten Hälfte des Bruchstücks. Ich glaube daher, daß ζωή schon an die Stelle von ύπνος zu setzen ist: „Tod ist was wir im Wachen sehen, was aber im Schlaf, Leben". Der Zustand des Wachseins steht symbolisch für das Individualleben auf Erden, der Schlaf entspricht dem physischen Tod, d. h. in Herakleitos Sinn dem Aufgehen in das Allleben des Feuers. Vgl. fr. 62 (72): άθάνατοι θνητοί, θνητοί αθάνατοι, Ζώντες τον εκείνων θάνατον, τον δέ εκείνων βίον τεθνεώτες. Im fr. 26 erscheint allerdings der Schlaf als Mittelzustand zwischen Tod und Leben. Aber dieser Gedanke einer dreifachen Stufenfolge paßt nicht zu dem ersten Teil des fr. 21, das vielmehr Gegensatzpaare verlangt wie fr. 8 8 : ταύτό τ' ενε ζών και τεθνηκός και το έγρηγορος και το καθευδον και νέον καί γηραιόν · τάδε γαρ μεταπεσόντα εκείνα έστι κάκεϊνα πάλιν μεταπεσόντα ταΰτα. Fr. 39 (117) Βίας, où πλέων λόγος ή των άλλων: „von dem mehr die Rede ist" D. ; „der mehr bedeutet" N. Vgl. hierzu und zu meiner Auffassung von λόγος in den Fragmenten des H. Philol. LXIY (1905) S. 375 ff. Fr. 57 (20) handelt von Hesiod, όστις ήμέρην και εύφρόνην ουκ έγίνωσκεν · εστι γαρ εν. Hierzu Diels (S.A. S. 14): „Mit Verachtung sieht der Philosoph auf den Aberglauben der guten und bösen Tage in Hesiods ,Werken und Tagen' herab". Gewiß tut das der Ephesier, aber nicht hier sondern mit dem Sätzchen, das in Senekas Uebertragung lautet : „unus dies par omni est" (fr. 106 D.), wo Diels auf fr. 57 zurückverweist. Es ist doch aber unmöglich, daß die Bestreitung des Glaubens an Glücks- und Unglückstage sich in die Form kleidet: „Tag und Nacht ist eins", d. h. einunddasselbe. Vielmehr mußte die Entgegnung etwa lauten, wie bei Plut. Cam. 19 steht: άγνοεί φύσιν ήμέρας άπάσης μίαν ουσαν. Die Betonung der tatsächlichen Identität von Tag und Nacht kann sich nur gegen eine Trennung und Isolierung dieser beiden Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Yorsokratikern und Sophisten. 535 Begriffe richten, wie sie in den Personifikationen der Hesiodischen Theogonie vorliegt v. 123: Νυκτός δ' αότ' Αίθ-ήρ τε και Ήμερη έξεγένοντο, wo also die Nacht als die Mutter des Tages erscheint und sie gar (748 f.) im Vorbeigehen mit einander reden. Fr. 5S (69). Der Philosoph sucht die Identität von αγαθόν und κακόν mit einem Beispiel zu illustrieren: die Aerzte, welche die Kranken schneiden und brennen, beanspruchen einen Lohn und verdienen doch keinen ταυτά εργαζόμενοι. Das erklärt Hippolyt mit τά άγαθ-ά και τάς νόσους. Diels 2 S. 663 : „schlecht paraphrasiert. Sie heben durch ihre Guttaten nur die Krankheiten auf. Heraklit meint, sie fügen ja auch Böses zu, tun also dasselbe wie die Krankheit und brauchen daher keinen besonderen Lohn. Prächter: „Man tilge das Komma nach έργαζόμενοι, ,da sie das Gute als das Nämliche wirken wie die Krankheiten', d. h. die Heilung, die schmerzvolle, ist nicht besser als die Krankheit. Daher sollen die Aerzte nichts bekommen". — Prächters Vorschlag scheint mir nicht glücklich. Das Heraklitwort muß mit ταύτα εργαζόμενοι. aufhören, das dem άγαθ-òv καί κακόν εν έστιν am Anfang entspricht. Außerdem aber sind alle drei Erklärungen insofern schief, als sie in dem ταυτά die Identität von Krankheit und Operation oder die Aufhebung der einen durch die andere suchen. Heraklit will aber beweisen, daß Gut und Schlimm dasselbe ist und dafür dient ihm die operative Tätigkeit des Chirurgen als Beispiel: diese selbst, das τέμνειν und καίε tv, ist sowohl αγαθόν, weil es Heilung bringt, als auch κακόν, weil es Schmerz verursacht. Das ist der Hauptgedanke. Dazu kommt erst der Nebengedanke: die Heilung ist die billige Entschädigung der Kranken für den Schmerz; also hat der Arzt, der ihm wohl und weh zugleich tat, von Rechts wegen nichts zu beanspruchen. Fr. 67 (76). Gott, das Absolute, ist die Zusammenfassung der relativen, empirischen Gegensätze. „Er wandelt sich aber wie das Feuer, das, wenn es mit Räucherwerk vermengt wird, nach eines jeglichen Wohlgefallen so oder so benannt wird. " D. Ich meine : die Bracbylogie δκωσπερ πυρ . . . . ονομάζεται ist nicht auf das άλλο'.οΰσθαι zu beschränBrought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 536 W. N e s t l e , ken, sondern auch auf das όνομάζεσθ·αι auszudehnen: „ E r verwandelt sich wie das Feuer und wird gleich diesem, wenn es sich mit Rauchwerk vermengt, nach jedermanns Belieben benannt". D i e l s 2 S. 663 verweist zu ονομάζεται auf fr. 48 (βώς — βίο?)· Noch mehr gehört hierher fr. 32 : εν το σοφον μοΰνον λέγεσθαι οόκ έθ-έλει και έθέλεί Ζηνος ονομα. Ferner verweise ich auf Plato, Krat. p. 400 Ε : ώσπερ έν ταίς εύχαίς νόμος έστίν ήμΐν εδχεσθαι, οΐ'τη/ές τε και οπόθεν χαίρουσιν ονομαζόμενοι (se. oí θεοί), ταΰτα και ήμας αύτους καλεϊν. Vgl. Phileb. 12 C; Euthyd. 288 Β ; Protag. 358 A ; Symp. 212 0 ; Parin. 133 D. Auch Eur. Bacch. 275 f. (Δημήτηρ — γη : όνομα δ' όπότερον βούλε:, κάλεt) und fr. 912, 2 f. (Ζευς ε?τ' Άΐδης ονομαζόμενος στέργεις) bewegen sich in dem Heraklitischen Gedankengang, der seinerseits an die vielen im Kultus gebräuchlichen Beinamen der Götter anknüpft. S. Usener, Götternamen S. 336; Gomperz, Griecli. Denker I 53 und 64. Fr. 78 (96). ή8ος γαρ άνθρώπειον μεν ουκ εχει γνώμας, •ö-είον δέ εχε:. Hier übersetzt Diels γνώμας mit ,Zwecke'. Dies scheint mir aus zwei Gründen unrichtig: denn 1) wie kann man behaupten, dass der Mensch keine Zwecke h a b e ? Man müßte das W o r t dann schon in prägnantem Sinn nehmen: er kennt die w a h r e n Zwecke der Dinge und Vorgänge nicht; 2) aber ist es wenigstens fraglich, ob wir Heraklit eine teleologische Weltansicht zuschreiben d ü r f e n , obwohl allerdings z. B. fr. 41 nach dieser Richtung zu weisen soheint. Denn sein Gott ist doch streng genommen der Weltprozeß selbst und er gebraucht f ü r sein Walten das Bild eines spielenden Kindes (fr. 52). E s ist mir überhaupt fraglich, ob γνώμη ,Zweck' bedeuten kann. Diels selbst übersetzt es fr. 41 mit ,Vernunft' und Gomperz (Apol. der Heilkunst in dem Wiener Sitz.-Ber. 120. 1890 S. 5 if.) hat auf die erkenntnistheoretische Bedeutung des Wortes in der älteren griechischen Philosophie aufmerksam gemacht. Ich halte daher hier etwa ,Einsicht' für passender. Der Mensch ermangelt der Einsicht in den Weltverlauf, Gott besitzt sie : „ der Mensch wird, gegen Gott gehalten, wie ein Affe erscheinen in Weisheit, Schönheit und allem andern" (fr. 38 D). Vgl. auch Kinkel, Gesch. der Phil. I 83 und A. 70. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 537 F r . 117 (99). Ein Trunkener wird von einem Kinde gef ü h r t σφαλλόμενος, ουκ επαΐων δκη βαίνει : „ er taumelt und m e r k t nicht, wohin er g e h t " D. Diels zieht also in der Uebersetzung das σφαλλόμενος zum Folgenden, während er es im griechischen Text durch das K o m m a als zu άγεται gehörig bezeichnet. Und es gibt in der T a t einen besseren Sinn, wenn m a n αγεται σφαλλόμενος zusammennimmt. Diese W o r t e werden durch die folgenden ουκ — βαίνει näher e r k l ä r t : „er m e r k t nicht, wohin er g e h t " . Dies ist aber nicht ein Grund zum ,taumeln', sondern zum ,getäuscht werdeil', was ich in σφάλλεσθαι finde: „Einen trunkenen Mann kann ein kleines Kind leiten und irre f ü h r e n ; denn er merkt nicht" etc. Die T r u n k e n h e i t h a t ihm so das Bewußtsein verdunkelt, daß ihn selbst ein Kind am Narrenseil f ü h r e n kann. F r . 121 (118). Bei dem H e r m o d o r f r a g m e n t glaube ich in ήβηδόν — άνήβοις einen beabsichtigten Gegensatz h e r a u s zuhören, der in der Uebersetzung „Mann f ü r Mann" (D.) nicht zur Geltung kommt. Der Sinn ist doch: die ältere, törichte Generation sollte einer jüngeren, verständigeren Platz machen. D a r u m h a b e ich geschrieben: „Die Ephesier sollten sich, so viele ihrer erwachsen sind, insgesamt a u f h ä n g e n " . Parmenides. F r . 1, 1. δσον „soweit" D. Ich ziehe „so o f t " v o r : v. Wilamowitz im Hermes 34 (1899) S. 203 f. s. Melissos. F r . 8, 5. Das g r o ß e Bruchstück handelt von der T r ü g lichkeit der Sinneswahrnehmung. E s wird die Möglichkeit der Veränderlichkeit des Seienden bestritten und dann fortgef a h r e n : ού γαρ αν μετέπιπτεν, ει άληθη ήν · αλλ' ήν οιόν περ έδόκει εκαστον τοιούτον, του γαρ έόντος αληθινού κρεϊσσον ουδέν. „ D e n n wären sie (sc. die Einzeldinge) wirklich, so schlügen sie nicht um, sondern jedes bliebe so wie es vordem aussah. Denn stärker als die wirklich vorhandene W a h r h e i t ist nichts". D. Diels sieht also, wie auch Kinkel (Gesch. der Phil. I 167), in den W o r t e n του — οόδέν den Gegensatz der Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 538 W. N e s t l e , durch das Denken erfaßten Wahrheit gegenüber dem Schein der Sinneswahrnehmung. Ohne diese Ergänzung steht der Gedanke ganz isoliert, da es unmittelbar weitergeht: „Schlägt aber etwas um, so geht das Vorhandene zu Grande" etc. Ich möchte daher in den Worten του Ιόντος άληθ-tvoö eine Rückbeziehuug auf άληθη ήν sehen und übersetzen: „denn nichts ist stärker als das, was wahrhaft ist", d. h. dem wirkliches Sein zukommt; und deswegen ist das wahrhaft Seiende auch keiner Veränderung zugänglich. Diese wäre ein Zeichen von Schwäche, von unvollkommenem Sein und nur denkbar bei etwas, dem kein παν είναι (fr. 2) zukommt. Empedokles. Fr. 4 (7), 9 fi'. Hier nimmt sich v. 10 μήτε τι δ ψ tv εχων πίστει πλέον ή κατ' άκουήν (sc. á-θ-ρει) wie eine direkte Polemik gegen Herakleitos fr. 101a (D2) aus: οφθαλμοί γαρ των ώτων ακριβέστεροι μάρτυρες. Mit der ganzen Stelle, die eine Aufforderung zum gleichmäßigen, wenn auch vorsichtigen Gebrauch der Sinne zum Zweck der Erkenntnis enthält, setzt sich Empedokles, offenbar bewußt und absichtlich, in Widerspruch gegen Parmenides fr. 1, 33 ff., wo eben der Gebrauch der Sinne als Irrweg der Forschung bezeichnet wird. Dagegen ist Empedokles in der Verwerfung des absoluten Werdens und Vergehens mit Parmenides so einig, daß auch seine diesbezüglichen Verse stark an die betreffenden des Eleaten anklingen: vgl. Emp. fr. 11, 12 und 17, 30 ff. mit Parm. fr. 8 v. 7 ff.; 12 f.; 21; endlich Emp. 2 (5), 8 f. und 17 (16), 26 mit Parm. 1, 27 f. und 8, 52. Fr. 5 (4), 1 : άλλα κακοίς μεν κάρτα μέλει κρατέουσιν άπιστειν. Diels gibt κρατέουσι mit „Herrschern" wieder und erklärt dies (2. Aufl. S. 683): „entweder den Philosophen oder den Weltprinzipien", Begriffe, die man schwerlich in dem deutschen Wort ,Herrscher' finden würde. Für die Deutung ,Weltprinzipien' könnte man auf 17, 29 verweisen, wo von den Elementen gesagt ist, daß sie κρατέουσι. Aber einmal scheint doch in κρατέουσι ein Gegensatz zu κακοί zu stecken und dann verlangt die Anwendung der warnenden Gewohnheit auf den Fall des Empedokles und seinen Schüler Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 539 Pausamas, die im Folgenden gemacht wird, doch ein persönliches Objekt der άπιστίη. So schrieb ich : „den Starken am Geist", dem Sinne nach gleich mit Dielsens erster Erklärung. Zu άπιστείν vgl. Herakl. fr. 86. Fi·. 17 (16), 19. Νείκός τ' ουλόμενον δίχα των άτάλαντον άπάντν). Die beiden letzten Worte lauten bei Diels : „ der überall gleich wuchtige". S. 684 wird auf Parm. 8, 24 und Arat Pkain. 22 verwiesen. Die erstere Stelle, wo von der absolut gleichmäßigen Fülle des eleatischen All-Einen die Rede ist, paßt kaum hierher. In dem Vers des Arat aber, wo es von der Himmelsachse heißt: εχει δ' ατάλαντος άπάντγ) μεσσηγυς γαίαν, kann ατάλαντος nichts anderes bedeuten als ,im Gleichgewicht schwebend'. Also : der Haß schwebt oder hält sich nach allen Seiten hin im Gleichgewicht. Er hat noch weder ein Uebergewicht über die Φιλότης bekommen, noch ihr gegenüber etwas eingebüßt, wie dies für den Zustand der Ruhe im σφαίρας durchaus paßt. Und nun vergleiche man dazu fr. 35 (21). Hier hat schon der Prozeß der Scheidung begonnen, er befindet sich aber noch im Anfangsstadium: der Haß ist in die unterste Tiefe des Wirbels gelangt, die Liebe befindet sich in der Mitte. Die Elemente sind teils schon gemischt, teils noch ungemischt, da der Haß noch nicht ganz in die äußersten Grenzen gewichen ist : πολλά δ' αμεικτ' εστηκε κεραιομένοισίν εναλλάξ, οσσ' ετι Νείκος ερυκε μετάρσιον (ν. 8 f.). „Doch blieb noch vieles ungemischt unter dem Gemischten, soviel der Streit noch davon in der Schwebe hielt." D. Diese Uebersetzung gibt zweifellos einen guten Sinn. Aber ist sie grammatisch möglich ? Bei der Beziehung auf πολλά kann doch δσσ' nur Apostrophierung von δσσα, nicht etwa von δσσον, sein; bezieht sich aber der Schwebezustand auf die πολλά, so müßte man doch eigentlich μετάρσια erwarten. Oder soll μετάρσιον ein sog. Accusativ des Inhalts oder adverbialer Accusativ sein, der als solcher unveränderlich wäre ? Karsten bezog μετάρσιον auf Νείκος und übersetzte — freilich höchst gewaltsam — : „ quotquot Discordia contumax adhuc teneret", wozu er im Kommentar S. 214 bemerkt: „Νείκος μετάρσιον, oppositum huic ήπιίφρων Φιλότης v. 177" (Diels ν. 13). Das ist sicher falsch, weil nun einmal Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 540 W. N e s t l e , μετάρσιον nicht ,contumax' heißen kann. Aber man k a n n übersetzen: „soviele der schwebende H a ß noch zurückhielt woran sich das Folgende g u t anschließt: „denn noch war er nicht völlig tadellos an die äußersten Grenzen des Kreises hinausgetreten". D a n n wäre damit der Zwischenzustand des Νεϊκος zwischen dem ένεμιμνε und έξεβεβήκει (ν. 11) bezeichnet. Zugleich wäre die Stelle eine Parallele zu 17, 19 n u r mit dem Unterschied, daß in fr. 17 das Gleichgewicht des Νείκος noch vollständig, nach allen Seiten (άπάντγ]) erhalten, während es fr. 35, 9 ff. schon ins W a n k e n geraten ist. Mit der W e n d u n g „das dort schwebend der H a ß festhielt" ließ ich absichtlich beide Auffassungen offen. F r . 77 und 78 (80), wo von immer grünen und immer F r u c h t tragenden Bäumen die Rede ist, die κατ' ήέρα das ganze J a h r hindurch in der Fülle der F r ü c h t e prangen, h a t t e Karsten (366 f.) auf die Schilderung des goldenen Zeitalters bezogen wie fr. 130 (79; Karsten 374 f.), während Diels jene beiden dem Gedicht περί φύσεως und nur dieses, das von dem vertraulichen U m g a n g des Menschen mit den wilden Tieren handelt, den καΒ·αρμοί zuweist (S. 690 zu 211, 21). Ich k a n n mich auch jetzt des Eindrucks nicht entschlagen, daß Karsten recht gesehen hat, und der Z u s a m m e n h a n g bei Theophrast (Caus. plant. I 13, 2) scheint mir auch d a f ü r zu sprechen : εί δέ και συνεχώς δ αήρ άκολουθ-οίη τούτοις (sc. τοις δένδροις), "ίσως ουδέ τα παρά των ποιητών λεγόμενα δόξειεν αν άλόγως εχειν ούδ' ώς Έ . άείφυλλα και έμπεδόκαρπά φησι θ'άλλειν, καρπών άφθ'ονίηισι κατ' ήέρα πάντ' ένιαυτόν. E s ist doch klar, daß hier dem N a t u r f o r s c h e r m ä r c h e n h a f t e Ideallandschaften vorschweben wie δ 565 ff. (Elysion), ζ 42 ff. (Olymp) und besonders η 114 ff. (Gärten des Alkinoos) : εν8·α δέ δένδρεα μακρά πεφύκει τηλεθόωντα, ογχναι και £οιαί και μηλέαι άγλαόκαρποι, συκαί τε γλυκερά: και έλαια: τηλεθ'όωσαι. τάων ουποτε καρπός άπόλλυται ουδ' απολείπε: χείματος οάδέ θέρεος έπετήσιος • άλλα μάλ' αίεί Ζεφυρίη πνείουσα τα μέν φύει, αλλα δέ πέσσε:. Ganz ähnlich schildert Pindar (Ol. I I 70 ff.) die Insel der Seligen : έν-8-α μακαρων νασος ώκεανίδες αύρα: περιπνέοισιν · Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 541 ανθ-εμα δέ χρυσοϋ φλέγει, τα μέν χρυσόθ·εν άπ' άγλαων δενδρέων, ύδωρ δ' άλλα φέρβει etc. In diese Reihe stellt Theophrast auch die Verse des Empedokles, von dem er sagt, daß er eine bestimmte Mischung der Luft, wie sie im Frühling ist, voraussetze (όποτιθέμενάς τινα του αέρος κρασιν, την ήρινήν, κοινήν) : „fingens qiiandam aeris temperiem, qualis verus tempore esse solet" (Karsten S. 269). Kurz es handelt sich offenbar um die Schilderung eines ewigen Frühlings, der zugleich die Freuden des Herbstes bietet, also um etwas Uebernatürliches. Ist dies richtig, wie ich überzeugt bin, so gehört das Bruchstück in die καθαρμοί. Zu fr. 105 (50) und 107 (49) vgl. Parin, fr. 16. Der bei dem Eleaten vom νέος gebrauchte Ausdruck παρίστασ8·αι kehrt in ganz entsprechendem Zusammenhang (τό φρονείν) hei Empedokles fr. 108, 2 wieder. An beiden Stellen erscheint die körperliche Disposition als Grundlage der Gedanken und Vorstellungen. Fr. 108: Die hier gegebene natürliche Deutung der Träume ist dieselbe, die Herodot VII. 46 dem Artabanos leiht. Ihre Weiterbildung liegt de morb. sacr. 17 vor. Fredrich, Hippokr. Unters. S. 214, 3. Fr. 121 (62), 3. Empedokles ist aus der Geisterwelt herabgestürzt an den unerfreulichen Ort, wo Mord und Groll und Scharen anderer Unglücksmächte (Κηρες) auf der Unheilswiese ("Ατής λειμών) im Düstern hinhuschen, αυχμηραί τε νόσοι και σήψιες έργα τε ρευστά. Was ist unter diesen έργα ^ευστά zu verstehen? „Opera fluxa" übersetzt Karsten (v. 22), Diels dagegen „Ueberscliwemmung". Es wäre also ein ähnlicher Ausdruck wie fr. 111 (3), 8 ρεύματα δενδρεόθ-ρεπτα, nur daß hier im günstigen Sinne ,baumernährende Regengüsse' gemeint sind. Man müßte dann aber auch fast wie dort (v. 6) den Gegensatz, die verderbliche Dürre (des Landes) erwarten. Neben Begriffen, die sich alle auf das geistige oder körperliche Wesen des Menschen beziehen, stehen ,Ueberschwemmungen' ohne alle Vermittlung. Andererseits müßte dem ,dörrenden Siechtum' (αυχμηραί νόσοι) ein Begriff wie Wassersucht oder dgl. entgegenstehen, was die Worte aber nicht bedeuten können. Ich glaube daher doch, daß es noch das Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 542 W. N e s t l e , Beste ist, mit Karsten (S. 166 f.) und Rohde (Psyche 2 II 178, 1) unter der εργα ρευστά die „res fluxae", „die unstäten, vergänglichen Werke der Menschen auf Erden" zu verstehen im Sinne des Heraklitischen Flusses aller Dinge und so kam ich zu der freilich etwas weitläufigen Uebersetzung : „Taten der Menschen dazu im Strom der Vergänglichkeit schwindend". Denn daß Empodokles hier von der Erde und nicht vom Hades redet, hin ich mit Rohde fest überzeugt. Er schildert eben das Jammertal der Erde mit den düstern Farben der Hölle. Fr. 125 und 126 habe ich mit einander verbunden (66), ohne zu übersehen, daß in dem ersten das Subjekt masculinum ist (αμείβων), im zweiten femininum (περιστέλλουσα). Der Sache nach ist es jedenfalls beidemal dasselbe, wie ja Porphyrios bei Stob. Erl. I 49, 60 zu fr. 126 sagt, daß Empedokles die είμαρμένη und φύσις auch δαίμων benenne. Er identifiziert ja auch fr. 134 (82) Apollo mit der weltregierenden φρήν Εερή. Ebenso könnte man an die 'Ανάγκη fr. 115 (57), 1 denken. — Karsten S. 271 f. Fr. 129 (71), 6 hat Diels seine in der 1. Auflage gegebene Uebersetzung der Worte καί τε δέκ' άνθ·ρώπων καί τ' εϊκοσιν αίώνεσσιν „auf zehn und zwanzig Menschengeschlechter hin" in der 2. geändert: „in seinen zehn und zwanzig Menschenleben". Sicher ist gemeint, daß Pythagoras nicht nur das persönlich von ihm Erlebte schaute, sondern überhaupt alles, was während dieser zehn oder zwanzig Generationen (αιώνες) in der Welt vorgegangen war. Bei fr. 137 (73) bin ich der Lesart Karstens (410 ff.) gefolgt, von der Stein (430 ff.) nur v. 3 mit ·9·ύοντος· ó δ' αρ νήκουστος statt {t-uovx' * ó δ' άνηκούστησεν abweicht. Die Konjektur Bergks οί δ' έπορεΰνται bei der Ueberlieferung οί δε πορευνταί hat zwar sehr viel Bestechendes ; aber wird dadurch die Situation klarer ? Ein Opfer wird geschlachtet und nun — „drängen die Opfer (οί δ') sich hinzu und flehen die Schlächter (-8'ύοντας, Sext : ·8·ύοντες) an". Man sollte doch eher denken, daß die Opfertiere zu fliehen suchen, wenn sie eines geschlachtet sehen, als daß sie sich hinzudrängen. Und woher kommt auf einmal die Mehrzahl der Schlächter und Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 543 der Opfertiere, während doch vorher nur von einem Vater und seinem in der Gestalt eines Tieres von ihm zu schlachtenden Sohn die Rede war? Allerdings ergibt sich bei Karstens Lesart ein Hysteronproteron ; aber das ist nichts Unerhörtes : mit σφάζει wird die ganze Opferhandlung kurz antizipiert; dann folgen die Einzelheiten, der Gang zum Altar und die vergeblich um Erbarmen bittenden Blicke. Auch braucht άείρειν nicht notwendig in dem engen Sinn emporheben' gefaßt zu werden: es heißt auch ,gewaltsam wegführen': z. Β. φ 18; daher Karsten: „trahens". Anaxagoras. Fr. 4 (9). In allem was sich verbindet, sind σπέρματα πάντων χρημάτων καί ιδέας παντοίας έχοντα και χροιάς καί ήδονάς. Das letzte Wort gibt Diels mit „Gerüche" und er verweist dazu S. 796 (zu 315, 8) auf Zeller 5 I S. 264, 4. Dazu hätte aber noch S. 984, 3 desselben Werkes angeführt werden sollen, wo Zeller zu unserer Stelle sagt: „Auch hier Hesse sich ihm (sc. dem Wort ήδονή) zwar die Bedeutung ,Geruch' geben; doch passt ,Geschmack' noch besser; das wahrscheinlichste ist aber, daß das Wort, ähnlich wie das deutsche ,Schmecken' in einzelnen Dialekten, beide Bedeutungen ohne schärfere Unterscheidung vereinigt". Dies halte ich für durchaus richtig und ich habe daher „Geruch und Geschmack" gesetzt. Bei Diogenes von Apollonia fr. 5 (7), wo das Wort genau in derselben Verbindung (ήδονής καί χροιης) erscheint, hat Diels selbst „Geschmack" übersetzt. Ganz zweifellos hat ήδονή die Bedeutung ,Geschmack' bei Xen. An. II 3, 16: die Soldaten essen das Mark der Dattelpalme und wundern sich τό τ' ε?δος καί την ιδιότητα της ήδονης. Ebenso ήδύς ,wohlschmeckend' ib. I 5, 3 (κρέα) und I 9, 25 (οίνος) ; ferner τούτοις ήσθ-η Κύρος I 9, 26 : dies schmeckte dem Kyros; ebenso Herodot I 119. Selbst in dem oben besprochenen Bruchstück des Herakleitos fr. 67 (76) könnte man κα·θ·' ήδονήν εκάστου übersetzen: ,jeder nach seinem Geschmack'. Vgl. Zeller 5 I 664, 1. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM W. N e s t l e , 544 Hippon 2 P l u t . De Is. et Os. 3 3 p. 364 Β : τον δέ "Οσιριν αό π ά λιν μελάγχρουν γεγονέναι μυθ-ολογοΰσιν, δτι παν δδωρ και γην και ίμάτια και νέφη μελαίνει μιγνύμενον, και τ ω ν ν έ ω ν υ γ ρ ó τ η ς ένοΰσα π α ρέχει τάς τρίχας μέλαινας, ή δε π ο λ ί ω σ ι ς οίον ώ χρ ί α σ ι ς ίιπό ξηρότητος έπιγίγνεται τοις παρ α κ μ ά ζ ο υ σ α ν. D a ß bei P l u t a r c h hier die gleiche Lehre vorliegt wie bei Hippon (oder Hipponax oder Hippokrates ? D i e l s 2 S. 693), springt in die Augen. Im folgenden 34. Kapitel (S. 364 D.) b r i n g t er einige Etymologien, darunter die des Osiris = "Γσιρις von υειν, die sich bei Hellanikos (fr. 154 Müller) f a n d und womit die von den „Hellenen" vollzogene Gleichsetzung dieses Gottes mit Dionysos (Herod. I I 42. 47 f. 123. 144 f. 156) motiviert wird. Die V e r m u t u n g liegt nahe, daß diese Etymologie έκ του αύτοϋ γυμνασίου s t a m m t wie die Feuchtigkeitstheorie und daß sie Hellanikos von Hippon übernommen hat. Diels S. 224 N r . 1 1 : Ί π π . . . . οιεται έν ήμϊν oìκείαν είναι υ γ ρ ό τ η τ α , κ α θ ' ήν και αισθ-ανόμεθ-α και η ζ ώ μεν · δταν μεν oöv οίκείως εχη ή τοιαύτη ύγρότης υγιαίνει το ζψον, δταν δε άναξηρ• άνθη, άναισθητεί τό ζψον και αποθνήσκει, δ ι à δ ή τ ο ύ το ο ί γ έ ρ ο ν τ ε ς ξηροί, δτι χωρίς όγροτητος. Pliilolaos. F r . 1 mit seinem seltsamen A n f a n g ά φύσις δ' έν τω κόσμιο („bei der W e l t o r d n u n g " D.) άρμόχϋ-η halte ich mit ß ö c k h f ü r ein ungenaues Zitat aus fr. 2 (1). F r . 2 (1). Ά π ε ι ρ α und περαίνοντα („Unbegrenztes — Begrenzendes" D.) habe ich mit „Unbestimmtes und Bestimmtes" (infinita — finita: Boëtliius bei Boeckh S. 56) übertragen. Diese Bedeutung h a t auch bei P l a t o im Philebos άπειρον und πέρας εχον. (Vgl. Ritter, Bemerkungen zum Philebos im Philol. 62. 1903 S. 509 ff.). — Das schwierigste Sätzchen in diesem Bruchstück lautet : δηλοί δέ και τα έν τοις εργοις. E s soll dies einen Beleg dafür e n t h a l t e n , δτι έκ περαινόντων δέ και απείρων δ τε κόσμος και τα έν αότψ συναρBrought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 545 μόχθη. Böckh (S. 50) verstand unter den 'έργα. Bauwerke. Diels 1 : „ Damit stimmt auch die Beobachtung an den Werken über ein". D i e l s 2 : „Aeckern" (statt „Werken"). Ich kann mir mm keine Vorstellung davon machen, was ein Acker sein soll, der „aus begrenzenden und unbegrenzten Linien" oder gar nur „aus unbegrenzten Linien gebildet wird". Vielmehr glaube ich, daß mit έργα. nichts anderes gemeint ist als was vorher τα έν αάτω (sc. τω κόσμω) hieß. Genauer erläutert wird dies durch fr. 11 (5), wo die Bedeutung der Zahl auseinandergesetzt wird, deren Natur und Kraft wirksam ist ου μόνον έν τοις δαομονίοις καί iteiotç πράγμασι . . . . άλλα και έν τοις άν&ρωπικοίς εργοις και λόγοις πάσο. Alles dies ist in fr. 2 (1) in έργα zusammengefaßt: es sind dies die sowohl von der Gottheit b e w i r k t e n G e b i l d e wie ζ. B. die Gestirne als auch „das Gebild der Menschenhand", ja nach pythagoreischer Anschauung sogar die sittlichen Leistungen und Begriffe, die ja auch auf Zahlenwerte zurückgeführt werden. Somit wäre der Sinn: Wohin man sehen mag im κόσμος, überall trifft man auf Gebilde teils göttlicher teils menschlicher Art, die aus περαίνοντα und απε:ρα bestehen : die περαίνοντα sind das substantiell gedachte formende Prinzip der Zahlen, die απεφα die qualitätslose Masse, das aus beidem Zusammengesetzte die einzelnen Dinge und Begriffe, die eben nur deshalb erkennbar sind, weil sie an der Zahl teilhaben : fr. Β (2), 4 (3) und 6 (7). Demokrit. Fr. 191 (47). Am Schluß des großen Bruchstücks über die εύδυμίη heißt es : ταύτης γαρ έχόμενος της γνώμης εύθ-υμάτερον τε δίαξεις etc. Diels: „hältst du dich an diesen Spruch, so wirst du wohlgemuter leben ". Ich glaube nicht, daß γνώμη hier ,Spruch' heißen kann. Dazu ist die vorhergehende Ausführung schon viel zu lang. Wichtiger ist aber, daß γνώμη bei Demokrit noch eine stark intellektualistiscke, erkenntnistheoretische Färbung h a t : unterscheidet er doch die γνησίη und σκοτίη γνώμη fr. 11 (13). Also wird auch hier zu übersetzen sein: „wenn du dich an diese Erkenntnis hältst". Da es sich aber um die Durchführung von als richtig erkannten Pliilologus LXVII (N. F. XXI), 4. 35 Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 546 W. N e s t l e , ethischen Grundsätzen handelt, dürfte dem Sinne etwa entsprechen: „in dieser Gesinnung". Vgl. in demselben F r a g ment: επί τοις δυνατοίς δει εχειν την γνώμην „auf das Mögliche muß man seinen Sinn (,Denken' D.) richten". Zu Fr. 195 (126) β ϊ δ ω λ α έσθ-ητι καί κόσμω διαπρεπέα προς θεωρίην, άλλα. καρδίης κενεά verweist Dyroff (Demokritstud. S. 140, 1) auf Isola - . 1, 27 : είναι βούλου τα περί την έσθ·ητα φιλόκαλος, άλλα μή καλλωπιστής, εστι δέ φιλακάλου μέν τδ μεγαλοπρεπές, καλλωπιστοΰ δέ τδ περίεργον. Die Aehnlichkeit heider Sprüche ist sehr vag. Viel näher liegt der Vergleich mit Sim. Amorg. fr. 7, 67 (Crus.) : καλδν μέν οδν θ έ η μ α τοιαύτη γυνή und Eurip. Hipp. 631 f., wo es von dem Mann einer eitlen Frau h e i ß t : γέγηθ-ε κόσμον προστιθ·είς ά γ ά λ μ α τ ι καλδν κακίστω, wozu man noch die ατυχία κοσμουμένα des Epicharm (Lor. Fr. Β 35, 5) vergleichen mag. Denn auch Demokrit hat offenbar die eitlen Frauen im Auge. Fr. 196 (86) : λήθ-η των ιδίων κακών θρασύτητα γεννά. „Vergessen der eigenen L e i d e n erzeugt Frechheit" D. Dieser Vorgang ist mir psychologisch unerklärlich. Dagegen ist ein klarer Gedanke : Wer seine eigenen F e h l e r vergißt, wird frech". Mit Recht hat Kinkel (Gesch. der Phil. I 226, 64) den Spruch mit fr. 43 (88) zusammengestellt, obgleich hier nicht dasselbe, sondern ein synonymes W o r t gebraucht ist: μεταμέλεια έπ' αίσχροϊσιν έργμασι βίου σωτηρίη. Vgl. übrigens unten Fr. 297. Fr. 228 (75). Demokrit vergleicht die ungebildeten Kinder geiziger Leute mit Akrobaten, die zwischen Schwertern (eigentlich: unter Schwerter hinein ές μαχαίρας) Sprünge machen: ήν ένδς μούνου μή τύχωσι καταφερόμενοι, έν·Θ·α δει τους πόδας έρεϊσαι, άπόλλυνται * χαλεπδν δέ τυχείν ενός · τδ γαρ ίχνιον μοϋνον λέλειπται των ποδών. Ebenso gehen auch jene Kinder zugrunde, ήν άμάρτωσι του πατρικού τύπου. Diels übersetzt ένδς μούνου etc. „nur ein einziges Mal den Fleck nicht treffen" . . . . „es ist aber schwierig den Fleck auch nur einmal zu treffen". E r muß also, weil er ένδς μούνου zeitlich faßt, den Hauptbegriff ,Fleck' aus ίχνιον μοϋνον των ποδών zweimal antizipierend ergänzen. Dadurch wird Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 547 aber auch der Vergleich schief: die Kinder können doch offenbar die Art ihres Vaters nur entweder verfehlen oder nicht, nicht aber einmal sie verfehlen, das andere Mal nicht. Dem muß das Bild entsprechen: nicht darauf kommt es au, ob die Akrobaten „auch nur einmal" die Stelle verfehlen, sondern darauf, daß sie die Eine Stelle verfehlen, wie die Kinder den Einen τύπος. E s wird also zu ενός μούνου als substantivischer Begriff etwa χωρίου allerdings zu ergänzen, aber darauf das durch μούνου verstärkte ενός zu beziehen sein: „wenn sie . . . . die Eine Stelle verfehlen" . . . es ist aber schwer, diese Eine zu treffen". Die aus Plut. Ages. 38 beigebrachte Parallele (S. 725) scheint mir daher nicht zu stimmen: seine Gesundheit kann man allerdings durch einen einzigen Unglücksfall oder groben Verstoß gegen eine richtige Diät für immer schädigen und ebenso einen Staat durch einen einzigen F e h l schlag in der Politik ; aber wie soll man durch eine einzige Handlung den Charaktertypus seines Vaters „verfehlen"? Sondern daß die Kinder g e n a u so werden wie der Vater, das ist fast unmöglich, und es ist in dem vorliegenden Fall verhängnisvoll, wenn diese kaum zu verwirklichende Möglichkeit nicht eintritt. — Zu dem Akrobatenkunststück selbst vgl. Xen. Symp. 2, 11. Fr. 258 (156) und 263 (151). Wer schädliche Menschen (oder Tiere : vgl. fr. 257) tötet, δίκης και θ-άρσεος και κτήσεως . . . . μείζω μοίραν μεθ-έξες und ebenso, wer die größten Auszeichnungen an die Würdigsten verteilt, δίκης καί αρετής μέγίστην μετέχε: μοίραν. Diels übersetzt μοίραν μετέχε:ν mit „beanspruchen dürfen" (258) und „Anspruch haben" (263). Gehen wir vom zweiten Fall aus ! Anspruch auf Gerechtigkeit hat doch wohl jeder Bürger und „Anspruch auf Tüchtigkeit" gibt überhaupt keinen rechten Sinn. W i r müßten bei dieser Uebersetzung hineindenken : auf das Lob (der Gerechtigkeit und Tüchtigkeit). Uebersetzen wir aber wörtlich: ,Anteil haben an', so ist das soviel als ein Vertreter von Gerechtigkeit und Tüchtigkeit sein. In einer gewissenhaften Amtsführung kommen diese Tugenden zum Ausdruck als Regierungsgrundsätze und dies trägt zur Pflege der δίκη und αρετή auch von Seiten der nicht im Amt stehenden Bürger bei. 35 * Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM W. N e s t l e , 548 Also: Gerechtigkeit und Tüchtigkeit wird gewahrt. Ebenso wahrt, wer Schädliches beseitigt, nicht nur seine eigene εύθυμίη, δίκη, θάρσος, κτησις, sondern diejenige aller Bürger, wofür dasselbe Wort ,wahren' passt 1 ). Wer so handelt, hat eben dadurch den größten Anteil daran, das größte Verdienst darum, daß die wertvollen Güter dem Staate und seinen Bürgern erhalten werden. Fr. 297 (36). Ένιοι θνητης φύσεως διάλυσιν ούκ είδότες άνθρωποι, συνειδήσεο δέ της έν τω βίω κακοπραγμοσύνης, τον της βιοτης χρόνον έν ταραχαίς καί φόβοις ταλαιπωρέουσ', ψεύδεα περί του μετά την τελευτήν μυθ-οπλαστέοντες χρόνου. Diels: „sich dagegen des menschlichen Elends wohl bewußt sind". Das Wort συνείδησης erscheint hier zum ersten Mal in der griechischen Literatur (Diels, Preuß. Jahrb. 1906 Bd. 125 S. 404); was ist nun unter der συνείδησης κακοπραγμοσύνης zu verstehen ? Offenbar ist sie die Ursache der ταραχαί und φόβοι, die ihrerseits wieder die Quelle cíes μυ-9-οπλαστεόν (zu dieser Neubildung s. übrigens ßohde, Psyche 2 II 191 A.) sind. Somit gehen die Hadesfabeln aus Angst, die Angst aber aus der συνείδησις κακοπραγμοσύνης hervor. Würde nun dies letztere Wort ,Elend' bedeuten, so müßte man doch wohl erwarten, daß die Leute, die ein so elendes Leben führen, auf ein b e s s e r e s Jenseits h o f f e n . Damit wäre aber das A n g s t g e f ü h l ausgeschaltet. Dieses verlangt vielmehr κακοπραγμοσύνη nicht als Ersatz für κακώς πράττειν sondern für κακά πράττειν zu fassen : „ die ihrer schlechten Handlungen sich bewußt sind", d. h. über ihr böses Leben ein schlechtes Gewissen haben 2). Die Gewissensbisse projizieren sich in s c h r e c k e n d e Hadesfabeln. Diese Auffassung stimmt sowohl zu den Bruchstücken Demokrits, worin die Todesfurcht als grundlos bekämpft wird ( l a [87]; 199 [40]; 205 [39]), als auch zu seiner psychologischen Theorie vom Ursprung der Religion aus der Angst (Diels, Vorsokr. 2 S. 365 Nr. 75), wie, endlich zur ganzen Tendenz seiner Lebensanschauung, Vgl. Plat. Prot. 322 D : αίδοϋς καί δίκης μετέχειν. ) Κακοπραγμοσύνη („nequitia, improbitas" Stephanus) zuerst bei Dem. Or. 25, 101; d a n n P o l y b . IV. 23, 8; beidemal ,Bosheit'; ebenso κακοπράγμων boshaft Xen. Hell. V. 2, 36. 2 Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 549 dem Ziel der εύ&υμίη und άθαμβίη und der Befreiung von allem abergläubischen Wahn, die, durch Epikur vermittelt, uns noch in dem Gedicht des Lucretius (III. 991 if.) so pakkend entgegentritt. Besonders gehören hierher v. 1009 — 1021, wo „metus in vita poenarum pro male factis" (1012) und „ m e n s s i b i c o n s c i a f a c t i s praemetuens (1016 f.) als die eigentlichen Schöpfer der Hadesfabeln bezeichnet werden. Je weniger ich mich Rohdes radikaler Verwerfung der Ethika des Demokrit anschließen kann, um so mehr muß ich gestehen, daß mir seine Ausführungen über die Ungeschiclitlichkeit der Person des L e u k i p p o s durch die bisherigen Entgegnungen yon Diels und Dyroff 3 ) keineswegs widerlegt erscheinen, obwohl fast die ganze gelehrte Welt sich gegen ihn erklärt hat. Was hauptsächlich für die Existenz des Leukippos vorgebracht wurde, die Abweichungen in der Meteorologie, z. B. der Erklärung des Gewitters, sind so irrelevante Dinge, daß ihnen gegenüber ein Hinweis auf die zahlreichen Widersprüche in der Seelen-, Ideen- und Staatslehre Piatos genügt, die doch keinen Menschen veranlaßt haben, ihm die Schriften, die solche von einander abweichende Ansichten enthalten, abzusprechen. Der Vergleich des Corpus Democriteum mit dem Corpus Hippocratenm aber (Diels, Vorsokr. 2 S. 711) hinkt gewaltig: bei dem letzteren liegt der oft wiederholte Fall vor, daß Schriften weniger bekannter Schüler unter dem Namen des berühmten Meisters laufen ; bei dem Corpus Democriteum aber wäre das umgekehrte anzunehmen : die epochemachenden Schriften des Meisters hätten sich unter denen seines Schülers verloren, wären von dessen umfangreicher Schriftstellerei sozusagen absorbiert worden. Doch mit solchen allgemeinen Erwägungen läßt sich freilich nicht viel ausrichten. Was aber von Rohdes Aufstellungen bis jetzt nicht widerlegt werden konnte, ist folgendes : 1) Die Lehre Demokrits, d. h. die atomistische Philosophie, wäre ein bloßer Abklatsch von der des Leukippos : ein Fall, der sonst nirgends, weder bei den Milesiern noch bei den Ele8 ) Diels, Verb, der 35. Philologenvers. 1880 S. 96 ff. und Rhein. Mus. 42 (1887) S. 1 ff. — Dyroff, Demokritstudien 1899. — Rohde, Kl. Sehr. I 205 ff.; 2-45 ff., wozu Crusius in der Biographie S. 134. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 550 W. N e s t l e , aten noch etwa bei Anaxagoras und Archelaos im Verhältnis von Lehrern zu Schülern nachweisbar ist. 2) Leukippos ist als Persönlichkeit völlig unfaßbar; er ist nichts als ein zwischen Milet, Abdera und Elea hin und her flatternder Name. 3) Demokrit hat sich nirgends auf Leukippos berufen und — was zu diesem negativen testimonium ex silentio das positive Komplement bildet — an einer Stelle, wo man die Nennung seines Lehrers erwarten müßte, vergleicht er sein Alter nicht mit dem des Leukippos sondern mit dem des AnaXagoras (fr. 5 Diels). 4) Es ist nicht darüber hinwegzukommen, daß der der peripatetischen Schule angehörige Verfasser der Schrift De Melisso, Xenophane, Gorgia mit dem Ausdruck „έν τοις Λευκίππου κχλουμένοις λόγοις" (cap. 6 p. 980a 7) Schriften unter dem Namen des Leukippos meinte und seinen Zweifel an dessen Autorschaft ausdrücken wollte, daß also die Leukippfrage schon in der peripatetischen Schule selbst kontrovers war. 5) Die Atomistik ist jünger als das System des Anaxagoras. Diesen letzten Satz hat namentlich Brieger mit sehr einleuchtenden Gründen gestützt, indem er nachwies, daß „das atomistische System durch Korrektur des anaxagoreischen entstanden" sei. Diese Korrektur lag 1) in der Anerkennung des Leeren, 2) in der Aufhebung der unendlichen Kleinheit der Urkörper und ihrer qualitativen Bestimmtheit, 3) in der Ersetzung des Urbewegers Νους durch den Wirbel der Atome 4 ). Zu dem dritten Punkte möchte ich, ohne die Leukippfrage in ihrem ganzen Umfang hier aufzurollen, noch etwas hinzufügen. Diels weist dem Leukippos die beiden Schriften Μέγας διάκοσμος und Περί Νου zu und hält die letztere für eine Psychologie. In der Tetralogienordnung des Thrasyllos (Diels S. 357 Z. 24) steht ja auch bei der Gruppe, in der περί Νου als Schrift des Demokrit erscheint : ταϋτά τίνες ¿μου γράφοντες περί ψυχής έπιγράφουσι. Zunächst erscheint es mir aber fraglich, ob jemals in der griechischen Litteratur eine Schrift über die Seele ,περί νου' betitelt worden ist und be4 ) Brieger im Hermes 36 (1901) S. 161 ff., der S. 168 auch die künstliehe Erklärung der Λευκίππου καλούμενοι λόγοι unter Heranziehung von Aristot. Gen. et cori·. I 8 p. 325 a 1 zurückweist. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 55J titelt werden konnte. Das einzige Fragment ans dieser Schrift lautet: ουδέν χρήμα μάτην γίνεται άλλα πάντα έκ λόγου τε και υπ' ανάγκης (Lenk. fr. 2. Diels). Dies ist ein Gedanke, den man viel weniger in einer Psychologie als in einer Kosmologie suchen würde, und in solch kosmologischem Zusammenhang wird er auch zwar nicht dem Leukippos, wohl aber dem Demokrit zugeschrieben (Diels, Vors. 2 S. 364 Nr. 68). Dazu kommt noch die Nachricht des Diog. L. IX. 34 f. (Demokr. f r . 5), daß Demokrit, der den Anaxagoras in seinen astronomischen Theorien des Plagiats beschuldigte 5 ), διασύρειν τε αότοΰ τα περί της διακοσμήσεως και του νου εχ-9-ρώς έχοντα προς αύτον, ότι δή μή προσήκατο αυτόν. Sehen wir von der Begründung ab, die wie Klatsch aussieht, so geht daraus hervor, daß Demokrit gegen die Nuslehre des Anaxagoras polemisierte. Auf Grund dessen wage ich die Vermutung : die Schrift Περί Νου war das Werk des Demokrit und richtete sich gegen den Νους des Anaxagoras. Sie muß daher vorwiegend kosmologischen Inhalts gewesen sein, wozu das erhaltene Bruchstück stimmt, und wird Psychologie nur soweit enthalten haben, als die Nuslehre des Anaxagoras solche auch enthielt und daher zur Widerlegung herausforderte. Dies war wohl der Grund, warum sie Thrasyllos unter die anthropologischen Schriften einordnete. Gerade der Ausdruck διακάσμησις (nicht διάκοσμος), den Diogenes gebraucht, weist ja auf die Tätigkeit des Anaxagoreischen Νους (fr. 12) deutlich hin. Mir, der ich an die Existenz des Leukipp nicht glaube und in diesem Namen mit Tannery 6 ) ein Pseudonym des jungen Demokrit zu sehen geneigt bin, stellt »sich also das Verhältnis der 3 atomistischen Hauptschriften folgendermaßen dar. Alle drei sind von Demokrit verfaßt und zwar enthielt 1) der Μέγας διάκοσμος die positive Darlegung des atomistischen Systems, insbesondere die Kosmologie, 2) der Μικρός διάκοσμος nach Rohdes Vermutung die Psychologie, 3) Περί 5 ) Daß Anaxagoras dies Plagiat gerade an Leukippos begangen habe, wird nirgends gesagt. Der Ausdruck άρχαΐαι scheint eher auf ein höheres Alter der in Frage stehenden Lehren περί ηλίου καΐ σελήνης hinzuweisen. 6 ) Revue des Éludes Grecques X (1897) p. 127 ss. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 552 W. N e s t l e , Νου eine speziell gegen die Nuslehre des Anaxagoras gerichtete Polemik. Protagoras. Von Beziehungen der beiden Abderiten Demokrit und Protagoras zu einander wußte schon das Altertum. Die Legende machte den Sophisten zum Schüler des Philosophen (Diog. L. IX. 53). Viel mehr Glauben verdient eine Notiz des Plutarch und Sextus Empirikus (Diels 2 531, 15 und fr. 156), daß Demokrit mit seiner Erkenntnistheorie dem Skeptizismus des Protagoras entgegengetreten sei. Dies hat Brochard im Einzelnen verfolgt 7 ). Aber nicht nur polemische Beziehungen lassen sich zwischen beiden Männern feststellen, sondern auch einige auffallende Uebereinstimmungen in ihren ethisch-politischen Anschauungen. Schon Chiapelli hat außer auf den Gebrauch von κατάστασις bei Protagoras (D. L. IX. 55) und Demokrit (fr. 278) auf einige Aehnlicbkeiten in der Schätzung von Naturanlage und Erziehung hingewiesen : vgl. Demokr. fr. 33. 182. 242 mit Protagoras bei Plato S. 323 D. 8 ). Und wenn Protagoras seine Wissenschaft als ευβουλία περί τε των οικείων, δπως αν αρ:στα την αύτοΰ ΰίκίαν διοικοί και περί των της πόλεως κτλ. bezeichnet (S. 318 Ε), so füllte zwar das erstere den Begriff der Demokritischen ευθ-υμίη oder εύεστώ nicht aus, gehörte aber doch auch dazu: sonst hätte man das letztere W o r t nicht mit εδ έστάνac τον οίκον (fr. 140) umschreiben können. Endlich schließt Protagoras, der dem νόμος zwar keine absolute, aber eine relative Gültigkeit innerhalb des einzelnen Gemeinwesens zuschrieb (Plat. Theaet. p. 167 C; 172 A), den ihm in den Mund gelegen Mythus mit dem Auftrag des Zeus an Hermes : κα: νόμον γε θες παρ' έμοΰ τον μή δυνάμενον αίδοΰς και δίκης μετέχειν κτείνειν ώς νόσον πόλεως (Prot. S. 322 D). Diese leidenschaftliche Aufforderung zur Vernichtung des Verbrechers als eines gemeinschädlichen Wesens hat ihre schlagende Paral') Archiv für Philosophie II (1889) S. 368 ff. Vgl. noch den Gebrauch von Καταβϋίλλε!ν bei Demokrit fr. 125 mit den ,Καταβάλλοντες' des Protagoras. 8 ) Per la storia della Sofistica Greca im Archiv f ü r Philos. III (1890) S. 15 Α. 43—45. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Yorsokratikern und Sophisten. 553 lele in einigen politischen Fragmenten Demokrits, wo der Verbrecher, den Protagoras als eine Pest oder ein Krebsgeschwür des Staates bezeichnet, mit schädlichem Getier auf eine Stufe gestellt wird, das zu vernichten jedermann das Recht haben sollte (fr. 257—260). Sogar der Ausdruck μετέχειν kehrt bei Deinokrit in diesem sozialen Sinn mehrfach wieder (vgl. oben fr. 258. 263)°). Daß aber der platonische Mythus wirklich Gedanken des Protagoras enthält, wenn er auch keine sklavische Nachahmung seiner Schrift περί της έν άρχη καταστάσεως ist, dafür bietet sich noch von anderer Seite ein Anhaltspunkt. Meines Wissens hat noch niemand auf folgende frappante Aelinlichkeit einer Stelle dieses Mythus mit einer solchen bei Herodot geachtet: Prot. 321 B. Herod. III. 108. εστι δ' ο!ς εδωκεν (sc. der καί κως του ·8·είου ή προvon Prometheus überwachte νοίη, ώσπερ καί οίκος έστι, Epimetheus) είναο τροφήν ζώ- έοΰσα σοφή, οσα μεν γε ψυχήν ων άλλων βοράν. και τοις μεν τε δειλά και εδώδιμα, ταΰτα ό λ ι γ ο γ ο ν ί α ν προσήψε, τοις μεν πάντα π ο λ ύ γ ο ν α πεδ' άναλισκομένοις όπο τούτων ποίηκε, ίνα μη έπιλίπη καπ ο λ υ γ ο ν ί α ν σωτηρίαν τω τεσθ-ιομένα, οσα δέ σχέτλια γένεο πορίζων. καί άνιηρά, ολιγόγονα. Daß die Uebereinstimmung dieser beiden Stellen auf Zufall beruhe, wird schwerlich jemand annehmen wollen. Wir stehen also vor folgenden Möglichkeiten : 1) entweder hat Plato aus Herodot geschöpft oder 2) Plato und Herodot haben beide aus Protagoras geschöpft oder 3) Plato-Protagoras und Herodot gehen auf eine gemeinsame dritte Quelle zurück. Die erste Möglichkeit hat wenig Wahrscheinlichkeit, da Plato kein historisches Interesse hatte. Die dritte ist nicht ganz ausgeschlossen, wie denn Stein bei der Herodotstelle an Anaxagoras gedacht hat 1 0 ). Aber obwohl Anaxagoras die teleologische Idee aufgebracht hat, macht es ihm ja eben der platonische Sokrates (Phaid. p. 98 B) zum Vorwurf, daß er diese nicht im Einzelnen durchgeführt habe, und gerade darauf kommt 9 ) Zu αιδώς vgl. auch die fr. 84. 244. 264 des Demokrit über αίσχύνεσθ-οα und αΐδεΐσθαι. 10 ) Stein, Vorrede 5 p. XXX, 8 und z. St. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM W. N e s t l e , 554 es liier ail. Eher könnte man daher an Diogenes von Apollonia erinnern, von dessen Teleologie (fr. 3—5. 8) Dümmler Spuren in Xenophons Memorabilien (I. 4, 6 if.) finden wollte 11 ). Auch hat ja Herodot (II. 24 f.) seine Erklärung der Nilschwelle (Diels S. 830, 18) übernommen 12 ). Trotzdem liegt aber die zweite Möglichkeit am nächsten; denn Herodot hat ganz gewiß den abderitischen Sophisten im Kreis des Perikles kennen gelernt, haben sich doch beide Männer an der von diesem unternommenen Ivoionisation von Thurioi beteiligt. Wir werden also die Herodotstelle als ein Zeugnis für die Genuinität der in dem platonischen Mythus vorgetragenen Gedanken des Protagoras betrachten dürfen 1 2 3 ). Fr. 1. Das berühmte erste Bruchstück des Protagoras ist uns in doppelter Form überliefert. Bei Sext. Emp. Adv. Math. VII. 60 lautet es: πάντων χρημάτων μέτρον εστίν άνθρωπος, των μεν όντων ώς εστίν, των δέ ο υ κ δντων ως ουκ εστίν. Bei Plato Theaet. S. 152 A dagegen lesen wir: των δέ μ ή όντων, ώς ούκ εστον. Ganz entsprechend der platonischen Ausdrucksweise sagt Xenophon (An. IV. 4, 15) von einem zur Rekognoszierung ausgesandten Manne : οδτος γαρ έδόκεί και πρότερον πολλά ήδη άληθευσαι τοιαύτα, τά δντα τε ώς δντα και τα μγ] όντα ώς ούκ όντα. Dieser Satz bildet auch insofern ein Seitenstück zu dem des Protagoras, als hier wie dort in dem ώς die Bedeutung ,daß' und ,wie' zugleich steckt (Zeller, Phil. d. Gr. 5 I. 1094, 1). Denn wie der rekognoszierende Offizier sowohl über das Vorhandensein als auch über die Beschaffenheit dessen, was er wahrnimmt, zu berichten hat, so geht auch der Satz des Protagoras zugleich auf die Existenz und die Eigenschaften der Dinge. Ich suchte in der Uebersetzung diesem Doppelsinn gerecht zu werden durch den Ausdruck: „wofern sie sind", bezw. nicht sind, und glaube damit auch dem hypothetisch beschränkenden Charakter des Partizipiums των δέ μη όντων, das ich für die getreuere und richtigere Ueberlieferung halte, ") Akademika S. 103; III ff. y ') B. Meyer, Gesch. des Altertums IV. 108. ) Weiteres in meinem Programm: Herodots Verhältnis zur Philosophie und Sophistik. Sehöntal 1908 S. 16 f. 1?a Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 555 R e c h n u n g getragen zu haben. ·— Endlich darf zu diesem Satz des Protagoras auch Demokrit f r . 165 (Diels 2 ) verglichen werden : λέγω τάδε περί των ξυμπάντων . . . . άνθρωπος έστιν, ο πάντες ϊδμεν. Der Ausgangspunkt der E r k e n n t n i s theorie ist f ü r die beiden Abderiten der Mensch, das eigene Ich als das, was uns zunächst bekannt ist. D a n n aber scheiden sich die W e g e : Protagoras verfällt einem individualistischen Skeptizismus, D e m o k r i t findet im Denken, der γνησίη γνώμη, das Regulativ der sinnlichen W a h r n e h m u n g , der σκοτίη γ ν ώ μ η , deren Objektivität er mit Protagoras preisgibt. F r . 11. Zu dem aus der syrischen Uebersetzung der pseudoplutarchischen Schrift περί ασκήσεως stammenden B r u c h s t ü c k : „Nicht sproßt Bildung in der Seele, wenn man nicht zu vieler Tiefe k o m m t " , bemerkt Diels 2 S. 5 4 0 : - B i l d vom G ä r t n e r " . W a r u m nicht ü b e r h a u p t vom L a n d m a n n ? Das Bild vom L a n d m a n n g e b r a u c h t P r o t a g o r a s vom Weisen bei Plato Theaet, 167 Β C : και του; σοφούς, ώ φίλε Σώκρατες, πολλού δέω βατράχους λέγειν (vgl. 161 C), άλλα κατά μεν σώματα ιατρούς λέγω, κατά δε φυτά γ ε ω ρ γ ο ύ ς · φημί γαρ και τούτους τοις φυτοϊς αντί πονηρών αισθήσεων, όταν τι αύτων άσθ-ενη, χρηστάς και ύγιεινάς αισθήσεις τε και άληθείας έμποιείν, τους δέ γε σοφούς τε και αγαθούς Ρήτορας ταϊς πόλεσι τα χρηστά αντί των πονηρών δίκαια δοκεϊν [είναι] ποιεϊν. Dasselbe Bild bei Antiphon fr. 60 : και γαρ τη γη οίον αν τις το σπέρμα έναρόση, τοιαύτα και τα εκφορά δει προσδοκαν και έν νέω σώματι οταν τις τήν παίδευσιν γενναίαν έναρόση, ζη τοϋτο και θάλλει δια παντός του βίου και αύτό ούτε ομβρος ουτε ανομβρία αφαιρείται. A u c h P l a t o Resp. VI. 6 S. 492 A geh ö r t h i e r h e r : ην τοίνυν εθεμεν του φιλοσόφου φύσιν, άν μέν, οίμαι, μαθήσεως προσηκούσης τύχη, εις πδσαν άρετήν άνάγκη αυξανομένην άφικνείσθαι, εάν δέ μή έν προσηκούση σπαρεϊσά τε και φυτευθεϊσα τρέφηται, είς πάντα τάναντία αδ, εάν μήτις αύτη βοηθήσας θεών τύχη. Gomperz (Griech. Denker 1 S. 354) verweist noch auf Matth. 13, 5. Prodikos. Aus dem Mythus von Herakles (Xen. Mem. II. 1, 2 1 ff.) liest D ö r i n g (Gesch. der griech. Phil. I. 832 und 334) eine Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 556 W. N e s t l e , Verherrlichung der Geburtsaristokratie heraus und auf den ersten Blick kann man ja dies in den Worten der 'Αρετή, die sich für ihre auf Herakles gesetzten Hoffnungen u. a. auch auf dessen Eltern beruft (είδυΐα τους γεννήσαντάς σε) finden; aber sie setzt gleich hinzu : καί τήν φύσιν τήν σήν έν παιδεία καταμαθοϋσα (27). Also hätten wir mindestens drei Voraussetzungen für die Tüchtigkeit: Abstammung, eigene Naturanlage und Erziehung. Doch abgesehen davon widerstrebt der ganze Mythus dieser Auffassung. Das Wesentliche der aristokratischen Gesinnung, wie sie etwa Theognis vertrat, ist ja eben, daß die αρετή dem Edlen in die Wiege gelegt wird und zwar eine αρετή, deren Ausübung zur Voraussetzung hat, daß andere, die Menge, zugunsten der adligen Minorität und ihrer Muße arbeiten. Die Hauptidee des Mythus aber liegt, ganz im Sinne Hesiods, des Gegners der adeligen Müßiggänger und Herolds der Arbeit, darin, daß των όντων άγα·9·ών και καλών ουδέν ανευ πόνου και επιμελείας οί ·9·εοϊ διδόασιν άνθ-ρώποις (28). Dazu kommt, daß sich die Αρετή ausdrücklich auch als εύμενής παραστάτις οικέταις (32) bezeichnet; m. a. W. auch die Sklaven können άρετή haben. Von der in weiten Kreisen verbreiteten Verachtung der βάναυσοι haben wir hier keine Spur. Darum braucht freilich Prodikos kein Gegner oder Verächter der Vornehmen gewesen zu sein und er mag wohl, wie Philostratos berichtet (Vit. soph. 12), dier vornehme Jugend an sich zu ziehen gesucht haben. Aber wenn ihm, wie der Sophistik überhaupt, die Tugend „lehrbar" war (Eur. Hik. 913 f., wozu Welcker Kl. Sehr. II. 509), so wird ihm auch adlige Geburt, wie es ihn der ,Eryxias' an dem Beispiel des Reichtums dartun läßt, ein Gut gewesen sein, das erst durch den Gebrauch, den man davon macht, gut oder schlimm, nützlich oder schädlich wird (Eryx 13 f. S. 397 E bis 398 D). Fr. 5. Cicero und Sextus Empiricus schreiben dem Prodikos die Lehre zu, die Religion sei in ihrer frühesten Form Verehrung der das menschliche Leben fördernden Dinge. Weniger deutlich drücken sich Themistios und Persaios aus. Ueber letzteren sagt Philodem, de piet. 9, 9 : ΙΙερσαΐος δέ δήλος έστιν . . . . άφανίζων το δαιμόνιον f¡ μηθ·εν υπέρ αύτοΰ Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 557 γινώσκων, όταν έν τώ Περί θεών μη άπίθανα λέγη φαίνεσθ·αι τά περί [του] τά τρέφοντα και ώφελοΰντα θ-εους νενομίσ9·αι και τετειμησθ-at. πρώτον υπό ΙΙροδίκου γεγραμμένα, μετά δέ ταΰτα τούς ευρόντας η τροφάς ή σκέπας ή τάς ά'λλας τέχνας ώς Δήμητρα και Διόνυσον και τους . . . . E s f r a g t sich hier, bis wohin die Worte des Persaios sich auf Prodikos beziehen, und dies hängt wieder ab von der Auffassung der W o r t e πρώτον und μετά ταΰτα. E s scheint zunächst, als habe man hier eine Kombination der vorhin bezeichneten Lehre des Prodikos mit dem späteren Euliemerismus vor sich und in der Tat schreibt Cicero (De nat. deor. I. 42, 118) nur den ersten Teil des Inhalts des obigen Satzes dem Prodikos, den zweiten (ib. I. 15, 38) dem Persaios zu. Demnach wäre πρώτον zu γεγραμμένα zu ziehen, mit μετά ταΰτα wäre eine spätere Zeit als die des Prodikos gemeint und der damit beginnende Satzteil wäre nicht mehr von γεγραμμένα abhängig zu denken, sondern man müßte ein zweites dem γεγραμμένα entsprechendes Partizip mit davon abhängigen Infinitiven ergänzen, also etwa υπ' άλλων είρημένα sc. νενομίσ8·αο και τετειμησΰ-αι. Allein dann wäre eine ganz andere Gliederung der Periode zu erwarten : E s müßte dann heißen : και τά μετά ταΰτα (oder vielmehr τοΰτον) περί του τους εύρόντας . . . . νενομίσ-θ-αι . . . . είρημένα. Wie künstlich das alles ist, liegt auf der Hand. Jede Schwierigkeit aber schwindet, wenn wir πρώτον zu den vorhergehenden Infinitiven ziehen und μετά ταΰτα auf τά τρέφοντα beziehen ; höchstens könnte man bei πρώτον ein μέν vermissen. Also : „ Es ist klar, daß Persaios die Gottheit znnichte macht oder auf jede Erkenntnis von ihr verzichtet, wenn er in seinem Buch über die Götter sagt : nicht unglaubhaft erscheine die Darstellung des Prodikos, daß zuerst die zur N a h r u n g dienenden und nützlichen Dinge f ü r Götter gehalten und verehrt worden seien, danach aber die Erfinder von Nahrungs- und Schutzmitteln oder Kunstfertigkeiten, wie Demeter, Dionysos und . . . Diese durch den griechischen Text, wie mir scheint, unausweichlich gegebene Auffassung wird nun aber gestützt durch die von Diels nicht angeführte lateinische Paraphrase bei Minucius Felix Oct. 21, 2 : „Prodicus adsumptos in deos loquitur, qui errando inventis novis Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM W. Ν e s 11 e , 558 frugibus utilitati hominum profuerunt. In eandem sententiani et Persaeus philosophatur et adnectit inventas fruges et frugum ipsarum repertores isdem nominibus" etc. Krisclie (Die theolog. Lehren der griech. Denker S. 440 ff.) zeiht zwar den Minucius einer Verwechslung mit Euhemeros, aber, wie gezeigt, mit Unrecht. Das ,errando' deutet auf Götter hin, die ihre Wohltaten durch Uniherstreifen auf der Erde verbreiteten, wie Demeter, Dionysos und wohl auch der große ευεργέτης Herakles. Prodikos hat also zwei Religionsstufen unterschieden : 1) die fetischistische Verehrung der τρέφοντα und ώφέλιμα selbst und 2) die Verehrung ihrer mutmaßlichen ,Erfinder' als persönlich gedachter Götter. Für den ersten Teil seiner Theorie mag ihm der seit Alters übliche metonymische Gebrauch zahlreicher Götternamen, für den zweiten der griechische Heroenkult die Anregung gegeben haben. Prodikos ist demnach der erste Denker, bei dem wir nicht nur eine Theorie über den Ursprung der Religion finden, wie bei Xenophanes, Kritias und Demokrit, sondern auch den Gedanken einer Entwicklung der religiösen Vorstellungen. Nach Sextus berief er sich für den ersten Teil seiner Theorie u. a. auf die Verehrung des Nils in Aegypten, von der Herodot II. 72 und 90 berichtet. Aber auch der zweite Teil scheint mir an einer Stelle desselben Buches durchzuschimmern (II. 146) : ει μέν γαρ φανεροί εγένοντο και κατεγήρασαν και οΰτοι εν τη Ελλάδι, κατά περ Ήρακλέης ó εξ Άμφιτρύωνος γενόμενος, και δή και Διόνυσος ó έκ Σεμέλης καI Πάν ó έκ Πηνελόπης γενόμενος, εφη αν τις καί τούτους άλλους γενομένους άνδρας εχειν τά εκείνων ούνόματα των προγεγονότων θεών. Herodot sieht in Herakles hier und sonst nur einen Menschen, nicht den Sohn des Zeus (Wipprecht, Rationalist. Mythendeutung I 89), und er kombiniert hier seine Theorie von der Herkunft der griech. Götternamen aus Aegypten wenigstens hypothetisch mit der Lehre des Prodikos von Dionysos u. a. zu Göttern erhobenen Menschen. Auch Euripides verwendet (Bacch. 274 ff.) die Lehre des Prodikos 13). 13 ) Vgl. Norwood, The riddle of the Bacchae (1908) S. 27 f.; 109 f.; 122 f. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 559 Thrasymachos. Fr. 1. In dem Bruchstück einer symbuleutischen Musterrede, welche die Kämpfe um die πάτριος πολιτεία zum Hintergründe hat, betont Thrasymachos, der sich (nach Cic. De or. III. 32, 128) auch mit Physik beschäftigte, sehr scharf die Verantwortung der Regierung für die vorgefallenen συμφοραί und will nichts davon wissen, daß die Götter oder die τύχη vorgeschoben werden. Dies ist ganz genau im Sinn des Demokrit (fr. 119 und 175 D. = 50. 58 N.) gesprochen. Ob eine bewußte Beziehung vorliegt, lasse ich dahingestellt, ebenso wie bei der auffallenden Aehnlichkeit der Schlußworte unseres Bruchstücks mit Antiphon von Rhammus, De mort. Her. 71, wofern an ersterer Stelle mit Sauppe μνήμης statt des überlieferten γνώμης zu lesen ist. Fr. 8. Οι θεοί ούχ όρώσι τα ανθρώπινα · ού γαρ αν το μέγιστον των έν άνθρώποις αγαθών παρεΐδον την δικαιοσύνην · όρώ μέν γαρ τους ανθρώπους ταύτη μή χρωμένους. Gegen diese Meinung, die έφα τις, protestierte schon Aischylos Ag. 369 ff. Vgl. ferner Eurip. El. 583 f., Xenoph. Mem. I. 4, 11 ff. und Plato Ges. X. 889 f., wo 714 C ohne Namen auch die aus dem Staat (I. 338 C ; 344 C) bekannte thrasymacheische Definition des δίκαιον als το του κρείττονος συμφέρον erscheint. Im Staat I. 350 E läßt Plato den Sophisten die Mythen für Altweibergerede erklären. Gorgias. Fr. 8. Kai το αγώνισμα ημών κατά τον Αεοντϊνον Γοργίαν διττών [δέ] αρετών δείται, τόλμης και σοφίας · τόλμης μέν τον κίνδυνον υπομεΐναι, σοφίας δέ το πλίγμα (αίνιγμα Hss.) γνώναι. ó γάρ τοι λόγος καθάπερ το κήρυγμα το Όλυμπίασι καλεί μέν τον βουλόμενον, στέφανοι δέ τον δυνάμενον. Welche Worte gehören in diesem Zitat des Clemens Alexandrinus dem Gorgias ? Hat dieser in seiner olympischen Rede nur von dem wirklichen Wettkampf geredet oder von dem λόγος, bezw. άγών λόγων, der Redekunst, die er mit einem Wettkampf verglich (wie bei Plato Gorg. 456 Β—457 C) ? Nimmt man das letztere an und versteht unter dem λόγος die Rhetorik, so muß man folgerichtig auch ήμών zu dem gorgianischen Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 560 W. N e s t l e , Text ziehen und τό αγώνισμα ήμών wäre dann eben der άγων λόγων. Gegen diese Erklärung spricht aber die Erwägung, daß die Rhetorik wohl allenfalls mit άγώνισμα, nicht aber mit einem κήρυγμα, wie im letzten Satze, verglichen werden kann. Dagegen paßt dieser Vergleich gut auf den christlichen λόγος, von dem Clemens in dem auf das Zitat folgenden Satze redet. Dann gehört aber auch ήμών nicht dem Gorgias sondern dem Clemens. Somit kann Gorgias nur von dem wirklichen Wettkampf geredet und die ihm entlehnten Worte werden so gelautet haben : και tò άγώνισμα διττών αρετών δείται — γνώναι · τό γαρ κήρυγμα το Όλυμπίασι καλεί — δυνάμενον. Die Echtheit des erhaltenen ,Lobes der H e l e n a ' wird heute von den meisten Gelehrten anerkannt 1 4 ). Besonders ist Blaß (Att. Bereds. 2 I. 72 ff.) dafür eingetreten, während Gomperz (Apol. der Heilkunst in den Wien. Sitzungsber. 120. 1890 S. 165 f. und Griech. Denker 1 I. 475 f.) allerdings noch immer den von Spengel (Συναγ. τεχν. 1828 p. 73 s.) geltend gemachten Einwand für ausschlaggebend hält, wonach in der ,Helena' des Isokrates zwischen dem Prooimion, wo Gorgias (§ 3) zu den längst verstorbenen Sophisten gerechnet werde, und § 14, wo der Verfasser der uns vorliegenden , Helena' als noch lebend gedacht werde, ein unlösbarer Widerspruch entstehen würde, wenn dieser Gorgias wäre. Diese Schwierigkeit ist aber nicht unüberwindlich. Gewiß waren zur Zeit, da Isokrates seine Helena schrieb, Protagoras, Zeno und Melissos tot. Gorgias aber, der 109 Jahre alt geworden sein soll und den Plato (Phaidr. 261 C) mit dem greisen Nestor vergleicht, braucht darum noch nicht gestorben gewesen zu sein, weil er •— und zwar wegen seiner radikalen Skepsis, deren litterarische Darlegung in seine Jugend fiel — mit jenen andern Männern zusammen genannt wird. Isokrates spricht ferner gar nicht bloß von dem Unterschied zwischen Einst und Jetzt in der Sophistik sondern auch von der Wahl würdiger und unwürdiger Gegenstände für die rhetorischen l4 ) Vgl. E. Maaß im Hermes 22 (1887) S. 572 ff., dem ich freilieh darin nicht beistimmen kann, daß isokrates die Helena des Gorgias ignoriere. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 561 Musterleistungen und es erscheint als eine pietätvolle Anerkennung gegenüber seinem Lehrer, wenn er ihn zu den ευ λέγειν βουληθέντες rechnet und nur an dem Titel des Werkchens eine leichte Kritik übt, da es ihm eher als eine V e r teidigung' oder (um mit Lessing zu reden) „Rettung" denn als eine Lobrede der Helena sich darstellt : eine Charakteristik, die auf die erhaltene Rede genau paßt. Ich möchte nun aus dem Inhalt der Rede eine Anzahl von Gründen für ihre Echtheit nachweisen. Zunächst entspricht die Art der Beweisführung genau der Schilderung, die Plato (Pkaidr. 267 A) von der Kunst des Tisias und Gorgias gibt: τά τε σμικρά μεγάλα και τά μεγάλα σμικρά φαίνεσθ-αι ποιοΰσι δια ρώμην λόγου16). Denn der Verfasser sucht die delHelena zugeschriebene Schuld möglichst zu verkleinern, die sie bestimmenden Motive als möglichst stark und unüberwindlich hinzustellen: εγώ δέ βούλομαι λογισμόν τινα τω λόγω δους την μέν κακώς άκούουσαν παϋσαι της αιτίας , τους δέ μεμφομένους ψευδομένους έπιδείξας και δειξας τάληθές παΰσαι της άμαθ-ίας (§ 2). Wie der Abschnitt vom λόγος (§ 8 ff.) dem Gorgias aus der Seele gesprochen ist, hat schon Maaß (a. a. 0 . 574 f.) in aller Kürze angedeutet; doch läßt sich gerade hiezu noch Weiteres beibringen. Diese Ausführungen sind das schlagendste Beispiel für die gorgianische Definition der Rhetorik als der πειθ·οΰς δημιουργός (Plat. Gorg. 453 A). Denn es wird hier gezeigt, wie die Rede die Seele nach ihren Absichten formt (τυποΰται 13. 15). Es wird ferner mit dem Begriff der άπατη (8. 10. 11) operiert. Nun wissen wir, daß Gorgias diesen tatsächlich behandelt hat (fr. 23), und wenn wir die Dialexeis (3, 10) hinzunehmen, so läßt sich sein Grandgedanke ganz genau rekonstruieren. Es gibt danach eine απάτη δικαία, ζ. Β. die poetische Illusion, und eine απάτη άδικος, den wirklichen Betrug. Gerade diese beiden Wirkungen des λόγος werden nun in der Helena hervorgehoben : zuerst die durch die Poesie erregten Affekte (8 f.), wobei offenbar insbesondere die Tragödie vorschwebt und der Gedanke, daß Mitleid und Furcht in der Seele um fremder Erlebnisse willen erregt wer15 ) Vgl. Isokr. Paneg. (4) 8; Cic. Brut. 12, 47. 82 (1897) S. 343 f. Philologue L X V I I (ÎT.P. XXI), 4. Gercke im Hermes 36 Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 562 W. N e s t l e , den, schon wie ein Ansatz zu der bekannten Katharsislehre des Aristoteles erscheint 10 ); dann im Anschluß daran die έπωδαί und ihre auf dem ihnen entgegengebrachten Glauben beruhende Wirkung ; schließlich die άπάτη, die sich die Unwissenheit der Menschen zunütze macht. Für die Worte έπωοαί, γοητεύειν (10. 14) haben auch die Gorgiasschüler Kallikles und Menon bei Plato (Gorg. 483 E ; 484 A ; Men. 80 A) eine Vorliebe 17 ) und man mag sich dabei erinnern, daß Gorgias Schüler des Empedokles war, den er angeblich selbst zaubern sah (Diog. L. VIII. 59; Emped. fr. 111. 112, 10 f.). Ebenso sind die ανάγκης ψηφίσματα (Hei. 6) eine bei Empedokles (fr. 115, 1) vorkommende Wendung, worauf schon Dümmler (Ak. 36, 1) aufmerksam gemacht hat 1 8 ). Doch, um zum λόγος zurückzukehren, so muß sich ein Redner, der wirken will, auch auf allgemeine Bildung verstehen: er muß in der Astronomie bewandert sein, wie denn Gorgias auf dem Grabmal des Isokrates in einen Himmelsglobus vertieft dargestellt war (Diels 2 S. 548 Nr. 17), ferner auf αγοραίους (conj. Diels, αναγκαίους Hss.) δια λόγων αγώνας, öffentliche Reden vor dem Gericht und in der Volksversammlung, und auf philosophische Dialektik (13). Bei dem letzteren Punkt denkt Dümmler an die „Dialoge der Sokratiker" (Ak. 35); aber es kann ebensogut die Eristik eines Zeno, die Gorgias die Waffen für seinen Angriff auf die Naturphilosophie widmete, und der Sophisten, wie Protagoras und anderer, gemeint sein ; zu dem zweiten Punkt vgl. Plato Gorg. 452 E, 456 Β und besonders 458 E und 459 A, wo das von Gorgias gebrauchte Wort όχλον (είς λόγος πολύν δχλον ετερψε Hei. 13) geflissentlich dreimal wiederholt wird 19 ). Zum Schluß wird die Wirkung der Rhetorik mit derjenigen von offizinellen Giften (φάρμακα) verglichen, die ebenso heilen wie töten können, ein Bild das im Theaet. 167 A wiederkehrt. So kann auch die Rhetorik gute und schlimme Wirkungen 1β ) Ygl. De diaet. I 24 (Diels8 S. 85), wozu Fredrich, Hippokr. Unters. S. 150. " ) Ygl. auch Plat. Menex. 234 C; Symp. 203 D. Dümmler, Akademika S. 31. 22. 39. 18 ) Ueber die Stilverwandtschaft des Gorgias und Empedokles vgl. Diels, Beri. Sitzber. 1884 S. 362 ff. 19 ) Vgl. Plato Soph. 232 Β—D. Gercke a. a. O. S. 352. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 563 ausüben. Genau diesen Gedanken, daß man von der Rhetorik sowohl guten als schlimmen Gebrauch machen könne, läßt auch Plato den Gorgias (456 C—457 C) ausführen und gibt damit offenbar die eigene Ansicht des Sophisten wieder, die der platonische Sokrates freilich nicht konsequent findet (457 E) und die in den ethisch so verschiedenen Richtungen seiner Schüler sich spiegelt : Proxenos will nur συν τω δοκαίω zu Einfluß, Ehre und Reichtum gelangen (Xen. An. II. 6, 16 ff.), Alkidamas spricht sich grundsätzlich gegen die Berechtigung der Sklaverei, Lykophron gegen die Privilegien des Adels aus, während Menon (Xen. An. II. 6, 21 ff.)20), Kritias und Kallikles skrupellos das Recht des Stärkeren proklamieren und ausüben. J a die 'Helena' zeigt uns gerade deutlich, daß Plato in der Tat im Recht war, wenn er den Gorgias wenigstens mittelbar für diese Theorie verantwortlich machte. Denn auch sie verkündet dies Naturgesetz (6) : πέφυκε γαρ ού το κρεόσσον οπό του ήσσονον κωλύεσθ'Οα άλλα το ήσσον όπό του κρείσσονος άρχεσ&αι και άγεσ&αο και το μεν κρείσσον ήγείσθαι, τό δέ ήσσον επεσθαι 21 ). Ein Unterschied zwischen der physikalischen und moralischen Welt wird nicht gemacht. Vielmehr wird am Schluß der ganzen Rede (15—19) noch ausgeführt, wie auch die Leidenschaft der Liebe eine Krankheit sei, der man ebensogut wie gewissen psychopathischen Halluzinationen unterliege: ει δ' εστίν ανθρώπων νόσημα και ψυχής άγνόημα, ο5χ ώς αμάρτημα μεμπτέον άλλ' ώς άτόχημα νομιστέον (19). Da nun auch die Redekunst zugestandenermaßen ein μέγας δυνάστης ist und die Beherrschung anderer Menschen zum Ziel hat (Men. 78 C; Gorg. 452 E), wer sie besitzt somit κρείττων των άλλων ist, so ist von da aus nur noch ein Schritt zu der Weltanschauung, die Kallikles mit Worten verkündigt, welche denen in der Helena merkwürdig ähnlich sehen : δίκαιον έστι τον άμείνω τοΰ χείρονος πλέον εχεον και τον δυνατώτερον του άδυνατωτέρου (483 D). Daß auch die von Gorgias in der Helena vorgetragene Lehre von den έρωτικαί άνάγκαι von Plato zurückgewiesen wird (Symp. 196 Β ; 197 Β) hat Dümmler (Ak. 2°) Sorof im Hermes 34 (1889) S. 568 ff. ->1) Zu θεών βουλεύμασι vgl. ψ 222: eine Stelle, die die Anregung zu diesem Gedanken gegeben haben könnte. 36* Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 564 W. N e s t l e , S. 38 f.) gezeigt. Mochte Gorgias diese Rede immerhin als ein παίγνιον (21) bezeichnen 22 ), er hat doch viel von seinen wirklichen Anschauungen für dieses Scherzspiel verwendet. Dem Ernst Piatons mußte ein solches Spiel als ein Spielen mit dem Feuer erscheinen. Alle diese Parallelen, welche die Helena zu Anschauungen des Gorgias bietet, die uns sonst, teils in den Resten seiner eigenen Schriften, teils bei Piaton und Xenophon bezeugt sind, sind m. E. geradezu entscheidend für die Echtheit der Rede. Es müßte wahrhaftig ein Künstler von einem Fälscher gewesen sein, der alle diese Beziehungen auf die wirklichen Ansichten des Gorgias mit solcher Geschicklichkeit und solch spielender Anmut in dieses παίγνιον zu verweben gewußt hätte. Und wie sollte man sich bei einer Fälschung vollends die von Dümmler erwiesenen Anspielungen Piatons erklären? Je mehr man in die Gedankensphäre eindringt, in der sich die Rede bewegt, desto mehr wird man von ihrer Echtheit überzeugt. Die Verteidigungsrede des P a l a m e d e s bildet zu der Helena das ergänzende Seitenstück. Gab Gorgias in der letzteren ein Beispiel vom άδίκως χρησθ·αι vT¡ ^ητορικη, so ist diese Rede eines unschuldig Angeklagten ein Exempel für das δικαίως χρήσθ-αι. Hier stehen sich δίκη und βία gegenüber (§ 2) und ringen um den Sieg. Auch Palamedes will πείθειν, aber es soll geschehen τω σαφεστάτω δικαίω, διδάξαντα τάληθ'ές, ούκ άπατήσαντα (§ 33; vgl. Hei. 10 f.); nur so trachtet er nach seiner Freisprechung. Dazu vergleiche man Plato, Phileb. 58 Α : ήκουον . . . εκάστοτε Γοργίου πολλάκις, ώς ή του πείθ-ειν πολυ διαφέροι πασών τεχνών πάντα γαρ ύφ' αυτή δοϋλα δι' έκόντων αλλ' ού δια βίας ποιοίτο. So heißen auch Pal. 30 die Gesetze φύλακες του δικαίου, das Gorgias nach Plato (Gorg. 456 E ; 457 Β ; 461 Α Β) nicht durch Gewalt (ίσχύϊ) vermittelst der Rhetorik verletzt wissen wollte. Ueber die Durchschnittsmoral erhebt sich Gorgias freilich nicht: übereinstimmend bezeichnen Plato (Men. 71 E), Xenophon (Anab. II. 6, 17 in der Charakteristik des Proxenos) und Palamedes in der vorliegenden Rede (§ 18. 25) als natürlichen Lebensgrundsatz, zu 22 ) Hiezu Gereke a. a. O. S. 355 f. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 555 dessen Durchführung man fähig werden will, τους μεν φίλους εύ ποιεϊν, χους δ' έχθ-ρους κακώς (bezw. ώφελείν — βλάπτειν), womit auch Gorg. fr. 21 im Einklang steht (vgl. Plato Staat I 16 p. 843 E) 2 3 ). Daß Palamedes als der ,Erfinder' aller möglicher Kulturgüter erscheint (30 f.), war zwar ein durch die Sage gegebenes Motiv; aber es verdient doch Beachtung, daß der Gorgiasschüler Polos seine rhetorische τέχνη mit einer Bemerkung über die Bedeutung der Erfindungen eröffnete, auf die Piaton (Gorg. 448 C ; 462 B) anspielt (Diimmler, Akad. S. 75). Nun haben schon Maaß (a. a. 0 . S. 579 f.) und Blaß (Att. Ber. 2 I S. 77) auf die Wiederkehr einiger τόποι des Palamedes in Antiphons Rede De morte Herodis aufmerksam gemacht. Auch derjenige über den Kulturfortschritt auf Grund der zunehmenden Erfahrung wird von Antiphon zweimal verwendet: Polos bei Plat. Gorg. Antiphon, De mort. Antiphon, De saltat. 448 C: Her. 14: 2: πολλαί τέχναι άνθρώποις είσίν έκ των εμπειριών έμπείρως ευρημέναι- εμπειρία μεν γάρ ποιεί τον αιώνα ó γάρ χρόνος καί ή ó χρόνος γαρ και ή ήμών πορεύεσθαι κα- εμπειρία τά μή κα- εμπειρία τά μή κατά τέχνην, απειρία δέ λώς έχοντα έκδιδά- λώς έχοντα διδάσκει σκει τους άν-θ-ρώπους τους ανθρώπους. κατά τύχην. Taucht dieser Gedanke vereinzelt auclL schon früher, ζ. Β. bei Xenophanes fr. 18, auf, so hat ihn doch erst die Sophistenzeit in's Einzelne verfolgt und gerade in der Schule des Gorgias muß er besonders gepflegt worden sein: Kritias (fr. 2) hat ihn in einer Elegie behandelt und im Sisyphos (fr. 25) auf Recht und Religion übertragen; auch in der dem Alkidamas zugeschriebenen Anklagerede des Odysseus gegen Palamedes (22 ff.) nimmt die Erfindungstheorie einen breiten Raum ein. Es paßt also vorzüglich, daß Gorgias seinen Palamedes sich rühmen läßt, er habe durch seine Erfindungen τον άνθρώπινον βίον πόριμον έξ άπορου καί κεκοσμημένον έξ ακόσμου ge2S ) Leopold Schmidt, Ethik der Griechen II 353. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 566 W. N e s t l e , macht (30) 2i ). Endlich spricht noch für die Verfasserschaft des Gorgias die Art, wie Palamedes mit den Begriffen Wissen (είδέναι) und Glauben (δόξα) operiert. Er sucht das angebliche Wissen des Odysseus um seine Schuld als eine bloße δόξα zu erweisen (5. 24). Wie in der Helena die Macht der απάτη darauf zurückgeführt wird, daß οί πλείστοι την δόξαν σύμβουλον τη ψ υ χ·?] παρέχονται (13), so bedauert hier Palamedes (24): αλλά μην το δοξάσαι κοινόν απασι περί πάντων. Und auch der grundsätzliche Skeptizismus des Gorgias, den er in der Schrift περί φύσεως ή περί του μη δντος begründet hatte, tritt in beiden Reden zu Tage: in der Helena da, wo es von den Astronomen heißt, daß sie nur eine δόξα beseitigen, um eine andere dafür an die Stelle zu setzen (13) und im Palamedes da, wo dieser das Nichtwissen des Odysseus um seine Schuld mit den Worten begründet: ούδ' εοιχ' δπως αν ειδείη τις είναι το μη γενόμενον (5). Dies ist dieselbe Art der Beweisführung, wie in der Schrift περί του μη οντος das angebliche Wissen von einem Seienden damit widerlegt wird, daß man auch Dinge sich vorstellen könne, die gar nicht existieren: Wagen, die auf dem Meere fahren, und allerlei Fabelwesen; denken (φρονείν) kann man das, aber nicht wissen (Sext. Emp. adv. math. VII. 79 f. bei Diels 2 S. 554). Auch fr. 26 gehört in diesen Gedankenkreis: το μεν είναι άφανές μη τυχόν του δοκεϊν, τδ οε δοκειν άσθ-ενές μή τυχόν του είναι. Selbstverständlich konnte Gorgias dem Palamedes, der ja die Wahrheit seiner Unschuld erweisen will, nicht seinen eigenen Skeptizismus leihen, dem das Wahrscheinliche über das Wahre ging (Plat. Phaidr. 167 A); aber an den angeführten Stellen schimmert er doch ein wenig durch. Es wird also auch hier bei der Anerkennung der Gorgianischen Autorschaft sein Verbleiben haben (Blaß, Att. Ber. 2 I 78 f.). Hippias. Fr. 4 (2). Das Bruchstück aus der Συναγωγή über die 14mal verheiratete Thargelia, die ihre Ehen zu politischem Einfluß benützte, ist wie ein Beispiel zu dem allgemeinen Satz -4) Zu dem Gedanken der sittlichen Wirkung wissenschaftlicher Tätigkeit (Pal. 31) vgl. Eurip. fr. 910. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 567 des Demokrit fr. 214 (129) : ενιοι δέ πολίων μεν δεσπόζουσι, γυναιξί δέ δουλεύόυσιν. Vgl. Dem. fr. I l l (128). Fr. 10: παραθήκή. Zu dem Gebrauch dieses Wortes, dessen sich Hippias statt des sonst üblichen παρακαταθήκη bediente, hat schon Diels auf Herod. VI. 73 und IX. 45 verwiesen; ein drittes Mal erscheint es VI. 86. Uebrigens gebraucht Herodot hier unmittelbar darauf mehrmals παρακαταθήκη, ebenso V. 92 η ; παρακαταθήκη auch bei Demokrit fr. 265 (152). Sollte es Zufall sein, daß der Begriff in seiner gewöhnlichen Form παρακαταθέσ&αι auch bei Xenophon in dem Gespräch des Sokrates mit Hippias figuriert unter den Beispielen, womit der Philosoph dem Sophisten den Nutzen der Gesetze und der Gesetzlichkeit beweisen will? Die Vermutung liegt n a h e , daß Hippias in einem von uns freilich nicht mehr zu erratenden Zusammenhang sich eingehender mit dem Begriff der παρα&ήκη befaßt hat. Herodot läßt den spartanischen König Leotychides die Verletzung der Pflicht, eine παραθήκη einzulösen, durch eine sehr moralische Geschichte illustrieren. Obwohl bei dem Geschichtschreiber gerade hier (VI. 86) die alte Form des Wortes steht und die Geschichte in Sparta spielt, wo Hippias viel verkehrte, ist es mir doch zu gewagt, einen Zusammenhang zu konstruieren. Fr. 17 (9). Auch hier h a t schon Diels auf Herod. VII. 1 0 η verwiesen; doch ist auch noch eine Stelle des Isokrates, der in zweiter E h e mit Hippias Tochter Plathane verheiratet war, zu vergleichen und es ist nicht ohne Interesse, diese drei Aeußerungen über die διαβολή zusammenzustellen. Hippias fr. 17: Herodot VII. 10 η : Isocrat. De antid. 18: δ ε ινό ν έστιν ή διαβο- διαβολή γάρ έστι δει- εστι μέγιστον κακόν λία, ότι οόδέ τιμωρία νότατον · έν δύο διαβολή 1 τί γαρ αν τις κατ' αύτών γέ- μεν είσιν οι άδικέον- γένοιτο ταύτης καγραπται έν τοις νόμοις τες, ε!ς δέ ó άδικε- κουργότερον, r] ποιεί ώσπερ των κλεπτών · όμενος. δ μέν γαρ τους μέν ψευδομένους καίτοι άριστον δν διαβάλλων άδικέει ου εύδοκιμείν, τους δέ κτήμα την φιλίαν παρεόντος κατηγο- μηδέν ήμαρτηκότας κλέπτουσιν, ώστε ή ρέων, 6 δέ άδικέει δοκείν άδικείν, τους ΰβρις κακούργος ουσα άναπειθόμενος, πριν δέ δικάζοντας επιορδικαιότερα εστί της ή άτρεκέως έκμάθη · κεΐν, δλως δέ την Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM W. N e s t l e , 568 διαβολής δια το άφανής ehtxi. ο Οε οή απεων του άλήθειαν αφανίζει, λόγου τάδε έν αύτοίσι ψευδή δέ δόξαν, παράδικέεται, διαβληθείς αστήσασα τοις άκούτε υπό του έτερου και ουσιν 8ν αν τύχη των νομισθείς προς του πολιτών άδίκως άτζετέρου κακός είναι. όλλυσιν ; Daß hier nur verschiedene Nuancierung desselben Gedankens vorliegt, dürfte kaum zu bestreiten sein. Bei Herodot steht die Stelle in einem der kunstvollsten und gedankenreichsten Gesprächscyklen seines Werkes. Der Gedanke ist etwas künstlich angebracht (s. A. Schöll im Philol. X. 1855 S. 78); denn es handelt sich hier nicht um die Verleumdung eines Einzelnen, sondern um die Herabsetzung des gesamten hellenischen Volkes in den A.ugen des Xerxes durch Mardonios. Man kann die ganze Ausführung über die διαβολή herausnehmen und dem Zusammenhang fehlt nicht das Geringste. Sie sieht aus wie ein nachträglich auf das Gemälde aufgesetztes Licht. Den Anknüpfungspunkt bot das W o r t διαβάλλων (vgl. V. 97). Ich bin daher überzeugt, daß wir es hier mit einem sophistischen τόπος zu tun haben, den Herodot und Isokrates in Anlehnung an Hippias jeder in seiner Art verwendet haben 2 5 ). Vielleicht hat ihn der Geschichtschreiber erst bei der Schlußredaktion seines Werks eingesetzt. Ich kann mich daher auch nicht dem Urteil von Diels (Vors. 2 S. 586) anschließen, der die Vermutung Dümmlers (Ak. 249 ff.), Hippias sei die Quelle f ü r die bei Herod. III. 38 entwickelte N o m o s t h e o r i e , eine „vage Hypothese" nennt. D a ß sie in den Dialexeis 2, 18 (26) wiederkehrt, ist eher ein Beweis dafür als dagegen ; denn dies bietet eine genaue P a rallele zu dem Gorgianisclien τόπος von der gerechten und ungerechten απάτη (Dial. 3 , 10. Gorg. fr. 2 3 ; Hei. 8 ff.). Das merkwürdigste aber ist, daß „diese wohl sprichwörtliche Sentenz" (Stein z. St.) mit etwas verändertem I n h a l t , aber in derselben Form bei Herodot VII. 152 in einem ziemlich erzwungenen Zusammenhang auf die κακά (d. h. αισχρά) der 2δ ) Vgl. Quint. III. 1, 12, wozu Gereke a. a. 0 . S. 852, 2. Doch könnte sich das „tractasse affectus" auch auf Untersuchungen beziehen, wie sie Plato Hipp. min. 370 f. erörtert. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 569 Menschen angewandt nochmals erscheint, was Dümmler nicht anführt. Als ein Apophthegma Solons über die Leiden der Menschen erwähnt sie Val. Max. VII. cap. II ext. 2 (A. Schöll im Philol. X, 1855 S. 39): Unser Sophist kann dieses ganz wohl wie das Pindarzitat in seinem Sinn verwendet haben. Wie viele Wandlungen mag sie durchgemacht haben, bis ihr Horaz die Form gab, die wir Sat. I. 1, 15 ff. vorfinden! Daß Hippias dem Begriff des νόμος besondere Aufmerksamkeit schenkte, kann angesichts der übereinstimmenden Darstellung Piatons (Protag. 337 C D ; Hipp, major 284 C D) und Xenophons (Mem. IV. 4) nicht bezweifelt werden. Und von beiden führen Verbindungslinien zu den Διαλέξεις, deren Verfasser ganz von den Gedanken der großen Sophisten lebt. Was dieser über das καλόν und αίσχρόν der αφροδίσια sagt (Dial. 2, 4), hat seine genaue Parallele in der Unterhaltung des Sokrates und Hippias bei Plato (Hipp. maj. 299 A). Auch daß 2, 11 ein ganz untergeordneter Sizilischer Brauch im Gegensatz zu einer Thessalisclien Sitte erwähnt wird, läßt sich gut aus der Bekanntschaft des Verfassers mit dem in Sizilien besonders heimischen Hippias erklären (Hipp. maj. 282 Ε) 2e). Xenophon Mem. IV. 4, 14 f. stimmt seinerseits genau mit Plato Hipp. maj. 284 C D überein, und Mem. IV. 4, 20 erfordert das ,πανταχού νομίζεται' der vorhergehenden Frage notwendig, bei denen, die den νόμος μήτε γονέας παισί μίγνυσθ·αι μήτε παίοας γονεΰσιν übertreten, nicht etwa an einzelne verhängnisvolle Verirrungen wie die des Oidipus zu denken, sondern an Völker, die diesen νόμος verletzen, d. h. nicht haben. Die Dial. 2, 15 führen dafür die Perser an und wiederum ist es Herodot, der, obwohl er den Kambyses eines Verbrechens zeihen möchte, doch berichten m u ß , daß in Persien kein νόμος die bei den Griechen verpönte Geschwisterehe (Eurip. Androni. 173 ff. ; Aiol. fr. 19) verbiete (III. 31): eine Nachricht, die Xanthos fr. 28 ergänzt: μίγνυνται δε oí μάγοι μητράσι και •9-υγατράσι · και άδελφαϊς μ:γνυσθ·αι ·9·εμιτον είναι. Auf solche Bräuche muß Hippias sein Augenmerk gerichtet und sie zur Vgl. Eurip. El. 815 ff. ; Dümmie , Proleg. zu Platona Staat S. 50. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 570 W. Nestle Grundlage seiner Naturrechtstheorie, die eine ganz andere war, als die des Kallikles, verwendet haben 27 ). Mit seiner αυτάρκεια (Diels 2 S. 579 No. 1; 582 No. 12) war Hippias eine Art Vorläufer der Kyniker und Stoiker und suchte seine Lehre, δτι τον σοφόν αότόν αυτω μάλιστα δει σοφόν είναι (Hipp. maj. 283 Β ; vgl. Eurip. fr. 905 ; Alex. fr. 61 ; Ennius fr. 50 Ribbeck), im Leben zu verwirklichen, indem ermöglichst unabhängig auf sich selbst zu stehen suchte. Das Wort αυτάρκης kommt, so viel ich sehe, vor Plato (Phileb. 67 Α αυτάρκεια) nur viermal vor: Aisch. Choeph. 757; Soph. Oed. Col. 1057; Eur. Aiol. fr. 29. In einem philosophischen Zusammenhang erscheint es erstmals wiederum bei Herodot I 32 in Solons Mund : άν·8·ρώπου σώμα εν ούδέν αυταρκές εστι · το μεν γαρ εχει, άλλου δε ενδεές έστον. Ich möchte vermuten, daß Hippias es war, der diesen Begriif in die Anthropologie eingeführt hat und daß das Wort Solons seine Spitze gegen ihn richtet. Diese Vermutung erhält noch dadurch eine Stütze, daß auch hier wie bei der διαβολή Isokrates eine Parallele zu Herodot bietet: Herod. I 32: ώσπερ χώρη ουδεμία καταρκέει πάντα εαυτή παρέχουσα, άλλα άλλο μέν εχει, έτέρου δέ έπιδέεται etc. Isokr. Paneg. 42 : ετι δέ την χώραν ουκ αυτάρκη κεκτημένων εκάστων, άλλα τά μεν έλλείπουσαν, τα δέ πλείω των Εκανών φέρουσαν ταύταις ταΐς συμφοραϊς έπήμυνεν. Herodot und Isokrates reden beide von der nicht vorhandenen αυτάρκεια eines Landes, das daher auf den Verkehr mit andern angewiesen ist, ζ. B. Attika (vgl. Eurip. Hik. 209 f.). Der Geschichtschreiber verwendet dies als Bild für den einzelnen Menschen, der auch die Unterstützung anderer braucht. Hippias dagegen wollte den Einzelnen möglichst auf sich selbst stellen. Nichtsdestoweniger kann der umstrittene Begriff von ihm stammen. Fr. 7 und 12. Hippias beschäftigte sich u. a. auch mit der Geschichte der Wissenschaften, der Philosophie (Diog. L. 27 ) U e b e r d a s V e r h ä l t n i s des σοφός zum νόμος vgl. Xen. Mein. IV. 4, 14 m i t D e m o k r i t bei Diog. L. IX. 45 u n d D i e l s 2 S. 401 Nr. 166; f r . 248. D ü m m l e r a. a. O. S. 52 f. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. I 24) und Mathematik. bis 282 A. 571 Hiezu vgl. Plato Hipp. maj. 281 C Antiphon. Fr. 32. Antiphon war Eklektiker. Wie er sich in seinem Gottesbegriff (fr. 1. 10 D ; 1 . 4 N.) an die Eleaten Xenophanes (Diels 2 S. 41 No. 32) und Parmenides (fr. 8, 32 f.) anschließt, so wandelt er in den Spuren des Empedokles, wenn er wie dieser (fr. 55 D; 31 N) das Meer als „Schweiß" (der Erde) bezeichnete (vgl. Emp. bei Diels 2 S. 163 No. 66). Fr. 49 (12). In dem großen Bruchstück über die Ehe schrieb Diels in der ersten Auflage S. 567: χαλεπαί μεν έκπομπαί [καί] τους φίλους εχθρούς ποίησα:, ισα φρονοϋντας ϊοα πνέοντας [ζην] άξιώσαντα και άξιωθ'έντα. In der zweiten S. 598 hat er das nicht überlieferte καί und ζην wieder gestrichen, durch Kommata ersetzt und erklärt τους φίλους für „die Eheleute (vgl. Aesch. Ag. 1236)". Ich halte das für keine Verbesserung. Zunächst ist bei Aischylos, wo Klytaimestra ασπουδον Ά ρ η φίλοις πνέουσα heißt, unter φίλοι nicht das Ehepaar Agamemnon und Klytaimestra zu verstehen, sondern der erstere und seine Kinder. Ferner müßte man, wenn mit φίλοι die sich scheidenden Eheleute gemeint wären, doch mindestens ein αλλήλοις erwarten; drittens, wie kann man Eheleute, die im Begriff stehen, sich zu scheiden, ,ίσα φρονοϋντας, ίσα πνέοντας' nennen? Der überlieferte Text ist daher unhaltbar. Mit den φίλο: müssen die Angehörigen der geschiedenen Frau gemeint sein; άξιώσαντα und άξιωθ'έντα beziehe ich auf den Mann, ίσα — πνέοντας ζην davon abhängig auf Mann und Frau zusammen: „Es ist schwer sich zu scheiden und dadurch Freunde sich zu Feinden zu machen, nachdem man die Erwartung (den Anspruch) gehegt und erregt hatte, gleichgesinnt und gleichgestimmt mit einander zu leben". Töpfers (Die sog. Fragmente des Sophisten Antiphon bei Jamblichos. Progr. Arnau 1902 S. 46) Heranziehung von Plato Gorg. 510 Β ist äußerst gesucht. Fr. 50 (10): το ζην έοικε φρουρά έφημέρω τό τε μήκος του βίου ή μέρα μια, ώς επος ειπείν, f¡v άναβλέψαντες προς το φως παρεγγυώμεν τοις έπιγιγνομένοις έτέροις. In der Deutung Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 572 W. N e s t l e , dieses Bruchstücks hat sich Diels (2. Aufl. S. 599) und Töpfer (a. a. 0. S. 42 und 45) an Sauppe angeschlossen: ,das Leben gleicht einem eintägigen Wachestehen, unsere Nachkommen sind die Ablösung'. Diese Erklärung hat den Vorzug, daß die Worte άναβλέψαντες προς το φως dadurch einen außerordentlich einfachen Sinn bekommen: ,das Morgenrot zeigt die Ablösung an (Diels)28). Andererseits erheben sich dagegen verschiedene Schwierigkeiten. Bedeutet φρουρά Ιφήμερος ,statio diurna', so sollte man die Ablösung von Rechts wegen am Abend und nicht am Morgen erwarten. Ferner eignet sich eine Wache, die einen ganzen Tag oder eine ganze Nacht dauert, nicht zu einem Vergleich, der die kurze Dauer einer andern Sache, hier des Lebens, veranschaulichen soll. Denn beim griechischen Heerwesen war bekanntlich das Postenstellen auf mehrere Nachtwachen verteilt (ζ. B. Xen. An. IV. 1, 5) 2 9 ). Dazu kommt, daß φρουρά, wie Buresch (Consolationum a Grraecis Romanisque scriptarum hist. crit. Leipziger Stud, zur klass. Philol. IX. 1887 p. 80 ss. ; 131 s.) gezeigt hat, in Stellen ähnlichen Zusammenhangs (Plato, Phaid. 62 B; Gorg. 525 A) passivisch zu verstehen ist, und .Gefängnis' bedeutet. Ganz in demselben Gedankenkreis erscheint auch εφήμερος oder εφημέριος: Emped. fr. 4 (7), 4; 131 (70), 1; Pind. Pyth. VIII. 95; Aisch. Prom. 83. 253. 546. 945. Der Ausdruck ,das Licht wieder schauen' ist aber auch wohl am Platze, wenn man an das Kommen aus dem Dunkel des Kerkers (Virg. Aen. VI. 734) denkt. Die έπιγιγνόμενοι sind diejenigen, welche jetzt in die φρουρά, das Gefängnis des Leibes kommen; die άναβλέψαντες προς το φως die ins wahre Leben eingegangenen Verstorbenen. Dieser Seitenblick auf die orphisch-pytliagoreische σώμα-σημαLehre entspricht ganz dem wenig konsequenten Eklektizismus des Sophisten. Fr. 64 (20): αΕ νεχι φιλίαι άναγκαίαι μέν, ai δέ παλα'.αί άναγκαωτεραι. „Frisch geschlossene Freundschaften sind un28 ) Der Deutung von Diels folgen seine Schüler E. Jafeoby, De Antiphontis sophistae Περί όμονοία;; libro (Berolini 1908) S. 39, 91 und W. Altwegg, De Antipbonte qui dicitur sophista quaestionum partícula I. De libro Περί δμονοία; scripto (ßasileae 1908) S. 31. 29 ) Vgl. die LXX Psalm 89 (90), 4 : χίλια ετη . . . ώς . . . φυλακή έν VUXXÍ. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkuugen zu den Vorsokratikern und Sophisten. 573 entbehrlich, unentbehrlicher aber die alten" (Töpfer a. a. 0 . S. 45 f.). Ich k a n n nicht verstehen, warum eine frisch geschlossene F r e u n d s c h a f t „unentbehrlich" sein soll. Offenbar h a t αναγκαίος wie das lateinische necessarius hier die Bedeut u n g ,eng verbindend' ; bei Personen heißt es j a ,eng verbunden' und es werden daher besonders gerne Verwandte damit bezeichnet, manchmal mit beigesetztem φίλοι : E u r . Androni. 6 7 1 ; Alk. 533 ; Xen. Hell. I . 7, 16 ; An. II. 4, 1 ; Lys. or. 19, 3 4 ; 20, 1;· 31, 2 3 ; Plato Resp. IX. 574 Β. Ich übersetze dah e r : „ J u n g e Freundschaften sind i n t i m , alte noch intimer". Bei den jungen b e r u h t die I n t i m i t ä t auf der ersten, frischen, begeisternden Freude der Freunde an einander, bei den alten ihr noch höherer Grad auf der langen E r p r o b u n g der F r e u n d schaft. Vgl. Demokrit f r . 100 (119). F r . 83 : άδυνασία. Vgl. αδυναμία bei Plato Gorg. 492 A synonym mit dem Schlagwort άνανδρία (ib.) oder δειλία (Anon. J a m b l . 6). S. meinen Euripides S. 487, 112. F r . 88 : βάσανος stand vermutlich in einem ähnlichen Z u s a m m e n h a n g wie Theogn. 450 ; Pind. P y t h . X. 67. Soph. Oid. Tyr. 510. Iiritias. F r . 2. Die E r f i n d u n g der Schrift w i r d , wie hier v. 10, so auch von Herodot V. 58 auf die Phoiniker z u r ü c k g e f ü h r t , während nicht nur Stesichoros, sondern auch Euripides und Gorgias dieses Verdienst dem Palamedes vindizierten und Aiscliylos es dem P r o m e t h e u s zuschrieb. Die Karer nennt K r i tias als Erfinder der Schiffe (v. 12), während Herodot I 171 von ihnen die E r f i n d u n g der Helmbüsche, Schildzeichen und Schildgriffe zu berichten weiß. Doch kennt er (II. 152. 154) und Thukydides (I 4) sie als mächtiges seefahrendes Volk. Die erste Schrift περί ευρημάτων v e r f a ß t e der Genealog Simonides von Keos, der E n k e l des gleichnamigen Dichters (Müller, F r a g i n . Hist. Graec. I I 42 f. ; W i p p r e c h t , Zur E n t w i c k l u n g der rationalistischen Mythendeutung bei den Griechen I 38); auch der Gorgiasschüler Polos (s. o.) b e r ü h r t e dieses Thema. F r . 9 (20). Der vereinzelte Pentameter έκ μελέτης πλείους η φύσεως αγαθοί deckt sich inhaltlich genau mit Demokrit fr. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 574 W. N e s t l e , 242: πλέονες εξ άσκήσιος άγαθ·οί γίνονται ή από φύσιος. Obgleich auch der Abderite die Bildung und Erziehung im Verhältnis zur Naturanlage sehr hoch einschätzte (fr. 33. 59. 182. 242 D), sieht doch der prosaische Satz wie eine (dann natürlich nicht von Dernokrit stammende) Paraphrase des Verses des Kritias aus. Uebrigens lag die Erörterung dieser Frage in der Sophistenzeit in der Luft: vgl. Krit. fr. 40 (21); Protag. fr. 3 (6); 10 (7); 11 (8) und besonders Antiphon fr. 60 (17). Fr. 19 (11). Unter dem hier angeredeten und als αυτοφυής bezeichneten Wesen verstand Clemens ΑΙ. τον δημιουργον νουν. Die Verse würden also die Lehre des Anaxagoras vom Nus und der von ihm eingeleiteten περιχώρησις [Anaxag. fr. 12 (14)] berücksichtigen. Doch könnte auch der fr. 18 (10) und 25 (16), 34 genannte Chronos τίκτων αυτόν εαυτόν gemeint sein. Vgl. Neue Jahrb. für klass. Philol. XI. 1903 S. 104 f. Fr. 22 (14). Zu den gegen die Rechtsverdrehungen der Redner gerichteten Versen vgl. das Verbot des Kritias λόγων τέχνην μή διδάσκειν Xen. Mem. I. 2, 31. Fr. 25 (16), 38. Diels erklärt hier τψ λόγψ „durch seine Fiktion" sc. habe der kluge Erfinder von Recht und Religion die Götter schön und passend im Himmel angesiedelt. Es wäre dann also so viel wie ψευδεΐ λόγω v. 26. Da aber ν. 38 eine nähere Bestimmung des λόγος fehlt, liegt es vielleicht doch näher, das Wort mehr in seiner Grundbedeutung zu fassen: ,mit überlegendem, berechnendem Sinn', gleichwertig etwa mit λογιζόμενος, so daß damit zum Schluß nochmals die Eigenschaft des Erfinders hervorgehoben würde, die ihm v. 12 mit so volltönenden Worten zugeschrieben wird, die Klugheit, Schlauheit. Zu γνώμην, das hier gewiß vor der Variante γνώναι zu bevorzugen ist, vgl. fr. 6 (5), 19; 39 (22. 23); 40 (21). Das Wort steht in erkenntnistheoretischem Sinn wie bei Antiphon fr. 2 , Demokrit fr. 11 (13), dem Verfasser der Apologie der Heilkunst (Gomperz, Wiener Sitzungsber. 120. 1890 S. 6 f.) und Euripides (s. mein Buch über diesen S. 414, 23). Fr. 28 (18): δεινον δ' δταν τις μή φρονων δοκ·7; φρονεΐν. Diels vergleicht dazu Aisch. Prom. 385 : κέρδιστον εδ φρονοΰντα Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 575 μή φρονεϊν δοκεϊν. Beide Verse sind Gegenstücke: „Schlau, wenn ein kluger Mann den Toren spielt"; umgekehrt Kritias: „Schlimm, wenn ein Dummkopf den Gescheiten spielt". Denn der Wortstellung nach muß μή notwendig zu dem Partizip φρονων und kann nicht zu dem Infinitiv δοκεϊν gezogen werden. Fr. 48 (25): κάλλιστον είδος έν τοις αρρεσι τό θήλυ, έν δ' αδ ταίς -θ-ηλείαις τέναντίον. Diese Behauptung einer perversen Sinnlichkeit erinnert an den Konflikt des Kritias mit Sokrates Xen. Mem. I. 2, 29 f. Anonymus Jamblichi. Zu den von Blaß entdeckten und dem Antiphon zugeschriebenen interessanten Bruchstücken bei Jambl. Protr. 20 hat Töpfer (Die sog. Fragmente des Sophisten Antiphon bei Jamblichos. Progr. Arnau 1902) viele feine Bemerkungen gemacht. Einige Abweichungen von seiner Uebersetzung und Erklärung sollen hier begründet und einige weitere Beiträge zum Verständnis der Fragmente hinzugefügt werden. Fr. 1. Hier wird die άρετή in Teile zerlegt: ein Verfahren, gegen das Sokrates energisch zu protestieren pflegte. Der Verfasser gehört also zu den έξαριθμοΰντες τάς αρετάς wie G o r g i a s : Plato, Men. 71 E (Diels 2 fr. 19 S. 560 f.). In demselben Bruchstück wird dann φύσιν und μάθησιν unterschieden und möglichst frühes Beginnen mit dem Lernen empfohlen ganz wie bei Ρ r o t a g o r a s fr. 3 (6) und in den oben zu Kritias fr. 9 angeführten Stellen. Fr. 3 (2) handelt vom Gebrauch der erlangten σοφία und ισχύς zu guten und schlechten Zwecken : eine Theorie, die sich gleichermaßen bei G o r g i a s (p. 456 f.), wie, auch auf materielle Güter angewendet, bei Ρ r o d i k o s (Eryx. 396 A; 397 E) findet. Ganz merkwürdig ist die z. T. wörtliche Uebereinstimmung des Schlußsatzes dieses Bruchstücks mit E u r i p i d e s Hik. 312 f. (s. Eurip. S. 525, 73 und 73 a). Fr. 5 (4). τότε γήρας κάκιον δν άνθρώποις „das minderwertige Alter " T. Ich ergänze zu κάκιον den Begriff θανάτου : „das Alter, das schlimmer ist als der Tod". — συνήθεια πονηρών αλόγων και έπιθυμημάτων „Vertrautheit mit niedrigen Grundsätzen und Begierden" Τ. συνήθεια ist hier ein aktiver Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 576 W. N e s t l e , Begriff: das Sichgewöhnen an, fast wie lateinisch indulgentia: Nachgiebigkeit gegen schlimme M o t i v e und Begierden, λόγοι in dieser Bedeutung bei P h i l o l a o s fr. 16. Zum Schlußgedanken vgl. Η e r a k l . fr. 29 (113): αίρευνται γαρ εν άντία πάντων oí άριστοι, κλέος άέναον θνητών, und den Schluß von Ρ r o d i k o s Herakles : μετά μνήμης τον άεί χρόνον υμνοΰμενοι θάλλουσι. Fr. 6 (5). Dieses Bruchstück verteidigt aufs nachdrücklichste die Herrschaft des νόμος und polemisiert gegen die von Kallikles (Gorg. 484 B) und Thrasymachos vertretene Lehre vom Recht des Stärkeren mit deutlicher Anspielung auf das bei Plato zitierte Pindarfragment, das auch dem Hippias im Protag. 337 D vorschwebt, und will den Gehorsam gegen die Gesetze nicht als δειλία in Verruf gebracht wissen (s. o. zu Antiphon fr. 83). Der Verfasser glaubt auch gar nicht an die Möglichkeit des von Kallikles gezeichneten Uebermenschen, der es nach seiner Meinung mit einer loyalen Majorität niemals aufnehmen könnte. Um zu beweisen, daß δ νόμος und TO δίκαιον φύσει seien, benützt er einen Gedanken, den Plato den P r o t a g o r a s in dem Mythus περί της εν αρχή καταστάσεως aussprechen läßt (322 Β C): wenn die Menschen nicht ini Besitz der πολιτική τέχνη sind, die auf αιδώς und δίκη beruht, so entsteht ein Kampf aller gegen alle, der zum Untergang führen muß. Fr. 7 (6). Denselben Gedanken wiederholt in weiterer Ausführung das folgende Bruchstück. Wie A n t i p h o n f r . 61 (18) die αναρχία, so bezeichnet der Anonymus die ανομία als das größte Uebel. Sie allein bildet die Voraussetzung fiadas Aufkommen eines τύραννος, der aus eigener Kraft über einen gesetzlich regierten Staat nie Herr werden könnte. Diese Theorie von der Entstehung der Monarchie aus der Anarchie (ανομία) findet sich auch bei Η e r o d o t , weniger deutlich in der Verfassungsdebatte der sieben Perser (III. 82), wo übrigens doch ein κακοϋν τα κοινά durch den Demos als Anlaß für das Auftreten des προστάτης vorausgesetzt wird, der sich dann zum Monarchen macht, ganz überraschend ähnlich aber in der Erzählung von der Begründung der medischen Monarchie durch Dejokes I 96 f. : έούσης άνομίης πολλής ανά πα σαν τήν Μηδικήν Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Yorsokratikern und Sophisten. 577 beginnt er mit seiner Richtertätigkeit (96). So erscheint er als άνήρ μοΰνος κατά το òpitòv δικάζων und er läßt die ανομία absichtlich anwachsen, indem er erklärt, er könne nicht zugunsten seiner Richtertätigkeit seine eigenen Interessen vernachlässigen. Und έούσης ών αρπαγής και άνομίης s u πολλω μάλλον ανά τάς κώμας wählen ihn die Meder zum König, um nicht durch die άνομίη ruiniert zu werden. Ganz im Sinn des Anonymus 6 (5), Β hält er sich als Fürst τοις νόμοις συμμάχων και τω δικαίω. Mag man dabei auch an die von Herodot III. 80 berührte persische Sitte denken, nach dem Tod eines Königs fünf Tage άνομία herrschen zu lassen, um so den Segen des Königtums der Bevölkerung fühlbar zu machen (Sext. Emp. adv. rhet. 83), die Uebereinstimmung ist zu frappant, um nicht die Vermutung von Maaß (Hermes 22. 1887 S. 583) gegen Ed. Meyer (Forsch, zur alt. Gesch. I 201 f.) zu bestätigen, daß diese Ideen dort (III. 80—82) und ebenso hier (I 96 f., eine Stelle, die Maaß nicht berücksichtigt hat) sophistischen Ursprungs sind. — Das große Bruchstück stellt die Zustände der ευνομία und άνομία einander gegenüber. Die εύνομία ist vor allem die Grundlage für wirtschaftlichen Kredit (πίστις), wobei man auch mit verhältnismäßig wenig Geld, wenn es nur in Umlauf kommt, auskommen kann. Zum Ausdruck κυκλούμενα sc. χρήματα vgl. E u r i p i d e s Aiol. fr. 22: κύκλω γάρ ερπε;. Der Satz aber, daß τούς τε αδ δυστυχοϋντας έπικουρείσθαι έκ των εύτυχούντων ο·.ά την έπιμειξίαν τε και πίστιν, hat große Aehnlichkeit mit D e m o k r i t fr. 255 (139), der verlangt, daß die Besitzenden sich entschließen, den Besitzlosen zu borgen, sie zu unterstützen und ihnen wohlzutun, damit die Eintracht unter den Bürgern gefördert werde. Als Symptom der άνομία kennzeichnet der Verfasser das für den Einzelnen ungünstige Verhältnis, in das die εργα zu den πράγματα treten. Was bedeuten nun diese beiden nach prodikeischer Synonymik unterschiedenen Worte? Dümmler (Proleg. zu Piatons Staat 1891 S. 9, 3; vgl. 19, 1) hat auf dieselbe Unterscheidung bei E u r i p i d e s , Helena 286 hingewiesen : άλλα πάντ' έχουσα δυστυχή \| τοις πράγμασι τέθ-νηκα, τοις δ' εργοισιν ου. Hier ist in πράγματα all das Unheil zusammengefaßt, das Helena getroffen hat; an diesem geht sie, wie sie Philologue L X V I I (Ν. Ε. XXI), 4. 37 Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 578 W. Ν e s 11 e, meint, zugrunde, nicht an ihren έργα, d. h. an ihren eigenen Taten oder Handlungen, an dem, was man ihr gewöhnlich schuld gibt: es ist der Grundgedanke der Gorgianischen Helena hier gestreift. Man darf nur versuchen, diesen Gegensatz an den betreffenden Stellen in unser Bruchstück einzuführen, so wird man finden, daß er nicht in den Zusammenhang paßt. Die ευνομία ermöglicht es, εις μεν τά πράγματα άργόν γίγνεσθαι, εις δέ τά εργα της ζωης έργάσιμον (3) ; die φροντίς πραγμάτων ist αηδέστατη, diejenige έργων dagegen ήδίστη (4). Im Zustand der ανομία ist es Regel εν τε πράγμασο καθεστάναι άεΐ δια έπιβουλάς τάς έξ άλλήλων (10) und man hat keine φροντίδες ήδειαι, also keine Sorgen um die εργα (11). Töpfer (a. a. 0 . S. 23) übersetzt πράγματα mit „quälende Sorgen", εργα mit „höhere Lebensaufgaben". Nun ist ohne weiteres zuzugeben, daß πράγματα hier in einem verwandten Sinn gebraucht ist wie in der Formel πράγματα παρέχειν τινί. Aber, daß noch etwas weiteres, ein aktiver Begriff, darin stecken muß, zeigt der Ausdruck άργόν γίγνεσθαι εις, das eben nicht „frei" (Töpfer) heißt, sondern untätig. Man denke an den φθόνος αργίας: Eurip. Med. 296 f.; Aristopli. Frösche 1498; Thuk. II. 40, 2. Da nun freilich dem begeisterten Republikaner, der der Anonymus ist, nicht zugetraut werden kann, daß er die politische άπραγμοσύνη oder άργία wünscht oder empfiehlt, so kann πράγματα nicht die regelmäßige Beteiligung des Bürgers an der Politik bezeichnen, wohl aber die darüber hinausgehenden politischen Widerwärtigkeiten, die durch στάσεις mit ihren επιβουλαί und auch durch auswärtige Kriege herbeigeführt werden (10). Diesen πράγματα kann man sich um seiner eigenen Sicherheit willen nicht entziehen und dies geht auf Kosten der εργα, die somit nichts anderes als die eigene Arbeit des Bürgers bedeuten, mag er nun Bauer oder Kaufmann, Dichter oder Künstler, Gelehrter oder Philosoph sein. — Zu der Bezeichnung des Schlafes als ήδεϊα υποδοχή (11) vgl. Kritias fr. 6 (5), 20: καμάτων λιμήν. Töpfer (S. 34 ff.) hat aus stilistischen und sachlichen Gründen die Vermutung von Blaß, daß die Bruchstücke dem Sophisten Antiphon gehören, ablehnen zu müssen geglaubt, Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Vorsokratikern und Sophisten. 579 ohne selbst einen andern Verfasser nennen zu können 30 ). Ich glaube Spuren von der Benützung der Schriften älterer Sophisten Protagoras, Gorgias, Prodikos wahrzunehmen. Der erstere könnte recht wohl gemeinsame Quelle für Herodot und unsern Anonymus sein. Dies paßt aber auf Antiphon, der in seiner 'Αλήθεια den Protagoras, wenn auch polemisch, berücksichtigte. Ebenso lassen sich die Berührungen mit Euripides unter dieser Voraussetzung gut erklären, wie Dümmler (a. a. 0.) gezeigt h a t , und auch Demokrit steht zeitlich nicht im Wege. Daher scheint mir die Hypothese von Blaß noch immer am meisten für sich zu haben. Dialexeis. Der interessanteste Abschnitt dieser sonst recht oberflächlichen sophistischen Vorträge ist der zweite, worin die verschiedenen νόμοι einer Reihe von Völkern zusammengestellt werden. Schon Dümmler (Akad. S. 252) hat hier auf die Parallele zwischen § 18 und H e r o d o t III. 38 (cf. VII. 152) aufmerksam gemacht und Hippias als gemeinsame Quelle angenommen (s. o.). Diels verweist zu dem § 11 erwähnten thessalischen Brauch auf Eurip. El. 815 f. ; zu der skythischen Sitte, aus den Schädeln der erlegten Feinde Trinkbecher herzustellen, auf Herod. IV. 65 und Ρ 1 a t o E u t li y d e m 299 E. Dieselbe Gepflogenheit erwähnt Herodot IV. 26 auch von den Issedonen. Auch ist zu κόμιον (13) IV. 64 zu vergleichen. Hinzuzufügen ist noch der vom Verfasser der Dialexeis 2, 14 genannte Brauch der Massageten, ihre alten Eltern zu schlachten und zu verzehren, den ebenfalls Herod. I. 216 berichtet und den er III. 99 auch den indischen Padäern zuschreibt. Ob hier direkte Benützung des Herodot anzunehmen ist oder eine ihm und den Dialexeis gemeinsame Quelle, ist nicht wohl auszumachen. — Daß in § 28 nach αίσχρόν ein konkreter Begriff gestanden hat, mag wohl sein ; aber άνδρα, das Diels 2 S. 641 unter Hinweis auf 3, 14 eingesetzt hat, will nicht passen, da es sich doch um lauter Dinge handelt, die die Menschen herbeibringen und wieder mitfortnehmen; 30 ) Daß die Gedanken des Anonymus mit denen Antiphons übereinstimmen, stellt auch Jakoby (a. a. 0 . S. 8, 20) gegen Töpfer fest. 87* Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM 580 W. N e s t l e , Im dritten Abschnitt haben die Ausführungen über das unter Umständen berechtigte ψεύδεσθαι und έξάπαταν (2 ff.) so große Aehnlichkeit mit Xen. Mem. IV. 2, 14 ff., daß man geradezu an Entlehnung daher gedacht hat. Das Wahrscheinlichere ist aber entschieden, hier eine gemeinsame sophistische Quelle anzunehmen, führt doch Xenophon den E u t h y d e m ein als γράμματα πολλά συνειλεγμένον ποιητών τε καί σοφιστών των εύδοκιμωτάτων (2, 1). Vgl. Rohde, Kl. Sehr. I 331, 2. — Bei der in § 8 erwähnten Säkularisation von Tempelgütern mag man sich an den Vorschlag des Hekataios vor dem jonischen Aufstand erinnern: Herod. V. 36. — Zur άπάτη in Kunst und Dichtung (10) vgl. Gorg. fr. 23 (s. o.). — Es folgt (11) das R ä t s e l d e r K l e o b u l i n a . Seine Lösung hat man auf zweierlei Weise versucht. Bergk hat auf Aristot. Eth. Nie. V. 10 p. 1134a 16 ff. hingewiesen, wobei dann ein Fall gemeint wäre, wie kurz vorher (3 f.) : ein Mann, der einem schwermütigen Freund sein Schwert oder einen Strick entwendet und ihn so gewaltsam am Selbstmord verhindert, begeht zwar einen Diebstahl, aber einen gerechten. Ein ähnliches Beispiel für gerechte άπάτη enthält § 2. Aber eine Lösung wie ó του μαινομένου κλεπτών το ξίφος erscheint doch zu umständlich, zu wenig konzis für den bei einem Rätsel vorschwebenden Begriff, v. Wilamowitz hat dagegen (Herakl. 2 I 97, 179) unter Heranziehung von De diaet. I 24 (Vorsokr. 2 S. 85) : παιδοτριβίη τοιόνδε · διδάσκουσι παρανομεϊν κατά νόμον, άδικείν δικαίως, έξαπατάν, κλέπτειν, άρπάζεσθαι, βιάζεσθαι, τά χΐσχιστα κάλλιστα, die Lösung im „Ringkämpfer" gefunden und seinen Vorschlag (Hermes 34. 1899 S. 219, 2), in den auf das Rätsel folgenden Worten ήν πάλαι ταϋτα das ή ν in èv zu ändern, hat Diels in den Text aufgenommen, also: „dies geschieht im Ringkampf". Obwohl diese Lösung etwas künstlich erscheint, so wird sie außer durch die angeführte Stelle aus de diaeta auch dadurch noch gestützt, daß der platonische Gorgias ebenfalls die Gymnastik heranzieht, um das δικαίως und άδίκως χρησφαι seiner Kunst, der Rhetorik, zu veranschaulichen (456 D E). Daß κλέπτειν eine weitere Bedeutung hat als das deutsche ,stehlen', sieht man auch aus dem Scherz Xenophons (An. IV. 6, 15), wo es in absichtlichem Doppelsinn. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM Bemerkungen zu den Yorsokratikern und Sophisten. 581 für das heimliche Erschleichen einer Berghöhe durch eine Abteilung Soldaten gebraucht wird. Aehnlich können die Worte κλέπτειν und εξαπατάν auch hier Schliche und Kniffe in der Palästra andeuten. In § 14 schreibt Diels: καί το: <καττωυτό ó) πολλά άδικήσας άποθ-ανέτω (ysj). ó πολλά και δίκαια δια)> πραξάμενος. Den unmittelbar vorhergehenden Beispielen: „ein gerechter Mann ist auch ungerecht, ein großer auch klein" entspricht aber nicht der Gedanke: „es soll hingerichtet werden, wer viel unrecht getan hat und auch wer viel recht getan hat" sondern: „wer viel unrecht getan hat als einer der auch viel recht getan hat". Der hinzurichtende Verbrecher ist gemäß der bekämpften Anschauung zugleich gerecht, also seine Gerechtigkeit, die mit seinem Unrecht identisch ist, der Grund seiner Strafe. Daher würde ich schreiben: άποθ-ανέτω πολλά και δίκαια δ:απραξάμενος. Euenos. Bei dem der sophistischen Skepsis im Sinn des Protagoras ergebenen Elegiker (fr. 1. B) fallen mir zwei Berührungen mit demokritischen Gedanken auf : Fr. 6 (η δέος ή λύπη παις πατρί πάντα χρόνον) entspricht der Abneigung Demokrits gegen das Familienleben: fr. 275 (132); 276 (131); ähnlich auch Antiphon fr. 49 (12) Ende. Geradezu wie Entlehnung von Demokrit fr. 33 (95) erscheint der Gedanke des fr. 9, daß lange Uebung in einer Sache zur zweiten Natur werde. Schöntal a. d. Jagst. W. Nestle. Brought to you by \| Universitaetsbibliothek Frankfurt/Main Authenticated Download Date \| 12/20/17 10:08 PM
https://openalex.org/W2136659006	https://ms.copernicus.org/articles/4/21/2013/ms-4-21-2013.pdf	English	null	Sub-modeling approach for obtaining structural stress histories during dynamic analysis	Mechanical sciences	2,013	cc-by	9,592	T. T. Rantalainen, A. M. Mikkola, and T. J. Bj¨ork Lappeenranta University of Technology, Department of mechanical engineering, Lappeenranta, Finland Correspondence to: T. T. Rantalainen (tuomas.rantalainen@lut.ﬁ) Received: 15 November 2012 – Accepted: 26 January 2013 – Published: 11 February 2013 T. T. Rantalainen, A. M. Mikkola, and T. J. Bj¨ork Lappeenranta University of Technology, Department of mechanical engineering, Lappeenranta, Finland Correspondence to: T. T. Rantalainen (tuomas.rantalainen@lut.ﬁ) Received: 15 November 2012 – Accepted: 26 January 2013 – Published: 11 February 2013 Abstract. Modern machine structures are often fabricated by welding. From a fatigue point of view, the struc- tural details and especially, the welded details are the most prone to fatigue damage and failure. Design against fatigue requires information on the fatigue resistance of a structure’s critical details and the stress loads that act on each detail. Even though, dynamic simulation of ﬂexible bodies is already current method for analyz- ing structures, obtaining the stress history of a structural detail during dynamic simulation is a challenging task; especially when the detail has a complex geometry. In particular, analyzing the stress history of every structural detail within a single ﬁnite element model can be overwhelming since the amount of nodal degrees of freedom needed in the model may require an impractical amount of computational eﬀort. The purpose of computer simulation is to reduce amount of prototypes and speed up the product development process. Also, to take operator inﬂuence into account, real time models, i.e. simpliﬁed and computationally eﬃcient models are required. This in turn, requires stress computation to be eﬃcient if it will be performed during dynamic simulation. The research looks back at the theoretical background of multibody simulation and ﬁnite element method to ﬁnd suitable parts to form a new approach for eﬃcient stress calculation. This study proposes that, the problem of stress calculation during dynamic simulation can be greatly simpliﬁed by using a combination of Floating Frame of Reference Formulation with modal superposition and a sub-modeling approach. In prac- tice, the proposed approach can be used to eﬃciently generate the relevant fatigue assessment stress history for a structural detail during or after dynamic simulation. Proposed approach is demonstrated in practice using one numerical example. Even though, examples are simpliﬁed the results show that approach is applicable and can be used as proposed. Mechanical Sciences Open Access Open Access Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ doi:10.5194/ms-4-21-2013 © Author(s) 2013. CC Attribution 3.0 License. Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ doi:10.5194/ms-4-21-2013 © Author(s) 2013. CC Attribution 3.0 License. Published by Copernicus Publications. 1 Introduction Workﬂow of a design process with integrated fatigue analysis. Figure 2. A crane with numerous discontinuities, which can be prone to fatigue. taken into account automatically. To perform mutual inte- gration, eﬃcient methods for performing fatigue analysis are required. diagnosis (Korkealaakso et al., 2006). Moreover, this stress history can further be used as initial data for the fatigue anal- ysis. In addition to eﬃcient stress history calculation, the pro- posed combined methods strategy oﬀers other beneﬁcial fea- tures, such as the possibility of attaching structural details to a simulation model without modifying the simulation model. Furthermore, since structural detail does not aﬀect the over- all behavior of a simulation, the number of details can be changed arbitrary and separately. Dynamically loaded structures such as booms are typically manufactured by welding. By deﬁnition, ﬂuctuating loads re- sult in fatigue damage to a structure. Without post-weld treat- ment, welds are prone to fatigue (Maddox, 1991; Haagensen and Maddox, 2011). In dynamic analysis, structural details that do not aﬀect the dynamic behavior of a structure are usu- ally neglected. Typically, this means that stress raisers are not analyzed in dynamic simulation even though they might be a possible location for fatigue. If treated separately, more work is required in fatigue analysis (Fig. 1 left). In Fig. 2, an ex- cavator crane is depicted to illustrate the possible locations that might be vulnerable to fatigue and should be taken into consideration when predicting the fatigue life of a crane. To approximate, eﬃciently, the fatigue life of a structure or structural detail, some simplifying assumptions must be made. These assumptions decrease the quality of fatigue life estimation, especially in high frequency loading. For this work, a linear strain-stress relationship is assumed. In ad- dition, presumptions have been made related to cumulative damage counting. For example, since fatigue damage can be linearly accumulated, structural failure is predicted when the entire available fatigue life is consumed. The focus here is on linear deformation. The strategy considers plastic defor- mation only local to the tip of a crack. Material nonlinearities fall outside the scope of this work. However, fatigue life esti- mation gives useful information when comparing alternative structures. For stress history prediction, a welded structure requires particular attention. In addition, as computational power in- creases, there is increasing interest in predicting fatigue life for dynamically loaded structures. 1 Introduction Using dynamic simulation to determine stresses for ﬂexi- ble bodies also provides an opportunity to predict the fatigue life of a structure in practical applications. Currently in engi- neering applications, the prediction of fatigue life for a struc- ture is thought to be a separate stage of design due its com- plexity and computational burden. If fatigue life prediction is implemented eﬃciently in multibody codes, it could be used throughout all stages of design without dedicating discrete designing stages to it. These two scenarios of the use of the fatigue analysis are depicted in Fig. 1. Multibody dynamic simulation represents a remarkable im- provement in predicting machine performance compared to previous methods, which are often based on very sim- pliﬁed analytical models in combination with large safety margins for model uncertainties or empirical testing. With the development of more computationally powerful comput- ers over recent decades, dynamic simulation has increas- ingly become a standard tool for comprehensive machine design. Furthermore, this continuously increasing computa- tional power, combined with the availability of increasingly advanced codes, oﬀers more possibilities for the dynamic analysis of complex structures. Instead of being an explicit design step, fatigue analysis can be integrated into the design process (Fig. 1 right). This makes fatigue analysis an integrated part of the design pro- cess, and from a design point of view, fatigue analysis is T. T. Rantalainen et al.: Simulation of structural stress 22 22 T. T. Rantalainen et al.: Simulation of structural stress Beginning Design step 1 Design step 2 Design step n Prototype Fatigue analysis OK? Prototype Design step n Design step 2 Design step 1 Beginning Fatigue analysis OK? Done Design Figure 1. Workﬂow of a design process with integrated fatigue analysis. Figure 2. A crane with numerous discontinuities, which can be Beginning Design step 1 Design step 2 Design step n Prototype Fatigue analysis OK? Prototype Design step n Design step 2 Design step 1 Beginning Fatigue analysis OK? Done Design Figure 1. Workﬂow of a design process with integrated fatigue analysis. Beginning Design step 1 Design step 2 Design step n Prototype Fatigue analysis OK? Prototype Design step n Design step 2 Design step 1 Beginning Fatigue analysis OK? Done Design Figure 1. Workﬂow of a design process with integrated fatigue analysis. Figure 2. A crane with numerous discontinuities, which can be prone to fatigue. Beginning Beginning Design Prototype Figure 1. 1 Introduction Even though computa- tional capabilities have been greatly increased, the need re- mains for using coordinate reduction methods, especially in the case of large and complicated structures. Multibody dy- namic simulation can be used to analyze the dynamics of complex mechanical systems. It can also be used to deter- mine dynamic loads or even stresses for further fatigue analy- sis. Obtaining stress data for fatigue analysis from multibody system simulation is a main component of this work. Mech. Sci., 4, 21–31, 2013 2 The ﬂoating frame of reference formulation Position of particle Pi in a ﬂexible body. the frame of reference, large displacements and rotations can be described. The deformations of a ﬂexible body in rela- tion to its frame of reference can be described with a number of methods, but in the present study, deformation can be de- picted with orthonormalized Craig-Bampton modes (Craig Jr. and Bampton, 1968). In it, eigenmodes are used together with static modes to describe structural deformation. The modes can be obtained using a ﬁnite element method. θi 0 2 + θi 1 2 + θi 2 2 + θi 3 2 = 1 (3) (3) The ﬁrst time-derivative of the Euler parameters ˙θ and the angular velocity vector ωi has the following linear connec- tion. The formulation separates the deformation of the body from the reference motion. The dynamics of the body can be generated using reference motion that is superposed by the deformation of the body. The interaction between the ref- erence motion and deformation is accounted for with a mass matrix and quadratic velocity vector. This permits even mass distribution and inertia modeling (Shabana, 1998). ωi = G iT ˙θi ωi = G iT ˙θi (4) (4) Matrices Ai, and G i depend on the selected generalized co- ordinates. Using Euler parameters, the matrix G i can be ex- pressed as. Matrices Ai, and G i depend on the selected generalized co- ordinates. Using Euler parameters, the matrix G i can be ex- pressed as. Figure 3 illustrates the position of particle Pi within a ﬂex- ible body i. In the undeformed state, the position of the parti- cle in the local reference frame of the body can be determined by vector uiP 0 . G i =  −θi 1 θi 0 θi 3 −θi 2 −θi 2 −θi 3 θi 0 θi 1 −θi 3 θi 2 −θi 1 θi 0 , (5) (5) As body i is deformed (Fig. 3), the position of particle Pi changes according to the vector uiP f . The global reference frame is represented (Fig. 3) using Cartesian coordinates X, Y, and Z. Respectively, the local reference frame of body i consists of coordinates xi, yi, and zi. 2 The ﬂoating frame of reference formulation A new strategy developed in the proposed method com- bines three commonly used engineering approaches: the ﬁ- nite element method, the ﬂoating frame of reference formu- lation, and the sub-modeling approach. This strategy can be used when carrying out dynamic simulation to obtain, eﬃ- ciently, the stress history of an arbitrary notch. In literature it is shown that computer simulations can be used for fault The ﬂoating frame of reference formulation is typically ap- plicable to systems with large displacements and rotations and small deformations, even though the method can be used for large deformation problems (Wallrapp and Wiedemann, 2003). The method is based on describing the deformations of a ﬂexible body with respect to a frame of reference. With www.mech-sci.net/4/21/2013/ T. T. Rantalainen et al.: Simulation of structural stress T. T. Rantalainen et al.: Simulation of structural stress 23 T. T. Rantalainen et al.: Simulation of structural stress Figure 3. Position of particle Pi in a ﬂexible body. if possible, the equation results in an inﬁnite series that de- scribes the deformations. For computational reasons, the inﬁ- nite series cannot be applied to the analysis of ﬂexible bodies. In practical application, the vector uiP f is described using the ﬁnite element method. i The rotation matrix Ai using Euler parameters can be for- mulated as follows. Ai =  1 2− θi 2 2 − θi 3 2 θi 1θi 2−θi 0θi 3 θi 1θi 3+θi 0θi 2 θi 1θi 2+θi 0θi 3 1 2− θi 1 2 − θi 3 2 θi 2θi 3−θi 0θi 1 θi 1θi 3−θi 0θi 2 θi 2θi 3+θi 0θi 1 1 2− θi 1 2 − θi 2 2 ,  (2) (2) where θi 0, θi 1, θi 2, and θi 3 are Euler parameters. In this study, Euler parameters are used to avoid singular conditions, which can occur when Euler or Bryant angles are used (Nikravesh and Chung, 1982). The following mathematical constraint must be taken into consideration when Euler parameters are applied. where θi 0, θi 1, θi 2, and θi 3 are Euler parameters. In this study, Euler parameters are used to avoid singular conditions, which can occur when Euler or Bryant angles are used (Nikravesh and Chung, 1982). The following mathematical constraint must be taken into consideration when Euler parameters are applied. Figure 3. www.mech-sci.net/4/21/2013/ 2 The ﬂoating frame of reference formulation By employing a modal reduction method, the deformation vector uiP f can be expressed in modal coor- dinates with a shape matrix. uiP f = ΦiP R pi (7) uiP f = ΦiP R pi (7) ˙qi = h ˙R iT ωiT ˙piT i (12) ΦiP R is the modal matrix whose columns describe the trans- lation of particle Pi within the assumed deformation modes of the ﬂexible body i (Shabana, 2005), and pi is a vector of elastic coordinates. In general, the complete modal matrix ΦiP for body i obtained from the ﬁnite element method con- tains the location translation and orientation of particle Pi. In multibody dynamics, the modal matrix should separate translation and orientation descriptions into their own com- ponents. ΦiP R is the modal matrix whose columns describe the trans- lation of particle Pi within the assumed deformation modes of the ﬂexible body i (Shabana, 2005), and pi is a vector of elastic coordinates. In general, the complete modal matrix ΦiP for body i obtained from the ﬁnite element method con- tains the location translation and orientation of particle Pi. In multibody dynamics, the modal matrix should separate translation and orientation descriptions into their own com- ponents. (12) The velocity of particle Pi can be determined by diﬀeren- tiating Eq. (1) with respect to time as follows. The velocity of particle Pi can be determined by diﬀeren- tiating Eq. (1) with respect to time as follows. ˙riP = h I −AiG i ˜uiP f AiΦiP R i  ˙R i ˙θ i ˙pi  (13) (13) Note the vector, in the right hand of Eq. (13), describes the velocity of the generalized coordinates of a ﬂexible body i. Diﬀerentiating the velocity of a particle Eq. (13) with respect to time, the acceleration of a particle can be written in this manner. The orthogonal shape matrix can be formulated from the eigenmodes of the body. Typically, the shape superposition technique yields acceptably accurate results even though only a few diﬀerential equations are applied. By approxi- mating Eq. (7) with an np number of modal coordinates, the deformation vector uiP f for a particle Pi can be written as fol- lows. 2 The ﬂoating frame of reference formulation ¨riP = h I −Ai ˜u iPG iAi ΦiP R RiP i  ¨R i ¨θ i ¨pi  + h 0 −Ai ˜ω i ˜u iPG i 2Ai ˜ω iΦiP R i  ˙R i ˙θ i ˙pi , (14) uiP f ≈ np X j=1 ϕi R, jpi j = ΦiP R pi (8) (14) (8) pi j is one modal coordinate in the modal coordinate vector that corresponds to the modal shape j. Rotations due to body deformation do not have any direct use in the ﬂoating frame of reference formulation, and therefore they are usually ig- nored. However, rotation modes can be used in the descrip- tion of constraint equations applied to rotational degrees of freedom (Korkealaakso et al., 2009). Rotational modes are used here to connect sub-models to large-scale models. With the rotational modal matrix ΦiP θ , the rotation change εiP f re- sulting from deformation can be approximated as follows. pi j is one modal coordinate in the modal coordinate vector that corresponds to the modal shape j. Rotations due to body deformation do not have any direct use in the ﬂoating frame of reference formulation, and therefore they are usually ig- nored. However, rotation modes can be used in the descrip- tion of constraint equations applied to rotational degrees of freedom (Korkealaakso et al., 2009). Rotational modes are used here to connect sub-models to large-scale models. With the rotational modal matrix ΦiP θ , the rotation change εiP f re- sulting from deformation can be approximated as follows. where ¨R i, ¨θ iE, and ¨pi are accelerations of translational coor- dinates, Euler parameters, and modal coordinates of body i. According to the D’Alembert principle, inertial forces can be treated as external forces, thus forces of the body i can be written as follows. Fi = Z Vi ρi¨riPdVi, (15) (15) εiP f ≈ np X j=1 ϕi θ, jpi j (9) where ρi is density and Vi is the volume of a body i, respec- tively. The virtual work done by the inertial forces can be represented as. (9) A rotation matrix AiP f that describes orientation due to de- formation at the location of particle Pi with respect to the reference frame can be composed like this. 2 The ﬂoating frame of reference formulation Therefore, the location of the particle in a global reference frame can be deﬁned with the vector riP as follows: where G i is the transformation matrix that relates the angular velocity ωi of a body and the ﬁrst time derivative of the Euler parameter. Using the model reduction method, the position of an arbitrary particle Pi in the global coordinate system can be expressed as. riP = Ri + Ai uiP 0 + ΦiP R pi (6) (6) riP = Ri + Ai uiP 0 + uiP f , (1) (1) Equation (6) is determined using a collection of modes. The vector uiP 0 and the modal matrix ΦiP R are constant with time. Consequently, they only need to be calculated once, at the beginning of the simulation. where Ri is translation of the local reference coordinate system of body i in the global coordinate system, and matrix Ai is the rotation matrix, which is expressed here in terms of four Euler parameters. The ﬁnite element model often consists of a large number of nodal degrees of freedom, and the use of large ﬁnite ele- ment models to describe ﬂexibility may be computationally ineﬃcient. For this reason, the ﬂoating frame of reference formulation is often used together with a modal reduction method in which the deformation is described with structural modes. Modes may be the presumed forms of deformation, In Eq. (1), uiP 0 is the position vector of particle Pi in the local reference coordinate system for the undeformed con- ﬁguration, and uiP f is the position vector in the local refer- ence coordinate system for the deformed conﬁguration. The behavior of the vector uiP f can be described with a series of parallel diﬀerential equations. By separating the variables, www.mech-sci.net/4/21/2013/ Mech. Sci., 4, 21–31, 2013 24 T. T. Rantalainen et al.: Simulation of structural stress has no direct use in formulating equations of motion, but the rotation may be needed to describe the constraint equations that are applied to rotational degrees of freedom. Taking into account the relation between ﬁrst time derivative of Euler an- gels and angular velocity of the body i, Eq. (4), the general- ized velocity vector of the ﬂexible body i can be written as follows. but most often, they are eigenmodes of structural vibrations. The eigenmodes can be obtained from a ﬁnite element model of the structure. 2.1 Fatigue Fatigue is a failure that occurs after cyclic loading, and it is a common cause of structural fracture. Fatigue damage is one of the most common faults in dynamically loaded structures. In principle, the entire development of fatigue damage can be described as follows: one or more cracks form in the material, and the cracks grow until fatigue failure takes place. The elastic forces can be described using modal coordi- nates and the stiﬀness matrix in modal coordinates Ki. The stiﬀness matrix in modal coordinates can be obtained using component mode synthesis. The virtual work of the elastic forces can be expressed as. A fundamental design objective for any dynamically loaded structure exposed to cyclic loading or vibration is to avoid fatigue failure throughout its service life. Welding is one of the most eﬃcient methods used to manufacture struc- tures. Cranes, vehicle frames, and machines are just some examples of welded structures that are dynamically loaded. As a structural detail, a weld is initially very prone to fa- tigue due to the notch eﬀect, high tensile residual stresses, and welding ﬂaws. High-strength steels, which are seeing in- creasing use, are even more sensitive to this phenomenon. In general, high strength steels are chosen to achieve larger payloads with more slender structural elements. As a trade oﬀ, the slender structures are subject to increased nominal stresses and welds become more prone to fatigue due to the higher stresses. Typically, structural welds are at or near areas of structural discontinuity. The weld itself is a local discon- tinuity. Furthermore, welding processes typically introduce ﬂaws in the weld or weld toe such as undercuts, the inclu- sion of impurities, and cold laps. These ﬂaws are sources of incipient cracking. δWi f = δpTKipi (21) δWi f = δpTKipi (21) The vector of elastic forces can be represented as follows. Qi f =  0 0 Kipi  (22) (22) In multibody dynamics, diﬀerent types of joints between bodies are accounted for with kinematic constraints applied on generalized coordinates. Algebraic equations are used for the description of constraints between bodies. By examining only holonomic constraints, constraint equations can be ex- pressed as follows. C(q) = 0 (23) (23) C(q) = 0 2 The ﬂoating frame of reference formulation δWi = Z Vi ρiδriPT¨riPdVi (16) (16) AiP f = I + ˜εiP f (10) AiP f = I + ˜εiP f The virtual displacement of the position vector δriP can be expressed as. (10) I is (3 × 3) identity matrix. The ˜ symbol above a variable indicates the skew-symmetric form. The orientation at the location of particle Pi within the frame of reference can be expressed as follows. δriP = δriP δqi δqi = h I −Ai ˜u iPG iAi ΦiP R i δqi, (17) (17) By substituting the virtual displacement Eq. (17) into the equation for virtual work (Eq. 16) and separating terms, the following equation can be obtained. viP f = AiP f viP 0 (11) viP f = AiP f viP 0 (11) viP 0 is the orientation of the location of particle Pi in the un- deformed state. The description of the rotation of the node viP 0 is the orientation of the location of particle Pi in the un- deformed state. The description of the rotation of the node δWi = δqi h Mi ¨qi+Qivi , (18) (18) Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013 Mech. Sci., 4, 21–31, 2013 T. T. Rantalainen et al.: Simulation of structural stress 25 T. T. Rantalainen et al.: Simulation of structural stress where δqi is virtual change of the generalized coordinates, Qiv is quadratic velocity vector of body i, and Mi is the mass matrix. where δqi is virtual change of the generalized coordinates, Qiv is quadratic velocity vector of body i, and Mi is the mass matrix. Equations (23) and (24) form a set of diﬀerential algebraic equations, which can be converted to ordinary diﬀerential equations (ODE) to solve for the dynamic response of the multibody system in the time domain. To be able to apply traditional ODE solvers to the system of equations, the con- straint equations must be diﬀerentiated twice with respect to time. The virtual work of externally applied forces can be de- ﬁned as. δWie = Z Vi δriPTFiPdVi= δqiTQie, (19) (19) ¨C(q, ˙q, ¨q) = Cq ¨q + Cq ˙q q ˙q = 0, (25) (25) where FiP is externally applied force per unit volume, and Qie is the vector of generalized forces, which can be expressed as follows. where Qc = − Cq ˙q q ˙q is the constraint force vector for the system. As a result, the ﬁnal matrix form of equations of mo- tion describing the system dynamics looks as following. Qie =  Qie R Qie θ Qie p , (20) (20) " M CT q Cq 0 #" ¨q λ # = " Qe + Qv −Q f Qc # (26) (26) where Qie R is translational components, Qie θ is rotational com- ponents, and Qie p is elastic components of the generalized force vector, respectively. 2.1 Fatigue C(q) = 0 In some applications, such as rotating axles, fatigue life is equivalent to the duration of the crack initialization stage. Because of the notch eﬀect, the crack in these applications quickly results in failure. For larger struc- tures, such as machine frames or many welded structures, cracks are present from the beginning, so fatigue life is the determined by the length of time it takes the initial cracks to propagate. rical shapes and their combinations combined with complex loading cases cannot be predeﬁned, especially if a speciﬁc but arbitrary level of accuracy is needed. Obviously, the gen- eral approach covering all kinds of geometrical combinations in the ﬁnite element method is to model them as they appear in the structure. Since stress values change drastically in the neighborhood of a structural discontinuity, a reﬁned element mesh is required, which will lead to a large number of de- grees of freedom. This approach is impractical due to the computational burden, especially in the case of multibody simulation. The fatigue life of a structure under dynamic load can be estimated by assuming it to have some initial amount of fa- tigue endurance and then assuming that one load cycle will result in fatigue damage of some amount. This is commonly known as Palmgren-Miner’s rule (Miner, 1945). It was sug- gested that fatigue damage could be accumulated linearly for a certain amplitude value. Finally, when all fatigue endurance has been depleted, failure is expected. A large amount of fa- tigue test data can be found in the literature, and stress histo- ries can be obtained through experimentation of by simula- tion. The sub-modeling approach is commonly used, and it can overcome the previously mentioned problems. With the ap- proach some new problems arise, but they will be discussed later. In principal, a sub-model is a model inside of or on top of a large scale model that describes a certain portion of the large scale model. It can be used to attach a locally re- ﬁned element mesh to the larger scale model, which does not need to be changed. In sub-modeling, the simpliﬁed struc- tural model is complemented by a more reﬁned sub-model of structural details. The sub-models do not inﬂuence the opera- tion of the system but get their boundary conditions and load- ing data from the larger simpliﬁed model. C(q) = 0 where, C is the constraint vector for the system. Equations of motion may be formulated using the widely known Lagrange method, in which kinematic constraints are accounted for as supplementary algebraic equations with the help of Lagrange multipliers. The method is called global formulation since it does not diﬀerentiate between open and closed kinematic chains, as topological methods do. After employing the con- cept of virtual work to externally applied forces and then in- troducing constraints with help of Lagrange multipliers, the equation of motion can be written in the form of a diﬀerential algebraic equation (DAE). Empirical testing with actual parts or complete systems is time consuming and laborious. Traditionally, the approach to avoiding fatigue failures in a new system is to fatigue test speciﬁc structural details before integrating them into the system design. In fact, many kinds of typical welded de- tails can be found commonly in the literature including ﬁl- let welds, corner joints, and butt joints. This approach has several weaknesses. For example, all pertinent structural de- tails to be used should be tested under various loading condi- tions if the fatigue evaluation is to be comprehensive, and the M¨q + Kq + CT qλ = Qe + Qv −Qf (24) (24) Mech. Sci., 4, 21–31, 2013 26 T. T. Rantalainen et al.: Simulation of structural stress T. T. Rantalainen et al.: Simulation of structural stress Figure 4. The sub-modeling approach for attaching dissimilar meshes. approach ignores other parameters that may relate to a par- ticular joint, such as the number of weld beads or other geo- metrical and technical details. Obviously, for a complex sys- tem with an arbitrary number of structural details subjected to various multi-axial loading scenarios, it is practically im- possible to use this traditional approach. More recently, numerical methods have been developed to estimate fatigue life making empirical testing unneces- sary and allowing the designer to more eﬀectively consider the eﬀects of fatigue (Haagensen and Maddox, 2011). To- day, the ﬁnite element method has become a standard ap- proach for estimating the fatigue life of a structure. Nonethe- less, even though computational capacity is increasing all the time, applying the ﬁnite element approach to a complex structure subject to dynamic multi-axial loading presents an overwhelming computational burden. Figure 4. The sub-modeling approach for attaching dissimilar meshes. Fatigue design approaches can be diﬀerentiated according to how cracks initiate. C(q) = 0 The sub-modeling approach is also referred to in the literature as the global/local approach (Knight et al., 1991). The sub-modeling approach can be used to connect dissimilar meshes together as shown schematically in Fig. 4. www.mech-sci.net/4/21/2013/ Mech. Sci., 4, 21–31, 2013 3 Sub-modeling Arriving at an optimized design and understanding precisely how internal stresses vary over time leads to struc- tures with improved fatigue life and whole systems that are safer. In the Floating Frame of Reference Formulation, bod- ies are loaded by numerous unique loads and moments; ex- ternal forces, constraint forces, and inertial forces, for ex- ample. Forces produce deformation, and deformations set up internal stresses. The prediction of local stresses using dy- namic simulation reveals structural weaknesses in the early design stages. In addition, dynamic simulation can analyze stress peaks in extreme cases, such as random overloading or component failure. Coupling between the sub-model and the large-scale model is assumed to be one directional, i.e., it is assumed that the behavior of the reduced model in dynamic simulation is not aﬀected by the sub-model. That means, the large scale model is complemented with a sub-model of the desired de- tail and it does not aﬀect the system’s overall stiﬀness. This crucial simpliﬁcation makes dynamic simulation and stress calculation independent from each other. Therefore, the com- putation can be straight forwardly parallelized. Displacement boundary conditions of the sub-model, however, are acquired from the large-scale model. In the proposed approach, during dynamic analysis the general behavior of the structure is cal- culated with a simpliﬁed model and details are examined as a separate problem. For assessing fatigue loads on a structure, this assumption is suﬃcient since any signiﬁcant change in structural ﬂexibility due to crack growth occurs only very late in the total life of a structure. A stress history from a multibody dynamics simulation can be used as initial data for component dimensioning. Fur- thermore, it can provide loading data for the analyst that is otherwise diﬃcult to obtain. Finally, the stress history can be used as input for the fatigue analysis of the component. In such cases, one should make sure that simulated opera- tions describe the operating conditions of the machine with suﬃcient accuracy. With simulation, it is diﬃcult to describe the impact of statistical issues, such as component wear and operator usage habits, on component loading. On the other hand, simulation helps to understand the causes and eﬀects related to loading. This allows the use of optimization rou- tines in component dimensioning. 3 Sub-modeling A common cause of structural damage is local stress con- centration due to structural geometry. Practically all struc- tural damage occurs where one or more stress raisers are present. Problematic details are often combinations of sev- eral geometries that concentrate hazardous stresses. Stress raisers result from local discontinuities in real structural fea- tures; such as welds, attachments grooves or holes. Typically, small discontinuities are neglected in multibody simulation, since their eﬀect on overall behavior is relatively small com- pared to their contribution to computational burden. Tradi- tionally, the problem of stress raisers is solved by calculating nominal stress levels and then taking into account the eﬀect of stress raisers by applying predeﬁned stress concentration factors. The stress concentration factor concept cannot be a general approach, since it is obvious that all possible geomet- Sub-modeling is an approach that is used together with the ﬁnite element method to combine two diﬀerent ﬁnite el- ement meshes. There are several reasons the sub-modeling approach is powerful. It can be used to connect ﬁnite models into a larger assembly. The approach does not require meshes to be similar and even element types can diﬀer. Currently available methods do not require any coincident nodes. These beneﬁcial features can be utilized to combine separately con- structed models or reﬁne the element mesh in a certain area without taking care of reﬁning the mesh smoothly. In addi- tion, a sub-model can be changed easily without modifying Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ T. T. Rantalainen et al.: Simulation of structural stress 27 simulation. Both approaches have beneﬁts, but in general, the method based on ﬁnite-element post processing is more ac- curate (Arczewski and Fra¸czek, 2005). other parts of the model. Problems arise if the level of re- ﬁnement diﬀers signiﬁcantly between two diﬀerent models. A coarse mesh tends to be too stiﬀand displacements are un- derestimated, and if those underestimated displacements are used as boundary conditions for the reﬁned model, calculated stress levels will be non-conservative. For lightweight structures loaded by dynamic forces, ac- curate dynamic simulation is necessary to guarantee long- term reliability (fatigue life), the accuracy of control, and sys- tem usability. Deformation, even small deformations, must be taken into account to achieve the needed simulation pre- cision. 3 Sub-modeling Simulation and measure- ment on a real-life machine can thus be considered to support each other, and using them together can help to reach an op- timal solution. A sub-model can make the overall structure stiﬀer, or it can only carry boundary conditions without aﬀecting the stiﬀness of a large-scale structure. Sub-models describe structural de- tails, and only those that are interesting from a design point of view can be attached to a ﬂexible multibody model. Con- nections between a multibody model and its sub-models are one-directional, guaranteeing that the multibody simulation is not aﬀected by the sub-model. For instance, if a ﬂexible multibody model is a part of a real-time simulation, it will be a real-time simulation even if sub-models are active. In this paper, one way of attaching sub-models to the larger multi- body model is introduced, but the concept is general. The literature provides a number of alternative approaches to determining stress histories from multibody simulation. The ﬁrst to combine the multibody dynamic approach and stress calculation was Melzer (1996). Yim and Lee (1996) obtained dynamic stress histories by using constraint forces solved in a multibody system. Dietz et al. (1997) described an approach for using multibody simulation to obtain all forces for ﬁnite element analysis. They also selected the most severe time steps for which stresses of the complete structure were later analyzed in ﬁnite element code. Later, Dietz et al. (1998) combined multibody simulation and fa- tigue life prediction. They obtained the load history from multibody simulation and calculated stress histories for se- lected locations using a stress load matrix. Stress histories were analyzed in a post-processing stage to predict fatigue life. Dietz (1999) presented a systematic way of combin- ing multibody dynamic simulation and fatigue analysis us- ing stress component modes. Claus (2001) generalizes the deformation-based stress recovery approach to multipurpose www.mech-sci.net/4/21/2013/ 4 Stress in multibody dynamic simulation This work focuses on welded structures, which can be studied un- der the assumption of cumulative damage counting. Welded structures without any post-weld treatments have large ten- sile residual stresses, even nominally as large as the yield strength of the material. In this case, initial compressive load- ing closes an incipient crack and residual stresses open it again. This leads to a situation in which even a fully compres- sive loading cycle will result in full fatigue damage. For the proposed method, linear hot spot extrapolation (Poutiainen et al., 2004) is selected, for both simplicity and robustness in use. ﬁnite element codes. More recently, Jun et al. (2008) used the modal stress recovery approach to obtain stress for fa- tigue analysis. They also discussed the reliability of fatigue life calculation. Lee et al. (2009) studied the fatigue life for various parts of a guideway vehicle by coupling multibody dynamics and fatigue analysis. They determined stresses us- ing the modal stress method or quasi-static force method de- pending on the loading conditions of the part. Braccesi and Cianetti (2005) used a modal approach to recover stresses. Arczewski and Fra¸czek (2005) compared and discussed dif- ferences between force-based and deformation-based stress recovery methods in MBS. More recently, Tobias and Eber- hard (2011) obtained stresses using a reduced MBS model and stress modes. They concluded the stress state in any particular point of a ﬂexible body could be expressed as a linear combination of global shape functions for stresses and nodal coordinates.In experiments, fatigue life predic- tion is mainly related to uniaxial cyclic loading. This leads to discussion about damage hypothesis, and the question arises about which damage hypothesis should be used. This work focuses on welded structures, which can be studied un- der the assumption of cumulative damage counting. Welded structures without any post-weld treatments have large ten- sile residual stresses, even nominally as large as the yield strength of the material. In this case, initial compressive load- ing closes an incipient crack and residual stresses open it again. This leads to a situation in which even a fully compres- sive loading cycle will result in full fatigue damage. For the proposed method, linear hot spot extrapolation (Poutiainen et al., 2004) is selected, for both simplicity and robustness in use. Figure 5. 4 Stress in multibody dynamic simulation Crane parameters and the placement of strain gauges (D) – Letters A, B, and C refer to the joint locations and m is the added mass of 110 kg. Strain gauges are attached close to the welded notch so the ﬁrst is 0.4t and the second is 1.0t from the notch. Plate thick- ness t is 4 mm. A fundamental motivation of the introduced approach is to keep the dynamic simulation as numerically eﬃcient as pos- sible. Therefore, the hosting structural model is simpliﬁed and is then reduced with the Craig-Bampton method (Craig Jr. and Bampton, 1968). Even though the proposed approach is general, in this example boom-type structures that can be eﬃciently described with beam elements are studied. With beam models, obtaining boundary conditions for sub-models is straightforward. In the proposed approach, multibody sim- ulation is used for producing displacement data for the sub- model, which is then analyzed and fatigue data is obtained. Problem about computational burden is tried to overcome by combining sub-modeling and multibody dynamic ap- proach in order to join computationally eﬃciency of multi- body dynamic approach and accuracy of ﬁnite element method in observing damaging loads. Sub-model of the crane is shown in Fig. 6. Black circle shows the area where hot spot stress is obtained. Nodes used in hot spot extrapolation are show in Fig. 7. 4 Stress in multibody dynamic simulation Dimensioning components require information about load- ing and more accurately, stresses. Even though, stresses can somehow be obtained from rigid body dynamics by using simulated forces as force boundary conditions in ﬁnite ele- ment method. In general, to obtain structural stresses, struc- tural ﬂexibility should be taken into account. Concept of multibody dynamics gives attractive approach to simulate real operating conditions and thus obtain realistic loading conditions. Stress recovery methods for a ﬂexible multibody system can be divided into two main categories. One is the stress- mode-based method, which determines the body’s stress state using a linear combination of stress modes and elastic coordinates. The other is the ﬁnite-element post-processing method, in which stresses are calculated by a ﬁnite-element code using forces or displacements obtained from multibody Mech. Sci., 4, 21–31, 2013 T. T. Rantalainen et al.: Simulation of structural stress Figure 5. Crane parameters and the placement of strain gauges (D) – Letters A, B, and C refer to the joint locations and m is the added mass of 110 kg. Strain gauges are attached close to the welded notch so the ﬁrst is 0.4t and the second is 1.0t from the notch. Plate thick- ness t is 4 mm. T. T. Rantalainen et al.: Simulation of structural stress 28 T. T. Rantalainen et al.: Simulation of structural stress ﬁnite element codes. More recently, Jun et al. (2008) used the modal stress recovery approach to obtain stress for fa- tigue analysis. They also discussed the reliability of fatigue life calculation. Lee et al. (2009) studied the fatigue life for various parts of a guideway vehicle by coupling multibody dynamics and fatigue analysis. They determined stresses us- ing the modal stress method or quasi-static force method de- pending on the loading conditions of the part. Braccesi and Cianetti (2005) used a modal approach to recover stresses. Arczewski and Fra¸czek (2005) compared and discussed dif- ferences between force-based and deformation-based stress recovery methods in MBS. More recently, Tobias and Eber- hard (2011) obtained stresses using a reduced MBS model and stress modes. They concluded the stress state in any particular point of a ﬂexible body could be expressed as a linear combination of global shape functions for stresses and nodal coordinates.In experiments, fatigue life predic- tion is mainly related to uniaxial cyclic loading. This leads to discussion about damage hypothesis, and the question arises about which damage hypothesis should be used. Mech. Sci., 4, 21–31, 2013 Table 2. Geometrical properties of cross section of the crane and material properties of the crane. Table 2. Geometrical properties of cross section of the crane and material properties of the crane. Table 2. Geometrical properties of cross section of the crane and material properties of the crane. Property Value Unit Proﬁle height 0.15 m Proﬁle width 0.10 m Proﬁle area 1.9 × 10−3 m2 Plate thickness 4 mm Area moment of inertia (yy) 6.17 × 10−6 m4 Area moment of inertia (zz) 3.29 × 10−6 m4 Elastic modulus 210 GPa Poisson’s ratio 0.3 Density 7850 kg m−3 Figure 6. FE-model of the sub-model – the black circle indicates the location of the hot-spot nodes. Figure 6. FE-model of the sub-model – the black circle indicates the location of the hot-spot nodes. Node 0.4 Node 1.0 Figure 7. Hot spot extrapolation nodes on a sub-model. Node 0.4 Node 1.0 cal properties of cross section of the crane are presented in Table 2. cal properties of cross section of the crane are presented in Table 2. 5.2 Stress history This displacement data is then used as a boundary condition for the sub-model. The sub-model contains the stress concen- trations where fatigue damage can possibly occur. Nodal dis- placement history is applied as boundary conditions on the sub-model as a sequential set of static boundary conditions. The ﬁnite element mesh of the sub-model has 1800 linear, brick elements and 250 rigid, massless beams. Figure 7. Hot spot extrapolation nodes on a sub-model. The gray lines seen on right side, in Fig. 6, represent rigid and massless beam webs and connect the cross section of the sub-model to the dynamical model via attachment nodes. Two attachment nodes are located on the middle line of the cross section. The use of rigid beam webs keeps the boundary cross section of the sub-model planar. This clearly simplify- ing assumption is made because in the beam model the cross section is assumed to remain planar. Generally, the problems with beam elements are the higher order deformations of cross section, such as warping and distortion, which are not included in low order beam elements. www.mech-sci.net/4/21/2013/ 5 Empirical example Fatigue damage typically originates from the points of dis- continuity of the structure, especially if there is residual stress aﬀecting the discontinuity. In order for the machine system to be simulated in real time, it must often be sim- pliﬁed, and details irrelevant in terms of structural stress ne- glected. The modeling of small details, such as welded clamp for a hydraulic pipe of a boom, increases the need for compu- tational eﬃciency, and such details have only a localized im- portance. To illustrate the method for obtaining stress history of a structural detail, a practical example of a simple crane is studied. The crane and the structural detail are depicted in Fig. 5. In this example the simpliﬁed dynamic simulation and relative accurate model of notch are combined together using sub-modeling techniques. The beam model represents the center line of a structural component. The sub-model is attached to the interpolated lo- cations of the reduced model via rigid and massless beams. Due to the use of rigid beam webs, the cross-section is as- sumed to remain planar at the boundary condition points. The eﬀect of this assumption, with respect to stresses in notch, is negligible due to Saint-Venant’s principle (von Mises, 1945). In dynamic simulation, translational and rotational displace- ments are solved as boundary conditions for the sub-model. In general, sub-models are attached at arbitrary locations of the structural component, thus nodal displacement interpola- tion should be used. The crane in Fig. 5 is hydraulically driven crane with one degree of freedom. Hydraulic system is modeled using the theory of lumped ﬂuid (Watton, 1989), in which hydraulic system is divided into separate volumes in where pressure is equally distributed. Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ T. T. Rantalainen et al.: Simulation of structural stress 29 Figure 6. FE-model of the sub-model – the black circle indicates the location of the hot-spot nodes. Table 1. Dimensions of the crane. Item x coordinate [m] y coordinate [m] Revolute joint A 0 0 Revolute joint B 0.32 −0.125 Revolute joint C 0 −0.925 Mass m 2.5 0 Welded detail D 0.65 0.075 Gauge 0.4 0.672 0.075 Gauge 1.0 0.674 0.075 Item x coordinate [m] y coordinate [m] Table 2. Geometrical properties of cross section of the crane and material properties of the crane. 6 Conclusions In this paper one approach of making fatigue analysis more usable among multibody simulation based product develop- ment is presented. In order to combine dynamic simulation and fatigue design this study introduces a novel approach for eﬃciently obtaining stress history from dynamic simulation. For fatigue assessment the hot spot or structural stress is often used (Niemi et al., 2006). The crane was loaded using measured position of a hydraulic cylinder, shown in Fig. 8 A linear surface extrapolation for hot spot stress (Poutiainen et al., 2004) has been performed for obtaining stress histories shown in Fig. 9. The hot spot structural stress at the edge of the welded discontinuity, is based on a linear extrapolation of surface stresses at nodes 4 mm and 10 mm from the edge of the discontinuity. Axial direction (x-direction) is selected for extrapolating hot spot stress on the edge of the notch. The selection of the direction of hot spot extrapolation is made based on the assumption that the majority of stresses are act- ing in the axial direction. This paper presents an approach in which the stress history for fatigue life estimation of an arbitrary structural disconti- nuity in a large-scale structure can be eﬃciently obtained in multibody simulation. In the proposed approach the structure is modeled with structured elements (i.e. planes or beams) in order to get rid small structural details to minimize nodal degrees of freedom. After that model is further reduced us- ing component mode synthesis, in numerical example, Craig- Bampton method was used, this model is called as reduced model. Reduced model is used to represent ﬂexible body in multibody simulation. Small structural details are modeled separately and are attached to reduced model using suitable methods. In this paper, in numerical examples sub-models were attached to reduced model using rigid beam webs. Sub- models were analyzed quasi-statically within ﬁnite element codes using displacements, obtained from dynamic simu- lation, as boundary conditions. This analysis can be made during the dynamic simulation or in post-processing phase. Computations involving sub-modeling allow the fatigue as- sessment calculation to be separated from the dynamic sim- ulation and structural details can be analyzed independently. The stress history obtained for the welded detail can later be processed using the rainﬂow counting algorithm, fatigue assessment or for any other post processing action. 5.1 Crane model composition and work cycle Measured position of the shaft of the hydraulic sylinder. Figure 9. Stress history of a notch at the crane. made linearly between the nodes. Since the sub-model only uses displacements obtained from the dynamic simulation as boundary conditions, it does not interfere with the overall be- havior of the model. 5.1 Crane model composition and work cycle The crane model contains four bodies, which are the crane, support and hydraulic cylinder and shaft for a cylinder. The lenght L of the crane is 4.5 m. depicted in Fig. 5. In addi- tion to those parts model has added mass m = 100 kg and a hydraulic system. Hydraulic circuit can be neglected while comparing results, since simulation model was driven by pre- deﬁned movement of the cylinder that is shown in Fig. 8. In Table 1, relevant dimensions and distances from revolute joint A along coordinates x, y, z, are presented. Since the sub-model and the dynamic model have over- lapping nodes, boundary conditions for the sub-model can be ﬁxed based on nodal deformation from the dynamic simu- lation. In the case of non-overlapping nodes, interpolation of nodal translational deformation and rotation deformation be- tween nodes is required. In this case, interpolation could be The crane model consists of a beam model with 20 nodes. Craig-Bampton method is used to reduce coordinates of the beam model. Two connection nodes were selected as a masted nodes for Craig-Bampton. Structural ﬂexibility was described using 10 lowest deformation modes. Geometri- www.mech-sci.net/4/21/2013/ Mech. Sci., 4, 21–31, 2013 30 T. T. Rantalainen et al.: Simulation of structural stress Position [mm] −60 −40 −20 0 20 40 60 Time [s] 0 1 2 3 4 5 Position of the cylinder Figure 8. Measured position of the shaft of the hydraulic sylinder. Hot spot stress [MPa] 20 30 40 50 60 70 80 90 Time [s] 0 1 2 3 4 5 Measured stress Simulated stress Figure 9. Stress history of a notch at the crane. T. T. Rantalainen et al.: Simulation of structural stress Hot spot stress [MPa] 20 30 40 50 60 70 80 90 Time [s] 0 1 2 3 4 5 Measured stress Simulated stress Figure 9. Stress history of a notch at the crane. T. T. Rantalainen et al.: Simulation of structural stress 30 30 Position [mm] −60 −40 −20 0 20 40 60 Time [s] 0 1 2 3 4 5 Position of the cylinder Figure 8. Measured position of the shaft of the hydraulic sylinder. Position [mm] −60 −40 −20 0 20 40 60 Time [s] 0 1 2 3 4 5 Position of the cylinder Figure 8. Measured position of the shaft of the hydraulic sylinder. Figure 8. www.mech-sci.net/4/21/2013/ Mech. Sci., 4, 21–31, 2013 6 Conclusions The results of this numerical experiment show that the developed method can be used to determine the stresses of a structural detail using a real-time simulation model. This method enables a wide variety of uses from determining stresses from positions that cannot be measured from the real machine to determining the best practices for machine oper- ation. Measuring bearing housing stresses, for example, in a real machine during an operation cycle is next to impossible. This method enables the determination of stresses during the entire operation cycle instead of just a suspected peak value. The method could be used to improve estimations on the ma- chine durability as well as improving the machine durability already in the machine design phase. A practical example would be to use virtual prototyping in the machine product development phase. The model could then be used to run a series of reference operation cycles while recording displace- ment data from a structure. The recorded data could then be used to run analysis on several crucial parts of the structure in order to determine the life expectancy under operating con- ditions as well as diﬀerent operators. In future work the integration of fatigue analysis and pro- duced stress history could be improved. The way how stress data is analyzed and compared to real fatigue test results dif- fers from the stress results that can be straightforwardly ob- tained from dynamic simulation. In numerical examples, this aspect is taken into account using linear surface extrapolation to estimate hot-spot stress. The problem using that approach is the diﬃculty of knowing the most probable crack growth direction. In reality, crack may change the direction of grow- ing depending on geometry and/or loading conditions. Also, Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ T. T. Rantalainen et al.: Simulation of structural stress 31 in future work methods of attaching sub-model into simu- lation model should be studied carefully. Possibility to use coordinate reduction for sub-models and what kinds of limi- tations it provides to attachment for sub-models. Korkealaakso, P., Mikkola, A., and Rouvinen, A.: Multi-body sim- ulation approach for fault diagnosis of a reel, Proceedings of the Institution of Mechanical Engineers, Part K: Journal of Multi- body Dynamics, 220, 9–19, 2006. References Maddox, S. J.: Fatigue strength of welded structures, Abington Pub- lishing, Cambridge, 2nd Edn., 1991. Arczewski, K. and Fra¸czek, J.: Friction Models and Stress Recovery Methods in Vehicle Dynamics Modelling, Multibody Syst. Dyn., 14, 205–224, 2005. Melzer, F.: Symbolic computations in ﬂexible multibody systems, Nonlinear Dynam., 9, 147–163, 1996. Miner, M. A.: Cumulative damage in fatigue, J. Appl. Mech., 12, 159–164, 1945. Braccesi, C. and Cianetti, F.: A procedure for the virtual evaluation of the stress state of mechanical systems and components for the automotive industry: development and experimental validation, Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, 219, 633–643, 2005. Niemi, E., Fricke, W., and Maddox, S. J.: Fatigue analysis of welded components: Designer’s guide to the structural hot-spot stress ap- proach, Woodhead Publishing, Cambridge, 1st Edn., 2006. Nikravesh, P. E. and Chung, I. S.: Application of Euler Parameters to the Dynamic Analysis of Three-Dimensional Constrained Me- chanical Systems, J. Mech. Design, 104, 785–791, 1982. Claus, H.: A Deformation Approach to Stress Distribution in Flexi- ble Multibody Systems, Multibody Syst. Dyn., 6, 143–161, 2001. Craig Jr., R. R. and Bampton, M. C. C.: Coupling of Substructures for Dynamic Analyses, AIAA J., 6, 1313–1319, 1968. Poutiainen, I., Tanskanen, P., and Marquis, G.: Finite element meth- ods for structural hot spot stress determination - a comparison of procedures, I. J. Fatigue, 26, 1147–1157, 2004. Dietz, S.: Vibration and fatigue analysis of vehicle systems us- ing component modes, Doctoral thesis, Technische Universit¨at Berlin, 1999. Shabana, A. A.: Dynamics of Multibody Systems, Cambridge Uni- versity Press, Cambridge, 2nd Edn., 1998. Dietz, S., Netter, H., and Sachau, D.: Fatigue life predictions by cou- pling ﬁnite element and multibody systems calculations, in: Pro- ceedings of Design Engineering Technical Conferences, 1997. Shabana, A. A.: Dynamics of Multibody Systems, Cambridge Uni- versity Press, Cambridge, 3rd Edn., 2005. Tobias, C. and Eberhard, P.: Stress recovery with Krylov-subspaces in reduced elastic multibody systems, Multibody Syst. Dyn., 25, 377–393, 2011. Dietz, S., Netter, H., and Sachau, D.: Fatigue Life Prediction of a Railway Bogie under Dynamic Loads through Simulation, Vehi- cle Syst. Dyn., 29, 385–402, 1998. von Mises, R.: On Saint Venant’s principle, B. Am. Math. Soc., 51, 555–562, 1945. Haagensen, P. J. and Maddox, S. J.: IIW Recommendations on Post Weld Fatigue Life Improvement of Steel and Aluminium Struc- tures, International Institute of Welding, Paris, 2011. Wallrapp, O. 6 Conclusions Korkealaakso, P., Mikkola, A., Rantalainen, T., and Rouvinen, A.: Description of joint constraints in the ﬂoating frame of reference formulation, Proceedings of the Institution of Mechanical Engi- neers, Part K: Journal of Multi-body Dynamics, 223, 133–145, 2009. Acknowledgements. The research is supported by the Academy of Finland, project 133154. Lee, S.-H., Park, T.-W., Park, J.-K., Yoon, J.-W., Jun, K.-J., and Jung, S.-P.: Fatigue life analysis of wheels on guideway vehicle using multibody dynamics, Int. J. Precis. Eng. Man., 10, 79–84, 2009. Edited by: A. M¨uller Reviewed by: J. Cuadrado and A. Schwab References and Wiedemann, S.: Comparison of Results in Flexi- ble Multibody Dynamics Using Various Approaches, Nonlinear Dynam., 34, 189–206, 2003. Jun, K., Park, T., Lee, S., Jung, S., and Yoon, J.: Prediction of fa- tigue life and estimation of its reliability on the parts of an air suspension system, Int. J. Automot. Techn., 9, 741–747, 2008. Watton, J.: Fluid power systems: Modeling, simulation, analog and microcomputer control, Prentice Hall, New York, 1st Edn., 1989. Knight, N. F., Ransom, J. B., Griﬃn, O. H., and Thompson, D. M.: Global/local methods research using a common structural analy- sis framework, Finite Elem. Anal. Des., 9, 91–112, 1991. Yim, H. J. and Lee, S. B.: An integrated CAE system for dynamic stress and fatigue life prediction of mechanical systems, J. Mech. Sci. Technol., 10, 158–168, 1996. Mech. Sci., 4, 21–31, 2013 www.mech-sci.net/4/21/2013/ www.mech-sci.net/4/21/2013/
https://openalex.org/W3134672830	https://www.nature.com/articles/s41598-021-85065-0.pdf	English	null	Higher choroidal thickness and lower choriocapillaris blood flow signal density based on optical coherence tomography angiography in diabetics	Scientific reports	2,021	cc-by	5,661	Higher choroidal thickness and lower choriocapillaris blood flow signal density based on optical coherence tomography angiography in diabetics OPEN Yaoli Zhang1,4, Yuanjun Qin1,4, Shuaishuai Wang2, Yuyan Liu3, Xinyu Li1 & Xufang Sun1* Diabetes mellitus (DM) is one of the fastest growing chronic diseases in the world and one of the main causes of vision loss. Whether or not diabetic choroidopathy (DC) is involved in the initiation and progression of diabetic ocular complications needs to be explored. We included 54 diabetic eyes from 36 diabetic patients, and 54 healthy eyes from 32 control subjects after propensity scores matching. All of the subjects were given pupil light and dark adaptation examination and optical coherence tomography angiography (OCTA). Scotopic pupil diameter (SPD), pupil contraction amplitude, and velocity of pupil contraction of the diabetic group were significantly lower than that of the healthy control group (P < 0.05).Choroidal thickness at temporal quadrant (at 750 μm) and superior quadrant (at 1500 μm and 2250 μm) increased in diabetic group compared to control group(P < 0.05). In the diabetic group, choriocapillaris blood flow signal density (CCBFSD) in the macular area (diameter = 2000 μm) were significantly decreased compared with the healthy control group (P < 0.05). Apparent changes in pupil and choroidal blood flow were observed in the diabetic patients. Diabetes mellitus (DM) is one of the fastest growing chronic diseases in the world and one of the leading causes of vision loss1. According to the World Health Organization (WHO), the total number of people with DM will grow from 171 million in 2000 to 366 million in 20302. At present, the pathological mechanisms of diabetic retinopathy are mainly explained by the effects of vascular leakage, vascular abnormalities, neuronal cell dys- function, and inflammatory mechanisms3.h l y The concept of diabetic choroidopathy was proposed by Saracco et al. as early as 19824. Histopathological studies revealed a variety of choroidal changes secondary to DM5–7. The choroid accounts for more than 85% of the retinal blood supply and supplies all the nutrients for the retinal pigment epithelium and photoreceptors8,9. It has been reported that photoreceptors die rapidly when the choroidal blood flow decline drastically10. The regulation of choroidal blood flow includes neuromodulation, autoregulation, and local choroidal ganglion regulation11–13. Diabetic autonomic neuropathy (DAN) refers to a disorder of the autonomic nerves caused by DM, which involves diseases related to multiple systems including orthostatic hypotension, tachycardia, gastroparesis, diar- rhea, constipation, etc. Its ocular complications include a reduction of scotopic pupil diameter and an Argyll- Robertson-like pupil14. www.nature.com/scientificreports www.nature.com/scientificreports Scientific Reports \| (2021) 11:5799 Higher choroidal thickness and lower choriocapillaris blood flow signal density based on optical coherence tomography angiography in diabetics OPEN Studies have shown that nerve modulation of the uvea affects blood flow in the iris, ciliary body, and choroid15,16. y y Both the choroid and the iris are parts of the uvea, and they have an inseparable relationship in anatomy, structure, and development of each other. Although there is a lot of evidence confirming the presence of pupillary lesions and changes in choroidal thickness caused by diabetic autonomic nerve damage17–21, little research on dia- betic choroidal blood flow and its relation to pupil lesions is currently available. Previous studies have shown that 1Department of Ophthalmology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, 1095 Jie‑fang Road, Wuhan, Hubei, China. 2Department of Ophthalmology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China. 3University of Massachusetts Amherst, Amherst, MA, USA. 4These authors contributed equally: Yaoli Zhang and Yuanjun Qin. email: sunxufang2016@163.com Scientific Reports \| (2021) 11:5799 \| https://doi.org/10.1038/s41598-021-85065-0 www.nature.com/scientificreports/ Table 1. Demographic and clinical characteristics of patients by group. BCVA best corrected visual acuity, IOP intraocular pressure, SE spherical equivalent, AL axial length. Higher choroidal thickness and lower choriocapillaris blood flow signal density based on optical coherence tomography angiography in diabetics OPEN Before PSM After PSM Healthy group (eyes = 89) Diabetic group (eyes = 84) P Healthy group (eyes = 54) Diabetic group (eyes = 54) P Age (years) 40.94 ± 10.99 50.54 ± 8.94 < 0.01 44.70 ± 11.50 48.33 ± 9.35 > 0.05 Female/Male 49/40 29/55 < 0.01 26/28 23/31 > 0.05 BCVA (logMAR) − 0.02 ± 0.05 0.00 ± 0.08 < 0.05 − 0.03 ± 0.06 0.01 ± 0.09 < 0.05 IOP (mmHg) 15.15 ± 2.49 16.49 ± 2.63 < 0.01 15.12 ± 2.55 15.64 ± 2.55 > 0.05 SE (D) − 1.18 ± 1.70 − 0.32 ± 1.48 < 0.01 − 0.61 ± 1.52 − 0.51 ± 1.53 > 0.05 AL 23.59 ± 1.18 23.37 ± 1.04 > 0.05 23.42 ± 1.11 23.36 ± 1.09 > 0.05 Stage (NDR/mild NPDR/moderate NPDR) – 44/32/8 – – 31/19/4 –t Control group (n = 54) Diabetic group (n = 54) P value SPD (mm) 5.31 ± 0.89 4.96 ± 0.78 < 0.05 PD1 (mm) 3.82 ± 0.90 3.50 ± 0.76 > 0.05 PD2 (mm) 2.90 ± 0.48 2.83 ± 0.49 > 0.05 PD3 (mm) 2.52 ± 0.32 2.53 ± 0.38 > 0.05 PCA (mm) 1.53 ± 0.31 1.36 ± 0.32 < 0.05 PCV (mm/s) 5.17 ± 0.90 4.84 ± 0.68 < 0.05 PDV (mm/s) 2.01 ± 0.32 2.12 ± 0.54 > 0.05 Table 1. Demographic and clinical characteristics of patients by group. BCVA best corrected visual acuity, IOP intraocular pressure, SE spherical equivalent, AL axial length. Before PSM After PSM Healthy group (eyes = 89) Diabetic group (eyes = 84) P Healthy group (eyes = 54) Diabetic group (eyes = 54) P Age (years) 40.94 ± 10.99 50.54 ± 8.94 < 0.01 44.70 ± 11.50 48.33 ± 9.35 > 0.05 Female/Male 49/40 29/55 < 0.01 26/28 23/31 > 0.05 BCVA (logMAR) − 0.02 ± 0.05 0.00 ± 0.08 < 0.05 − 0.03 ± 0.06 0.01 ± 0.09 < 0.05 IOP (mmHg) 15.15 ± 2.49 16.49 ± 2.63 < 0.01 15.12 ± 2.55 15.64 ± 2.55 > 0.05 SE (D) − 1.18 ± 1.70 − 0.32 ± 1.48 < 0.01 − 0.61 ± 1.52 − 0.51 ± 1.53 > 0.05 AL 23.59 ± 1.18 23.37 ± 1.04 > 0.05 23.42 ± 1.11 23.36 ± 1.09 > 0.05 Stage (NDR/mild NPDR/moderate NPDR) – 44/32/8 – – 31/19/4 – Table 1. Demographic and clinical characteristics of patients by group. Higher choroidal thickness and lower choriocapillaris blood flow signal density based on optical coherence tomography angiography in diabetics OPEN BCVA best corrected visual acuity, IOP intraocular pressure, SE spherical equivalent, AL axial length. Table 2. Results of pupillometry for the patients by group after PSM. SPD scotopic pupil diameter, PD1 pupil diameter at light intensity of 1 cd/m2, PD2 pupil diameter at light intensity of 10 cd/ m2, PD3 pupil diameter at light intensity of 100 cd/m2. Control group (n = 54) Diabetic group (n = 54) P value SPD (mm) 5.31 ± 0.89 4.96 ± 0.78 < 0.05 PD1 (mm) 3.82 ± 0.90 3.50 ± 0.76 > 0.05 PD2 (mm) 2.90 ± 0.48 2.83 ± 0.49 > 0.05 PD3 (mm) 2.52 ± 0.32 2.53 ± 0.38 > 0.05 PCA (mm) 1.53 ± 0.31 1.36 ± 0.32 < 0.05 PCV (mm/s) 5.17 ± 0.90 4.84 ± 0.68 < 0.05 PDV (mm/s) 2.01 ± 0.32 2.12 ± 0.54 > 0.05 Table 2. Results of pupillometry for the patients by group after PSM. SPD scotopic pupil diameter, PD1 pupil iameter at light intensity of 1 cd/m2, PD2 pupil diameter at light intensity of 10 cd/ m2, PD3 pupil diameter at ght intensity of 100 cd/m2. Table 2. Results of pupillometry for the patients by group after PSM. SPD scotopic pupil diameter, PD1 pupil diameter at light intensity of 1 cd/m2, PD2 pupil diameter at light intensity of 10 cd/ m2, PD3 pupil diameter at light intensity of 100 cd/m2. the blood flow of choroidal capillaries in diabetic patients is significantly lower than that of healthy people22,23. Some studies have shown that their blood flow further decreases as the severity of the disease increases24. the blood flow of choroidal capillaries in diabetic patients is significantly lower than that of healthy people22,23. Some studies have shown that their blood flow further decreases as the severity of the disease increases24. d l h h ( ) h l h l d l In recent years, advances in optical coherence tomography (OCT) technology has revolutionized retinal research and clinical practice25. OCT angiography (OCTA) is a recently developed technique that allows noninva- sive visualization of vascular structures26–28. Moreover, the higher speed of the new OCT equipment has facilitated the optical reconstruction of complex vascular structures25. This feature of OCTA makes it an extremely valuable tool for studying choroidal blood flow29. Results P ti t d Patient demographics and clinical characteristics. A total of 108 eyes (no DR, 31 eyes; mild NPDR, 19 eyes; moderate NPDR, 4 eyes; control group, 54 eyes) were included for final analysis after PSM. The demo- graphic and clinical characteristic details can be identified in Table 1. Pupillometry. The details of the diabetic and healthy control groups for pupil diameters at light intensities of 0 cd/m2, 1 cd/m2, 10 cd/m2, and 100 cd/m2, pupil contraction amplitude (PCA), velocity of pupil contraction (PCV) and velocity of pupil dilation (PDV) are shown in Table 2. This study showed that the scotopic pupil diameter (SPD), PCA and PCV of the diabetic group were significantly lower than that of the healthy control group (4.96 ± 0.78 vs 5.31 ± 0.89, 1.36 ± 0.32 vs 1.53 ± 0.31, 4.84 ± 0.68 vs 5.17 ± 0.90, respectively; P < 0.05). The pupil size at light intensities of 1 cd/m2, 10 cd/m2, 100 cd/m2, and PDV were not significantly different from those in the healthy control group (P > 0.05). Optical coherence tomography. Choroidal thickness (CT) at different locations of the fundus of the diabetic and the control groups is shown in Fig. 1. Results showed that CT at temporal quadrant (at 750 μm) and superior quadrant (at 1500 μm and 2250 μm) increased in diabetic group compared to control group (P < 0.05). There were no statistically significant differences in mean subfoveal choroidal thickness (SFCT), mean choroidal thick- ness, and CT in other locations and between two groups (P > 0.05). Choriocapillaris blood flow signal density (CCBFSD). CCBFSD in the macular area in the diabetic group were significantly lower than that in the healthy control group (P < 0.05). The details are shown in Fig. 2. https://doi.org/10.1038/s41598-021-85065-0 Scientific Reports \| (2021) 11:5799 \| www.nature.com/scientificreports/ Figure 1. CT (μm) in 17 locations by group after PSM. Significantly related on the 0.05 level (both sides). Figure 1. CT (μm) in 17 locations by group after PSM. Significantly related on the 0.05 level (both sides). Figure 2. Choriocapillaris blood flow signal density by group after PSM. Significantly related on the 0.05 level (both sides). Figure 2. Choriocapillaris blood flow signal density by group after PSM. *Significantly related on the 0.05 le (both sides). www.nature.com/scientificreports/ www.nature.com/scientificreports/ Our study showed a lower CC perfusion in diabetic eyes compared with healthy eyes. Previous study has illustrated that eyes with NPDR are affected by macular hypoperfusion and photoreceptor damage and a lower CC perfusion may be associated with a greater damage of photoreceptor35. We speculate that there are two rea- sons for decreased CCBFSD in the macular area with diabetes. Firstly, decreased CCBFSD in diabetic patients is caused by vascular damage and blood flow decreasing due to long-term hyperglycemia. Secondly, although it is not realistic to completely exclude the effect of bleeding and exudation on signal acquisition, the patients we included were all patients with no DR, and mild or moderate DR whose bleeding and exudation were not severe. It should be noted that there is no evidence that slight retinal hemorrhage could significantly affect the collection of choroidal blood flow signals. l g At the same time, the increase of choroidal thickness and the decrease of SPD occurred in the diabetic group, and both were regulated by autonomic nerves, suggesting that this may be related to diabetic autonomic nerve damage. As we know, the decline of SPD is related to autonomic nerve damage. Furthermore, the middle vascular layer and large vascular layer of choroid also have autonomic and trigeminal innervation. The mechanism and their correlation deserve further research. A previous report by Borrelli et al. demonstrated that a further dilation of the area occupied by medium-sized and larger-sized choroidal vessels and/or choroidal stroma appears to be associated with a progressive reduction in CC perfusion in healthy eyes36. It might suggest that an increased choroidal thickness may be associated with a lower CC perfusion, which is not surprising. Increased choroidal thickness and decreased choriocapillaris perfusion in patients with central serous chorioretinopathy, similar to the results in our study of patients with diabetes37,38. This reminds us that there may be some similar mechanisms in these two diseases. Firstly, thicker choroid may be caused by increased choroidal vessel wall permeability in an inflammatory state in both diseases. Secondly, choriocapillaris hypoperfusion may be caused by mechanical stress resulting from compression based on focal or diffusely enlarged choroidal vessels37. f Compared with previous studies, we better matched certain factors, such as age, gender, intraocular pressure, axial length and refractive status, to minimize the effect of these confounding factors on CT. www.nature.com/scientificreports/ At the same time, the OCT examination is performed at a constant time period of a day, which reduces the influence of the daily rhythm on the choroidal measurement.f At the same time, we also need to consider the following points. First, differences in choroid assessment methods may lead to conflicting results. Diabetic choroidal vascular abnormalities are most common in the mid-circumference. However, OCT examinations focus on the macular/foveal area, so the results of these studies may not accurately reflect the actual situation. Second, OCTA has limitations as well: that is, only the velocity of blood flow within a certain range can be detected by the equipment. The velocity of choroidal capillaries may occasionally be too low to generate a disassociation signal, preventing OCTA from recording blood flow signals in capillaries. Thirdly, compared with the choroidal capillary layer, it is difficult to accurately evaluate blood flow of medium and large choroidal vessels because of the occlusion of retinal vessels and choroidal capillaries, and high blood flow velocity in these vessels. gl y In short, with the advent of the latest OCTA technology, DC has become one of the most compelling diseases to study for ophthalmological researchers. Nonetheless, the relationship between DC and DR, and the role of DC in diabetic ocular diseases remain unknown. This important field requires further research. Discussion Our study showed that SPD, PCA, and PCV in the diabetic group were all lower than those in the healthy control group, which is consistent with results of previous studies20,21. We consider that this difference may result from pupillary autonomic nerve disorders in diabetic patients, suggesting that hyperglycemia damages the autonomic nerves of the iris. We found that the CT at temporal quadrant (at 750 μm) and superior quadrant (at 1500 μm and 2250 μm) in the diabetic group significantly increased compared with the healthy control group. For this point, previous studies have had different results. Some clinical findings suggest choroidal thinning of early diabetic retinopathy (DR)17,18,30,31; however, some conflicting results suggest early thinning, late thickening19, or completely contrast choroidal thickening32–34. Some of these studies have their own characteristics that may explain the heterogene- ity of the results. Querques et al. only compared CT in the central region18, and we know that choroidal changes in diabetes not only affect the central region but, more importantly, affect the circumference. The Vujosevic’s study included all stages of DR, including some patients with PDR17. In Esmaeelpour’s study, some patients had a history of hypertension and were taking medications30, which could affect choroidal circulation. y yp gf OCTA, as a method for non-invasive visualization technique of fundus blood vessels, can provide a more comprehensive assessment of choroidal circulation. Compared with OCT, it provides a richer amount of blood flow information. Although OCTA has been frequently used clinically, it is rarely used as a tool for the study of DC. This is an innovative aspect of this research.l h p Compared with previous studies, choriocapillaris blood flow signal density assessment method has been further updated. We use the self-programmed code to position and analyze each image. First of all, we perform batch analysis of pictures based on code, which can reduce manual errors and have good repeatability. Secondly, the code can accurately calculate the blood flow signal of each picture, thus forming 512 B-scan pictures into a "map" representing the choriocapillaris blood flow volume in the macular area, and finally calculating the percentage of choriocapillaris capillary blood signal density in this area. Scientific Reports \| (2021) 11:5799 \| https://doi.org/10.1038/s41598-021-85065-0 Materials and methodsh (C) OCTA B-scan image with yellow blood flow signal after self-written code process. The two horizontal red lines are inner and outer boundaries of choriocapillaris layer respectively. These two vertical red lines indicate the boundary of the 2000-μm-diameter macular area we calculated in the B scan OCTA image. The area enclosed by the above four lines is the area and blood flow signal we need to calculate in the 256th OCTA image. Figure 3. The 256th OCTA image. Yellow signals indicate blood flow signals. (A) En face image of structural OCT. Green rectangle is the scanned 3 × 3 mm2 area. The red circle area indicates the 2000-μm-diameter macular area where we calculate CCBFSD. (B) OCT B-scan segmented image with yellow blood flow signal after manual adjustments. (C) OCTA B-scan image with yellow blood flow signal after self-written code process. The two horizontal red lines are inner and outer boundaries of choriocapillaris layer respectively. These two vertical red lines indicate the boundary of the 2000-μm-diameter macular area we calculated in the B scan OCTA image. The area enclosed by the above four lines is the area and blood flow signal we need to calculate in the 256th OCTA image. at different light intensities (0 cd/m2, 1 cd/m2, 10 cd/m2, and 100 cd/m2), pupil contraction amplitude (PCA), velocity of pupil contraction (PCV), velocity of pupil dilation (PDV).h y y The 3 × 3 mm2 parafoveal area with 512 B-scan OCT images per eye scanned by Heidelberg SPECTRALIS OCT device (Heidelberg Engineering GmbH, Heidelberg, Germany) was used to analyze the choriocapillaris blood flow signal in this region. The OCT B-scan was performed horizontally and vertically through the fovea. CT at the nasal, superior, temporal, and inferior choroid quadrants (at 750 μm, 1500 μm, 2250 μm and 3000 μm intervals from the center of the fovea) and subfoveal choroidal thickness (SFCT) were recorded. Mean CT is their average. Choroidal thickness was measured as the perpendicular distance between the hyperreflective outer bor- der of the retinal pigment epithelial/Bruch’s membrane layer and the sclero-choroidal interface, manually drawn blindly by the examiner. OCTA and OCT examinations were performed using the Heidelberg SPECTRALIS OCT equipment in 8:30 a.m.–13:00 p.m. for reducing choroidal changes with daily rhythm.t We used SPECTRALIS Software Version 6.9a to layer the acquired images. Materials and methodsh Subject groups. This study was performed according to the Declaration of Helsinki and was approved by the Ethics Committee of Tongji Hospital. Informed consent was obtained from all study participants before examination. Subjects. We included 47 diabetic patients (84 eyes) and 48 healthy people (89 eyes) from the Department of Ophthalmology and Department of Endocrinology of Tongji Hospital, Tongji Medical College of Huazhong University of Science and Technology. Because the baselines of age, intraocular pressure, sex and spherical equivalent (SE) in two groups were not well matched, propensity scores matching was used. 54 diabetic eyes from 36 diabetic patients and 54 healthy eyes from 32 control subjects were included for final analysis after propensity scores matching, which were used to balance the two groups according to several baseline variables, including age, sex, intraocular pressure, SE. DM was diagnosed according to the 1999 WHO diagnostic criteria. Inclusion criteria. Diabetic patients were included who had no signs of DR, or those diagnosed with mild or moderate non-proliferative diabetic retinopathy (NPDR). Exclusion criteria. These included: (1) hypertension, kidney disorders, and other systemic diseases that may affect fundus circulation; (2) glaucoma, cataract, and history of ocular surgery; (3) vitreous opacity; (4) high myopia; (5) age-related macular degeneration, choroidal neovascularization, central serous retinal choroidopa- thy, and other eye diseases that affect the fundus circulation; (6) DM patients treated with laser and intraocular anti-vascular endothelial growth factor (VEGF), or macular edema; (7) patients with a history of ocular trauma. Study procedure. All subjects underwent visual acuity (VA), intraocular pressure (IOP), slit lamp exami- nation, ophthalmoscopy, optometry, fundus photography, examination of pupil light and dark adaptation, para- foveal 3 × 3 mm2 OCTA scanning, and foveal OCT line scan. Subjects with B-scan OCT image quality lower than 30 and OCTA quality lower than 35 were excluded. q y Pupil function tests were performed using a comprehensive pupillometer (Vision Monlter Mon2013K, Metro- ision, France). Examinations were performed by the same ophthalmologist, including pupil diameter (PD) Scientific Reports \| (2021) 11:5799 \| https://doi.org/10.1038/s41598-021-85065-0 www.nature.com/scientificreports/ Figure 3. The 256th OCTA image. Yellow signals indicate blood flow signals. (A) En face image of structural OCT. Green rectangle is the scanned 3 × 3 mm2 area. The red circle area indicates the 2000-μm-diameter macular area where we calculate CCBFSD. (B) OCT B-scan segmented image with yellow blood flow signal after manual adjustments. Materials and methodsh We would make manual adjust- ments if the automatic segmentation was incorrect, and then used the self-written code to calculate choriocapil- laris blood flow signal density (CCBFSD) in the macular area (centered by the fovea, 2000 microns in diameter), which is the percentage of the yellow signal to the total area. Choriocapillaris was segmented with an inner boundary Bruch membrane and an outer boundary 49 μm below the Bruch membrane. The positioning method of the choroidal capillary area for calculation in the 256th image of 512 pictures is described detailly in Fig. 3. We took out the calculation area picture enclosed by the four red lines in Fig. 3C, converted it into a grayscale image, and then binarized it (cut off value is 30), and calculated the total pixels of the image and the pixels occupied by the blood flow signal. We performed the above operations on all 512 OCTA images in one eye, and finally obtain: CCBFSD = total blood flow signal pixels of 512 images /total pixels of 512 images. The image positioning and calculation operations were all completed in batches by self-edited code. Statistical analysis. All the data in this study were analyzed using IBM SPSS Statistics ver.26.0 (IBM, Armonk, NY, USA). The propensity score was estimated using a logistic regression model with 1:1 nearest neigh- bour matching without replacement, based on an acceptable caliper width of 0.25 times. Kolmogorov–Smirnov test was used for the normality test. Independent sample t-test was used to compare the data between these two groups that satisfied the normal distribution; if it did not, non-parametric test was used. Chi-square test was used for gender comparison between the two groups. Measurement data were expressed as mean ± standard deviation. P < 0.05 was considered statistically significant. Data availabilityh The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Received: 11 May 2020; Accepted: 24 February 2021 Received: 11 May 2020; Accepted: 24 February 2021 https://doi.org/10.1016/S01616420(98)960250 (1998). 2. Wild, S., Roglic, G., Green, A., Sicree, R. & King, H. Global prevalence of diabetes: Estimates for the year 2000 and projections for 2030. Diabetes Care 27, 1047–1053. https://doi.org/10.2337/diacare.27.5.1047 (2004). References p g 2. Wild, S., Roglic, G., Green, A., Sicree, R. & King, H. Global prevalence of diabetes: Estimates for the year 2000 and projections for 2030. Diabetes Care 27, 1047–1053. https://doi.org/10.2337/diacare.27.5.1047 (2004). Optical coherence tomography angiography quantitative assessment of choriocapillaris blood flow serous chorioretinopathy Am J Ophthalmol 194 26 34 https://doi org/10 1016/j ajo 2018 07 004 (2018) y y p p g j p 37. Rochepeau, C. et al. Optical coherence tomography angiography quantitative assessment of choriocapillaris blood flow in central serous chorioretinopathy. Am. J. Ophthalmol. 194, 26–34. https://doi.org/10.1016/j.ajo.2018.07.004 (2018). 8. Hanumunthadu, D. et al. Evaluation of choroidal layer thickness in central serous chorioretinopathy. J. Ophthal. Vis Res. 14 164–170. https://doi.org/10.4103/jovr.jovr_152_17 (2019). Scientific Reports \| (2021) 11:5799 \| https://doi.org/10.1038/s41598-021-85065-0 www.nature.com/scientificreports/ 3. Antonetti, D. A., Klein, R. & Gardner, T. W. Diabetic retinopathy. N. Engl. J. Med. 366, 1227–1239. https://doi.org/10.1056/NEJMr a1005073 (2012). 4. Saracco, J. B., Gastaud, P., Ridings, B. & Ubaud, C. A. Preliminary study on diabetic choroidopathy. Bull. Soc. Ophtalmol. Fr 82 451–454 (1982).i ( ) 5. Fryczkowski, A. W., Hodes, B. L. & Walker, J. Diabetic choroidal and iris vasculature scanning electron microscopy findings. In Ophthalmol 13, 269–279. https://doi.org/10.1007/bf02280087 (1989).i p p g 6. Fryczkowski, A. W., Sato, S. E. & Hodes, B. L. Changes in the diabetic choroidal vasculature: scanning electron microscopy find ings. Ann. Ophthalmol. 20, 299–305 (1988). ings. Ann. Ophthalmol. 20, 299 305 (1988). 7. Lutty, G. A. Diabetic choroidopathy. Vis. Res. 139, 161–167. https://doi.org/10.1016/j.visres.2017.04.011 (2017). g p , ( ) 7. Lutty, G. A. Diabetic choroidopathy. Vis. Res. 139, 161–167. https://doi.org/10.1016/j.visres.2017.04.011 (2017). y g . & Science, V. Some aspects of the ocular circulation. Friedenwald 8. Bill, A. J. I. O. & Science, V. Some aspects of the ocular circula 9. Hidayat, A. A. & Fine, B. S. Diabetic choroidopathy. Light and electron microscopic observations of seven cases. Ophthalmology 92, 512–522 (1985). 0. Gaudric, A., Coscas, G. & Bird, A. C. Choroidal ischemia. Am. J. Ophthalmol. 94, 489–498. https://doi.org/10.1016/0002- 9394(82)90242-2 (1982). 11. Ríos, J. D. et al. Immunolocalization of muscarinic and VIP receptor subtypes and their role in stimulating goblet cell secretion. 40, 1102–1111 (1999).h 2. Yamamoto, R., Bredt, D. S., Snyder, S. H. & Stone, R. A. The localization of nitric oxide synthase in the rat eye and related crania ganglia. Neuroscience 54, 189–200. https://doi.org/10.1016/0306-4522(93)90393-t (1993).l 3. Guglielmone, R. & Cantino, D. Autonomic innervation of the ocular choroid membrane in the chicken: A fluorescence-histo- chemical and electron-microscopic study. Cell Tissue Res. 222, 417–431. https://doi.org/10.1007/bf00213222 (1982). 4. Vinik, A. I., Freeman, R. & Erbas, T. Diabetic autonomic neuropathy. Semin. Neurol. 23, 365–372. https://doi org/10.1055/s-2004-817720 (2003). 5. Wikberg-Matsson, A. & Simonsen, U. Potent alpha(2A)-adrenoceptor-mediated vasoconstriction by brimonidine in porcine ciliary arteries. Invest. Ophthalmol. Vis. Sci. 42, 2049–2055 (2001).hf p 6. Alm, A. & Bill, A. The effect of stimulation of the cervical sympathetic chain on retinal oxygen tension and on uveal, retinal and 16. Alm, A. & Bill, A. The effect of stimulation of the cervical sy hf y p yg cerebral blood flow in cats. Acta Physiol. Scand. 88, 84–94. https://doi.org/10.1111/j.1748-1716.1973.tb05436.x (1973). V S M F C F P l E & M d E M l d ll h d l h k d b l y p g j 17. Vujosevic, S., Martini, F., Cavarzeran, F., Pilotto, E. & Midena, E. Macular and peripapillary choroidal thickness in diabetic patients. Retina 32 1781 1790 https://doi org/10 1097/IAE 0b013e31825db73d (2012) l 17. Vujosevic, S., Martini, F., Cavarzeran, F., Pilotto, E. & Midena, E. Macular and peripapillary c Retina 32, 1781–1790. https://doi.org/10.1097/IAE.0b013e31825db73d (2012). l 17. Vujosevic, S., Martini, F., Cavarzeran, F., Pilotto, E. & Midena, E. Macular and peripapillary choroidal thickness in diabetic patients. Retina 32, 1781–1790. https://doi.org/10.1097/IAE.0b013e31825db73d (2012). 8. Querques, G. et al. Enhanced depth imaging optical coherence tomography in type 2 diabetes. Invest. Ophthalmol. Vis. Sci. 53 6017–6024. https://doi.org/10.1167/iovs.12-9692 (2012). p g 9. Kase, S. et al. Choroidal thickness in diabetic retinopathy in relation to long-term systemic treatments for diabetes mellitus. Eur J. Ophthalmol. 26, 158–162. https://doi.org/10.5301/ejo.5000676 (2016). hthalmol. 26, 158–162. https://doi.org/10.5301/ejo.5000676 (2016 p p g j ( ) 0. Park, J. C. et al. Pupillary responses in non-proliferative diabetic retinopathy. Sci. Rep. 7, 44987. https://doi.org/10.1038/srep44987 (2017). 21. Piha, S. J. & Halonen, J. P. Infrared pupillometry in the assessment of autonomic function. Diabetes Res. Clin. Pract. 26, 61–66. https://doi.org/10.1016/0168-8227(94)90140-6 (1994). p g 2. Cao, D. et al. Optical coherence tomography angiography discerns preclinical diabetic retinopathy in eyes of patients with type 2 diabetes without clinical diabetic retinopathy. (2018).i diabetes without clinical diabetic retinopathy. (2018). p y 3. Zijing et al. Optical coherence tomography angiography findings of neurovascular changes in type 2 diabetes mellitus patients without clinical diabetic retinopathy. (2018). 24. Nesper, P. L. et al. Quantifying microvascular abnormalities with increasing severity of diabetic retinopathy using optical coherence tomography angiography. 58, BIO307-BIO315 (2017). g y g g y 5. Ferrara, D., Waheed, N. K. & Duker, J. S. Investigating the choriocapillaris and choroidal vasculature with new optical coherence tomography technologies. Prog. Retin. Eye Res. 52, 130–155. https://doi.org/10.1016/j.preteyeres.2015.10.002 (2016).h 26. Wang, R. K. et al. Three dimensional optical angiography. Opt. Express 15, 4083–4097. https://doi.org/10.1364/oe.15.004083 (2007). h 7. Zhang, A., Zhang, Q., Chen, C. L. & Wang, R. K. Methods and algorithms for optical coherence tomography-based angiography A review and comparison. J. Biomed. Opt. 20, 100901. https://doi.org/10.1117/1.JBO.20.10.100901 (2015). p p p g ( ) 28. Kashani, A. H. et al. Optical coherence tomography angiography: A comprehensive review of current methods and clinical applica- tions. Prog. Retin. Eye Res. 60, 66–100. https://doi.org/10.1016/j.preteyeres.2017.07.002 (2017).l g 8. Kashani, A. H. et al. Optical coherence tomography angiography: A comprehensive review of current methods and clinical applica- P R i E R 6 66 h //d / 6/ ( ) 9. Alm, A. & Bill, A. Ocular and optic nerve blood flow at normal and increased intraocular pressures in monkeys (Macaca irus): A study with radioactively labelled microspheres including flow determinations in brain and some other tissues. 15, 15–29. 29. Alm, A. & Bill, A. Ocular and optic nerve blood flow at normal and increased intraocular pressures in monkeys (Macaca irus): A study with radioactively labelled microspheres including flow determinations in brain and some other tissues. 15, 15–29. 30 Esmaeelpour M et al Mapping choroidal and retinal thickness variation in type 2 diabetes using three dimensional 1060 nm y y p gl 0. Esmaeelpour, M. et al. Mapping choroidal and retinal thickness variation in type 2 diabetes using three-dimensional 1060-nm optical coherence tomography. 52, 5311. p g p y 31. Regatieri, C. V., Branchini, L., Carmody, J., Fujimoto, J. G. & Duker, J. S. Choroidal thickness in patients with diabetic retinopathy analyzed by spectral-domain optical coherence tomography. Retina 32, 563–568. https://doi.org/10.1097/IAE.0b013e31822f5678 (2012). 32. Melancia, D., Vicente, A., Cunha, J. P., Abegao Pinto, L. & Ferreira, J. Diabetic choroidopathy: a review of the current litera Graefes Arch. Clin. Exp. Ophthalmol. 254, 1453–1461. doi:https://doi.org/10.1007/s00417-016-3360-8 (2016). l h d l h k d b h d b h J g J p y Graefes Arch. Clin. Exp. Ophthalmol. 254, 1453–1461. doi:https://doi.org/10.1007/s00417-016-3360- Ferreira, J. et al. Choroidal thickness in diabetic patients without d p p y 34. Rewbury, R., Want, A., Varughese, R. & Chong, V. Subfoveal choroidal thickness in patients with diabetic retinopathy and dia macular oedema. Eye (Lond) 30, 1568–1572. https://doi.org/10.1038/eye.2016.187 (2016). 5. Borrelli, E., Palmieri, M., Viggiano, P., Ferro, G. & Mastropasqua, R. Photoreceptor damage in diabetic choroidopathy. Retina 40 1062–1069. https://doi.org/10.1097/IAE.0000000000002538 (2020). p g 36. Borrelli, E. et al. Choroidal luminal and stromal areas and choriocapillaris perfusion are characterised by a non-linear quadratic relation in healthy eyes. Br. J. Ophthalmol. https://doi.org/10.1136/bjophthalmol-2020-316479 (2020).l y y p p g j p ( ) 37. Rochepeau, C. et al. Additional information Correspondence and requests for materials should be addressed to X.S. Correspondence and requests for materials should be addressed to X.S. Reprints and permissions information is available at www.nature.com/reprints. Acknowledgements Th k d b g This work was supported by grants from National Natural Science Foundation of China (81570868). g This work was supported by grants from National Natural Science Foundation of China (81570868). Author contributions Conception or design of the work: X.S. Data collection: Y.Z., Y.Q. Data analysis and interpretation: Y.Z., Y.Q. Drafting the article: Y.Z. Critical revision of the article: X.S., Y.Q., S.W., Y.L., X.L. Final approval of the version: Y.Z., Y.Q., S.W., Y.L., X.L., X.S. https://doi.org/10.1038/s41598-021-85065-0 Scientific Reports \| (2021) 11:5799 \| www.nature.com/scientificreports/ Competing interests The authors declare no competing interests. Additional information Correspondence and requests for materials should be addressed to X.S. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and nstitutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 https://doi.org/10.1038/s41598-021-85065-0 Scientific Reports \| (2021) 11:5799 \|
https://openalex.org/W1542816230	https://www.intechopen.com/citation-pdf-url/28216	English	null	Non-Linear Numerical Analysis of Earthquake- Induced Deformation of Earth-Fill Dams	InTech eBooks	2,012	cc-by	12,429	1. Introduction From January 2010 to March 2011, eight major earthquakes of magnitude 6.1 or greater, one of which was 9.0, have shaken the globe. These destructive events, such as January 2010 Haiti earthquake and March 2011 Tohoku Japan earthquake, have resulted in numerous impacts on different lifeline facilities such as dams (Takewaki, 2010; Eberhard et al., 2010; Rathje, 2010; Ashford et al., 2011; Harder et al., 2011; Matsumoto et al., 2011). During these recent earthquakes as well as the past strong ones, numerous earth-fill dams have failed and enormous loss of life and significant property damage have been caused (Sherard et al., 1963; Pinto, 1993; Ozkan, 1998; Krinitzsky & Hynes, 2002; Basudhar et al., 2010; Matsumoto et al., 2011). Therefore, evaluation of stability conditions of earth-fill dams is a major issue of concern in seismic areas. Earthquake-induced deformations in an earth-fill dam may lead to overtopping and consequently to severe losses in terms of property and human lives. Hence, an accurate evaluation of seismic stability of earth-fill dam should be employed to guarantee safer conditions for dams during earthquakes. The performance of earth-fill dams, subjected to seismic action, can be evaluated through different approaches including force-based pseudo-static methods, displacement-based sliding block methods and dynamic analysis (Sherard, 1967; Seed et al., 1975). Since the 1971 San Fernando earthquake in California, understanding the effects of earthquake actions on dams has been significantly progressed (USCOLD, 1992). Gazetas (1987) discussed the historical developments of theoretical methods for estimating the seismic response of earth- fill dams to earthquake. He outlined important features, advantages and limitations of the methods. Pseudo-static approach, the most common method, is widely used in engineering practice to assess the seismic stability of earth-fill dams. This approach is quite simplistic by which the complex aspects of seismic behaviour are represented in terms of static forces and the dam stability is expressed in terms of an overall factor of safety. The response of dam to earthquake may be related to several factors such as dam geometry, mechanical properties of construction soil materials, distributions of pre-seismic stresses and pore water pressures inside the dam body, and input motion characteristics. Most of these factors are partially or totally neglected by the approaches traditionally adopted for assessing the seismic safety of earth-fill dams. Selection of our books indexed in the Book Citation Index in Web of Science™ Core Collection (BKCI) Interested in publishing with us? Contact book.department@intechopen.com Numbers displayed above are based on latest data collected. For more information visit www.intechopen.com Open access books available Countries delivered to Contributors from top 500 universities International authors and editors Our authors are among the most cited scientists Downloads We are IntechOpen, the world’s leading publisher of Open Access books Built by scientists, for scientists 14% 191,000 210M TOP 1% 154 7,200 6 Non-Linear Numerical Analysis of Earthquake- Induced Deformation of Earth-Fill Dams Babak Ebrahimian University of Tehran Iran Babak Ebrahimian University of Tehran Iran www.intechopen.com 1. Introduction For instance, some earthquake parameters such as frequency content and duration which significantly affect the soil response are neglected in the pseudo-static approach (Ambraseys, 1960). On the other hand, studying the seismic response of earth-fill dam is complex and, in general, requires dynamic analysis methods with different levels of www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 140 sophistication in terms of proper problem formulation, characterization of material properties and modelling of soil stress–strain behaviour. In dynamic analysis methods using numerical simulation techniques, comprehensive analysis of earth-fill dam responses to dynamic loading is allowed. The development of geotechnical computation and numerical modelling offers interesting facilities for dam response analysis, considering complex issues such as soil non-linearity, evolution of pore water pressures and real earthquake records. In this regard, Prevost et al. (1985) presented 2D and 3D non-linear dynamic finite element (FE) analyses of an earth-fill dam, based on non-linear hysteretic analysis using multi-surface plasticity theory. Lacy & Prevost (1987) proposed a general and efficient numerical procedure for analyzing the seismic response of earth-fill dams. In their procedure, the dams were considered as non-linear two-phase systems. They outlined appropriate coupled dynamic field equations for the response of two-phase soil system. Abouseeda & Dakoulas (1998) studied the non-linear seismic behaviour in earth-fill dam-foundation interaction using boundary element (BE) and finite element (FE) methods. Chen & Harichandran (2001) studied the stochastic response of Santa Felicia earth-fill dam, in southern California, to spatially varying earthquake ground motion (SVEGM). They used SVEGM model in which the propagation of seismic waves is taken into account. Cascone & Rampello (2003) investigated the seismic stability of an earth-fill dam using decoupled displacement analysis. Ming & Li (2003) conducted a full coupled analysis of failure and considered the remediation of Lower San Fernando Dam. They used a critical state model, incorporating the concept of state dependent dilatancy for describing the soil behaviour over full loading ranges. Adalier & Sharp (2004) studied the seismic behaviour and remediation of an embankment on a liquefiable foundation. Papalou & Bielak (2004) studied the non-linear seismic response of earth-fill dams with canyon interaction. In their developed FE-based method, the dam was idealized as a shear beam and the surrounding medium as a half- space. The dam’s non-linearity was considered using multi-yield surface plasticity theory. www.intechopen.com 1. Introduction Ebrahimian & Vafaeian (2005) considered the seismic response of earth-fill dams during earthquake using 2D full non-linear dynamic analysis. They used finite difference (FD) method and adopted the Mohr-Coulomb elastic-perfectly plastic constitutive model to describe the stress-strain relation of the soil. They focused on the seismic behaviour of very high earth-fill dams. Wang et al. (2006) presented the dynamic analyses in which a non- linear, effective-stress-based soil model is employed. They used bounding surface hypoplasticity model for sand and implemented the model into a 2D finite difference (FD) program. The advantages of the proposed non-linear approach, conducted for a rock-fill dam, were illustrated by comparing the obtained results with those of equivalent linear approach. The model’s capability was demonstrated by evaluating the seismic performance of an earth-fill dam. Siyahi & Arslan (2008) carried out the transient dynamic time history FE simulations to investigate the performance of earth-fill dams under seismic excitation. Then, they studied different failure modes of earth-fill dams as the earthquake aftermath. Sica et al. (2008) studied the effect of loading history on the seismic response of earth-fill dams. They considered the static and seismic behaviours of a real case-history using coupled dynamic approach. The approach was solved numerically by FE method. Rampello et al. (2009) studied the response of an earth-fill dam to seismic loading using the displacement- based analysis and the FE effective-stress dynamic analysis. The FE analysis was carried out using a constitutive model which was capable to reproduce the soil non-linearity and calibrated against laboratory measurements. They also investigated the effects of assumed input motion and bedrock depth on the seismic response of earth-fill dam. Ebrahimian (2009) presented a numerical modelling of seismic behaviour of an earth-fill dam rested on www.intechopen.com 141 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams liquefiable foundation. The numerical simulation was carried out using effective-stress- based, full coupled non-linear dynamic analysis. In this regard, Finn-Byrne model with extended Masing rules was used to model the pore water pressure generation in the liquefied soils. Ebrahimian (2011) investigated the dynamic behavior of earth dams by using a full non-linear dynamic finite difference analysis. The effects of input motion characteristics and dam reservoir condition on the dynamic response of earth dams were identified in this study. For this purpose, three real earthquake records with different levels and PGAs were used as the input motions. 1. Introduction In many parts of the world, the repetition of medium–strong intensity earthquake ground motions at brief intervals of time has been observed. The design philosophies for dams in seismic regions are based on multi-level design approaches, which take into account more than a single damageability limit state. According to these approaches, a sequence of seismic actions may produce important consequences on the dam safety. In fact, dams have been among the first structures that have been designed systematically against different earthquake levels. Since 1989, the ICOLD guidelines have introduced several levels of seismic loading, namely the Operating Basis Earthquake (OBE), Maximum Credible Earthquake (MCE), Maximum Design Earthquake (MDE) and Safety Evaluation Earthquake (SEE). However, the terms MDE or SEE are used as substitutes for the MCE. In order to analyze the behaviour of dams for specified levels of seismic hazard, several requirements should be considered. The seismic input and performance levels associated with serviceability, damage control, and collapse prevention are also defined. A thorough review about the different earthquake levels for dam design has been given in Wieland (2008). Amadio et al. (2003) analyzed the effects of repeated earthquake ground motions on the response of single-degree-of-freedom systems (SDOF) with non-linear behaviour. Accordingly, a comparison study was performed to investigate the effect of a single seismic event on the originally non-damaged system for different hysteretic models in terms of pseudo-acceleration response spectra and damage parameters. They showed that the elastic–perfect plastic system is the most vulnerable one under repeated earthquake ground motions. Moustafa & Takewaki (2010) modelled ground motions of multiple sequences that produce the maximum damage in the structure. It was shown that critical repeated acceleration sequences produce larger structural damage compared to single critical earthquakes. Afterwards, Moustafa (2011) developed a new framework to model the design earthquake loads for inelastic structures. New measures of the structure performance that were based on energy concepts and damage indices were introduced in his paper. Concerning the seismic-resistant design of dams, several international standards have been developed by scientific communities in the past 25 years. However, only few countries have their own guidelines and regulations for seismic design of dams. Therefore, the ICOLD Bulletins and the local seismic building codes (e.g., Eurocode 8) are used as references. 1. Introduction Soil stiffness and hysteretic damping change with loading history. Firstly, the procedures of calibrating the constructed numerical models with centrifuge test data as well as real case history are presented and explained. Moreover, some important aspects of model calibration are discussed. Long Valley earth-fill dam, subjected to the 1980 Mammoth Lake earthquake, is analyzed for the real case history and the obtained numerical results are compared with the real ones, measured at the site in both time and frequency domains. The computed values show relative good agreements with the measured ones. It is shown that the Masing rules, combined with the simple elastic-plastic model, offer reasonable numerical predictions. A comprehensive parametric study is also carried out to identify the effects of dam height, input motion characteristics, soil behaviour and strength of shell materials on the seismic response of earth-fill dams. It is demonstrated that the fundamental aspects of seismic behaviour of earth-fill dams can accurately be captured by the current numerical procedure. It should be mentioned that this study does not consider the fluid-skeleton interaction, which may have significant effects on the seismic response of earth-fill dams. 1. Introduction In brief, other famous international codes are USCOLD (United States Committee on Large Dams), US Army Corps of Engineers (USACE), ANCOLD, (Australian National Committee on Large Dams), IITK-GSDMA Guidelines for Seismic Design of Earth Dams and Embankments and Canadian Dam Association Guidelines for Dam Safety. In this chapter, a numerical study of seismic behaviour of earth-fill dams overlaying bedrock subjected to real earthquake records is presented. For this purpose, full non-linear dynamic finite difference (FD) analysis is employed incorporating a simple elastic perfectly plastic constitutive model and Rayleigh damping. The former is used to describe the stress-strain response of the soil and the latter to increase the hysteretic damping level. The effect of non- www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 142 linear soil behaviour is then considered in the analysis from the very beginning of earthquake loading. In order to precisely explain the soil response under general cyclic loading, Masing rules (Masing, 1926) are implemented into the constitutive model. Soil stiffness and hysteretic damping change with loading history. Firstly, the procedures of calibrating the constructed numerical models with centrifuge test data as well as real case history are presented and explained. Moreover, some important aspects of model calibration are discussed. Long Valley earth-fill dam, subjected to the 1980 Mammoth Lake earthquake, is analyzed for the real case history and the obtained numerical results are compared with the real ones, measured at the site in both time and frequency domains. The computed values show relative good agreements with the measured ones. It is shown that the Masing rules, combined with the simple elastic-plastic model, offer reasonable numerical predictions. A comprehensive parametric study is also carried out to identify the effects of dam height, input motion characteristics, soil behaviour and strength of shell materials on the seismic response of earth-fill dams. It is demonstrated that the fundamental aspects of seismic behaviour of earth-fill dams can accurately be captured by the current numerical procedure. It should be mentioned that this study does not consider the fluid-skeleton interaction, which may have significant effects on the seismic response of earth-fill dams. linear soil behaviour is then considered in the analysis from the very beginning of earthquake loading. In order to precisely explain the soil response under general cyclic loading, Masing rules (Masing, 1926) are implemented into the constitutive model. 2.1 Numerical modeling procedure Here, numerical analysis is conducted using FLAC program, based on a continuum finite difference discretization applying Lagrangian approach (Itasca, 2004). Every derivative in the set of governing equations is directly replaced by algebraic expression written in terms of field variables (e.g., stress or displacement) at discrete point in space. Regarding dynamic analysis, explicit finite difference scheme is applied to solve the full equation of motion using the lumped grid point masses derived from the real density of surrounding zone. The calculation sequence first invokes the equations of motion for deriving new velocities and displacements from stresses and forces; then, strain rates are derived from velocities, and new stresses from strain rates. Every cycle around the loop corresponds to one time step. Each box updates all grid variables from known values which are fixed over the time step being executed (Fig. 1). Fig. 1. Basic explicit calculation cycle (Itasca, 2004) Fig. 1. Basic explicit calculation cycle (Itasca, 2004) www.intechopen.com 143 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams The equation of motion, in the simplest form, relates the acceleration ( du dt  ) of a mass (m) to the applied force (F) which may vary with time. Newton’s law of motion for the mass- spring system is: du m F dt   (1) (1) In a continuous solid body, Eq. (1) is generalized as follows: In a continuous solid body, Eq. (1) is generalized as follows: s generalized as follows: ij i i j u g t x           (2) (2) where, ρ = mass density; t = time; xj = components of coordinate vector; gi = components of gravitational acceleration (body forces); ǔij = components of stress tensor; i = components in a Cartesian coordinate frame. For problem analysis, the strain rate tensor and rotation rate tensor, having the velocity gradients, are calculated by the following equations: 1 2 j i ij j i u u e x x                 (3) 1 2 j i ij j i u u x x                  (4) (3) (4) where, ije = components of strain rate; ij  = components of rotation rate; iu = components of velocity. 2.1 Numerical modeling procedure The specific mechanical relationship is used in order to obtain the stress tensor as below:       , , ij ij ij M e (5) (5) where, M = specific rule of behaviour;  = history parameters (based on the specific rules which may or may not exist). where, M = specific rule of behaviour;  = history parameters (based on the specific rules which may or may not exist). The problem selected here is the simplified representation of typical earth-fill dam geometry. The dam section is a symmetric zone section with central clay core rested on bedrock,   0.5 5 H 3 Fig. 2. Geometry of dam Fig. 2. Geometry of dam Fig. 2. Geometry of dam Fig. 2. Geometry of dam www.intechopen.com Advances in Geotechnical Earthquake Engineering – Li f i d S i i S f f D d M Advances in Geotechnical Earthquake Engineering – uefaction and Seismic Safety of Dams and Monuments Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 144 d a ces Geotec ca a t qua e g ee g Soil Liquefaction and Seismic Safety of Dams and Monuments as shown in Fig. 2. Five earth-fill dam cross sections of different heights (H = 40, 80, 120, 200 and 280 m) are analyzed. as shown in Fig. 2. Five earth-fill dam cross sections of different heights (H = 40, 80, 120, 200 and 280 m) are analyzed. 2.1.1 Constitutive model Mohr-Coulomb constitutive relation is used to model the behaviour of soil. The failure envelope for this model corresponds to a Mohr-Coulomb criterion (shear yield function) with tension cutoff (tensile yield function). Stress-strain relationship is linear elastic- perfectly plastic. Linear behaviour is defined by elastic shear and bulk modulus. While, plastic behaviour is determined by the angle of internal friction and cohesion of the soil. Shear modulus of sandy shell materials is calculated from the following formula (Kokusho & Esashi, 1986):   2 0.6 max (2.17 ) 8400 1 m e G e      (6) (6) where, Gmax = maximum (small strain) shear modulus in kPa; e = void ratio; m  = mean effective confining stress in kPa; Poisson’s ratio is considered as 0.3 for the shell materials. Shear modulus of clayey core materials is calculated by the below formula (Hardin & Black, 1968):    2 0.5 max 2.973 3270 1 m e G e      (7) (7) Poisson’s ratio for the core materials is taken as 0.45. Here, the basic elastic-perfectly plastic model is modified in order to better fit with the curves of shear modulus and damping ratio derived from the experimental data. This modified model can predict the seismic behaviour and the associated deformations of earth- fill dams. Masing behaviour is implemented into FLAC via FISH subroutine (Itasca, 2004) in order to represent more accurately the non-linear stress-strain behaviour of soil that follows the actual stress-strain path during cyclic loading. Masing model consists of a backbone curve and several rules that describe the unload-reload behaviour of soil and the cyclic modulus degradation. Backbone curve can be constructed by the modulus reduction curves coupled with the small strain modulus (Gmax). Unload-reload rules can similarly be formulated to reproduce the hysteretic damping values expected from the standard curves of damping ratio versus shear strain (e.g., Seed et al., 1986; Vucetic & Dobry, 1991). These formulations are described later. In this study, shear modulus and damping ratio curves, proposed by Seed et al. (1986) for sandy soils and Vucetic & Dobry (1991) for clayey soils, are adopted as the references. The geotechnical properties of earth-fill dam materials, used in the analyses, are presented in Table 1. Region wet  sat  ǖ Porosity (n) C Ǘ K (kN/m3) (kN/m3) (kPa) degree (cm/s) Core 20 20.5 0.45 0.41 80 8 10-7 Shell 22 23.0 0.30 0.33 - 40 10-2 www.intechopen.com 145 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 2.1.2 Boundary conditions The geotechnical problems can be idealized by assuming that the regions far from the area of interest extend to infinity. The unbounded theoretical models should be truncated to the manageable size by using the artificial boundaries for minimizing the computation time as well as avoiding the outwards propagating waves form reflecting to the model. The viscous boundary, developed by Lysmer & Kuhlemeyer (1969), is used in the current calculations. In this case, independent dashpots are used in the normal and shear directions at the model boundaries, as shown in Fig. 3. During the static analysis, the bottom boundary is fixed in both the horizontal and the vertical directions and the lateral boundaries only in the horizontal direction. In dynamic analysis, when the dam is laid on the foundation (and not on the bedrock), lateral boundaries are changed into free-field boundaries, available in the FLAC library, in order to eliminate the wave reflections from the truncated boundaries. 2.1.3 Element size Numerical distortion of propagating wave can occur in dynamic analysis as a function of modelling condition. The numerical accuracy of wave transmission is affected by both the frequency content of input wave and the wave speed characteristics of system. Kuhlemeyer & Lysmer (1973) showed that for an accurate representation of wave transmission through the soil model, the spatial element size should be smaller than 1/10 to 1/8 of the wavelength associated with the highest frequency component of input wave i.e., 9 L    (8) (8) where, λ = wave length associated with the highest frequency component that contains significant energy. Considering the above mentioned criteria, the element size is defined small enough to allow the seismic wave propagation throughout the analysis. where, λ = wave length associated with the highest frequency component that contains significant energy. Considering the above mentioned criteria, the element size is defined small enough to allow the seismic wave propagation throughout the analysis. Fig. 3. Free field boundaries in dynamic analysis (Itasca, 2004) Fig. 3. Free field boundaries in dynamic analysis (Itasca, 2004) www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 146 2.1.4 Damping Material damping in soil is generally because of its viscosity, friction and plasticity development. Indeed, the role of damping in the numerical models is the reproduction of energy losses in the natural systems subjected to dynamic loads. The dynamic damping is provided in the model by Rayleigh damping option available in FLAC. Rayleigh damping was primarily used in analyzing the structures and elastic continua in order to damp the natural oscillation modes of system. Rayleigh damping Rd=5% is used to compensate the energy dissipation in the media (Itasca, 2004). In the dynamic analysis incorporated plasticity constitutive models, considerable amount of energy dissipation can occur during the plastic flow. In the calculations of such cases, minimal percentage of hysteretic damping (e.g. 2%) is considered as well. The dam’s natural frequency is determined as the Rayleigh damping parameter by Fourier analysis of its free response, as shown in Fig. 4. The fundamental frequency (f1 = 1.71 Hz) of the dam with 40 m height is shown in this figure and those of dams with different heights are tabulated in Table 2. 2.2 Input excitations Selecting dynamic input motion is an important task in the seismic evaluation processes. In non-linear dynamic analysis, the expected earthquakes should be expressed as a set of ground motion time histories. For more correct evaluation, the input motions which offer an appropriate range of dam responses should be selected in the adaptable time history realizations. This procedure may be intractable due to the number of time-history realizations. However, in reality, the level of earthquake responses, probably achieved by physical system, is limited. Quantifying such responses demands good understanding of the seismic response of the system as well as the ground motion parameters that characterize the damage potential of seismic input (USCOLD, 1999). Different parameters can be employed to identify the severity and damage potential of a certain acceleration time history, assumed as the representative of expected earthquake ground motion; peak ground acceleration value (PGA) is of such kind. The use of this descriptor is intuitively natural since accelerations and resulting inertial forces are directly related by Newton’s second law. However, there is no direct relation between PGA and structural response at the dominant natural frequencies of most typical dams. Moreover, large PGA values are not sufficient for generating response conditions which lead to significant damage. Despite these limitations, PGA is still the fundamental parameter used to judge the damage potential of certain acceleration histories. On the other hand, the seismic response of system is strongly affected by the frequency content of earthquake. Therefore, the better characterization of a given input motion can be achieved by using some forms of spectral representations. In particular, using Fourier amplitude spectrum is at the core of earthquake engineering practice. However, such characterizations do not provide direct description of the duration or time variation features of a given input motion. Based on above, in this chapter, three different real acceleration time histories are selected from a database of earthquake records: Tabas, PGA=0.93g in MCE level; Naghan, PGA=0.72g in MDE level; San Fernando, PGA=0.21g in DBE level. In the dynamic analysis of dams, the scaled records are filtered to the maximum frequency of 10 Hz, transferred to the “inside” bedrock formation by standard de-convolution analysis and applied at the base of numerical model. The information of earthquake records are summarized in Table 3, and their corresponding acceleration time histories and Fourier amplitude spectra are shown in Figs. 5 & 6, respectively. 2.1.5 Time step The governing equations in time should be integrated incrementally for completing the numerical solution. The solution time step should be small enough in order to accurately define the applied dynamic loads and guarantee the stability and convergence of solution. In this regard, time step is about 10-6 second in the current FLAC model. Fig. 4. Fourier amplitude spectrum of free horizontal motion at the dam crest First mode of vibration Dam Height (m) 40 80 120 200 280 f1 (Hz.) 1.7 0.88 0.60 0.36 0.26 Table 2. Fundamental frequency Fig. 4. Fourier amplitude spectrum of free horizontal motion at the dam crest First mode of vibration Dam Height (m) 40 80 120 200 280 f1 (Hz.) 1.7 0.88 0.60 0.36 0.26 Table 2. Fundamental frequency www.intechopen.com 147 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 2.2 Input excitations 6 0 0.2 0.4 0.6 0.8 0 5 10 15 20 25 Frequency (Hz) FAS (m/s) (c) San Fernando Fig. 6. Fourier amplitude spectra 2.3 Full non-linear dynamic analysis Frequency (Hz) (c) San Fernando Fig. 6. Fourier amplitude spectra 2.2 Input excitations Earthquake Station Date M Closest Distance PGA (g) PGV PGD (km) (cm/s) (cm) Tabas Tabas 1978 7.4 94 0.93 121.4 94.58 Naghan Naghan 1977 5.4 75 0.72 46.20 61.00 San Fernando Pasadena 1971 6.6 19 0.21 10.90 2.320 Table 3. Earthquake records data Table 3. Earthquake records data www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 148 -10 -5 0 5 10 0 5 10 15 20 25 Time (Sec.) Acceleration (m/s2) (a) Tabas -10 -5 0 5 10 0 5 10 15 20 25 Time (Sec.) Acceleration (m/s2) (b) Naghan -2.5 0 2.5 0 5 10 15 20 Time (sec.) Acceleration (m/s2) (c) San Fernando -10 -5 0 5 10 0 5 10 15 20 25 Time (Sec.) Acceleration (m/s2) (a) Tabas -10 -5 0 5 10 0 5 10 15 20 25 Time (Sec.) Acceleration (m/s2) (b) Naghan -2.5 0 2.5 0 5 10 15 20 Time (sec.) Acceleration (m/s2) (c) San Fernando Fig. 5. Input acceleration time histories www.intechopen.com 25 5 10 15 Time (sec.) (c) San Fernando 5 15 20 Fig. 5. Input acceleration time histories Fig. 5. Input acceleration time histories Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 149 0 2 4 6 0 5 10 15 20 25 Frequency (Hz) FAS (m/s) (a) Tabas 0 0.5 1 1.5 2 0 5 10 15 20 25 Frequency (Hz) FAS (m/s) (b) Naghan 0 0.2 0.4 0.6 0.8 0 5 10 15 20 25 Frequency (Hz) FAS (m/s) (c) San Fernando Fig. 6. Fourier amplitude spectra 2.3 Full non-linear dynamic analysis Equivalent linear analysis is the common method used for evaluating the seismic behaviour of earth structures. In this approach, first, the responses are linearly analyzed using the initial values of damping ratio and shear modulus. Then, the new values of damping ratio and shear modulus are estimated, using maximum value of shear strain and laboratory curves. These values are used for redoing the analysis. This procedure is repeated several times until the material properties show no variation. Therefore, no non-linear effect is directly captured by this method as it assumes linearity during the solution process. Strain- dependent modulus and damping functions are considered roughly in order to approximate some effects of non-linearity (damping and material softening). 2.3 Full non-linear dynamic analysis 2.3 Full non-linear dynamic analysis Equivalent linear analysis is the common method used for evaluating the seismic behaviour of earth structures. In this approach, first, the responses are linearly analyzed using the initial values of damping ratio and shear modulus. Then, the new values of damping ratio and shear modulus are estimated, using maximum value of shear strain and laboratory curves. These values are used for redoing the analysis. This procedure is repeated several times until the material properties show no variation. Therefore, no non-linear effect is directly captured by this method as it assumes linearity during the solution process. Strain- dependent modulus and damping functions are considered roughly in order to approximate some effects of non-linearity (damping and material softening). Equivalent linear analysis is the common method used for evaluating the seismic behaviour of earth structures. In this approach, first, the responses are linearly analyzed using the initial values of damping ratio and shear modulus. Then, the new values of damping ratio and shear modulus are estimated, using maximum value of shear strain and laboratory curves. These values are used for redoing the analysis. This procedure is repeated several times until the material properties show no variation. Therefore, no non-linear effect is directly captured by this method as it assumes linearity during the solution process. Strain- dependent modulus and damping functions are considered roughly in order to approximate some effects of non-linearity (damping and material softening). www.intechopen.com Advances in Geotechnical Earthquake Engineering – il Li f ti d S i i S f t f D d M t Advances in Geotechnical Earthquake Engineering – l Liquefaction and Seismic Safety of Dams and Monuments Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 150 Advances in Geotechnical Earthquake Engineering Soil Liquefaction and Seismic Safety of Dams and Monuments In the non-linear analysis, employed in this study, the non-linear stress-strain relationship is directly followed by each zone. Damping ratio and shear modulus of the materials are calculated automatically at different strain levels. The real behaviour of soil, under cyclic loading, is non-linear and hysteretic. Such behaviour can be simulated by Masing model (Masing, 1926), which can model the dynamic behaviour of soil. 2.3 Full non-linear dynamic analysis In this model, the shear behaviour of soil may be explained by a backbone curve as:    max max max 1 / bb G F G       (9) (9) (9) where, Fbb(γ) = backbone or skeleton function;  = shear strain amplitude; Gmax = initial shear modulus; Ǖmax = maximum shear stress amplitude. where, Fbb(γ) = backbone or skeleton function;  = shear strain amplitude; Gmax = initial shear modulus; Ǖmax = maximum shear stress amplitude. Stress-strain curve follows the backbone curve in the first loading, as shown in Fig. 7(a); however, for explaining the unload-reload process, the above equation should be modified. If load reversal occurs at the point (Ǖr, γr), stress-strain curve follows the path given by the below formula: 2 2 r r bb F              (10) (10) In other words, the shapes of unload-reload curves are similar to that of backbone curve (with the origin shifted to the loading reversal point) except they are enlarged by a factor of 2, as shown in Fig. 7(b). The Eqs. (9) & (10) describe the Masing behaviour (Masing, 1926). Masing rules seem not to be enough for precise explanation of soil response under general cyclic loading. Finn et al. (1977) developed modified rules to describe the irregular loading. They suggested that unloading and reloading curves follow the concerning two rules. If the new unloading or reloading curve exceeds the last maximum strain and cut the backbone curve, it will follow the backbone curve up to meeting the next returning point, as shown in Fig. 7(c). If a new unloading or reloading curve passes through the previous one, it will follow the former stress-strain curve, as shown in Fig. 7(d). 2.3 Full non-linear dynamic analysis According to this model, the tangent shear modulus can be defined at the points on the backbone and new reloading- unloading curves by the Formulas (11) & (12), respectively, as: 2 max max max 1 t G G G           (11) 2 max max max 1 2 t r G G G             (12) 2 max max max 1 t G G G           (11) (11) 2 max max max 1 2 t r G G G             (12) (12) Based on the results, obtained in this research, the shear stress decreases as the number of load cycles increases; it means that shear stress-strain curves are more inclined. In this study, Masing rules are implemented into FLAC via a series of FISH functions in order to simulate the non-linear stress-strain relationships. www.intechopen.com Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 151 ( r r) Gmax Gmax (a) Rule 1 (b) Rule 2 (c) Rule 3 (d) Rule 4 Fig. 7. General patterns of loading, unloading and reloading paths in Masing model ( r r) Gmax Gmax (a) Rule 1 (b) Rule 2 Gmax (b) Rule 2 (a) Rule 1 ( ) ( ) (c) Rule 3 (d) Rule 4 Fig. 7. General patterns of loading, unloading and reloading paths in Masing model (d) Rule 4 (c) Rule 3 (c) Rule 3 (d) Rule 4 Fig. 7. General patterns of loading, unloading and reloading paths in Masing model 3.1 Validation analysis In this research, one-zone sample is simulated using the unit cell as shown in Fig. 8, in order to validate the implementation of Masing rules in FLAC program. The one-zone sample consists of sandy soil and a periodic motion is exerted at its base. Vertical loading is established only by gravity and then, the Equilibrium stresses are installed in the soil. The stress/strain loops of the sample are shown in Fig. 9(a) for several cycles. According to the figure, shear modulus decreases as shear strain increases. It seems that the hysteretic model can reasonably handle the multiple nested loops. Energy dissipation and shear stiffness degradation are clearly observed during seismic loading, as shown in Fig. 9(b). Shear modulus reduction curve, obtained in this study, follows well the empirical relation proposed by Seed et al. (1986) and the test data. www.intechopen.com Advances in Geotechnical Earthquake Engineering – f i d S i i S f f D d M Advances in Geotechnical Earthquake Engineering – uefaction and Seismic Safety of Dams and Monuments 152 Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments q g g Soil Liquefaction and Seismic Safety of Dams and Monuments Fig. 8. One-zone model in FLAC for simulating cyclic simple shear test -300 -250 -200 -150 -100 -50 0 50 100 150 200 250 -0.0015 -0.0010 -0.0005 0.0000 0.0005 0.0010 Shear strain Shear stress (kPa) Numerical Prediction Seed et al. (1986) Test data (a) (b) Fig. 9. Simulation results of one-zone sample: (a) hysteresis loop, and (b) comparing the shear modulus reduction curves Fi 8 O d l i FLAC f i l ti li i l h t t Fig. 8. One-zone model in FLAC for simulating cyclic simple shear test Fig. 8. One-zone model in FLAC for simulating cyclic simple shear test Numerical Prediction Seed et al. (1986) Test data (b) -300 -250 -200 -150 -100 -50 0 50 100 150 200 250 -0.0015 -0.0010 -0.0005 0.0000 0.0005 0.0010 Shear strain Shear stress (kPa) (a) (a) (b) Fig. 9. Simulation results of one-zone sample: (a) hysteresis loop, and (b) comparing the shear modulus reduction curves The results obtained from the numerical analyses are compared with those of experimental ones in order to evaluate the capability of proposed model. www.intechopen.com 3.1 Validation analysis One of the centrifuge tests related to the embankment which was performed in VELACS project (VErification of Liquefaction Analysis using Centrifuge Studies (Arulanandan and Scott, 1993, 1994)) is chosen as a reference. It is attempted to create almost similar conditions for both laboratory model test and numerical model, as shown in Figs. 10(a) & (b), respectively. The numerical results are presented and compared with those obtained from the corresponding centrifuge test data. The computed (numerical simulations) and measured (centrifuge tests) results are shown in Figs. 11(a) & (b). According to these comparisons, the reference numerical model can rationally predict the seismic behaviour of earth-fill dam. Finally, the model’s ability in simulating the seismic behaviour of earth-fill dam during a real earthquake is verified by a real well-documented case history. In this regard, the results of non-linear dynamic analysis of Long Valley (LV) earth-fill dam in California subjected to the 1980 Mammoth Lake earthquake (Griffiths and Prevost, 1988) are presented. Then, the results are compared with the real measurements recorded at the site and also with the results presented by previous authors. LV dam is located in Mammoth Lake area www.intechopen.com 153 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams (California) in close proximity of active faults. The dam is a rolled earth-fill one formed mainly with the impervious zone. The dam has maximum height of 55 m, 182 m length at the crest, and upstream and downstream slopes of 3h/1v. The LV dam was instrumented in the 1970’s with a multiple-input-output array; it has 3 accelerometer stations to monitor the boundary conditions, and 5 stations to record the dam response (Fig. 12(a)). Thus, the array comprised a total of 22 accelerometers linked to a common triggering mechanism. (California) in close proximity of active faults. The dam is a rolled earth-fill one formed mainly with the impervious zone. The dam has maximum height of 55 m, 182 m length at the crest, and upstream and downstream slopes of 3h/1v. The LV dam was instrumented in the 1970’s with a multiple-input-output array; it has 3 accelerometer stations to monitor the boundary conditions, and 5 stations to record the dam response (Fig. 12(a)). Thus, the array comprised a total of 22 accelerometers linked to a common triggering mechanism. (a) (b) Fig. 10. 3.1 Validation analysis Model configuration: (a) schematic sketch of dam in centrifuge container box, and (b) numerical grid constructed in FLAC (a) (b) Fig. 11. Measured time histories versus calculated ones at the middle of dam height: (a) acceleration, and (b) vertical displacement (a) (b) Fig. 10. Model configuration: (a) schematic sketch of dam in centrifuge container box, and (b) numerical grid constructed in FLAC (a) (b) (a) (b) Fig. 10. Model configuration: (a) schematic sketch of dam in centrifuge container box, and (b) numerical grid constructed in FLAC (b) (a) (b) Fig. 11. Measured time histories versus calculated ones at the middle of dam height: (a) acceleration, and (b) vertical displacement (a) (a) (b) Fig. 11. Measured time histories versus calculated ones at the middle of dam height: (a) acceleration, and (b) vertical displacement In May 1980, a series of 6 earthquakes occurred in the Mammoth Lakes area. The magnitudes of these earthquakes were ML = 4.9 - 6.7, and the induced peak accelerations at the crest centre was 0.5g in the upstream-downstream direction (x direction, as shown in Fig. 13(a)) during the strongest event. Extensive arrays of 22 input-output (excitation- response) accelerations were recorded, providing a valuable information source of the dam seismic responses over a wide range of deformation levels. In this study, the dam is subjected to the input motion, recorded downstream at the outlet during Mammoth Lake earthquakes. The first 12 seconds of the recorded acceleration is used with data point at 0.02 second intervals and the peak acceleration is 0.135g in the upstream-downstream (x) direction and 0.084g in the vertical direction (y). The cross section of LV dam is shown in Fig. 12(b) and its detailed information is found in (Griffith and Prevost, 1988). The numerical grid constructed in FLAC is presented in Fig. 12(c). The input accelerations are applied in the horizontal and vertical directions of the model base. Free Field boundary conditions are exerted to the lateral boundaries of numerical model. 3.1 Validation analysis www.intechopen.com Advances in Geotechnical Earthquake Engineering – oil Liquefaction and Seismic Safety of Dams and Monuments Advances in Geotechnical Earthquake Engineering – f i d S i i S f f D d M Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 154 Advances in Geotechnical Earthquake Engineering Soil Liquefaction and Seismic Safety of Dams and Monuments This research focuses on the computed acceleration at the crest which can be compared directly with the measured values. Previously, LV dam has been analyzed by: Lai & Seed, 1985; Elgamal et al., 1987; Griffiths & Prevost, 1988; Yiagos & Prevost, 1991; Zeghal & Abdel- Ghaffar, 1992; Woodward & Griffiths, 1996. The first natural frequency, obtained in this This research focuses on the computed acceleration at the crest which can be compared directly with the measured values. Previously, LV dam has been analyzed by: Lai & Seed, 1985; Elgamal et al., 1987; Griffiths & Prevost, 1988; Yiagos & Prevost, 1991; Zeghal & Abdel- Ghaffar, 1992; Woodward & Griffiths, 1996. The first natural frequency, obtained in this (a) (b) (c) Fig. 12. Long Valley earth-fill dam: (a) schematic view of Long Valley canyon, earth-fill dam and installed instrumentation array acceleration, (b) cross section, and (c) numerical grid for dynamic analysis (a) (a) (b) (c) (b) (c) Fig. 12. Long Valley earth-fill dam: (a) schematic view of Long Valley canyon, earth-fill dam and installed instrumentation array acceleration, (b) cross section, and (c) numerical grid for dynamic analysis www.intechopen.com Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 155 study, is presented in Table 4 and compared with the other solutions available in the literature. The results of present study are reasonably in close agreement with those of other relevant numerical investigations. Here, the crest acceleration responses of LV dam are computed and compared with those recorded at the site in both time and frequency domains. Frequency Spectral Analysis 2D FE Analysis 3D FE Analysis Elasto- plastic FE Analysis Elasto- plastic FE Analysis Present Study Griffith & Prevost (1988) Griffith & Prevost (1988) Griffith & Prevost (1988) Yiagos & Prevost (1991) Woodward & Griffiths (1996) f (Hz.) 1.85 1.76 1.95 1.987 1.79 1.71 Table 4. First natural frequency of Long Valley dam Table 4. First natural frequency of Long Valley dam Fig. www.intechopen.com 3.1 Validation analysis 13(a) shows the computed horizontal acceleration of the crest; it indicates that the amplification occurs between the base and the crest. The magnification factor of peak amplitude at the crest is about 5.47 over the peak base amplitude. The crest response, computed in the horizontal direction, is compared with the measured values, as shown in Fig. 13(b); the dashed line corresponds to the computed response there. Excellent overall agreement is achieved between the computed and measured values; however, the computed values show higher amplitudes. The frequency contents of two time records are compared in the form of Fourier amplitude spectra (FAS), as shown in Fig. 13(c). Their peaks are in close agreement although the computed values show rather more energy associated with the fundamental frequency around 1.8 Hz. The frequency content of the up/down stream motion, presented in Fig. 13(c), shows that the energy is concentrated just at the frequencies below 2 Hz. In the vertical direction, the calculated acceleration shows low agreement with the measured values, as shown in Fig. 14(a). According to this figure, the plots of vertical acceleration are superimposed at the base and crest. This excitation is considerably noisier and less intensive in the vertical direction in compared with that of horizontal one. The maximum accelerations at the crest, recorded in the vertical and horizontal directions are 0.172g and 0.64g, respectively. The computed accelerations in the vertical direction are compared with the measured values in the crest of LV dam, as shown in Fig. 14(b). According to this figure, the computed values have generally lower amplitudes in compared with those of measured values. The Fourier amplitude spectra of these time histories are given in Fig. 14(c). The measured values show a broad band of frequencies, none of which is dominant. The computed values also contain a broad band of frequencies, but with clear peaks in the ranges of 2-3 Hz and 5-6 Hz. It should be mentioned that the time and frequency domain www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 156 (a) (b) (c) Fig. 13. Comparing computed and measured time histories in horizontal direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum (a) (b) (c) Fig. 13. 3.1 Validation analysis Comparing computed and measured time histories in horizontal direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum (a) (b) (c) Fig. 14. Comparing computed and measured time histories in vertical direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum (a) (b) (c) (c) Fig. 13. Comparing computed and measured time histories in horizontal direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum (a) (b) (c) Fig. 14. Comparing computed and measured time histories in vertical direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum (a) (b) (a) (b) (a) (b) (c) (a) (b) (c) (c) Fig. 14. Comparing computed and measured time histories in vertical direction: (a) crest and base acceleration, (b) crest acceleration, and (c) Fourier amplitude spectrum www.intechopen.com 157 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams results show more appropriate accordance in the horizontal direction in compared with those of the vertical direction. The results, obtained in the validation analysis of LV dam, in term of crest acceleration are given in Table 5 and compared with the other numerical results presented by previous authors. According to the comparisons, the numerical procedure, presented in this study, can properly capture the fundamental aspects of seismic behaviours of earth-fill dams. As mentioned earlier, the numerical model is then used for parametric studying of hypothetical earth-fill dams due to the satisfactory modelling of validation cases. Yiagos & Prevost (1991) Woodward & Griffiths (1996) Present Study Measured values Maximum horizontal Acceleration (g) 0.53 0.80 0.61 0.40 Minimum horizontal Acceleration (g) -0.65 -0.68 -0.50 -0.50 Table 5. Comparing the Numerical results obtained for Long Valley dam www.intechopen.com 3.2 Parametric study Here, the analyses are carried out to investigate the effects of dam height, input motion characteristics, soil behaviour and strength of shell materials on the seismic behaviour of earth-fill dams. The effects of different earthquakes are studied on the horizontal permanent deformations, permanent shear strains and maximum accelerations, as shown in Fig. 15. The values have been induced at the crests of dams with different heights. The displacements are shown in Fig. 15(a) and the relevant shear strains in Fig. 15(b). It is clear that the shear strain variation is similar to displacement. The horizontal displacements and shear strains in the dam body increase with dam height increasing. The calculated values are much higher in Tabas earthquake and the failure occurs in the dam body. According to Fig. 15(a), the maximum horizontal displacement computed at the crest of dam is about 94 cm at the end of Tabas earthquake. It can be observed in Figs. 15 (a) & (b), that increasing in the input motion energy leads to significant increase of displacements and shear strains. Fig. 15(c) illustrates the coupled effects of dam height and earthquake type on the maximum acceleration induced at the dam crest. According to the figure, the crest acceleration decreases as the dam height increases and no amplification is seen maybe due to more flexible behaviour, larger damping and larger developed plastic zones, observed in higher dams. Therefore, because of these factors, more energy is absorbed in higher dams in compared with that in the shorter ones. It can be seen in Fig. 15 (c) that the accelerations in the dam crest are more reduced in higher dams comparing with the smaller ones. It should be mentioned that PGA of Naghan earthquake (0.72g) is much higher than that of San Fernando earthquake (0.21g). However, the created displacements and shear strains in the www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 158 Advances in Geotechnical Earthquake Engineering Soil Liquefaction and Seismic Safety of Dams and Monuments dam crest caused by Naghan earthquake are close to those of San Fernando input motion. It can be concluded that using just PGA parameter is not sufficient for evaluating the effect of a certain earthquake time history on the dam response. Therefore, other earthquake parameters such as effective duration, magnitude and frequency content should be considered in the analysis. (a) (b) (c) Fig. 3.2 Parametric study 15. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations at dam crest versus dam height for different earthquakes Failure mechanism with permanent shear strain contour in the dam body is shown in Fig. 16, regarding two different heights at the end of Naghan earthquake. The slip surface is much deeper and more obvious in the dam with 280 m height (Fig. 16(b)) in compared with that of 120 m height (Fig. 16(a)). (b) (a) (b) (a) (c) (c) Fig. 15. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations at dam crest versus dam height for different earthquakes Failure mechanism with permanent shear strain contour in the dam body is shown in Fig. 16, regarding two different heights at the end of Naghan earthquake. The slip surface is much deeper and more obvious in the dam with 280 m height (Fig. 16(b)) in compared with that of 120 m height (Fig. 16(a)). Failure mechanism with permanent shear strain contour in the dam body is shown in Fig. 16, regarding two different heights at the end of Naghan earthquake. The slip surface is much deeper and more obvious in the dam with 280 m height (Fig. 16(b)) in compared with that of 120 m height (Fig. 16(a)). (a) (b) Fig. 16. Failure surface with shear strain contour in dam body at the end of Naghan earthquake: (a) dam height=120 m, and (b) dam height=280 m (a) (b) (b) (a) Fig. 16. Failure surface with shear strain contour in dam body at the end of Naghan earthquake: (a) dam height=120 m, and (b) dam height=280 m www.intechopen.com 159 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams A dam with 40 m height is subjected to the mentioned earthquakes and chosen as a reference with two different behaviours, elastic and elastic-perfectly plastic, in order to investigate the effect of soil behaviour on the seismic response of dam body. As it is expected, regarding linear elastic behaviour, smaller displacements and shear strains are observed along the dam height, Figs. 17(a) & (b). However, large amplification occurs especially for the strongest earthquake, Fig. 17(c). It means that plasticity reproduce more energy dissipation during dynamic loading. In such cases, the accelerations are reduced across the dam height and therefore become lower than the base acceleration. According to Fig. 3.2 Parametric study 17(a), maximum displacement occurs at about Z/H = 0.88 in linear elastic behaviour, while it happens at the crest of dam in linear elastic-perfectly plastic behaviour. Furthermore, in elastic-perfectly plastic behaviour, the dynamic induced residual (permanent) displacement increases largely in the upper part of dam, especially for Tabas and Naghan earthquakes, confirmed in the previous research works (Ohmachi and Kuwano, 1994; Ozkan et al., 2006). That is why the crest should especially be considered in designing the embankment dams, due to the stronger shaking at the upper parts, for avoiding undesirable deformations. The distribution of shear strain is extremely non-linear along the dam height in the stronger earthquakes, as shown in Fig. 17(b). In the elastic dams, maximum acceleration occurs in the dam crest, as shown in Fig. 17(c). In the acceleration profile of Tabas earthquake, a special increase is seen along the dam centerline at Z/H = 0.38. (a) (b) (c) Fig. 17. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations along the dam height for elastic and elastic- plastic behaviours subjected to different earthquakes (a) (b) (a) (b) (a) (b) ( ) ( ) (c) (c) Fig. 17. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations along the dam height for elastic and elastic- plastic behaviours subjected to different earthquakes www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 160 Strength of dam materials is an important parameter which can significantly affect the seismic response of dam. In this regard, different friction angles are assumed accompanying with different dam heights, subjected to Naghan earthquake, in order to clarify the above mentioned effect. Fig. 18(a) shows horizontal displacement values versus dam height for different friction angles of shell materials. The variations of shear strains in the crests of dams, with different heights, are shown in Fig. 18(b) for different friction angles of shell materials. As it is expected when the friction angle increases, the horizontal displacement and shear strain in the dam crest decrease. It can be seen in the above figure that the variation of friction angle causes no significant change in the displacement and shear strain regarding Ǘ ≤ 40˚. 3.2 Parametric study However, the variation is more significant in Ǘ = 45˚, compared with the lower friction angles; the highest displacement values correspond to Ǘ = 30˚. The horizontal displacements computed at the crests of dams with 40 and 120 m heights are about 16 and 13 cm, respectively, and their shear strains are about 3.5־³ and 2.5־³, respectively. The maximum acceleration induced at the top of dam decreases as the friction angle decreases or the dam height increases, as shown in Fig. 18(c). Considering larger friction angles for the shell materials (e.g., Ǘ = 45˚) leads to about 70% increase in the dynamic amplification. The computed maximum crest acceleration of dam with 40 m height is about 0.89g for Ǘ = 45˚, while that of 120 m height is 0.52g. When the dam height deceases, the horizontal displacement and shear strain increase but the acceleration decreases at the crest of dam. All variations are linear for Ǘ = 45˚, but for the other friction angles are slightly non-linear, as shown in Fig. 18. 0.00 0.05 0.10 0.15 0.20 0.25 30 50 70 90 110 130 Height (m) Horizontal Displacement (m)     0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.0040 0.0045 0.0050 30 50 70 90 110 130 Height (m) Shear Strain     (a) (b) 3 4 5 6 7 8 9 10 30 50 70 90 110 130 Height (m) Acceleration (m/s2)     (c) Fig. 18. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations at dam crest versus dam height for different friction angles of shell materials 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.0040 0.0045 0.0050 30 50 70 90 110 130 Height (m) Shear Strain     (b) 0.00 0.05 0.10 0.15 0.20 0.25 30 50 70 90 110 130 Height (m) Horizontal Displacement (m)     Sh St i (a) (b) (a) 3 4 5 6 7 8 9 10 30 50 70 90 110 130 Height (m) Acceleration (m/s2)     (c) (c) Fig. 18. Computed values of: (a) permanent horizontal displacements, (b) permanent shear strains, and (c) induced maximum accelerations at dam crest versus dam height for different friction angles of shell materials www.intechopen.com 161 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 3.3 Lessons learned The author experienced several interesting points and noteworthy items during numerical model calibration and numerous dynamic analyses which are listed below:  In full non-linear dynamic analysis, soil stiffness degradation is automatically taken into account upon constitutive model of soil and just the initial shear modulus is needed as an input parameter. Therefore, it is important to be sure that in the numerical model, the trend of shear modulus decrease and damping ratio increase are in agreement with those of laboratory test results during dynamic loading.  The poisson’s ratio of about 0.5 should not be used in the calculating of bulk modulus of undrained soil layers (such as clay) in the analysis. Otherwise, bulk modulus increases irrationally and the time step of analysis decreases rigorously and consequently the calculation time increases excessively. Therefore, the poisson’s ratio should not be more than 0.45 in such cases.  If a “raw” acceleration record from a site is used as a time history, then FLAC model may exhibit residual displacements once the motion is finished. This arises from the fact that the integral of complete time history may not be zero. Therefore, the process of baseline drift correction should be performed in such cases.  The input motion should be filtered before being applied to the FLAC grid in order to eliminate all high frequency components form it. g y  The stages of construction should be considered in the numerical simulation of earth-fill dams. In the present study, the stages are: initial state of foundation (if any); layer by layer dam replacement; applying the hydrostatic water pressure due to the replacement of dam reservoir; seepage analysis in the dam body; mechanical adjustment to new flow field; and finally dynamic analysis. Regarding the mentioned stages, one is run to equilibrium and then the next stage is started. However, construction sequences have much greater effects on static results than dynamic results. www.intechopen.com 4. Conclusions This chapter presents the non-linear seismic behaviour of earth-fill dams using explicit finite difference method. In this regard, a simple elastic-perfectly plastic constitutive model with Mohr-Coulomb failure criterion is used to describe the stress-strain response of the soil. Here, Rayleigh damping is used to promote the level of hysteretic damping during dynamic analysis. Masing rules are implemented into the constitutive model to precisely explain the non-linear soil response under general cyclic loading. The numerical model is then calibrated using centrifuge test data as well as field data. The field data are obtained in real measuring of Long Valley earth-fill dam subjected to the 1980 Mammoth Lake earthquake. The results of dynamic analysis, obtained in this study, are compared with the real measurements of Long Valley dam in terms of accelerations computed at the crest of dam in both time and frequency domains. The proposed numerical model can properly reproduce the overall seismic behaviours of earth-fill dams, their qualities and quantities, under earthquake loading conditions, confirmed by validation analyses. After validation, the effects of dam height, real earthquake loading, soil behaviour and strength of shell materials on the seismic response of earth-fill dams are evaluated through a comprehensive www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 162 parametric study. The effect of dam height on the non-linear seismic behaviour is particularly focused in this research. The following conclusions are obtained based on the performed parametric study:  If the dam materials keep their elastic behaviours during dynamic loading, then the horizontal acceleration increases along the dam height (from the base to the top). In this case, the higher dams show larger amplifications, especially if the natural periods of their bodies coincide with the periodical nature of earthquake waves.  When the dam body shows non-linearity or the materials go towards plastic behaviour during a strong shaking, the attenuation of acceleration waves in the dam body becomes more effective. Consequently, the amplitudes of earthquake accelerations decrease when moving from the base towards the top.  According to the non-linear elastic-plastic analyses, when the height of the dam increases, then the strongest dynamic loading (Tabas earthquake) induces plasticity in large parts of the dam body. In fact, strong earthquakes are more effective in changing the material behaviour from elastic to plastic condition in comparison with weak earthquakes. 4. Conclusions  The higher dams are more flexible than the smaller ones. Consequently, the flexibility affects the shear strains which influence the shear modulus degradation and attenuating coefficient. All these effects are on the trend of weakening the accelerations along the height. g g  Soils with less strength (suppose low friction angle) go towards yielding by small amount of dynamic force which cause the attenuation of acceleration along the dam height in the weak materials compared with the strong ones.  Regarding a dam subjected to the earthquake with lower energy, the dam body behaves as an elastic material. Therefore, the induced seismic accelerations inside the dam body become larger from the base of dam to its top. In this case, small plasticity zones are developed in the dam body and the dam remains safe during dynamic loading.  Finally, non-linear dynamic analysis shows that plasticity should be considered in the investigation of seismic response of earth-fill dams, because of which the acceleration of the dam crest decreases and the displacements and shear strains of dam body as well as the energy dissipation increase. All these can significantly affect the seismic response of earth-fill dams. 5. References Abouseeda, H. & Dakoulas, P. (1998). Non-linear Dynamic Earth Dam–foundation Interaction using a BE–FE Method. Journal of Earthquake Engineering and Structural Dynamics, Vol.27, pp. 917–936 Adalier, K. & Sharp, M.K. (2004). Embankment Dam on Liquefiable Foundation-Dynamic Behavior and Densification Remediation. Journal of Geotechnical Geoenvironmental Engineering, Vol.130, No.11, pp. 1214–1224 www.intechopen.com 163 Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams Amadio, C.; Fragiacomo, M. & Rajgelj, S. (2003). The Effects of Repeated Earthquake Ground Motions on the Non-linear Response of SDOF Systems. Earthquake Engineering & Structural Dynamics, Vol.32, No.2, pp. 291–308 Ambraseys, N.N. (1960). The Seismic Stability of Earth Dams, Proceedings of the Second World Conference on Earthquake Engineering, pp. 1345–1363, Tokyo, Japan Arulanandan, K. & Scott, R.F. (1993). Verification of Numerical Procedures for the Analysis of Soil Liquefaction Problems, Conference Proceedings of VELACS, Vol. 1, Balkema, Rotterdam Arulanandan, K. & Scott, R.F. (1994). Verification of Numerical Procedures for the Analysis of Soil Liquefaction Problems, Conference Proceedings of VELACS, Vol. 2, Balkema, Rotterdam Ashford, S.A.; Boulanger, R.W.; Donahue, J.L. & Stewart J.P. (2011). Geotechnical Quick Report on the Kanto Plain Region during the March 11, 2011, Off Pacific Coast of Tohoku Earthquake, Japan, Geotechnical Extreme Events Reconnaissance (GEER) Basudhar, P.K.; Kameswara Rao, N.S.V.; Bhookya, M. & Dey, A. (2010). 2D FEM Analysis of Earth and Rockfill Dams under Seismic Condition, Fifth International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics and Symposium in Honor of Professor I.M. Idriss, San Diego, California Cascone, E. & Rampello, S. (2003). Decoupled Seismic Analysis of an Earth Dam. Soil Dynamics and Earthquake Engineering, Vol.23, pp. 349–365 Chen, M. & Harichandran, R.S. (2001). Response of an Earth Dam to Spatially Varying Earthquake Ground Motion. Journal of Engineering Mechanics, Vol.127, pp. 932–939 Chen, M. & Harichandran, R.S. (2001). Response of an Earth Dam to Spatially Varying Earthquake Ground Motion. Journal of Engineering Mechanics, Vol.127, pp. 932–939 Eberhard, M.; Baldridge, S.; Marshall, J.; Mooney, W. & Rix, G. (2010). The Mw 7.0 Haiti Earthquake of January 12, 2010: USGS/EERI Advance Reconnaissance Team, Team Report Vol 1 Chen, M. & Harichandran, R.S. (2001). Response of an Earth Dam to Spatially Varying Earthquake Ground Motion. Journal of Engineering Mechanics, Vol.127, pp. 932–939 Eberhard, M.; Baldridge, S.; Marshall, J.; Mooney, W. & Rix, G. (2010). The Mw 7.0 Haiti q J f g g pp Eberhard, M.; Baldridge, S.; Marshall, J.; Mooney, W. & Rix, G. (2010). 5. References The Mw 7.0 Haiti Earthquake of January 12, 2010: USGS/EERI Advance Reconnaissance Team, Team Report, Vol.1 Ebrahimian, B. & Vafaeian, M. (2005). Effects of Dam Height on the Seismic Response of Earth Dam, Proceedings of 7th International Conference on Civil Engineering, Tarbiat Modarres University, Tehran, Iran Ebrahimian, B. (2009). Numerical Modeling of the Seismic Response of an Earth Dam Founded on Liquefiable Soils, Proceedings of 2nd International Conference on Long Term Behavior of Dams, E. Bauer, S. Semprich, G. Zenz, (Eds.), pp 610-615, Graz University of Technology, Graz, Austria Ebrahimian, B. (2011). Numerical Analysis of Nonlinear Dynamic Behavior of Earth Dams. Frontiers of Architecture and Civil Engineering in China, Vol.5, No.1, pp. 24–40 Elgamal, A.M.; Abdel-Ghaffar, A.M. & Prevost, J.H. (1987). 2-D Elastoplastic Seismic Shear Response of Earth Dams: Application. Journal of the Engineering Mechanics ASCE, Vol.113, No.5, pp. 702-719 Finn, W.D.L.; Lee, K.W. & Martin, G.R. (1977). An Effective Stress Model for Liquefaction. Journal of Geotechnical Engineering Division ASCE, Vol.103, No.6, pp. 517-553 Gazetas, G. (1987). Seismic Response of Earth Dams: Some Recent Developments. Soil Dynamics and Earthquake Engineering, Vol.6, No.1, pp. 1–48 Griffith, D.V. & Prevost, J.H. (1988). Two- and Three-Dimensional Dynamic Finite Element Analyses of the Long Valley Dam. Geotechnique, Vol.38, pp. 367-388 www.intechopen.com www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 164 Harder, L.F.; Kelson, K.I.; Kishida, T. & Kayen, R. (2011). Preliminary Observations of the Fujinuma Dam Failure Following the March 11, 2011 Tohoku Offshore Earthquake, Japan, Geotechnical Extreme Events Reconnaissance (GEER) Hardin, B.O. & Black, W.L. (1968). Vibration Modulus of Normally Consolidated Clay. Journal of Soil Mechanics & Foundation Engineering ASCE, Vol.84, No.2, pp. 1531-1537 Itasca Consulting Group, Inc. FLAC (Fast Lagrangian Analysis of Continua), Version 4. 2004, Minneapolis, MN Kokusho, T. & Esashi, Y. (1981). Cyclic Tri-Axial Test on Sands and Coarse Material, Proceedings of 10th International Conference on Soil Mechanics and Foundation Engineering, (Quoted by Ishihara 1986) Krinitzsky, E.L. & Hynes, M.E. (2002). The Bhuj, India, Earthquake: Lessons Learned for Earthquake Safety of Dams on Alluvium. Engineering Geology, Vol.66, pp. 163-196 Kuhlmeyer, R.L. & Lysmer, J. (1973). Finite Element Method Accuracy for Wave Propagation Problems. Journal of the Soil mechanics and Foundation Division ASCE, Vol.99, No. SM5, pp. 421-427 pp Lacy, S.J. & Prevost, J.H. (1987). Nonlinear Seismic Response Analysis of Earth Dams. Soil Dynamics and Earthquake Engineering, Vol.6, No.1, pp. 48–63 Lai, S.S. into, S.P. (1993). Soil Dynamics and Geotechnical Earthquake Engineering. Balkema, Rotterdm 5. References & Seed, H.B. (1985). Dynamic Response of Long Valley Dam in the Mammoth Lake Earthquake Series of May 25-27 1980, Report No. UCB/EERC-85/12, Earthquake Engineering Research Center Lysmer, J. & Kuhlmeyer, R.L. (1969). Finite Element Method for Infinite Media. Journal of Engineering Mechanics ASCE, Vol.95, No. EM4, pp. 859-877 Masing, G. (1926). Eigenspannungen und Verfestigung Beim Messing, Proceedings of 2nd International Congress on Applied Mechanics, Zurich. Matsumoto, N.; Sasaki, T. & Ohmachi, T. (2011). The 2011 Tohoku Earthquake and Dams, ICOLD 89th Annual Meeting, Lucerne, Switzerland Ming, H.Y. & Li, X.S. (2003). Fully Coupled Analysis of Failure and Remediation of Lower San Fernando Dam. Journal of Geotechnical and Geoenvironmental Engineering, Vol.129, No.4, pp. 336–349 pp Moustafa, A. & Takewaki, I. (2010). Modeling Critical Ground-Motion Sequences for Inelastic Structures. Advances in Structural Engineering, Vol. 3, No.4, pp. 665-679 Moustafa, A. (2011). Damage-based Design Earthquake Loads for Single-degree-of-freedom Inelastic Structures. Journal of Structural Engineering, Vol.137, No.3, pp. 456-467 Ohmachi, T. & Kuwano, J. (1994). Dynamic Safety of Earth and Rock Fill Dams, Proceedings of a course, T. Ohmachi, J. Kuwano, (Eds.), Balkema, Rotterdam Özkan, M.; Yözyazicioglu, M. & Aksar, U.D. (2006). An Evaluation of Güldürcek Dam Response during 6 June 2000 Orta Earthquake. Soil Dynamics and Earthquake Engineering, Vol.26, No.5, pp. 405–419 Özkan, M. (1998). A Review of Considerations on Seismic Safety of Embankments and Earth and Rock-Fill Dams. Soil Dynamics and Earthquake Engineering, Vol.17, pp. 439–458 Papalou, A. & Bielak, J. (2004). Nonlinear Seismic Response of Earth Dams with Canyon Interaction. Journal of Geotechnical and Geoenvironmental Engineering, Vol.130, No.1, pp. 103–110 Pinto, S.P. (1993). Soil Dynamics and Geotechnical Earthquake Engineering. Balkema, Rotterdm Pinto, S.P. (1993). Soil Dynamics and Geotechnical Earthquake Engineering. Balkema, Rotterdm www.intechopen.com Non-Linear Numerical Analysis of Earthquake-Induced Deformation of Earth-Fill Dams 165 Prevost, J.; Abdel-Ghaffar, A.M. & Lacy, S. (1985). Nonlinear Dynamic Analyses of an Earth Dam. Journal of Geotechnical Engineering, Vol.111, No.7, pp. 882–897 Rampello, S.; Cascone, E. & Grosso, N. (2009). Evaluation of the Seismic Response of a Homogeneous Earth Dam. Soil Dynamics and Earthquake Engineering, Vol.29, pp. 782–798 Rathje, E.M. (2010). Case History: The Geotechnical Aspects of the Haiti Earthquake, ISSMGE Bulletin, Vol.4, Issue 3 Seed, H.B.; Lee, K.L.; Idriss, I.M. & Makdisi, F.I. (1975). The Slides in the San Fernando Dams during the Earthquake of February 9, 1971. Journal of the Soil mechanics and Foundation Division ASCE, Vol.101, No. GT7, pp. 5. References 651–688 Seed, H.B.; Wong, R.T.; Idriss, I.M. & Tokimatsu, K. (1986). Moduli and Damping Factors for Dynamic Analyses of Cohesionless Soils. Journal of Geotechnical Engineering ASCE, Vol.112, No.11, pp. 1016-1032 Sherard, J.I.; Woodward, R.J.; Gizienski, S.F. & Clevencer, W.A. (1963). Earth and Earth-Rock Dams. John Wiely and Sons, Inc Sherard, J.L. (1967). Earthquake Considerations in Earth Dam Design. Journal of the Soil mechanics and Foundation Division ASCE, Vol.93, pp. 377–401 Sica, S.; Pagano, L. & Modaressi, A. (2008). Influence of Past Loading History on the Seismic Response of Earth Dams. Computers and Geotechnics, Vol.35, pp. 61–85 p p pp Siyahi, B. & Arslan, H. (2008). Nonlinear Dynamic Finite Element Simulation of Alibey Earth Dam. Environmental Geology, Vol.54, pp. 77–85 Siyahi, B. & Arslan, H. (2008). Nonlinear Dynamic Finite Elem Dam. Environmental Geology, Vol.54, pp. 77–85 Siyahi, B. & Arslan, H. (2008). Earthquake Induced Deformation of Earth Dams. Bulletin of Engineering Geology and the Environment, Vol.67, pp. 397–403 Takewaki, I. (2011). Preliminary Report of the 2011 off the Pacific Coast of Tohoku Earthquake. Journal of Zhejiang University-SCIENCE A (Applied Physics & Engineering), Vol.12, No.5, pp. 327-334 USCOLD (US Committee on Large Dams). Observed performance of dams during earthquakes. Committee on Earthquakes, 1992, July, Denver, CO q y USCOLD (US Committee on Large Dams). Updated guidelines for selecting seismic parameters for dam projects. Committee on Large Dams, 1999, Denver, CO Vucetic, M. & Dobry, R. (1991). Effects of the Soil Plasticity on Cyclic Response. Journal of Geotechnical Engineering ASCE, Vol.117, pp. 89–107 Wang, Z.L.; Makdisi, F.I. & Egan, J. (2006). Practical Applications of a Nonlinear Approach to Analysis of Earthquake-Induced Liquefaction and Deformation of Earth Structures. Soil Dynamics and Earthquake Engineering, Vol.26, pp. 231–252 Wieland, M. (2008). Large Dams the First Structures Designed Systematically against Earthquakes, The 14th World Conference on Earthquake Engineering, October 12-17, 2008, Beijing, China Woodward, P.K. & Griffiths, D.V. (1996). Non-linear Dynamic Analysis of the Long Valley Dam. Computer Methods and Advances in Geomechanics, Vol.11, No.6, pp. 635-644 Yiagos, A.N. & Prevost, J.H. (1991). Tow-phase Elasto-plastic Seismic Response of Earth Dams: Applications. Soil Dynamics and Earthquake Engineering, Vol.10, No.7, pp. 371- 381 www.intechopen.com www.intechopen.com Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 166 Zeghal, M. & Abdel-Ghaffar, A.M. (1992). Analysis of Behavior of Earth Dam Using Stron- Motion Earthquake Records. Journal of Geotechnical Engineering ASCE, Vol.118, No.2, pp. 266-277 Advances in Geotechnical Earthquake Engineering – Soil Liquefaction and Seismic Safety of Dams and Monuments 166 Zeghal, M. & Abdel-Ghaffar, A.M. (1992). Analysis of Behavior of Earth Dam Using Stron- Motion Earthquake Records. Journal of Geotechnical Engineering ASCE, Vol.118, No.2, pp. 266-277 www.intechopen.com www.intechopen.com www.intechopen.com Advances in Geotechnical Earthquake Engineering - Soil Liquefaction and Seismic Safety of Dams and Monuments Edited by Prof. Abbas Moustafa Advances in Geotechnical Earthquake Engineering - Soil Liquefaction and Seismic Safety of Dams and Monuments Edited by Prof. Abbas Moustafa ISBN 978-953-51-0025-6 Hard cover, 424 pages Publisher InTech Published online 10, February, 2012 Published in print edition February, 2012 This book sheds lights on recent advances in Geotechnical Earthquake Engineering with special emphasis on soil liquefaction, soil-structure interaction, seismic safety of dams and underground monuments, mitigation strategies against landslide and fire whirlwind resulting from earthquakes and vibration of a layered rotating plant and Bryan's effect. The book contains sixteen chapters covering several interesting research topics written by researchers and experts from several countries. The research reported in this book is useful to graduate students and researchers working in the fields of structural and earthquake engineering. The book will also be of considerable help to civil engineers working on construction and repair of engineering structures, such as buildings, roads, dams and monuments. InTech China Unit 405, Office Block, Hotel Equatorial Shanghai No.65, Yan An Road (West), Shanghai, 200040, China Phone: +86-21-62489820 Fax: +86-21-62489821 How to reference In order to correctly reference this scholarly work, feel free to copy and paste the following: Babak Ebrahimian (2012). Non-Linear Numerical Analysis of Earthquake- Induced Deformation of Earth-Fill Dams, Advances in Geotechnical Earthquake Engineering - Soil Liquefaction and Seismic Safety of Dams and Monuments, Prof. Abbas Moustafa (Ed.), ISBN: 978-953-51-0025-6, InTech, Available from: http://www.intechopen.com/books/advances-in-geotechnical-earthquake-engineering-soil-liquefaction-and- seismic-safety-of-dams-and-monuments/non-linear-numerical-analysis-of-earthquake-induced-deformation- and-liquefaction-of-earth-dams InTech Europe InTech Europe University Campus STeP Ri Slavka Krautzeka 83/A 51000 Rijeka, Croatia Phone: +385 (51) 770 447 Fax: +385 (51) 686 166 www.intechopen.com © 2012 The Author(s). Licensee IntechOpen. This is an open access artic stributed under the terms of the Creative Commons Attribution 3.0 cense, which permits unrestricted use, distribution, and reproduction in ny medium, provided the original work is properly cited. © 2012 The Author(s). Licensee IntechOpen. This is an open access article distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
https://openalex.org/W2588428660	https://www.research.ed.ac.uk/files/31215923/Fournie_et_al._2017._Early_animal_farming_and_zoonotic_disease_dynamics.pdf	English	null	Early animal farming and zoonotic disease dynamics: modelling brucellosis transmission in Neolithic goat populations	Royal Society open science	2,017	cc-by	10,124	Edinburgh Research Explorer Edinburgh Research Explorer General rights C i h f h General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Digital Object Identifier (DOI): 10.1098/rsos.160943 Document Version: Publisher's PDF, also known as Version of record Published In: Royal Society Open Science Early animal farming and zoonotic disease dynamics Modelling brucellosis transmission in Neolithic goat populations Citation for published version: Fournié, G, Pfeiffer, DU & Bendrey, R 2017, 'Early animal farming and zoonotic disease dynamics: Modelling brucellosis transmission in Neolithic goat populations', Royal Society Open Science, vol. 4, no. 2, 160943. https://doi.org/10.1098/rsos.160943 Citation for published version: Fournié, G, Pfeiffer, DU & Bendrey, R 2017, 'Early animal farming and zoonotic disease dynamics: Modelling brucellosis transmission in Neolithic goat populations', Royal Society Open Science, vol. 4, no. 2, 160943. https://doi.org/10.1098/rsos.160943 on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from rsos.royalsocietypublishing.org Research Cite this article: Fournié G, Pfeiffer DU, Bendrey R. 2017 Early animal farming and zoonotic disease dynamics: modelling brucellosis transmission in Neolithic goat populations. R.Soc.opensci. 4: 160943. http://dx.doi.org/10.1098/rsos.160943 Keywords: archaeology, epidemiology, mathematical modelling, animal domestication, one health, emerging disease Robin Bendrey3,4 1Veterinary Epidemiology, Economics and Public Health group, Department of Production and Population Health, Royal Veterinary College, University of London, Hawkshead Lane, North Mymms, Hatfield AL9 7TA, UK 2School of Veterinary Medicine, City University of Hong Kong, Kowloon, Hong Kong 3Department of Archaeology, University of Reading, Whiteknights Box 226, Reading RG6 6AB, UK 4School of History, Classics and Archaeology, University of Edinburgh, William Robertson Wing, Old Medical School, Teviot Place, Edinburgh EH8 9AG, UK 4School of History, Classics and Archaeology, University of Edinburgh, William Robertson Wing, Old Medical School, Teviot Place, Edinburgh EH8 9AG, UK GF, 0000-0002-6998-1201 GF, 0000-0002-6998-1201 Subject Category: Biology (whole organism) Zoonotic pathogens are frequently hypothesized as emerging with the origins of farming, but evidence of this is elusive in the archaeological records. To explore the potential impact of animal domestication on zoonotic disease dynamics and human infection risk, we developed a model simulating the transmission of Brucella melitensis within early domestic goat populations. The model was informed by archaeological data describing goat populations in Neolithic settlements in the Fertile Crescent, and used to assess the potential of these populations to sustain the circulation of Brucella. Results show that the pathogen could have been sustained even at low levels of transmission within these domestic goat populations. This resulted from the creation of dense populations and major changes in demographic characteristics. The selective harvesting of young male goats, likely aimed at improving the efﬁciency of food production, modiﬁed the age and sex structure of these populations, increasing the transmission potential of the pathogen within these populations. Probable interactions between Neolithic settlements would have further promoted pathogen maintenance. By fostering Subject Areas: health and disease and epidemiology/computational biology Research Cite this article: Fournié G, Pfeiffer DU, Bendrey R. 2017 Early animal farming and zoonotic disease dynamics: modelling brucellosis transmission in Neolithic goat populations. R.Soc.opensci. 4: 160943. http://dx.doi.org/10.1098/rsos.160943 Guillaume Fournié1, Dirk U. Pfeiffer1,2 and Robin Bendrey3,4 Take down policy Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 24. Oct. 2024 1. Background The shift from hunting and gathering wild food resources to the control and husbandry of domestic animals had fundamental and far-reaching repercussions for the evolution of infectious diseases in humans [1,2]. Through bringing animals together in larger, denser herds, in close association with human communities, a stable conduit for exposure of humans to infection in their animals was established [3,4]. Examination of the changing dynamics of human–animal relationships at the start of farming can not only advance understanding of the consequences of farming on human and animal health and wellbeing, but also contribute long-term perspectives to present and future concerns as animal management evolves to ensure sufﬁcient and reliable food supply for the ever-growing global human population which in turn has resulted in changing environments. However, while the origins of zoonoses as a consequence of the adoption of farming have been frequently hypothesized, there is little evidence in support of this supposition from archaeological records. Here, we discuss the origins of brucellosis as a zoonotic disease, a process that has been hypothesized as intensifying during the early period of animal domestication in the Near East [5,6]. Brucella melitensis is the main agent responsible for human brucellosis, today’s commonest bacterial zoonosis in the world [7]. Humans become infected through ingestion of unpasteurized dairy products and the management of infected animals, primarily sheep and goats, the main reservoir of the bacteria [6]. A recent review of early evidence for brucellosis in human (Homo sapiens) skeletons identiﬁes that the earliest probable cases reported come from the Bronze Age Near East [5], the region of domestication of goats and sheep, and also cattle and pigs, in multiple centres during the preceding Neolithic [8,9]. A further possible case derives from the early Neolithic Near East in association with evidence for early goat husbandry (an adult male skeleton (GD#22) from the site of Ganj Dareh exhibiting new woven bone on the anterior and lateral surfaces of a thoracic vertebral body and resorption of the superior anterior surface of a lumbar vertebral body removing a portion of the annular ring; both signs indicative of early brucellosis infection) [10]. A further possible case of brucellosis is also reported from a 2.4 to 2.8 Myr old hominin (Australopithecus africanus) skeleton [11]. Author for correspondence: Guillaume Fournié e-mail: gfournie@rvc.ac.uk Author for correspondence: Guillaume Fournié e-mail: gfournie@rvc.ac.uk Electronic supplementary material is available online at https://dx.doi.org/10.6084/m9. figshare.c.3677173. Electronic supplementary material is available online at https://dx.doi.org/10.6084/m9. figshare.c.3677173. 2017 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. 2017 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. 2017 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 2 conditions suitable for allowing domestic goats to become reservoirs of Brucella melitensis, the early stages of agricultural development were likely to promote the exposure of humans to this pathogen. conditions suitable for allowing domestic goats to become reservoirs of Brucella melitensis, the early stages of agricultural development were likely to promote the exposure of humans to this pathogen. 1. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 0 200 above 500 m above 1500 m above 2000 m above 3000 m km Figure 1. Location map showing Ganj Dareh, Ali Kosh and Jarmo in the eastern Fertile Crescent. Inset: region of early plant and animal domestications in the Near East known as the Fertile Crescent. ng.org R.Soc. opensci. 4: 160943 Figure 1. Location map showing Ganj Dareh, Ali Kosh and Jarmo in the eastern Fertile Crescent. Inset: region of early plant and animal domestications in the Near East known as the Fertile Crescent. (ﬁgure 1) [14,15], where possible indicators of brucellosis were identiﬁed in a human skeleton [10]. The spread of goat husbandry can then be followed to nearby lowland zones, reaching Ali Kosh by ca 7500 BC, and Jarmo a few centuries later still where it co-occurs with domestic sheep [16] (ﬁgure 1). Examining these three populations—from Ganj Dareh, Ali Kosh and Jarmo—allows an assessment of the diversity in management strategies during earlier phases of animal husbandry [15] and the impact of these strategies on the potential maintenance of Brucella melitensis within domestic goat populations. With unpasteurized milk being a particularly key mode of transmission to humans [6], the origins of dairying are highlighted as an important innovation. Although the precise antiquity of dairying is still debated, zooarchaeological studies of herd proﬁles provide indirect evidence to suggest that milking may have begun in the Near East during the eight millennium BC [12,17]; whereas the earliest direct evidence comes from organic residues preserved in pottery from 7th millennium BC Anatolia [18]. 2. Material and methods The modelled goat population demographic proﬁles were deﬁned using post-cranial remains found at the sites of Ganj Dareh, Ali Kosh and Jarmo as identiﬁed, recorded and reported by Zeder [16]. Although calculating demographic proﬁles from bone fusion provides less detail than from teeth eruption and attrition, importantly, it does allow calculation of sex-speciﬁc age proﬁles (not possible from dental data) due to sexual dimorphism in the goat post-cranial skeleton [14,16]. In the following, goats less than 1 year old are referred to as young, goats between 1 and 2 years old as yearlings, and goats of 2 years old or more as adults. 1. Background While it indicates that humans may have been infected through contact with wildlife prior to animal domestication, the development of animal farming is likely to have further enhanced the risk of human infection by increasing (i) the prevalence of infection among in-contact animal populations and (ii) the frequency of contact between humans and infected animals through, for instance, the emergence of milk exploitation. To explore the potential impact of the development of animal farming on brucellosis dynamics in domestic goats, and, therefore, on the risk of human infection, we consider the dynamics of Brucella melitensis infection in early domestic goat herds through a stochastic and age-structured mathematical model simulating its spread within village goat populations. The aim is to gain understanding of when in the evolution of goat husbandry conditions were reached for these animal populations to have the potential to sustain bacterial circulation within a settlement and to become a permanent reservoir for human infection. Current evidence indicates the emergence of goat husbandry in potentially multiple centres across the Near East during the late ninth/early eighth millennia BC [8,9]. Across the region, there is a range of evidence for increasing levels of management of goats at this time. For example, morphologically wild goats were transported to Cyprus, appearing as early as 8400 BC at Shillourokambos [12]. The early management of goats is reported at Nevalı Çori (ca 8200–8000 BC) in the upper Euphrates basin in the northern Fertile Crescent on the basis of size changes and demographic proﬁles [13]. Further east, early managed goats are also identiﬁed at the site of Ganj Dareh in the Zagros mountains of the eastern Fertile Crescent at ca 7900 BC, where the demographic proﬁle indicates a population under human management that are morphologically unaltered from wild animals [14,15]. Archaeological sites from the Zagros region offer an ideal case-study for articulating a demographic model of early domestic goat populations to investigate the possible dynamics of brucellosis in the early stages of husbandry due to the well-dated and characterized sequence of site assemblages and the fact that for around a millennium goat was the only domestic food animal in the region [16]. The area has produced the earliest and most accurately dated demographic proﬁle suggestive of a managed population at Ganj Dareh in the highland Zagros on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 3 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . 2.2. Brucellamelitensis infection and transmission Homogeneous mixing within the goat population of a village was assumed. Infected kids were assumed to be non-infectious, to recover from infection and be fully susceptible when reaching 1 year of age [19]. The infectious material excreted from the vaginal tract of infected goats following abortion or full- term parturition is generally considered to be the main source of infection for susceptible hosts [19]. While Brucella can also be shed in the semen, transmission is uncommon during natural mating [19,20]. Therefore, only female goats were considered to be potentially infectious, with transmission occurring through contacts with infectious material excreted following abortion or full-term parturition. It has been reported that infected goats may either be infectious for one abortion or parturition, or remain persistently infectious, with intermittent shedding [19]. Models published in the literature assumed either that infected hosts were infectious for only a couple of months [21], or remained infectious until their death [22]. In order to reﬂect this variability and uncertainty related to the course of infection of Brucella melitensis in goats, two scenarios were modelled. Under the lifelong infectiousness scenario, the goat population was divided into three mutually exclusive health states: Susceptible, Latent and Infectious. Infected goats entered into the latent state, and only females became infectious from their ﬁrst abortion or full-term parturition since infection. Infectious females shed bacteria in fetal ﬂuid and vaginal discharges for a period of ε days each year, following each abortion or full-term parturition. Under the transitory infectiousness scenario, the population was divided into four mutually exclusive health states: Susceptible, Latent, Infectious and Non-Infectious. Infectious goats only shed bacteria for a period of ε days following their ﬁrst abortion or full-term parturition since infection. They then became non-infectious, and could not become susceptible or infectious again. The non-infectious health state included goats that recovered from Brucella infection and became immune, as well as goats that were permanently infected but did not shed the bacteria any longer. Brucella transmission was assumed to be frequency-dependent, as this mode of transmission seems to be the most suitable for describing transmission in extensive production systems. Results assuming density-dependence transmission are also provided in the electronic supplementary material. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from . . . . 2.1. Population dynamics Individual goats were the unit of analysis, and the model was run in discrete time with a daily time-step. New goats entered into the population through births. Births were seasonal, with the kidding season lasting θ days per year. While seasonal births meant that the size of the population varied over time, the average size of a goat population was stable from one year to the next. All goats born within the same season deﬁned a cohort. Goats could leave the population at any time through harvesting or death due to other causes. The probability δas of a goat dying between days d and d + 1 depended on its sex on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from y y . . . . . . . . . . . . . . . . . . s and the age a, in years, of its cohort. Let Nasd be the number of goats of sex s in a cohort of age a on day d. The number of goats dying on day d + 1 was simulated by a binomial process with Nasd as the number of trials, and δas as the probability of a success. When a cohort reached the maximum age Ω, it was removed from the population. The number of new born kids joining the population over a day of a given birth season was simulated by a Poisson process with ωΠ/θ as the average number of events. ω was the average litter size, i.e. the average number of kids per year and per female of kid bearing age, θ was the length, in days, of the kidding season and Π referred to the average number of females of kid bearing age (more than or equal to 2 years old) during a kidding season. s and the age a, in years, of its cohort. Let Nasd be the number of goats of sex s in a cohort of age a on day d. The number of goats dying on day d + 1 was simulated by a binomial process with Nasd as the number of trials, and δas as the probability of a success. When a cohort reached the maximum age Ω, it was removed from the population. The number of new born kids joining the population over a day of a given birth season was simulated by a Poisson process with ωΠ/θ as the average number of events. ω was the average litter size, i.e. the average number of kids per year and per female of kid bearing age, θ was the length, in days, of the kidding season and Π referred to the average number of females of kid bearing age (more than or equal to 2 years old) during a kidding season. 4 4 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Outcome A simulation started by setting a random goat, on a random day, as infected. The probability of disease invasion was the proportion of simulations resulting in a substantial outbreak, deﬁned as the infection of at least 50 goats. The disease was said to be endemic if there was at least one infected goat in the population after a period of 200 years. To calculate the probability of disease endemicity, only simulations that resulted in a substantial outbreak were taken into account. In the metapopulation model, the infection was seeded in a single village, and the disease was considered to be endemic if there was at least one infected goat among the n villages after 200 years. A thousand simulations were run for each parameter combination. The basic reproduction number R0 was estimated by calculating the dominant eigenvalue of the next-generation matrix M [28]. R0 informs about the potential for the bacteria to spread within a goat population: the disease may invade the population if R0 is higher than 1, while it will not if R0 is lower than 1. The entry mij was interpreted as the expected number of newly infected goats of age i (in days) produced by one goat which became infected at age j, in an initially fully susceptible population. The calculation of R0 was checked numerically. As shown in ﬁgure 3a, the probability of disease invasion in the ‘modern’ goat population increased sharply when R0 became higher than one. 2.3. Parameters The demographic proﬁle of goat populations was modelled using four daily probabilities of mortality: for young goats regardless of their sex, δa = 0, for male and female yearlings, δa=1,s=M and δa=1,s=F, and adults, δa≥2. The model was ﬁrst ﬁtted to the survival probabilities assessed at each of the three archaeological sites (further details in the electronic supplementary material). This resulted in values of δa=0 which were lower than expected [23] and which varied across sites. It may have been due to systematic errors—smaller and less dense remains of the youngest animals may have been preferentially destroyed by processes of taphonomic attrition—or variation in survival probabilities across sites, but was unlikely to reﬂect major differences in harvesting practices. Also, this ﬁrst parametrization meant that the litter size (i.e. average number of kids by females at an age for bearing kids) required for a population size to remain stable from a year to the next differed between populations. As we were interested in assessing the impact of goat population management practices on disease dynamics, we adjusted all probabilities of mortality so that δa=0 and the litter size remained constant across all sites (further details in the electronic supplementary material). A hypothetical, extensive, ‘modern’ goat population characterized by a lower ASR than for the three Neolithic populations, but a similar probability of mortality of young goats and average litter size was designed to assess the impact of increased male-biased harvesting on disease dynamics. For this ‘modern’ goat population, the sex ratio of 0.28 among goats more than 1 year old was comparable to the one reported in contemporary populations [24]. The range of explored values of β was selected so that, when applied to the ‘modern’ goat population, simulated seroprevalences (electronic supplementary material) covered the range of within-village seroprevalences reported by cross-sectional serological surveys conducted in the Middle East and Africa (5–35%) [21,25–27]. Other parameter values and estimation of goat population sizes at each site were assessed based on a review of the literature. Parameter values and details about their selection or calculations are provided in the electronic supplementary material. 2.2. Brucellamelitensis infection and transmission The probability pad of any susceptible goat in a cohort of age a becoming infected on a day d to d + 1 was expressed as follows: If a = 0, pad = 0 and and if a ≥1, pad = 1 −exp −β l IS ld l,s Nlsd , where IS ld was the number of infectious (female) goats in a cohort of age l that shed bacteria on day d, i.e. within the ε-day period following abortion or parturition. β was the per capita number of effective contacts per unit of time (i.e. a contact resulting in infection if it involved an infectious goat). The number of goats in a cohort of age a becoming infected between day d and d + 1 was then simulated using a binomial process with the number of susceptible goats in that cohort as the number of events and pad as the probability of a success. In the metapopulation model, the probability pavd of any goat in a cohort of age a, in village v, becoming infected between day d and d + 1 was expressed as: If a = 0, pavd = 0 and and if a ≥1, pavd = 1 −exp ⎡ ⎣−β ⎛ ⎝(1 −α) l IS lvd l,s Nlsvd + α (n −1) i,i̸=v l IS lid l,s Nlsid ⎞ ⎠ ⎤ ⎦. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from With α the proportion of contacts that a goat makes with goats from other villages, and n the number of villages. The ﬁrst component of the exponent in this equation was the within-village infection process, and the second component was the between-village infection process. . . . With α the proportion of contacts that a goat makes with goats from other villages, and n the number of villages. The ﬁrst component of the exponent in this equation was the within-village infection process, and the second component was the between-village infection process. 5 3. Results and discussion Demographic proﬁles in the three selected Neolithic sites were all characterized by a male-biased mortality of yearlings (ﬁgure 2). However, it was more pronounced in Ganj Dareh, resulting in a lower adult sex ratio (ASR = 0.24)—the ratio between the number of adult males and females—than at both Ali Kosh (ASR = 0.61) and Jarmo (ASR = 0.76). The higher mortality of males compared with females is likely to result from the selective harvesting of males. It is a common demographic structure for domestic livestock populations where only a few males are needed to ensure the reproductive continuity of the herd [29], and a feature of farming systems specialized in meat production (with most males slaughtered for consumption, while females are kept for reproduction), but may potentially also be associated with early dairy production strategies [30]. The low ASR in Ganj Dareh suggests the practice of such advanced levels of management aimed at increasing food production may have been applied at an early stage of the development of goat farming, although this pattern was by no means consistent across the Fertile Crescent [17]. A hypothetical, extensive, ‘modern’ goat population characterized by an even lower ASR (0.15) than Ganj Dareh was designed to further explore the impact of increased male-biased harvesting of yearlings on disease dynamics (ﬁgure 2). y g y g As shown in ﬁgure 3 for the lifelong infectiousness scenario (and electronic supplementary material, ﬁgure S2, for the transitory infectiousness scenario), Brucella melitensis could invade (ﬁgure 3a) and be maintained for low levels of transmission in population sizes that were within the estimated ranges for the investigated Neolithic sites (grey and black bars in ﬁgure 3c–e; electronic supplementary material). These results suggest that conditions were present in these early domestic goat populations for the establishment of endemicity of the pathogen, which could thus have acted as a potential permanent reservoir for human infection. The probability of mortality of male yearlings was the demographic parameter showing the highest level of variation across all four above-mentioned demographic proﬁles. 3. Results and discussion It was also a highly inﬂuential parameter on the disease dynamics under both infectiousness scenarios (electronic supplementary material): for a given population size, the vulnerability to pathogen invasion and the probability of sustaining its circulation were higher in populations with high male-biased mortality, as in the Ganj Dareh and ‘modern’ proﬁles, compared with populations with low sex-biased mortality, as at Jarmo and Ali Kosh (lifelong infectiousness scenario: ﬁgure 3b; transitory infectiousness scenario: electronic supplementary material, ﬁgure S2b). In other words, the pathogen could be transmitted in such populations at disease transmission levels that did not allow it to be transmitted in comparable populations of the same size but for which harvesting was not biased towards males (further details about the impact of variations in the ASR on Brucella invasion and maintenance are presented in the electronic supplementary material). Likewise, the pathogen would circulate at a higher prevalence level in populations with high male-biased mortality. For a given population size and disease transmission level, the prevalence of infection was the highest in the ‘modern’ proﬁle, and the lowest in Jarmo (electronic supplementary material, ﬁgure S5). However, the potential for the bacteria to become endemic in the Jarmo goat population may have been underestimated as sheep were also present [18] and likely to contribute to endemicity. Preferential harvesting of males increased the proportion of females in the population, and, therefore, their proportion among newly infected goats. As adult females are responsible for pathogen transmission, such a population structure would promote the transmission of Brucella melitensis (i.e. higher value of R0 for a given transmission rate). The potential for Brucella melitensis to become endemic was substantially increased if goats from different villages mixed together, even if the level of inter-village mixing was low (ﬁgure 4). A small metapopulation of 10 identical villages with demographic proﬁles similar to Ganj Dareh was simulated, with varying levels of contact between villages. In a metapopulation where goats made 0.5% of their contacts with goats from other villages (and, therefore, 99.5% of their contacts with goats from the same village), a probability of disease endemicity of, for instance, 0.2 was achieved for village populations 1.5 times smaller than villages in a metapopulation composed of isolated villages (ﬁgure 4). While disease extinction was more likely in these smaller populations, inter-village mixing could lead to the re-introduction of the pathogen in populations in which it faded out. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from All analyses were run using R v. 3.1.0. The package ‘sensitivity’ was used to conduct the sensitivity analysis. 6 2.5. Sensitivity analysis A global sensitivity analysis was conducted to assess the impact of variations in age- and sex- speciﬁc mortality probabilities on the minimum population size required for the probability of disease endemicity to be equal to or higher than an arbitrary value of 0.2. Parameter values were sampled using the Latin hypercube sampling scheme, and the partial rank correlation coefﬁcients (PRCCs) were calculated, providing a measure of the inﬂuence of each parameter on the outcome. For each parameter, the range over which their value was varied was deﬁned by [r −0.1r, r + 0.1r], where r was the value of the corresponding parameter for Ganj Dareh demographic proﬁle. PRCCs are provided in the electronic supplementary material. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 3. Results and discussion This rescue effect meant that the pathogen could persist at the metapopulation level even if it could not persist at the population level [31]. Interactions between goats belonging to different villages were likely to on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 7 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.0 Jarmo >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo Ali Kosh Ganj Dareh modern 0.8 0.6 0.4 survival probability survival probability 0.2 0 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 age (in years) age (in years) age (in years) age (in years) 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 1.0 0.8 0.6 0.4 0.2 0 1.0 0.8 0.6 0.4 0.2 0 1.0 0.8 0.6 0.4 0.2 0 (b) (a) (c) (d) Figure2. Villagegoatdemographicprofiles.Panels(a),(b)and(c)wereinferredfrompost-cranialremainsfoundinJarmo,AliKoshand Ganj Dareh, respectively. Panel (d) refers to a hypothetical population reproducing features of ‘modern’ goat populations. Probabilities of survival as a function of age for the overall population (solid line), males (dashed line) and females (dotted line) are shown. 1.0 Jarmo >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo 0.8 0.6 0.4 survival probability 0.2 0 0 1 2 3 4 5 6 7 8 9 age (in years) (a) 7 rsos.royalsocietypublishing.org R.Soc. opensci. 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Results and discussion >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo Ali Kosh 0 1 2 3 4 5 6 7 8 9 age (in years) 1.0 0.8 0.6 0.4 0.2 0 (b) 7 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 (a) survival probability ng.org R.Soc. opensci. 4: 160943 >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo modern age (in years) age (in years) 0 1 2 3 4 5 6 7 8 9 1.0 0.8 0.6 0.4 0.2 0 (d) >2 yo <1 yo 0.3 0.3 0.2 0.2 females females males males all 0.1 0.1 0 0 1–2 yo Ganj Dareh survival probability 0 1 2 3 4 age (in years) age (in years) 5 6 7 8 9 1.0 0.8 0.6 0.4 0.2 0 (c) (d) (c) Figure2. Villagegoatdemographicprofiles.Panels(a),(b)and(c)wereinferredfrompost-cranialremainsfoundinJarmo,AliKoshand Ganj Dareh, respectively. Panel (d) refers to a hypothetical population reproducing features of ‘modern’ goat populations. Probabilities of survival as a function of age for the overall population (solid line), males (dashed line) and females (dotted line) are shown. occur during this period as Neolithic communities are known to have been linked by a broad range of regional interactions, including exchange networks amongst diverse other social relationships [32]. Such inter-village contacts may have resulted from the introduction of goats from one village to another, during seasonal transhumance practices [33], or the use of common pastures or water points. In the vicinity of Ganj Dareh, for example, several contemporaneous sites are situated within 2 or 3 h walking distance [34]. Moreover, contacts between human settlements may have further promoted the spread of certain population management practices, and especially the selective harvesting of young males, among these settlements. The study has several limitations due to the nature of the data informing the model. Assumptions about the routes and modes of transmission of Brucella in Neolithic goats were based on our current understanding of the epidemiology of the disease. Infection of wild goat species by Brucella is common, and can reach high levels of prevalence [6,35]. 3. Results and discussion As present-day domestic goats derive from Neolithic populations, we assumed that the main features of Brucella pathology, and especially the restriction of Brucella transmission to females, remained the same. While the explored values of β were selected to allow the ampliﬁcation of the pathogen in the ‘modern’ goat population, it is unknown whether infectious contact rates in the Neolithic period were comparable to those observed nowadays. Further archaeological investigations and genetic analyses of ancient Brucella DNA would help in assessing the validity of these assumptions. Remains of goats that were not slaughtered and consumed, but died on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 8 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.4 5000 1000 100 10 3 2 R0 R* 0 1 0 0 1 2 3 0.3 0.2 0.1 prob. of disease invasion population size (p = 0.2) 0 3.0 2.5 2.0 1.5 1.2 10 100 population size 1000 5000 10 100 population size 1000 5000 10 0 0.1 0.2 0.3 0.4 probability of disease endemicity (p) 0.5 0.6 0.7 0.8 0.9 1.0 100 population size 1000 5000 10 100 population size 1000 5000 3.0 2.5 2.0 1.5 1.2 0 0.5 1.0 1.5 Jarmo Ganj Dareh modern Ali Kosh 2.0 2.5 3.0 1.2 1.5 2.0 R* 0 R* 0 R* 0 R* 0 3.0 2.5 2.0 1.5 1.2 3.0 2.5 2.0 1.5 1.2 R* 0 R* 0 2.5 3.0 (e) ( f ) (b) (a) (c) (d) Jarmo Ali Kosh Ganj Dareh modern igure 3. The probability of disease invasion and endemicity in a village goat population. Infectiousness was lifelong. R∗ 0 is the value of he basic reproduction number R0 for the hypothetical ‘modern’ demographic profile. R0 informs about the potential for the bacteria o spread within the goat population: the disease may invade the population if R0 is higher than 1, while it will not if R0 is lower han 1. As population-specific R0 were linearly dependent (a), R∗ 0 was chosen as a reference, and was reported on the x- or y-axes to llow comparisons between populations. (a) Probability of disease invasion as a function of R∗ 0 . 3. Results and discussion 3.0 2.5 2.0 1.5 1.2 10 100 population size 1000 5000 10 100 population size 1000 5000 3.0 2.5 2.0 1.5 1.2 Jarmo Ali Kosh 0 R* 0 R* 0 0 (c) (d) Jarmo Ali Kosh Ganj Dareh modern 0 10 100 population size 1000 5000 3.0 2.5 2.0 1.5 1.2 Ali Kosh R* 0 (d) Ganj Dareh modern (c) 10 0 0.1 0.2 0.3 0.4 probability of disease endemicity (p) 0.5 0.6 0.7 0.8 0.9 1.0 100 population size 1000 5000 10 100 population size 1000 5000 Ganj Dareh modern R* 0 3.0 2.5 2.0 1.5 1.2 3.0 2.5 2.0 1.5 1.2 R* 0 (e) ( f ) (e) ( f ) Figure 3. The probability of disease invasion and endemicity in a village goat population. Infectiousness was lifelong. R∗ 0 is the value of the basic reproduction number R0 for the hypothetical ‘modern’ demographic profile. R0 informs about the potential for the bacteria to spread within the goat population: the disease may invade the population if R0 is higher than 1, while it will not if R0 is lower than 1. As population-specific R0 were linearly dependent (a), R∗ 0 was chosen as a reference, and was reported on the x- or y-axes to allow comparisons between populations. (a) Probability of disease invasion as a function of R∗ 0 . (b) Minimum population size required for reaching a probability of disease endemicity p = 0.2 as a function of R∗ 0 . (c–f) Probability of disease endemicity as a function of populationsize and R∗ 0. Greyand black barsshow the rangesofestimated populationsizesateachsite assuming 100and 300people per hectare, respectively (electronic supplementary material). for other reasons (e.g. diseases), might have been discarded at a distance from the settlement, and smaller and less dense remains of the youngest animals may have been preferentially destroyed by processes of taphonomic attrition. This may have led to systematic errors in the estimation of survival probabilities. For comparison purposes, a hypothetical modern demographic proﬁle was developed. Certain features, such as average litter size [36,37] and age at slaughter may vary between breeds and husbandry practices, and these variations were not captured here. Our main aim was to represent the impact of the sex-biased harvesting of goats observed in modern extensive goat ﬂocks, and the resulting ASR, in order to allow a comparison with the different observed Neolithic demographic goat population proﬁles. 3. Results and discussion (b) Minimum population size required or reaching a probability of disease endemicity p = 0.2 as a function of R∗ 0 . (c–f) Probability of disease endemicity as a function of populationsize and R∗ 0. Greyand black barsshow the rangesofestimated populationsizesateachsite assuming 100and 300people per ectare, respectively (electronic supplementary material). 8 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.4 5000 1000 100 10 3 2 R0 R* 0 1 0 0 1 2 3 0.3 0.2 0.1 prob. of disease invasion population size (p = 0.2) 0 3.0 2.5 2.0 1.5 1.2 10 100 population size 1000 5000 10 100 population size 1000 5000 3.0 2.5 2.0 1.5 1.2 0 0.5 1.0 1.5 Jarmo Ali Kosh 2.0 2.5 3.0 1.2 1.5 2.0 R* 0 R* 0 R* 0 R* 0 2.5 3.0 (b) (a) (c) (d) Jarmo Ali Kosh Ganj Dareh modern 0.4 5000 1000 100 10 3 2 R0 R* 0 1 0 0 1 2 3 0.3 0.2 0.1 prob. of disease invasion population size (p = 0.2) 0 0 0.5 1.0 1.5 2.0 2.5 3.0 1.2 1.5 2.0 R* 0 R* 0 2.5 3.0 (b) (a) 8 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (a) 8 . . . . . . . . . . . . . . . . . . . . . . . . 3. Results and discussion Such a practice could have resulted in multiple introductions of the pathogen into managed ﬂocks, and could have, therefore, further promoted the likelihood of the pathogen being maintained in village goat populations. In conclusion, the increase in livestock densities may not be the only feature resulting from the early development of farming that promoted disease invasion and maintenance. The alteration of goat population demographic proﬁles, probably associated with management decisions to increase productivity of herds, and likely interactions between settlements further increased the potential for these populations to spread and maintain infection. Through these changes in goat population dynamics and contact patterns, conditions promoting the exposure of humans to a zoonotic pathogen emerged at an early stage of farming development. In the earliest period of caprine husbandry across the Near East a diversity of management strategies were practiced, as communities experimented with differing herd proﬁles, with only a minority of Early Neolithic sites demonstrating pronounced young male kill- off [15]. This situation changes from the mid-7th millennium BC, after which the majority of sites produce clear evidence for young male kill-off of domestic caprines at the same time as a new emphasis on intensive and large-scale mixed sheep and goat pastoralism emerges [15]. Thus, zoonotic brucellosis had the potential for emergence in some geographical areas of the Early Neolithic, but may not have become widespread until the relevant management strategies were in wider use. Transmission to human communities would have been further enabled by the development of dairying practices, although whether or not milking was practiced in the Zagros Neolithic sites is currently inconclusive from the herd demographics. Understanding of the interrelationship between disease dynamics and population characteristics within this broader regional narrative will allow future osteological and genetic research to focus on those areas most likely to produce direct evidence for the emergence of livestock-related zoonotic disease. To date, the study of the role of animal domestication in the emergence of brucellosis has been limited by what is identiﬁable from the archaeological records. Palaeopathological studies of human and animal remains can often indicate only non-speciﬁc infections, as identifying the causative agent based on structural changes within bone alone is problematic given the lack of speciﬁcity of these changes for the various pathogens and also only a proportion of individuals infected by an infectious organism might show evidence of skeletal changes [39,40]. 3. Results and discussion We estimated possible sizes of goat populations at Ganj Dareh, Ali Kosh and Jarmo. Such calculations are a challenge for archaeological research given the nature of the evidence [38], but was attempted here to add perspective to the discussion of the relationship between population on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 9 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.0 0.8 0.6 0.4 probability of disease endemicity 0.2 0 10 100 population size 1000 R0* = 1.5, a = 0 R0* = 1.5, a = 0.005 R0* = 1.5, a = 0.01 R0* = 1.5, a = 0.05 R0* = 2.5, a = 0 R0* = 2.5, a = 0.005 R0* = 2.5, a = 0.01 R0* = 2.5, a = 0.05 5000 Figure 4. Impact of inter-population mixing on the potential for the disease to become endemic in a metapopulation of goats. Infectiousness was assumed to be life-long, and Ganj Dareh demographic profile was selected. α refers to the proportion of contacts thatagoatmadewithgoatsfromothervillages.R∗ 0 referstothereferencevalueofthebasicreproductionnumberR0 forthehypothetical ‘modern’demographicprofile.Greyandblackbarsshowtherangesofestimatedgoatpopulationsizesateachsiteassuming100and300 people per hectare, respectively (electronic supplementary material). 1.0 0.8 0.6 0.4 probability of disease endemicity 0.2 0 10 100 population size 1000 R0* = 1.5, a = 0 R0* = 1.5, a = 0.005 R0* = 1.5, a = 0.01 R0* = 1.5, a = 0.05 R0* = 2.5, a = 0 R0* = 2.5, a = 0.005 R0* = 2.5, a = 0.01 R0* = 2.5, a = 0.05 5000 9 Figure 4. Impact of inter-population mixing on the potential for the disease to become endemic in a metapopulation of goats. Infectiousness was assumed to be life-long, and Ganj Dareh demographic profile was selected. α refers to the proportion of contacts thatagoatmadewithgoatsfromothervillages.R∗ 0 referstothereferencevalueofthebasicreproductionnumberR0 forthehypothetical ‘modern’demographicprofile.Greyandblackbarsshowtherangesofestimatedgoatpopulationsizesateachsiteassuming100and300 people per hectare, respectively (electronic supplementary material). size and disease endemicity. The wide range of parameter values we used aimed to capture the uncertainty associated with these estimates, with the population size estimates spanning several orders of magnitude (electronic supplementary material). We assumed that managed goat population dynamics was independent from wild goat populations. However, early farmers might have regularly recruited wild animals and introduced them into their ﬂocks. 10 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . shing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . These ﬁndings further support the view that the transition from food collection to production during the Neolithic transition while allowing for larger human population sizes resulted in signiﬁcant adverse effects on human health and wellbeing [2,47]. It further demonstrates the importance of recognizing the complexity of eco-social systems, where it is often very difﬁcult to obtain a holistic impression of the different types of impacts that a particular change in the system has. In this case, early farmers discovered that they could improve the efﬁciency of food production while maintaining herd reproductive continuity by selectively culling young male goats, a cause–effect relationship that must have been clear to them. But they were unlikely to have realized that this led to increased risk of human brucellosis, due to the cause–effect relationship not being directly observable. Arguably, even if they had recognized the link, the perceived beneﬁts of more effective food production and associated outputs may still have resulted in choosing the same goat herd production management approach. Data accessibility. This study uses data from the literature. Authors’ contributions. G.F. and R.B. designed the study; R.B. collated archaeological data; G.F. and D.U.P. designed the model. All authors discussed the results and contributed to the paper. C i i W d l h Data accessibility. This study uses data from the literature. Authors’ contributions. G.F. and R.B. designed the study; R.B. collated archaeological data; G.F. and D.U.P. des model. All authors discussed the results and contributed to the paper. Authors’ contributions. G.F. and R.B. designed the study; R.B. collated archaeological data; G.F. and D.U.P. designed the model. All authors discussed the results and contributed to the paper. Competing interests. We declare we have no competing interests. ompeting interests. We declare we have no competing interests. Funding. R.B. was supported by the Central Zagros Archaeological Project (AHRC: AH/H0343125 Funding. R.B. was supported by the Central Zagros Archaeological Project (AHRC: AH/H0343125/2: PI R Matthews). References 10 000 years ago. Science 287, 2254–2257. (doi:10.1126/science.287.5461.2254) 1. Diamond J. 2002 Evolution, consequences and future of plant and animal domestication. Nature 418, 700–707. (doi:10.1038/nature01019) 8. Arbuckle BS. 2014 Pace and process in the emergence of animal husbandry in Neolithic Southwest Asia. Bioarchaeol.NearEast 8, 53–81. 1. Diamond J. 2002 Evolution, consequences and future of plant and animal domestication. Nature 418, 700–707. (doi:10.1038/nature01019) 8. Arbuckle BS. 2014 Pace and process in the emergence of animal husbandry in Neolithic Southwest Asia. Bioarchaeol.NearEast 8, 53–81. 15. Arbuckle BS, Atici L. 2013 Initial diversity in sheep and goat management in Neolithic south-western Asia. Levant 45, 219–235. (doi:10.1179/007589 1413Z.00000000026) 9. Zeder MA. 2011 The origins of agriculture in the Near East. Curr.Anthropol. 52, S221–S235. (doi:10.1086/ 659307) 2. Bocquet-Appel JP. 2011 When the world’s population took off: the springboard of the Neolithic Demographic Transition. Science 333, 560–561. (doi:10.1126/science.1208880) 2. Bocquet-Appel JP. 2011 When the world’s population took off: the springboard of the Neolithic Demographic Transition. Science 333, 560–561. (doi:10.1126/science.1208880) 16. Zeder MA. 2008 Animal domestication in the Zagros: an update and directions for future research. In ArchaeozoologyoftheNearEastVIII TravauxdelaMaisondel’OrientetdelaMéditerranée 49 (eds E Vila, L Gourichon, AM Choyke, H Buitenhuis), pp. 243–277. Lyon, France: Maison de l’Orient et de la Méditerranée. 10. Merrett DC. 2004 BioarchaeologyinEarlyNeolithic Iran:assessmentofhealthstatusandsubsistence strategy. Winnipeg, Canada: University of Manitoba. 10. Merrett DC. 2004 BioarchaeologyinEarlyNeolithic Iran:assessmentofhealthstatusandsubsistence strategy. Winnipeg, Canada: University of Manitoba. 3. Wolfe ND, Dunavan CP, Diamond J. 2007 Origins of major human infectious diseases. Nature 447, 279–283. (doi:10.1038/nature05775) 3. Wolfe ND, Dunavan CP, Diamond J. 2007 Origins of major human infectious diseases. Nature 447, 279–283. (doi:10.1038/nature05775) 11. D’Anastasio R, Zipfel B, Moggi-Cecchi J, Stanyon R, Capasso L. 2009 Possible brucellosis in an early hominin skeleton from Sterkfontein, South Africa. PLoSONE 4, e6439. (doi:10.1371/journal.pone. 0006439) 4. Pearce-Duvet JM. 2006 The origin of human pathogens: evaluating the role of agriculture and domestic animals in the evolution of human disease. Biol.Rev.Camb.Philos.Soc. 81, 369–382. (doi:10.1017/S1464793106007020) 4. Pearce-Duvet JM. 2006 The origin of human pathogens: evaluating the role of agriculture and domestic animals in the evolution of human disease. Biol.Rev.Camb.Philos.Soc. 81, 369–382. (doi:10.1017/S1464793106007020) 17. Helmer D, Gourichon L, Vila E. 2007 The development of the exploitation of products from Capra and Ovis (meat, milk and fleece) from the PPNB to the Early Bronze in the northern Near East (8700 to 2000 BC cal.). Anthropozoologica 42, 41–69. 12. 3. Results and discussion Furthermore, the taphonomic histories of most animal bone assemblages (butchered, fragmented and cooked) act against the identiﬁcation of diseases [41]. Analyses of ancient DNA have increasing potential for the identiﬁcation of infectious agents, although they have been challenged to date by detrimental environmental conditions in the Near East to DNA survival in archaeological bones [42,43]. They are, however, beginning to contribute to the identiﬁcation on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from of brucellosis for later archaeological periods [44], and recent advances represent signiﬁcant potential for the recovery of pathogen DNA from the ancient Near East [45]. Future directions should focus on analyses to investigate the presence and distribution of brucellosis in the archaeological populations, particularly through reﬁnement and application of palaeopathological criteria for identifying brucellosis in goat skeletons, and application of ancient DNA analytical protocols. Such future analyses are needed to test the model results and validate conclusions. For example, we would expect to see higher prevalence rates at sites with population demographics promoting infection. Future work should also further reﬁne and validate the parameters employed in the model, for example, data on timings and changes in birth seasonality can be generated from stable isotope analysis of archaeological teeth [46] and extend the simulations to later periods where we witness the development of more intensive husbandry systems [15] or the increase in occurrence of reported cases of brucellosis in humans [5,6]. Stable isotope analyses on archaeological remains could also provide information on the spatial mobility of goats, and, therefore, on the likelihood of interactions between goat ﬂocks belonging to distinct settlements. of brucellosis for later archaeological periods [44], and recent advances represent signiﬁcant potential for the recovery of pathogen DNA from the ancient Near East [45]. Future directions should focus on analyses to investigate the presence and distribution of brucellosis in the archaeological populations, particularly through reﬁnement and application of palaeopathological criteria for identifying brucellosis in goat skeletons, and application of ancient DNA analytical protocols. Such future analyses are needed to test the model results and validate conclusions. For example, we would expect to see higher prevalence rates at sites with population demographics promoting infection. Future work should also further reﬁne and validate the parameters employed in the model, for example, data on timings and changes in birth seasonality can be generated from stable isotope analysis of archaeological teeth [46] and extend the simulations to later periods where we witness the development of more intensive husbandry systems [15] or the increase in occurrence of reported cases of brucellosis in humans [5,6]. Stable isotope analyses on archaeological remains could also provide information on the spatial mobility of goats, and, therefore, on the likelihood of interactions between goat ﬂocks belonging to distinct settlements. 10 11 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40. Aufderheide AC, Rodríguez-Martín C. 1998 The Cambridgeencyclopediaofhumanpalaeopathology. Cambridge, UK: Cambridge University Press. 31. Keeling MJ, Rohani P. 2008 Modellinginfectious diseasesinhumansandanimals. Princeton, NJ: Princeton University Press. 41. Bartosiewicz L. 2008 Taphonomy and palaeopathology in archaeozoology. Geobios 41, 69–77. (doi:10.1016/j.geobios.2006. 02.004) 22. Zinsstag J, Roth F, Orkhon D, Chimed-Ochir G, Nansalmaa M, Kolar J, Vounatsou P. 2005 A model of animal-human brucellosis transmission in Mongolia. Prev.Vet.Med. 69, 77–95. (doi:10.1016/ j.prevetmed.2005.01.017) 32. Asouti E. 2006 Beyond the pre-pottery Neolithic B interaction sphere. J.WorldPrehistory 20, 87–126. (doi:10.1007/s10963-007-9008-1) 42. Smith CI, Chamberlain AT, Riley MS, Stringer C, Collins MJ. 2003 The thermal history of human fossils and the likelihood of successful DNA amplification. J.Hum.Evol. 45, 203–217. (doi:10.1016/S0047-2484(03)00106-4) 23. Redding RW. 1981 Decisionmakinginsubsistence herdingofsheepandgoatsintheMiddleEast. Ann Arbor, MI: University of Michigan. 33. Sutliff DJ. 2015 On nomadic transhumance at Neolithic Tepe Tula’i, Iran: a re-analysis of findings. J.Archaeol.Sci.Rep. 3, 392–397. (doi:10.1016/j. jasrep.2015.06.014) 24. Central Agricultural Census Commission. 2003 Ethiopia Federal Democratic Republic, agricultural census 2001/02 main results. 34. Smith PEL, Mortensen P. 1980 Three new Early Neolithic sites in western Iran. Curr.Anthropol. 21, 511–512. (doi:10.1086/202503) 43. Bollongino R, Vigne JD. 2008 Temperature monitoring in archaeological animal bone samples in the Near East arid area, before, during and after excavation. J.Archaeol.Sci. 35, 873–881. (doi:10.1016/j.jas.2007.06.023) 25. Asmare K etal. 2013 A study on seroprevalence of caprine brucellosis under three livestock production systems in southern and central Ethiopia. Trop. Anim.HealthProd. 45, 555–560. (doi:10.1007/ s11250-012-0258-2) 35. Mick V, Le Carrou G, Corde Y, Game Y, Jay M, Garin-Bastuji B. 2014 Brucellamelitensis in France: persistence in wildlife and probable spillover from Alpine ibex to domestic animals. PLoSONE 9, e94168. (doi:10.1371/journal.pone.0094168) 44. Kay GL, Sergeant MJ, Giuffra V, Bandiera P, Milanese M, Bramanti B, Bianucci R, Pallen MJ. 2014 Recovery of a medieval Brucellamelitensis genome using shotgun metagenomics. mBio 5, e01337-14. (doi:10.1128/mBio. 01337-14) 26. Teklue T, Tolosa T, Tuli G, Beyene B, Hailu B. 2013 Sero-prevalence and risk factors study of brucellosis in small ruminants in Southern Zone of Tigray Region, Northern Ethiopia. Trop.Anim.Health Prod. 45, 1809–1815. (doi:10.1007/s11250-013- 0439-7) 36. Gall C. 1996 Goatbreedsoftheworld. Technical Centre for Agricultural and Rural Cooperation (CTA). Weikersheim, Germany: Margraf. 37. Tuncel E, Rehber E. 1995 Goat production systems in Turkey. In Goatproductionsystemsinthe Mediterranean (eds A El Aich, S Landau, A Bourbouze, R Rubino, P Morand-Fehr), pp. 111–135. Wageningen, The Netherlands: Wageningen Pers. 45. Warinner C, Speller C, Collins MJ. References VigneJD,CarrèreI,BrioisF,GuilaineJ.2011Theearly process of the mammal domestication in the Near East: new evidence from the Pre-Neolithic and Pre-Pottery Neolithic in Cyprus. Curr.Anthropol. 52, S255–S271. (doi:10.1086/659306) 12. VigneJD,CarrèreI,BrioisF,GuilaineJ.2011Theearly process of the mammal domestication in the Near East: new evidence from the Pre-Neolithic and Pre-Pottery Neolithic in Cyprus. Curr.Anthropol. 52, S255–S271. (doi:10.1086/659306) 5. D’Anastasio R, Staniscia T, Milia ML, Manzoli L, Capasso L. 2011 Origin, evolution and paleoepidemiology of brucellosis. Epidemiol. Infect. 139, 149–156. (doi:10.1017/S0950268810 00097X) 13. Peters J, Von den Driesch A, Helmer D. 2005 The upper Euphrates-Tigris basin: cradle of agro-pastoralism. In Thefirststepsofanimal domestication,newarchaeozoologicalapproachesto tracethefirststepsofanimaldomestication (eds JD Vigne, J Peters, D Helmer), pp. 1–16. Oxford, UK: Oxbow Books. 13. Peters J, Von den Driesch A, Helmer D. 2005 The upper Euphrates-Tigris basin: cradle of agro-pastoralism. In Thefirststepsofanimal domestication,newarchaeozoologicalapproachesto tracethefirststepsofanimaldomestication (eds JD Vigne, J Peters, D Helmer), pp. 1–16. Oxford, UK: Oxbow Books. 18. Evershed RP etal. 2008 Earliest date for milk use in the Near East and southeastern Europe linked to cattle herding. Nature 455, 528–531. (doi:10.1038/ nature07180) 6. Moreno E. 2014 Retrospective and prospective perspectives on zoonotic brucellosis. Front. Microbiol. 5, 213. (doi:10.3389/fmicb.2014. 00213) 19. European Commission Scientific Committee on Animal Health and Animal Welfare. 2001 Brucellosis in sheep and goats. 19. European Commission Scientific Committee on Animal Health and Animal Welfare. 2001 Brucellosis in sheep and goats. 7. Pappas G, Papadimitriou P, Akritidis N, Christou L, Tsianos EV. 2006 The new global map of human brucellosis. LancetInfect.Dis. 6, 91–99. (doi:10.1016/ S1473-3099(06)70382-6) 14. Zeder MA, Hesse B. 2000 The initial domestication of goats (Caprahircus) in the Zagros mountains 20. World Organization for Animal Health (OIE). 2016 TerrestrialAnimalHealthCode. Paris, France. on February 15, 2017 http://rsos.royalsocietypublishing.org/ Downloaded from 11 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . studies. J.Archaeol.Sci. 35, 1581–1590. (doi:10.1016/j.jas.2007.11.002) 40. Aufderheide AC, Rodríguez-Martín C. 1998 The Cambridgeencyclopediaofhumanpalaeopathology. Cambridge, UK: Cambridge University Press. 41. Bartosiewicz L. 2008 Taphonomy and palaeopathology in archaeozoology. Geobios 41, 69–77. (doi:10.1016/j.geobios.2006. 02.004) 42. Smith CI, Chamberlain AT, Riley MS, Stringer C, Collins MJ. 2003 The thermal history of human fossils and the likelihood of successful DNA amplification. J.Hum.Evol. 45, 203–217. (doi:10.1016/S0047-2484(03)00106-4) 43. Bollongino R, Vigne JD. 2008 Temperature monitoring in archaeological animal bone samples in the Near East arid area, before, during and after excavation. J.Archaeol.Sci. 35, 873–881. (doi:10.1016/j.jas.2007.06.023) 44. Kay GL, Sergeant MJ, Giuffra V, Bandiera P, Milanese M, Bramanti B, Bianucci R, Pallen MJ. 2014 Recovery of a medieval Brucellamelitensis genome using shotgun metagenomics. mBio 5, e01337-14. (doi:10.1128/mBio. 01337-14) 45. Warinner C, Speller C, Collins MJ. 2015 A new era in palaeomicrobiology: prospects for ancient dental calculus as a long-term record of the human oral microbiome. Phil.Trans.R.Soc.B 370, 20130376. (doi:10.1098/rstb.2013.0376) 46. Balasse M, Obein G, Ughetto-Monfrin J, Mainland I. 2012 Investigating seasonality and season of birth in past herds: a reference set of sheep enamel stable oxygen isotope ratios. Archaeometry 54, 349–368. (doi:10.1111/j.1475-4754.2011. 00624.x) 47. Larsen CS. 2006 The agricultural revolution as environmental catastrophe: implications for health and lifestyle in the Holocene. Quat.Int. 150, 12–20. (doi:10.1016/j.quaint.2006.01.004) 30. Vigne JD, Helmer D. 2007 Was milk a ‘secondary product’ in the Old World Neolithisation process? Its role in the domestication of cattle, sheep and goats. Anthropozoologica 42, 9–40. studies. J.Archaeol.Sci. 35, 1581–1590. (doi:10.1016/j.jas.2007.11.002) studies. J.Archaeol.Sci. 35, 1581–1590. (doi:10.1016/j.jas.2007.11.002) 21. Hegazy YM, Moawad A, Osman S, Ridler A, Guitian J. 2011 Ruminant brucellosis in the Kafr El Sheikh Governorate of the Nile Delta, Egypt: prevalence of a neglected zoonosis. PLoSNegl.Trop.Dis. 5, e944. (doi:10.1371/journal.pntd.0000944) 11 11 rsos.royalsocietypublishing.org R.Soc. opensci. 4: 160943 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2015 A new era in palaeomicrobiology: prospects for ancient dental calculus as a long-term record of the human oral microbiome. Phil.Trans.R.Soc.B 370, 20130376. (doi:10.1098/rstb.2013.0376) 45. Warinner C, Speller C, Collins MJ. 2015 A new era in palaeomicrobiology: prospects for ancient dental calculus as a long-term record of the human oral microbiome. Phil.Trans.R.Soc.B 370, 20130376. (doi:10.1098/rstb.2013.0376) 27. Gabli A, Agabou A, Gabli Z. 2015 Brucellosis in nomadic pastoralists and their goats in two provinces of the eastern Algerian high plateaus. Trop.Anim.HealthProd. 47, 1043–1048. (doi:10.1007/s11250-015- 0825-4) 27. Gabli A, Agabou A, Gabli Z. 2015 Brucellosis in nomadic pastoralists and their goats in two provinces of the eastern Algerian high plateaus. Trop.Anim.HealthProd. 47, 1043–1048. (doi:10.1007/s11250-015- 0825-4) 46. Balasse M, Obein G, Ughetto-Monfrin J, Mainland I. 2012 Investigating seasonality and season of birth in past herds: a reference set of sheep enamel stable oxygen isotope ratios. Archaeometry 54, 349–368. (doi:10.1111/j.1475-4754.2011. 00624.x) 38. Ebersbach R. 2013 Quantitative approaches to reconstructing prehistoric stock breeding. In Economicarchaeology:fromstructuretoperformance inEuropeanarchaeology (eds T Kerig, A Zimmermann), pp. 143–160. Bonn, Germany: Habelt. 38. Ebersbach R. 2013 Quantitative approaches to reconstructing prehistoric stock breeding. In Economicarchaeology:fromstructuretoperformance inEuropeanarchaeology (eds T Kerig, A Zimmermann), pp. 143–160. Bonn, Germany: Habelt. 28. Diekmann O, Heesterbeek JAP. 2000 Mathematical epidemiologyofinfectiousdiseases:modelbuilding, analysisandinterpretation. New York, NY: John Wiley & Sons. 28. Diekmann O, Heesterbeek JAP. 2000 Mathematical epidemiologyofinfectiousdiseases:modelbuilding, analysisandinterpretation. New York, NY: John Wiley & Sons. 47. Larsen CS. 2006 The agricultural revolution as environmental catastrophe: implications for health and lifestyle in the Holocene. Quat.Int. 150, 12–20. (doi:10.1016/j.quaint.2006.01.004) 29. Zohary D, Tchernov E, Horwitz L. 1998 The role of unconscious selection in the domestication of sheep and goats. J.Zool. 245, 129–135. (doi:10.1111/j.1469- 7998.1998.tb00082.x) 39. Bendrey R, Taylor GM, Bouwman AS, Cassidy JP. 2008 Suspected bacterial disease in two archaeological horse skeletons from southern England: palaeopathological and biomolecular 39. Bendrey R, Taylor GM, Bouwman AS, Cassidy JP. 2008 Suspected bacterial disease in two archaeological horse skeletons from southern England: palaeopathological and biomolecular
https://openalex.org/W2961730384	https://www.frontiersin.org/articles/10.3389/fpsyg.2019.01653/pdf	English	null	Psychometric Properties of a Measure of Borderline Personality Organization in a Spanish Court-Referred Partner-Violent Male Sample	Frontiers in psychology	2,019	cc-by	8,383	Edited by: Connie J. Beck, The University of Arizona, United States Reviewed by: Beata Pastwa-Wojciechowska, University of Gda ´nsk, Poland Kathryn Sharratt, University of Huddersﬁeld, United Kingdom Correspondence: Natalia Redondo Rodríguez natalia.redondor@uam.es Specialty section: This article was submitted to Forensic and Legal Psychology, a section of the journal Frontiers in Psychology Keywords: validation, psychometric properties, borderline personality organization, borderline personality, intimate partner violence Psychometric Properties of a Measure of Borderline Personality Organization in a Spanish Court-Referred Partner-Violent Male Sample Natalia Redondo Rodríguez1, José Luis Graña Gómez2, María Luisa Cuenca Montesino3 and Marina Julia Muñoz-Rivas1 Borderline personality organization (BPO) is a key personality component of some but not all partner-violent men. The study described in this paper examines the psychometric properties of the borderline personality organization scale (BPO Scale; Oldham et al., 1985) in a Spanish sample of 643 men undergoing court-mandated psychological treatment after conviction for episodes of intimate-partner violence. Three conﬁrmatory factor analyses were carried out ﬁrst, and the three-factor structure of the BPO scale was then tested. Results for concurrent validity show positive and signiﬁcant correlations between the subscales and the overall BPO scale, and with other instruments that measure borderline and antisocial personality disorders (ASPDs), and impulsivity. The BPO scale also presents evidence of known-groups validity, since BPO scores decrease with age, and of discriminant validity, as the scale discriminates between participants who do and do not exceed the cutoff point on a borderline personality scale. The BPO Scale is a suitable instrument for evaluating BPO in partner-violent men. INTRODUCTION Received: 24 January 2019 Accepted: 01 July 2019 Published: 12 July 2019 Intimate partner violence (IPV) has become a matter of acute social concern and it is now seen as one of the principal public health issues worldwide, due not only to the scale of the problem but also to the seriousness of its personal, family, social, and legal consequences (DeBoard-Lucas and Grych, 2011; Okuda et al., 2011; Stylianou, 2018). Chief among the adverse outcomes caused by IPV are its eﬀects on victims, including impacts on both physical and mental health, increased risk of depression, anxiety, substance abuse, and suicide (World Health Organization [WHO], 2002). As a phenomenon, IPV takes in many behaviors and is deﬁned by the United Nations (1994) as “any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suﬀering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public, or private life” (Article 1 of the United Nations Declaration Keywords: validation, psychometric properties, borderline personality organization, borderline personality, intimate partner violence ORIGINAL RESEARCH published: 12 July 2019 doi: 10.3389/fpsyg.2019.01653 Citation: Redondo Rodríguez N, Graña Gómez JL, Cuenca Montesino ML and Muñoz-Rivas MJ (2019) Psychometric Properties of a Measure of Borderline Personality Organization in a Spanish Court-Referred Partner-Violent Male Sample. Front. Psychol. 10:1653. doi: 10.3389/fpsyg.2019.01653 July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org 1 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. correlating the scores of the BPO scale with the borderline pathological personality C scale of the Millon Clinical Multiaxial Inventory-Version 2 (MCMI-II; Millon, 1987), ﬁnding positive correlations of 0.71 (Dutton and Starzomski, 1993). They also analyzed the psychological and physical relationship of the BPO with IPV, as measured using the psychological maltreatment of women inventory (PMWI; Tolman, 1989), and the conﬂict tactics scale (CTS; Straus, 1979). The PMWI assesses diﬀerent forms of emotional/verbal abuse (e.g., withholding emotional resources, verbal attacks, and behavior that degrades women) and dominance/isolation (e.g., sex roles, demands for subservience, and isolation from resources). Dutton and Starzomski (1993) found signiﬁcant correlations between the three BPO subscales (identity diﬀusion, reality testing, and primitive defenses) and the overall BPO scale, and emotional abuse (correlations from 0.50 to 0.55). These authors also used the CTS to assess the frequency and intensity of diﬀerent dyadic tactics used to resolve conﬂict, including physical aggression, ﬁnding signiﬁcant correlations between BPO and physical aggression (correlations from 0.21 to 0.33). The reliability of the BPO scale in partner-violent men is reﬂected in the Cronbach’s alpha values for identity diﬀusion, primitive defenses, and reality testing, which were 0.85, 0.87, and 0.80, respectively (Tweed and Dutton, 1998). on the elimination of violence against women, 1994). Countless studies have been undertaken in an eﬀort to identify the individual factors distinguishing aggressors, which might help both in predicting aggressive behaviors of this kind, and also provide targets of intervention for treatment programs (Graña et al., 2014, 2017; Jose et al., 2014; Carbajosa et al., 2017; Loinaz et al., 2018). Among these factors, personality disorders have gained prominence in recent years, as they increase both the risk of an individual’s committing violent acts (Logan and Blackburn, 2009) and of recidivism (Hiscoke et al., 2003). Among all the personality disorders mentioned in the scientiﬁc literature on IPV, it is borderline personality disorder (BPD) which is associated with partner-violent men (Holtzworth-Munroe et al., 2000; Jose et al., 2014; Romero-Martínez et al., 2016; Davoren et al., 2017; Mackay et al., 2018). Citation: According to the diagnostic criteria established in DSM-5 (American Psychiatric Association, 2014), BPD is theoretically deﬁned by the presence of the following symptoms, among others: (a) alteration of an individual’s personality structure, including disturbance of the sense of identity, intense distortions of self-image, and chronic feelings of emptiness; (b) alteration of aﬀective states, including intense and inappropriate emotional outbursts and anger management issues, as well as general aﬀective instability; (c) behavioral alterations, including suicide attempts and self-harm, as well as extreme impulsivity leading to potentially unsafe behaviors; and (d) alteration of interpersonal relations, including a pattern of unstable yet very intense relationships, constant striving to avoid abandonment, and transitory paranoid ideation. Numerous researchers have drawn on these early studies to apply the scale to assess BPO in partner-violent men (Dutton et al., 1996; Taft et al., 2004; Stoops et al., 2010), ﬁnding that BPO is associated with a greater risk of IPV. However, no study has so far used a conﬁrmatory factor analysis (CFA) to examine the internal structure of the instrument in relation to aggressors of this type. Hence, the main aim of the present study is to examine the characteristics of the BPO scale at the psychometric level among a sample of men convicted of gender violence oﬀenses in Spain. The Spanish comprehensive gender violence protection act (Ley Orgánica 1/2004, 2004) deﬁnes gender violence as any act or acts of violence perpetrated by a man against a woman within the scope of an intimate relationship, and as a manifestation of the discrimination, inequality and asymmetrical power relations brought to bear on women by cohabiting, and non-cohabiting intimate partners. Thus, the term “gender violence” will be used for the term IPV in this manuscript as that is the oﬃcial term used in Spanish law. Meanwhile, the clinical category of borderline personality organization (BPO) is one of the most widely researched risk factors for IPV. BPO was initially described outside the scope of IPV by Gunderson (1984), who listed the following key characteristics: (1) a tendency to maintain intense yet unstable interpersonal relationships; (2) labile sense of self, displaying high levels of anxiety in the face of possible abandonment and very low tolerance to being alone; and (3) intense anger and impulsiveness (Dutton and Starzomski, 1993). Frontiers in Psychology \| www.frontiersin.org Citation: In later studies by Dutton (1994, 2007), BPO was identiﬁed as a central component of the “abusive personality,” insofar as BPO is a continuum of problems and BPD is merely the most extreme form, while BPO presents a less severe and categorical clinical picture. The hypotheses guiding this study were: (1) The BPO scale will maintain the same factorial structure as the original study (Oldham et al., 1985) for this sample, consisting of three correlated factors; (2) The three subscales will display adequate reliability; (3) In terms of convergent validity, the BPO scale will correlate with other measures of borderline personality disorder, since BPD is the most severe presentation of BPO (Dutton, 1994, 2007), and with a measure of impulsivity, which is a key aspect of the clinical deﬁnition of BPO (Fossati et al., 2004; Liu et al., 2012). It is further expected that the BPO scale will correlate with a measure of antisocial personality disorder (ASPD), since BPD and ASPD are both Cluster B personality disorders and share common characteristics, including disproportionate impulsivity and highly unstable interpersonal relations (Liu et al., 2012). However, it is anticipated that the correlation will be weaker in this case because ASPD and BPD are distinct disorders; (4) In One of the most extensively used instruments to assess BPO in partner-violent men is the Borderline personality organization scale (Oldham et al., 1985). The scale consists of 30 items distributed across three subscales: Identity Diﬀusion (10 items), Reality Testing (10 items), and use of Primitive Defenses (10 items). Identity Diﬀusion is deﬁned as a subjective feeling of inconsistency and problems of self-deﬁnition that can in turn lead to emotional disturbance in the context of intimate relationships. Reality Testing refers to an individual’s capacity adequately to perceive reality. Lastly, the Primitive Defenses subscale concerns the defensive mechanisms that an individual unconsciously deploys to minimize the consequences of overly intense emotional situations. Dutton and Starzomski (1993) evaluated the psychometric characteristics of this instrument for a sample of 120 batterers, July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org 2 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. Citation: terms of known-groups validity, there will be signiﬁcant age- based diﬀerences in BPO (Jackson and Burgess, 2000; Cohen et al., 2005); and (5) In terms of discriminant validity, it is expected that the scores obtained on the BPO scale will correctly distinguish between individuals who meet the diagnostic criteria for BPD and those who do not. to me by either overwhelming me with love or abandoning me. Uncontrollable events are the cause of my diﬃculties. In the original studies, the mean score of individuals diagnosed with BPD was 73 on the overall scale compared to 59 for those not diagnosed with BPD. Meanwhile, the reliability scores were 0.92 for identity diﬀusion, 0.84 for reality testing, and 0.87 for primitive defenses (Oldham et al., 1985). Borderline and Antisocial Personality Disorders The structured clinical interview for DSM-IV Axis II Personality Disorders (SCID-II; First et al., 1999) was used. This structured interview format is designed to establish the presence or absence of the symptoms listed for diﬀerent personality disorders in the DSM-IV. In this study, only the items related to the borderline and antisocial personality scales were administered. The test- retest reliability obtained was 0.84 for the antisocial disorder and 0.37 for the borderline disorder. Socio-Demographic Questionnaire Questions were included to gather participant’s socio- demographic information including age, marital status, nationality, education, and occupation. The information was collected using a semi-structured interview format. Information relating to the oﬀenses committed were gathered from court records. Instruments Borderline Personality Organization Scale (Oldham et al., 1985) The BPO scale comprises 30 items and uses a 5-point Likert- type response format with answers ranging from “never true” to “always true.” There are three subscales: identity diﬀusion (10 items), reality testing (10 items), and use of primitive defenses (10 items). The identity diﬀusion subscale refers to the fragmentation of mental representations of the self, presenting in the form of grave distortions in the sphere of interpersonal relations, in particular intimate relationships. Examples of the items included in this subscale include: I see myself in totally diﬀerent ways at diﬀerent times; I’m afraid of losing myself when I get sexually involved; and I feel empty inside. The reality testing subscale concerns an individual’s ability to maintain a satisfactory perception of reality. BPO individuals often alternate between a clear understanding of reality and distorted or confused perceptions. Examples of the items used in the form to assess this aspect include: I hear things that other people claim are not really there; I’ve had relationships in which I couldn’t feel whether I or the other person was thinking or feeling something; and I believe that things will happen simply by thinking about them. Finally, the primitive defenses subscale deals with the defense mechanisms that the individual unconsciously uses to minimize the repercussions of emotionally charged situations, which might otherwise shake their psychological stability. Items from this subscale include: It is hard for me to trust people because they so often turn against me or betray me. People tend to respond Instruments Instruments Borderline Personality Organization Scale (Oldham et al., 1985) Frontiers in Psychology \| www.frontiersin.org Impulsivity Characteristics The plutchik impulsivity scale (Plutchik and Van Praag, 1989; Spanish adaptation by Rubio et al., 1998) was used. This scale is made up of 15 items and uses a Likert-type response format based on four alternatives ranging from never to almost always. The scale consists of 4 subscales: Planning capacity (e.g., Do you plan in advance? Are you careful or cautious?). Emotional control (e.g., Do you often become impatient? Do you ﬁnd it diﬃcult to control your emotions?). Behavior control in relating to eating spending and sex (e.g., Do you spend money impulsively? Do you ﬁnd it diﬃcult to control your sexual impulses? Do you eat even when you are not hungry?). Other behavior control (e.g., Do you ﬁnd queuing diﬃcult? Do you often say the ﬁrst thing that comes into your head?). Reliability was 0.73 in the original study (Plutchik and Van Praag, 1989) and 0.90 in the Spanish adaptation (Rubio et al., 1998). Participants The sample is made up of court-referred partner-violent men from the Madrid Region of Spain convicted of gender violence oﬀenses carrying a prison term of less than 2 years. Though their sentences were suspended, the oﬀenders were ordered by the courts to undergo a psychological treatment program, in accordance with Part IV of Spanish Law 1/2004 on comprehensive gender violence protection measures (Redondo et al., 2017, p.585). The sample included 643 partner-violent men ranging from 18 to 74 years of age (Mean 38.45, SD = 10.36). In terms of educational attainment, 41.2% of the sample had completed primary and 42.8% secondary education, while 16% had obtained a university degree. With regard to marital status, 31% of the men in the sample were married or cohabiting, 36.7% were single (i.e., they had no partner at the time of the evaluation), 31.9% were separated or divorced, and 0.6% were widowers. Lastly, 60.5% were Spanish, 29.2% were Central or South American, and 10.2% belonged to other nationalities. Conﬁrmatory Factor Analysis Co ato y acto a ys s Five hypotheses were proposed. The ﬁrst was that the BPO scale will maintain the same factorial structure as the original study for this sample, consisting of three correlated factors. Three factorial analyses were carried out to test this initial hypothesis: We carried out three CFAs using the maximum likelihood estimation method. Table 1 shows the goodness-of-ﬁt indices of the 3 models. The three-factor model yielded the best goodness-of- ﬁt indices (CMIN/df = 2.444; GFI = 0.917; AGFI = 0.901; and RMSEA = 0.047). The GFI and AGFI indices are above 0.90, while the RMSEA index is below 0.05. These data mean that the model that best ﬁts this scale is the three-factor model established in the original study. Therefore, the structure based on three correlated factors is the one that best ﬁts the data obtained from the study. The standardized regression coeﬃcients (standard factor Known-Groups Validity The forth hypothesis was that there will be signiﬁcant age- based diﬀerences in BPO scale. We established three age groups following the criteria described in research on the development of personality over an individual’s lifetime. These studies recommend using certain key decades as transition points to deﬁne age groups (Roberts and DelVecchio, 2000; Ullrich and Coid, 2009): men up to 29 years of age (early adulthood), from 30 to 50, and above 50. Signiﬁcant diﬀerences were observed in (a) the overall BPO scale depending on the age group, F(2,640) = 5.402, p < 0.01, and speciﬁcally between the early adult group and the middle-aged group, and the early adult group and the oldest group; (b) identity diﬀusion, F(2,640) = 3.493, p < 0.05, and (c) primitive defenses, F(2,640) = 3.783, p < 0.05. Signiﬁcant diﬀerences were found in both subscales between Groups 1 (≤29 years) and 2 (30–50 years). In all cases, Group 1 (≤29 years) presented higher levels of BPO (see Table 4). Data Analysis The statistical analyses were carried out using SPSS and AMOS (version 23.0). Cronbach’s alpha was calculated to establish the reliability of the instruments. Pearson’s correlation coeﬃcient was used to evaluate the concurrent validity of the BPO scale. Known- groups validity was assessed by means of a variance analysis to identify age-based diﬀerences in the BPO scores and to determine whether or not the borderline personality threshold point had been reached. We also used eta-squared (η2) to calculate the eﬀect size of the diﬀerences found. Procedure Prior to beginning this study a Spanish translation and cultural adaptation of the BPO scale was prepared following the guidance of the international tests commission (ITC; Hambleton, 2001). This involved separate translation by two groups of IPV experts with the appropriate cultural background and language skills, before a ﬁnal consensus version was agreed by both groups. The complete assessment was then carried out before the start of the court-mandated psychological treatment program to which oﬀenders had been assigned. This was the ﬁrst time that the participants had taken part in a program of this kind. The oﬀense data was collected using court records. From the court, they sent us the sentence of the participants where the facts that had been proven during the trial and for which July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org 3 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. they had been condemned were recorded. The type of IPV crime was analyzed: psychological, physical or sexual. Each participant attended a 60-min individual session weekly for between 4 and 8 weeks, during which (a) the conditions and goals of the research were explained, and informed consent was obtained in writing (all participants were over the age of 16); (b) socio-demographic data was collected; and (c) the participants completed the questionnaires for the scales described in the Instruments section (counterbalanced self-administration) (Redondo et al., 2017, pp. 585–586). loadings) are shown in Table 2. Scores above 0.40 are considered adequate (Redondo et al., 2017). However, three items were found below this value, although all of them are above 0.37 (i.e., item 9 “I hear things that other people claim are not really there,” item 26 “I am not sure whether a voice I have heard, or something that I have seen, is my imagination or not,” and item 27 “I have heard or seen things when there is no apparent reason for it”). It was therefore decided that these three items should not be removed from the ﬁnal model given that the other ﬁt indices (see Table 1) were all adequate. The second hypothesis was that the three subscales will display adequate reliability. The reliability (Cronbach’s alpha) of the total scale and the three subscales was satisfactory (0.78 for identity diﬀusion, 0.77 for reality testing, 0.80 for primitive defenses, and 0.91 for the total scale). Concurrent Validity In terms of convergent validity, the third hypothesis of this study is that the BPO scale will correlate with other measures of borderline and antisocial personality, and with a measure of impulsivity. We calculated the correlations between identity diﬀusion, primitive defenses, and reality testing and diﬀerent measurements for borderline personality, antisocial personality, and impulsivity. As may be observed in Table 3, all the correlations were either small-to-moderate or moderate, and all were statistically signiﬁcant (p < 0.001). We performed a CFA using the AMOS program to analyze the factorial structure of the BPO scale. The goodness-of- ﬁt indices used were CMIN/df, goodness of ﬁt index (GFI), adjusted goodness of ﬁt index (AGFI), and root mean-square error of approximation (RMSEA). Values equal to or higher than 0.90 are considered acceptable for GFI and AGFI, whereas values equal to or lower than 0.05 are considered excellent for RMSEA, and values lower than 0.08 are acceptable (Redondo et al., 2017, p. 586). Frontiers in Psychology \| www.frontiersin.org Discriminant Validity The last hypothesis of this study was that the scores obtained on the BPO scale will correctly distinguish between individuals who meet the diagnostic criteria for BPD and those who do not. To assess the BPO scale’s ability to distinguish participants who meet the majority of the diagnostic criteria for BPD, as determined from the scores obtained in the SCID II screening questionnaire, we divided the participants into two groups, one including those who passed the instrument’s cut-oﬀpoint (score equal to or higher than 5) and those who were below the cut-oﬀpoint. As may be observed in Table 5, the group of participants who passed the cut-oﬀpoint for BPD diagnosis presented statistically higher TABLE 1 \| Goodness-of-ﬁt indices for each model. CMIN/df AGFI RMSEA GFI 1-factor model 3.479 0.858 0.062 0.878 Model with 3 correlated ﬁrst-order factors 2.444 0.901 0.047 0.917 Hierarchical 3-factor model 3.508 0.857 0.063 0.878 Frontiers in Psychology \| www.frontiersin.org 4 TABLE 1 \| Goodness-of-ﬁt indices for each model. July 2019 \| Volume 10 \| Article 1653 4 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. DISCUSSION 16 I’ve had relationships in which I couldn’t feel whether I or the other person was thinking or feeling something 0.2 19 People see me as being rude or inconsiderate and I don’t know why 0.2 20 I can’t tell whether certain physical sensations I’m having are real, or whether I am imagining them 0.2 26 I am not sure whether a voice I have heard, or something that I have seen, is my imagination or not 0. 27 I have heard or seen things when there is no apparent reason for it 0. 30 Somehow, I never know quite how to conduct myself with people 0.4 Primitive defenses subscale 3 It is hard for me to trust people because they so often turn against me or betray me 0.3 4 People tend to respond to me by either overwhelming me with love or abandoning me 0.3 6 I act in ways that strike others as unpredictable and erratic 0.3 8 Uncontrollable events are the cause of my difﬁculties 0.2 11 I tend to feel things in a somewhat extreme way, experiencing either great joy or intense despair 0.3 14 I feel that certain episodes of my life do not count and are better erased from my mind 0. 18 I feel people don’t give me the respect I deserve unless I put pressure on them 0.3 22 I ﬁnd myself doing things which feel okay while I am doing them but which I later ﬁnd hard to believe I did 0.3 28 I feel I don’t get what I want 0.3 TABLE 2 \| Three-factor model: Standardized factor loadings for the borderline personality organization (BPO) scale. DISCUSSION Squared multiple correlations Factor loading Critical ratio Item Identity diffusion subscale 1 I feel like a fake or an imposter, that others see me as quite different at times 0.269 0.518 10.115 5 I see myself in totally different ways at different times 0.426 0.653 11.599 10 I feel empty inside 0.346 0.588 10.936 12 It is hard for me to be sure about what others think of me, even people who have known me well 0.350 0.591 10.975 15 I ﬁnd it hard to describe myself 0.286 0.535 10.320 17 I don’t feel like myself unless exciting things are going on around me 0.339 0.582 10.864 21 Some of my friends would be surprised if they knew how differently I behave in different situations 0.197 0.443 9.091 24 When I want something from someone else, I can’t ask for it directly 0.187 0.432 8.924 25 I feel I’m a different person at home as compared to how I am at work or school 0.252 0.502 – Reality testing subscale 2 I feel almost as if I’m someone else, like a friend or relative or even someone I don’t know 0.202 0.449 10.770 7 I ﬁnd I do things which get other people upset, and I don’t know why such things upset them 0.437 0.661 15.565 9 I hear things that other people claim are not really there 0.150 0.387 9.314 16 I’ve had relationships in which I couldn’t feel whether I or the other person was thinking or feeling something 0.261 0.511 12.195 19 People see me as being rude or inconsiderate and I don’t know why 0.267 0.517 12.324 20 I can’t tell whether certain physical sensations I’m having are real, or whether I am imagining them 0.235 0.485 11.597 26 I am not sure whether a voice I have heard, or something that I have seen, is my imagination or not 0.139 0.373 8.992 27 I have heard or seen things when there is no apparent reason for it 0.138 0.372 8.950 30 Somehow, I never know quite how to conduct myself with people 0.467 0.683 – Primitive defenses subscale 3 It is hard for me to trust people because they so often turn against me or betray me 0.326 0.571 – 4 People tend to respond to me by either overwhelming me with love or abandoning me 0.317 0.563 13.145 6 I act in ways that strike others as unpredictable and erratic 0.358 0.599 12.354 8 Uncontrollable events are the cause of my difﬁculties 0.297 0.545 11.537 11 I tend to feel things in a somewhat extreme way, experiencing either great joy or intense despair 0.360 0.600 12.371 14 I feel that certain episodes of my life do not count and are better erased from my mind 0.169 0.411 9.189 18 I feel people don’t give me the respect I deserve unless I put pressure on them 0.362 0.601 12.387 22 I ﬁnd myself doing things which feel okay while I am doing them but which I later ﬁnd hard to believe I did 0.337 0.580 12.079 28 I feel I don’t get what I want 0.302 0.550 11.605 TABLE 2 \| Three-factor model: Standardized factor loadings for the borderline personality organization (BPO) scale. July 2019 \| Volume 10 \| Article 1653 DISCUSSION scores on the overall BPO scale, t(641) = −17.483, p < 0.001. Likewise, statistically higher scores were observed in the BPD group in identity diﬀusion, t(641) = −16.982, p < 0.001, reality testing, t(641) = −14.572, p < 0.001, and primitive defenses, t(641) = −15.032, p < 0.001 (see Table 5). Even though the BPO scale is a widely used as an instrument in research involving partner-violent men, no prior research has analyzed the structure of the scale in this population. Our ﬁrst ed the structure of the scale in this population. Our ﬁrst nization (BPO) scale. uared multiple correlations Factor loading Critical ratio 0.269 0.518 10.115 0.426 0.653 11.599 0.346 0.588 10.936 0.350 0.591 10.975 0.286 0.535 10.320 0.339 0.582 10.864 0.197 0.443 9.091 0.187 0.432 8.924 0.252 0.502 – 0.202 0.449 10.770 0.437 0.661 15.565 0.150 0.387 9.314 0.261 0.511 12.195 0.267 0.517 12.324 0.235 0.485 11.597 0.139 0.373 8.992 0.138 0.372 8.950 0.467 0.683 – 0.326 0.571 – 0.317 0.563 13.145 0.358 0.599 12.354 0.297 0.545 11.537 0.360 0.600 12.371 0.169 0.411 9.189 0.362 0.601 12.387 0.337 0.580 12.079 0.302 0.550 11.605 TABLE 2 \| Three-factor model: Standardized factor loadings for the borderline personality organizati Squared corre Item Identity diffusion subscale 1 I feel like a fake or an imposter, that others see me as quite different at times 0.2 5 I see myself in totally different ways at different times 0.4 10 I feel empty inside 0.3 12 It is hard for me to be sure about what others think of me, even people who have known me well 0.3 15 I ﬁnd it hard to describe myself 0.2 17 I don’t feel like myself unless exciting things are going on around me 0.3 21 Some of my friends would be surprised if they knew how differently I behave in different situations 0. 24 When I want something from someone else, I can’t ask for it directly 0. 25 I feel I’m a different person at home as compared to how I am at work or school 0.2 Reality testing subscale 2 I feel almost as if I’m someone else, like a friend or relative or even someone I don’t know 0.2 7 I ﬁnd I do things which get other people upset, and I don’t know why such things upset them 0.4 9 I hear things that other people claim are not really there 0. DISCUSSION July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. was similar to the scores observed in other studies of batterers (Tweed and Dutton, 1998). hypothesis was that the optimum data ﬁt would be achieved in the model proposed in the original study (Oldham et al., 1985). However, we also considered that it would be a necessary ﬁrst step in adapting the scale for use with a sample of batterers in Spain to establish the goodness-of-ﬁt between the data obtained from this study with all possible factorial models in order to verify that the original study does in fact have the best ﬁt based on empirical data. In the case of the hierarchical model, the hypothesis was that there is an overall BPO score which contains the three subscales proposed in the original study. Therefore, in the present study, we ran the CFA in the three possible models, ﬁnding the best ﬁt in the three-factor model, as proposed in the original study. The reliability of the three BPO subscales and the overall scale was above 0.77, and therefore satisfactory in all cases, verifying the second hypothesis regarding the reliability of the scale. Reliability The third hypothesis tested in the study, regarding concurrent validity, was also veriﬁed. The BPO scale correlated signiﬁcantly and positively with BPD, which is the most extreme form of BPO (Dutton, 1994, 2007). We also found signiﬁcant correlations with impulsivity, a key construct in the clinical characteristics of BPO (Fossati et al., 2004; Liu et al., 2012). Positive and signiﬁcant correlations were also found between BPO and ASPD, although the scores observed were somewhat lower because these are independent disorders. The association between BPO and ASPD is explained by the fact that both personality disorders belong to Cluster B despite their diﬀering clinical presentations, placing them among the disorders most frequently associated with IPV, with which they share clinical features like emotional instability, TABLE 3 \| Means, standard deviations, alpha reliability coefﬁcients, and Pearson’s correlations between BPO subscales and measures of antisocial and borderline personality, and impulsiveness. July 2019 \| Volume 10 \| Article 1653 DISCUSSION Measure Identity diffusion Reality testing Primitive defenses Total M SD α BPO Identity diffusion – 5.582 5.384 0.777 Reality testing 0.763∗∗∗ – 3.899 4.342 0.772 Primitive defenses 0.783∗∗∗ 0.756∗∗∗ – 8.159 6.181 0.798 Total 0.924∗∗∗ 0.896∗∗∗ 0.934∗∗∗ – 17.639 14.643 0.913 SCID II measures of personality Borderline 0.656∗∗∗ 0.603∗∗∗ 0.641∗∗∗ 0.691∗∗∗ 4.063 3.152 0.805 Antisocial 0.292∗∗∗ 0.297∗∗∗ 0.299∗∗∗ 0.312∗∗∗ 1.700 2.379 0.798 Measures of impulsivity Plutchnik 0.549∗∗∗ 0.521∗∗∗ 0.535∗∗∗ 0.582∗∗∗ 11.412 5.859 0.753 SCID II, self-report assessment of the diagnostic and statistical manual of mental disorders-IV R personality disorders; Plutchik, plutchik impulsive control scale. ∗∗∗p < 0.001. TABLE 4 \| Means, standard deviations, and differences by age in the subscales of the borderline personality organization (BPO) scale. Group 1 (≤29 years) (n = 137, 21.3%) M (SD) Group 2 (30–50 years) (n = 422, 65.6%) M (SD) Group 3 (>50 years) (n = 84, 13.1%) M (SD) Total (n = 643) M (SD) F(2.640)/ η2 Bonferroni BPO total 20.679 (16.427) 16.808 (13.914) 16.857 (14.668) 17.639 (14.668) 5.402∗∗0.02 1>2∗∗1>3∗ Identity diffusion 6.898 (6.264) 5.279 (5.039) 4.952 (5.243) 5.582 (5.385) 3.493∗0.01 1>2∗ Reality testing 4.752 (4.759) 3.628 (4.156) 3.869 (4.420) 3.899 (4.342) 1.748 0.01 Primitive defenses 9.029 (6.660) 7.900 (6.039) 8.036 (6.017) 8.159 (6.181) 3.783∗0.01 1>2∗ η2, eta-square; ∗p < 0.05. ∗∗p < 0.01. TABLE 5 \| Means, standard deviations, and differences in the subscales of the borderline personality organization (BPO) scale according to the cut-off point for BPD as measured with the SCID II questionnaire. BPD SCID II below the cut-off point (n = 455; 70.8%) M (SD) BPD SCID II exceeding the cut-off point (n = 188; 29.2%) M (SD) Total (n = 643) M (SD) t(641) η2 BPO total 12.295 (10.316) 30.575 (15.497) 17.639 (14.644) −17.483∗∗∗ 0.323 Identity diffusion 3.655 (3.789) 10.245 (5.815) 5.582 (5.385) −16.982∗∗∗ 0.310 Reality testing 2.508 (2.508) 7.266 (5.272) 3.899 (4.342) −14.572∗∗∗ 0.249 Primitive defenses 6.132 (6.132) 13.064 (6.140) 8.159 (6.181) −15.032∗∗∗ 0.261 SCID II, self-report assessment of the DSM-IV R personality disorders. ∗∗∗p < 0.001. TABLE 3 \| Means, standard deviations, alpha reliability coefﬁcients, and Pearson’s correlations between BPO subscales and measures of antisocial and borderline personality, and impulsiveness. DISCUSSION Measure Identity diffusion Reality testing Primitive defenses Total M SD α BPO Identity diffusion – 5.582 5.384 0.777 Reality testing 0.763∗∗∗ – 3.899 4.342 0.772 Primitive defenses 0.783∗∗∗ 0.756∗∗∗ – 8.159 6.181 0.798 Total 0.924∗∗∗ 0.896∗∗∗ 0.934∗∗∗ – 17.639 14.643 0.913 SCID II measures of personality Borderline 0.656∗∗∗ 0.603∗∗∗ 0.641∗∗∗ 0.691∗∗∗ 4.063 3.152 0.805 Antisocial 0.292∗∗∗ 0.297∗∗∗ 0.299∗∗∗ 0.312∗∗∗ 1.700 2.379 0.798 Measures of impulsivity Plutchnik 0.549∗∗∗ 0.521∗∗∗ 0.535∗∗∗ 0.582∗∗∗ 11.412 5.859 0.753 SCID II, self-report assessment of the diagnostic and statistical manual of mental disorders-IV R personality disorders; Plutchik, plutchik impulsive control scale. ∗∗∗p < 0.001. TABLE 4 \| Means, standard deviations, and differences by age in the subscales of the borderline personality organization (BPO) scale. Group 1 (≤29 years) (n = 137, 21.3%) M (SD) Group 2 (30–50 years) (n = 422, 65.6%) M (SD) Group 3 (>50 years) (n = 84, 13.1%) M (SD) Total (n = 643) M (SD) F(2.640)/ η2 Bonferroni BPO total 20.679 (16.427) 16.808 (13.914) 16.857 (14.668) 17.639 (14.668) 5.402∗∗0.02 1>2∗∗1>3∗ Identity diffusion 6.898 (6.264) 5.279 (5.039) 4.952 (5.243) 5.582 (5.385) 3.493∗0.01 1>2∗ Reality testing 4.752 (4.759) 3.628 (4.156) 3.869 (4.420) 3.899 (4.342) 1.748 0.01 Primitive defenses 9.029 (6.660) 7.900 (6.039) 8.036 (6.017) 8.159 (6.181) 3.783∗0.01 1>2∗ η2, eta-square; ∗p < 0.05. ∗∗p < 0.01. TABLE 5 \| M t d d d i ti d diff i th b l f th b d li lit i ti (BPO) l di t th t ff i t f BPD TABLE 5 \| Means, standard deviations, and differences in the subscales of the borderline personality organization (BPO) scale according to the cut-off point for BPD as measured with the SCID II questionnaire. BPD SCID II below the cut-off point (n = 455; 70.8%) M (SD) BPD SCID II exceeding the cut-off point (n = 188; 29.2%) M (SD) Total (n = 643) M (SD) t(641) η2 BPO total 12.295 (10.316) 30.575 (15.497) 17.639 (14.644) −17.483∗∗∗ 0.323 Identity diffusion 3.655 (3.789) 10.245 (5.815) 5.582 (5.385) −16.982∗∗∗ 0.310 Reality testing 2.508 (2.508) 7.266 (5.272) 3.899 (4.342) −14.572∗∗∗ 0.249 Primitive defenses 6.132 (6.132) 13.064 (6.140) 8.159 (6.181) −15.032∗∗∗ 0.261 SCID II, self-report assessment of the DSM-IV R personality disorders. ∗∗∗p < 0.001. July 2019 \| Volume 10 \| Article 1653 6 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. Graña et al., 2014). DISCUSSION The availability of valid, reliable instruments to measure these characteristics should help with the early identiﬁcation of the most severe partner-assaultive men, who present with disturbed personality traits and disorders, thereby allowing the adaptation of intervention programs to address their speciﬁc needs (Cavanaugh and Gelles, 2005; Stoops et al., 2010). Where aggressors present disturbed personality traits or diagnosed personality disorders, the clinical formulation of each speciﬁc case is especially relevant, as is the existence of psychometrically guaranteed assessment instruments to allow evaluation of the core intervention goals. The present study thus underscores the importance of instruments to assess BPO for a range of purposes, including (a) research to establish the risk factors involved in IPV and the proﬁles and subtypes of aggressors with diﬀerential personality traits in order to design and hone diﬀerent intervention strategies; (b) evaluation of the eﬀectiveness of the psychological treatment programs undertaken with aggressors; and (c) forensic assessments and evaluations, given that the population concerned are convicted oﬀenders serving suspended prison sentences. impulsivity, and unstable personal relations (Liu et al., 2012; González et al., 2016). These signiﬁcant and positive correlations between the BPO subscales and other external variables indicate that the BPO scale has good concurrent validity, as had already been shown in previous studies (Dutton and Starzomski, 1993; Dutton, 1994). Given the importance of age as a variable that explains many personality traits (Zanarini et al., 2005, 2006), we analyzed the relationship between the age of the study participants and the BPO scale. We found no research focusing speciﬁcally on age-related changes in the levels of BPO, but we did ﬁnd studies that analyzed the lifetime course of BPD. Historically, the supposed stability of personality traits and disorders has been considered a key feature. However, more recent studies increasingly suggest that personality disorders vary over a person’s lifetime, undergoing periods of stability and change (Clark, 2005). In this light, it would seem that the general tendency is for the prevalence of personality disorders to decrease with age according to both longitudinal studies (Cohen et al., 2005) and cross-sectional studies (Jackson and Burgess, 2000). Furthermore, the relationship between younger age and personality disorders seems to be especially relevant in the case of Cluster B disorders (Samuels et al., 2002; Coid et al., 2006), in particular BPD (Engels et al., 2003). AUTHOR CONTRIBUTIONS All authors contributed to revision, read, and approved the submitted version of the manuscript. All authors contributed to revision, read, and approved the submitted version of the manuscript. ETHICS STATEMENT This study was approved by the Ethics Commission of the Faculty of Psychology of the University Complutense of Madrid, on May 30, 2009. This study constitutes an important advance in research into partner-assaultive men. In recent years, numerous papers have shown that aggressors of this kind do not form a single homogeneous group. Rather, various diﬀerent subtypes exist depending on the severity, frequency and generality of the IPV exhibited, levels of anger and the presence of associated psychopathology (Holtzworth-Munroe and Stuart, 1994; Holtzworth-Munroe et al., 2000; Cavanaugh and Gelles, 2005; DISCUSSION Our results regarding the relationship between BPO and age provide evidence for the existence of known-groups validity, and they are likewise consistent with earlier ﬁndings indicating that BPD traits are more acute before the age of 30 and stabilize increasingly between the ages of 50 and 70 (Roberts and DelVecchio, 2000; Terracciano et al., 2006). The fourth hypothesis proposed, regarding age, was therefore veriﬁed. This study has certain limitations which should be addressed. In the ﬁrst place, there are three items in the model with factor scores of less than 0.40, although it was decided to retain them given that the scores for the remaining indices were satisfactory. Also, we only analyzed the data of partner-violent men sentenced to prison terms of less than 2 years. The legal situation of the participants in this study and the possible social desirability attendant upon their responses may therefore have aﬀected their responses. Meanwhile, future research might explore the capacity of the BPO scale to predict recidivism, given the association between BPD and reoﬀending among partner-violent men (Taft et al., 2004; Romero-Martínez et al., 2016). Lastly, it would be interesting to complete the results of these self-reported scales with interviews designed directly to analyze the diagnostic criteria for BPD and the clinical features of BPO. Finally, the ﬁfth hypothesis, regarding discriminant validity, was also veriﬁed. We found that participants who met the diagnostic criteria for BPD scored signiﬁcantly higher on the three BPO subscales and on the total scale compared to the group of participants who did not meet the diagnostic criteria for BPD. BPO is a continuum of disturbances in which BPD is the pole of greatest severity (Dutton, 1994, 2007), so these results are plausible from a theoretical viewpoint. REFERENCES Violence and Control in Intimate Relationships, (2nd Edn). New York, NY: Guilford Press. Mackay, J., Bowen, E., Walker, K., and O’Doherty, L. (2018). Risk factors for female perpetrators of intimate partner violence within criminal justice settings: a systematic review. Aggress. Violent Behav. 41, 128–146. doi: 10.1016/j.avb.2018. 06.004 Dutton, D. G., and Starzomski, A. J. (1993). Borderline personality in perpetrators of psychological and physical abuse. Violence Vict. 8, 327–337. doi: 10.1891/ 0886-6708.8.4.327 Millon, T. (1987). Manual for the MCMI-II. Minneapolis, MN: National Computer Systems. Dutton, D. G., Starzomski, A. J., and Ryan, L. (1996). Antecedents of abusive personality and abusive behavior in wife assaulters. J. Fam. Violence 11, 113–132. doi: 10.1007/bf02336665 Okuda, M., Olfson, M., Hasin, D., Grant, B. F., Lin, K., and Blanco, C. (2011). Mental healthof victims of intimate partner violence: results from a national epidemiologic survey. Psychiatr. Serv. 62, 959–962. doi: 10.1176/ps.62. 8.pss6208_0959 Engels, G. I., Duijsens, I. J., Haringsma, R., and van Putten, C. M. (2003). Personality disorders in the elderly compared to four younger age groups: a cross-sectional study of community residents and mental health patients. J. Pers. Disord. 17, 447–459. doi: 10.1521/pedi.17.5.447.22971 Oldham, J., Clarkin, J., Appelbaum, A., Carr, A., Kernberg, P., Lotterman, A., et al. (1985). “A self-report instrument for borderline personality organization,” in The Borderline: Current Empirical Research. The Progress in Psychiatry Series, ed. H. McGlashan (Washington, DC: American Psychiatric Press), 1–18. First, M. B., Gibbon, M., Spitzer, R. L., Williams, J. B. W., and Smith Benjamin, L. (1999). Guía Del Usuario de la Entrevista Clínica Estructurada Para los Trastornos de Personalidad del Eje II del DSM-IV- SCID-II [Structured Clinical Interview User Guide for DSM-IV Axis II Personality Disorders - SCID-II]. Paris: Masson. Plutchik, R., and Van Praag, H. (1989). The measurement of suicidality, agressivity and impulsivity. Prog. Neuro Pshychopharmacol. Biol. Psychiatry 13, 523–534. Fossati, A., Barratt, E. S., Carretta, I., Leonardi, B., Grazioli, F., and Maﬀei, C. (2004). Predicting borderline and antisocial personality disorder features in nonclinical subjects using measures of impulsivity and aggressiveness. Psychiatr. Res. 125, 161–170. doi: 10.1016/j.psychres.2003.12.001 Redondo, N., Peña, M. E., Graña, J. L., and Andreu, J. M. (2017). Psychometric properties of the aggression questionnaire: a replication in a sample of partner-assaultive men in psychological treatment. Psicothema 29, 584–589. doi: 10.7334/psicothema2016.390 González, R. A., Igoumenou, A., Kallis, C., and Coid, J. W. (2016). Borderline personality disorder and violence in the UK population: categorical and dimensional trait assessment. REFERENCES BMC Psychiatr. 16:180. doi: 10.1186/s12888-016- 0885-7 Roberts, B. W., and DelVecchio, W. F. (2000). The rank-order consistency of personality traits from childhood to old age: a quantitative review of longitudinal studies. Psychol. Bull. 126, 3–25. doi: 10.1037//0033-2909. 126.1.3 Graña, J. L., Redondo, N., Muñoz-Rivas, M. J., and Cantos, A. L. (2014). Subtypes of batterers in treatment: empirical support for a distinction between Type I. Type II and Type III. PLoS One 9:e110651. doi: 10.1371/journal.pone.0110651 Romero-Martínez, Á., Lila, M., and Moya-Albiol, L. (2016). Empathy impairments in intimate partner violence perpetrators with antisocial andborderline traits: a key factor in the risk of recidivism. Violence Vict. 31, 347–360. doi: 10.1891/ 0886-6708.VV-D-14-00149 Graña, J. L., Redondo, N., Muñoz-Rivas, M. J., and Cuenca, M. L. (2017). Short- term treatment response, attrition and recidivism in a partner violent men typology compared with a control group. Behav. Psychol. 25, 465–482. Rubio, G., Montero, I., Jáuregui, J., Martínez, M. L., Álvarez, S., and Marín, J. J. (1998). Validación de la escala de impulsividad de plutchik en población española. Arch. Neurobiol. 61, 223–232. Gunderson, J. G. (1984). Borderline Personality Disorder. Washington, DC: American Psychiatric Press. Samuels, J., Eaton, W. W., Bienvenu, O. J., Brown, C. H., Costa, P. T., and Neustadt, G. (2002). Prevalence and correlates of personality disorders in a community sample. Br. J. Psychiatry 180, 536–542. doi: 10.1192/bjp.180.6.536 Hambleton, R. K. (2001). The next generation of the ITC test translation and adaptation guidelines. Eur. J. Psychol. Assess. 17, 164–172. doi: 10.1027//1015- 5759.17.3.164 Hiscoke, U. L., Langström, N., Ottosson, H., and Grann, M. (2003). Self-reported personality traits and disorders (DSM–IV) and risk of criminal recidivism: a prospective study. J. Pers. Disord. 17, 293–305. doi: 10.1521/pedi.17.4.293. 23966 Stoops, C., Bennett, L., and Vincent, N. (2010). Development and predictive ability of a behavior-based typology of men who batter. J. Fam. Violence 25, 325–335. doi: 10.1007/s10896-009-9294-8 Straus, M. A. (1979). Measuring intrafamily conﬂict and violence: the conﬂict tactics scales. J. Marriage Fam. 41, 75–88. Holtzworth-Munroe, A., Meehan, J. C., Herron, K., Rehman, U., and Stuart, G. L. (2000). Testing the holtzworth-munroe and stuart batterer typology. J. Consult. Clin. Psychol. 68, 1000–1019. doi: 10.1037/0022-006X.68.6.1000 Stylianou, A. (2018). Economic abuse experiences and depressive symptoms among victims of intimate partner violence. J. Fam. Violence 33, 381–392. doi: 10.1007/s10896-018-9973-4 Holtzworth-Munroe, A., and Stuart, G. L. (1994). Typologies of male batterers: three subtypes and the diﬀerences among them. Psychol. Bull. 116, 476–497. doi: 10.1037//0033-2909.116.3.476 Taft, C. REFERENCES Cavanaugh, M. M., and Gelles, R. J. (2005). The utility of male domestic violence oﬀender typologies: new directions for research, policy, and practice. J. Int. Violence 20, 155–166. doi: 10.1177/0886260504268763 American Psychiatric Association (2014). Diagnostic and Statistical Manual of Mental Disorders (DSM-5). 5th Edn. Arlington, VA: American Psychiatric Association. Clark, L. A. (2005). Stability and change in personality pathology: revelations of three longitudinal studies. J. Pers. Disord. 19, 524–532. doi: 10.1521/pedi.2005. 19.5.524 Carbajosa, P., Catalá-Miñana, A., Lila, M., and Gracia, E. (2017). Diﬀerences in treatment adherence, program completion, and recidivism among batterer subtypes. Eur. J. Psychol. Appl. Legal Context 9, 93–101. doi: 10.1016/j.ejpal. 2017.04.001 Cohen, P., Crawford, T. N., Johnson, J. G., and Kasen, S. (2005). The children in the community study of developmental course of personality disorder. J. Pers. Disord. 19, 466–486. doi: 10.1521/pedi.2005.19.5.466 July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org Frontiers in Psychology \| www.frontiersin.org 7 Borderline Personality in Partner-Violent Men Redondo Rodríguez et al. Coid, J., Yang, M., Tyrer, P., Roberts, A., and Ullrich, S. (2006). Prevalence and correlates of personality disorder in Great Britain. Br. J. Psychiatr. 188, 423–431. doi: 10.1192/bjp.188.5.423 Gender-Based Violence. Oﬃcial State Bulletin, 313. Available at: http://www.boe. es/boe/dias/2004/12/29/pdfs/A42166-42197.pdf (accessed December 29, 2004). Liu, N., Zhang, Y., Brady, H. J., Cao, Y., He, Y., and Zhang, Y. (2012). Relation between childhood maltreatment and severe intrafamilial male-perpetrated physical violence in Chinese community: the mediating role of borderline and antisocial personality disorder features. Aggress. Behav. 38, 64–76. doi: 10.1002/ ab.20417 Davoren, M., Constantinos, K., González, R. A., Freestone, M., and Coid, J. W. (2017). Anxiety disorders and intimate partner violence: can the association be explained by coexisting conditions or borderline personality traits? J. Forensic Psychiatr. Psychol. 28, 639–658. doi: 10.1080/14789949.2016.1172659 Logan, C., and Blackburn, R. (2009). Mental disorder in violent women in secure settings: potential relevance to risk for future violence. Int. J. Law Psychiatry 32, 31–38. doi: 10.1016/j.ijlp.2008.11.010 DeBoard-Lucas, R., and Grych, J. (2011). Children’s perceptions of intimate partner violence: causes, consequences, and coping. J. Fam. Violence 26, 343–354. doi: 10.1007/s10896-011-9368-2 Dutton, D. G. (1994). Behavioral and aﬀective correlates of borderline personality organization in wife assaulters. Int. J. Law Psychiatr. 17, 265–279. doi: 10.1016/ 0160-2527(94)90030-2 Loinaz, I., Marzabal, I., and Andrés-Pueyo, A. (2018). Risk factors of female intimate partner and non-intimate partner homicides. Eur. Psychol. Appl. Legal Context 10, 49–55. doi: 10.5093/ejpalc2018a4 Dutton, D. G. (2007). The Abusive Personality. REFERENCES T., Murphy, C. M., Musser, P. H., and Remington, N. A. (2004). Personality, interpersonal and motivational predictors of the working alliance in group cognitive-behavioral therapy for partner violent men. J. Consult. Clin. Psychol. 72, 349–354. doi: 10.1037/0022-006X.72.2.349 Jackson, H. J., and Burgess, P. M. (2000). Personality disorders in the community: a report from the Australian national survey of mental health and wellbeing. Soc. Psychiatry Psychiatr. Epidemiol. 35, 531–538. doi: 10.1007/s001270050276 Terracciano, A., Costa, P. T., and McCrae, R. R. (2006). Personality plasticity after age 30. Pers. Soc. Psychol. Bull. 32, 999–1009. doi: 10.1177/014616720628 8599 Jose, A., O’Leary, K. D., Graña, J. L., and Foran, H. M. (2014). Risk factors for men’s intimate physical aggression in Spain. J. Fam. Violence 29, 287–297. doi: 10.1007/s10896-014-9589-2 Tolman, R. M. (1989). The development of a measure of psychological maltreatment of women by their male partners. Violence Vict. 4, 159–177. doi: 10.1891/0886-6708.4.3.159 Ley Orgánica 1/2004 (2004). Medidas de Protección Integral Contra la Violencia de Género. Boletín Oﬁcial del Estado Núm. [Measures of Integral Protection Against July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org 8 Redondo Rodríguez et al. Borderline Personality in Partner-Violent Men Zanarini, M. C., Frankenburg, F. R., Hennen, J., Reich, B., and Silk, K. R. (2006). Prediction of the 10-year course of borderline personality disorder. Am. J. Psychiatry 163, 827–832. doi: 10.1176/ajp.2006.163.5.827 Tweed, R. G., and Dutton, D. G. (1998). A comparison of impulsive and instrumental subgroups of batterers. Violence Vict. 13, 217–230. doi: 10.1891/ 0886-6708.13.3.217 Ullrich, S., and Coid, J. (2009). The age distribution of self-reported personality disorder traits in a household population. J. Pers. Disord. 23, 187–200. doi: 10.1521/pedi.2009.23.2.187 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. United Nations (1994). Declaration on the Elimination of Violence Against Women (A/RES/48/104). New York, NY: United Nations General Assembly. Copyright © 2019 Redondo Rodríguez, Graña Gómez, Cuenca Montesino and Muñoz-Rivas. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. Frontiers in Psychology \| www.frontiersin.org July 2019 \| Volume 10 \| Article 1653 REFERENCES No use, distribution or reproduction is permitted which does not comply with these terms. World Health Organization [WHO] (2002). Informe Mundial Sobre Violencia y Salud [World Report on Violence and Health]. Geneva: World Health Organization. Zanarini, M. C., Frankenburg, F. R., Hennen, J., Reich, B., and Silk, K. R. (2005). The mclean study of adult development (msad): overview and implications of the ﬁrst six years of prospective follow-up. J. Pers. Disord. 19, 505–523. doi: 10.1521/pedi.2005.19.5.505 July 2019 \| Volume 10 \| Article 1653 Frontiers in Psychology \| www.frontiersin.org 9
https://openalex.org/W2564007830	https://www.scielo.br/j/rsp/a/ThGbJp54VGJ85zXNBXMfxQQ/?lang=en&format=pdf	English	null	Primary Health Care: care coordinator in regionalized networks?	Revista de saúde pública/Revista de Saúde Pública	2,016	cc-by	8,453	Correspondência: Patty Fidelis de Almeida Rua Marquês de Paraná, 303 3º andar Prédio Anexo ao Hospital Antonio Pedro 24030-210 Niterói, RJ, Brasil E-mail: pattyfidelis@id.uff.br http://www.rsp.fsp.usp.br/ Original Article Original Article Rev Saúde Pública 2016;50:80 Patty Fidelis de AlmeidaI, Adriano Maia dos SantosII Patty Fidelis de AlmeidaI, Adriano Maia dos SantosII I Departamento de Planejamento em Saúde. Instituto de Saúde Coletiva. Universidade Federal Fluminense. Niterói, RJ, Brasil II Instituto Multidisciplinar em Saúde. Campus Anísio Teixeira. Universidade Federal da Bahia. Vitória da Conquista, BA, Brasil Primary Health Care: care coordinator in regionalized networks? Patty Fidelis de AlmeidaI, Adriano Maia dos SantosII INTRODUCTION Lack of care coordination is identified as a major cause of poor quality in health services, associated with higher costs, duplication and overuse of diagnostic procedures, use of multiple medicines, and conflicting therapies, with chronic conditions suffering the greatest negative impact6,12,14,a. Coordination presupposes organizing patient care, which may involve two or more providers and users themselves, in order to facilitate the timely provision of services, involving planning related to staff and other resources and instruments for information exchange between providers13,17. Organizational elements to ensure coordination should include the definition of shared goals for the health system; financial incentives via disbursement and allocation of resources; communication mechanisms between health professionals; development of a common culture and leadership oriented towards teamwork, collaboration and better performance; and strengthening of a care model based on Primary Health Care (PHC)a. Different contexts must be considered, since there is no broadly developed definition13. The Política Nacional de Atenção Básica (PNAB – National Primary Care Policy)b defines coordination as “coordination of integrality,” one of the basis of PHC, which should be enabled by means of horizontal integration strategies (programmatic action and spontaneous demand, surveillance and care initiatives, multidisciplinary and interdisciplinary work) and vertical integration strategies between different levels of the Healthcare Networks. In Brazil, studies suggest that integration of Healthcare Networks, one of the dimensions of coordination, has been strengthened by expansion of Family Health Strategy; creation of specialized services in health districts; introduction of a regulatory system; computerization of medical records; development of management and clinical protocols; and initiatives in communication and technical support1,8,9,15. In this article, care coordination is understood as interaction between various services, actions and professionals related to health care to ensure it is always synchronized and focused on achieving a common goal, regardless of where it is provided11. It is supported by the existence of integrated action between providers at different levels or within a same level, so that different interventions are perceived and experienced by users in a manner that is continuous and appropriate to their health needs5. a Schang L, Waibel S, Thomson S. Measuring care coordination: health system and patient perspectives: report prepared for the Main Association of Austrian Social Security Institutions. London: LSE Health; 2013 [cited 2015 Jul 5]. Available from: http://eprints. lse.ac.uk/59573/1/__lse.ac.uk_ storage_LIBRARY_Secondary_ libfile_shared_repository_ Content_Schang%2C%20L_ Schang_Measuring%20care%20 coordination_2014.pdf This article sought to examine the breadth of care coordination by PHC in regionalized networks. b Ministério da Saúde (BR). Portaria 2.488, de 21 de outubro de 2011. Aprova a Política Nacional de Atenção Básica, estabelecendo a revisão de diretrizes e normas para a organização da Atenção Básica, para a Estratégia Saúde da Família (ESF) e o Programa de Agentes Comunitários de Saúde (PACS). Brasília (DF); 2011 [cited 2015 Jul 5]. Available from: http://bvsms.saude.gov. br/bvs/saudelegis/gm/2011/ prt2488_21_10_2011.html INTRODUCTION It aims to contribute new elements by debating the issue in the context of health regions, given the lack of research proposing to investigate coordination in circumstances that require horizontal integration between same-level professionals and services providers, based on PHC with strong essential attributes, and vertical integration between network services managed by different state agencies. Therefore, analyzing facilitating devices and barriers to coordination in regional areas may indicate paths to achieve timelier and higher quality access to the Brazilian Unified Health System (SUS). b Ministério da Saúde (BR). Portaria 2.488, de 21 de outubro de 2011. Aprova a Política Nacional de Atenção Básica, estabelecendo a revisão de diretrizes e normas para a organização da Atenção Básica, para a Estratégia Saúde da Família (ESF) e o Programa de Agentes Comunitários de Saúde (PACS). Brasília (DF); 2011 [cited 2015 Jul 5]. Available from: http://bvsms.saude.gov. br/bvs/saudelegis/gm/2011/ prt2488_21_10_2011.html a Schang L, Waibel S, Thomson S. Measuring care coordination: health system and patient perspectives: report prepared for the Main Association of Austrian Social Security Institutions. London: LSE Health; 2013 [cited 2015 Jul 5]. Available from: http://eprints. lse.ac.uk/59573/1/__lse.ac.uk_ storage_LIBRARY_Secondary_ libfile_shared_repository_ Content_Schang%2C%20L_ Schang_Measuring%20care%20 coordination_2014.pdf RESUMO OBJECTIVE: To analyze the breadth of care coordination by Primary Health Care in three health regions. METHODS: This is a quantitative and qualitative case study. Thirty-one semi-structured interviews with municipal, regional and state managers were carried out, besides a cross-sectional survey with the administration of questionnaires to physicians (74), nurses (127), and a representative sample of users (1,590) of Estratégia Saúde da Família (Family Health Strategy) in three municipal centers of health regions in the state of Bahia. RESULTS: Primary Health Care as first contact of preference faced strong competition from hospital outpatient and emergency services outside the network. Issues related to access to and provision of specialized care were aggravated by dependence on the private sector in the regions, despite progress observed in institutionalizing flows starting out from Primary Health Care. The counter-referral system was deficient and interprofessional communication was scarce, especially concerning services provided by the contracted network. CONCLUSIONS: Coordination capacity is affected both by the fragmentation of the regional network and intrinsic problems in Primary Health Care, which poorly supported in its essential attributes. Although the health regions have common problems, Primary Health Care remains a subject confined to municipal boundaries. DESCRIPTORS: Primary Health Care, organization & administration. Health Services, supply & distribution. Health Services Coverage. Regional Health Planning. Systems Integration. How to cite: Almeida PF, Santos AM. Primary Health Care: care coordinator in regionalized networks? Rev Saude Publica. 2016;50:80. Copyright: This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided that the original author and source are credited. Care coordination in regionalized networks Almeida PF and Santos AM Care coordination in regionalized networks Almeida PF and Santos AM METHODS This is a case study carried out in three municipal centers of the health regions of Feira de Santana, Santo Antônio de Jesus and Vitória da Conquista, in the state of Bahia. It combined qualitative and quantitative approaches – “mixed methods”10, using semi-structured interviews and surveys. The qualitative analysis was performed based on 31 semi-structured interviews with managers at municipal, state and regional level, in 2012. Care coordination in regionalized networks Almeida PF and Santos AM The experiences of professionals were assessed based on self-administered questionnaires with physicians and nurses of Equipes de Saúde da Família (EqSF – Family Health Teams) in activity in October 2012: 84 in Feira de Santana, 21 in Santo Antônio de Jesus and 38 in Vitória da Conquista, with estimated population coverage of 51.5%, 78.7% and 41.2%, respectivelyc. Physicians and nurses were chosen because they are more directly involved in coordinating activities. The instrument used was adapted from a study by Giovanella et al.d (2008). The survey administered 106 questionnaires in Feira de Santana, 31 in Santo Antônio de Jesus, and 64 in Vitória da Conquista, i.e., 201 of the 286 planned questionnaires. Losses (30%) are unevenly distributed between physicians (48%) and nurses (11%), justified not by refusal, but by precarious labor contracts, deteriorated during the period of municipal elections. The experiences of professionals were assessed based on self-administered questionnaires with physicians and nurses of Equipes de Saúde da Família (EqSF – Family Health Teams) in activity in October 2012: 84 in Feira de Santana, 21 in Santo Antônio de Jesus and 38 in Vitória da Conquista, with estimated population coverage of 51.5%, 78.7% and 41.2%, respectivelyc. Physicians and nurses were chosen because they are more directly involved in coordinating activities. The instrument used was adapted from a study by Giovanella et al.d (2008). The survey administered 106 questionnaires in Feira de Santana, 31 in Santo Antônio de Jesus, and 64 in Vitória da Conquista, i.e., 201 of the 286 planned questionnaires. Losses (30%) are unevenly distributed between physicians (48%) and nurses (11%), justified not by refusal, but by precarious labor contracts, deteriorated during the period of municipal elections. A household-based survey was carried out, with a questionnaire administered to a representative sample of families enrolled in EqSF in each municipality. The family respondent was the head of household or spouse, interviewed at home. METHODS The questionnaire for users was also adapted from Giovanella et al.c Calculation of the users sample considered the percentage of population covered by EqSF (p), according to Caderno de Avaliação e Monitoramento da Atenção Básicae (Primary Care Monitoring and Assessment Register), assuming 95% confidence level (z2) (represented by the value 1.96, of the normal distribution abscissa [0.1]) and 4% accuracy (w – π)2. The number of users, assuming a simple random sample, was 596, 430 and 576 for Feira de Santana, Santo Antônio de Jesus and Vitória da Conquista, respectively, for a total sample of 1,602 users. Users were selected by cluster sampling in three stages. The first selected the number of users to be interviewed in each EqSF, dividing the sample of each municipality by the number of existing teams in October 2012. In the second stage, a community health worker was selected for each EqSF by simple random sampling. Finally, the number of families of the selected community health worker was drawn. The questionnaires for professionals and users were administered between November 2012 and February 2013. For users, 1,590 questionnaires were administered. The concept of coordination used in the study was operationalized by means of dimensions and indicators that incorporate components to strengthen the essential attributes of PHC, such as organization as first contact of preference and breadth/resolvability; and integration of Healthcare Networks, such as provision of and access to specialized care and communication between professionals (Table 1). The qualitative results of the semi-structured interviews are presented in narrative synthesis. Data of the different methods were triangulated, seeking to integrate the perspective of managers, professionals and families, in addition to performing a comparative analysis between the cases. Quantitative data were encoded and computed in Epidata, and Epidata Stat was used for descriptive analysis. The project was approved by the Research Ethics Committee of Faculdade Maria Milza (Opinion 323/2011) and authorized by the Municipal Health Department. c Ministério da Saúde (BR), Departamento de Atenção Básica. Histórico de cobertura da Saúde da Família. Brasília (DF); 2015 [cited 2015 nov 5]. Available from: http://dab.saude. gov.br/portaldab/historico_ cobertura_sf.php e Ministério da Saúde (BR), Diretoria de Atenção Básica. Caderno de Avaliação e Monitoramento da Atenção Básica. Brasília (DF); 2012 [cited 2013 Oct 1]. Available from: http://www.saude.ba.gov.br/ dab/index.php?option=com_ content&id=450&Itemid=145 d Giovanella L, Moraes SME, Mendonça MHM. Estudo de caso sobre implementação da Estratégia Saúde da Família em quatro grandes centros urbanos: relatório final. Rio de Janeiro: Fiocruz; 2008. c Ministério da Saúde (BR), Departamento de Atenção Básica. Histórico de cobertura da Saúde da Família. Brasília (DF); 2015 [cited 2015 nov 5]. Available from: http://dab.saude. gov.br/portaldab/historico_ cobertura_sf.php c Ministério da Saúde (BR), Departamento de Atenção Básica. Histórico de cobertura da Saúde da Família. Brasília (DF); 2015 [cited 2015 nov 5]. Available from: http://dab.saude. gov.br/portaldab/historico_ cobertura_sf.php d Giovanella L, Moraes SME, Mendonça MHM. Estudo de caso sobre implementação da Estratégia Saúde da Família em quatro grandes centros urbanos: relatório final. Rio de Janeiro: Fiocruz; 2008. é d úd Care coordination in regionalized networks Almeida PF and Santos AM Table 1. Matrix structure for analysis of care coordination by Primary Health Care in Health Regions. Dimension Indicator Organization of PHC as first contact of preference Awareness of ESF (U) Awareness of and easy access to ESF facility (U) Knowledge of and home visit by CHW (U) Private health insurance coverage (U) USF as regularly sought service and first contact (U/P) Average waiting time for medical appointment at USF (P) Scheduling medical appointments (U) Spontaneous demand care (U/P) Regional strategies to strengthen PHC (M) PHC breadth and resolvability Strategies to increase PHC breadth and resolvability (techincal support, oral health coverage, physical infrastructure and facilities, logistics and communication support) (M) Availability of physicians (M) EqSF professional responsible for care (U) Satisfaction with care (U) Resolution of health problem in EqSF care (U/P) Actions executed when health problem is not resolved (U) Sample collection for lab tests at USF (P) Performance of clinical pathology tests required by EqSF (U) Access to medicines prescribed by EqSF (U/P) Provision of and access to specialized care in Healthcare Network Adequate provision of specialized care in HN (M) Provision of specialized care appointments (U) Forms of access to specialized care appointments referred by EqSF (U/P/M) Scheduling of specialized care appointments and tests and hospitalization by EqSF (P) Performance of specialized lab tests requested by EqSF (U) Forms of access to specialized lab tests (U) Main characteristic of specialized services (M) Regional strategies to provide specialized care (M) Communication between professionals Use of referral and counter-referral instruments (P/M) Computerization of health services (M) Use of clinical protocols (M) Completion of medical records following appointments at USF (P) Follow-up care of users of other health services (U/P) U: users/families; P: professionals (physicians and nurses) of family health teams; M: health managers; ESF: Family Health Strategy; USF: Family Health Unit; CIR: Comissão Intergestores Regional (Regional Inter-Managerial Commission); CHW: Community Health Worker; EqSF: Equipe de Saúde da Família (Family Health Team); HN: Healthcare Networks Table 1. Matrix structure for analysis of care coordination by Primary Health Care in Health Regions. Dimension Indicator Table 1. Matrix structure for analysis of care coordination by Primary Health Care in Health Regions. 4 DOI:10.1590/S1518-8787.2016050006602 RESULTS The results were organized according to dimensions and indicators featured in Table 1, which together represent strategies to achieve better care coordination based on strengthening PHC and integrating Healthcare Networks, based on the results of the cross-sectional study and interviews with managers in the health regions studied. e Ministério da Saúde (BR), Diretoria de Atenção Básica. Caderno de Avaliação e Monitoramento da Atenção Básica. Brasília (DF); 2012 [cited 2013 Oct 1]. Available from: http://www.saude.ba.gov.br/ dab/index.php?option=com_ content&id=450&Itemid=145 The organization of PHC as first contact of preference was one of the dimensions used to analyze how to strengthen coordination. The survey with families showed a high percentage of registered users who were aware of ESF, with the lowest rates in Vitória da Conquista. Many of them possibly did not use Unidade de Saúde da Família (USF – Family Health Unit), DOI:10.1590/S1518-8787.2016050006602 Care coordination in regionalized networks Almeida PF and Santos AM Another reason for users to seek different means of first contact, in the three municipalities, was poor or inexistent reception services, although professionals evaluated that the team incorporated care into meeting spontaneous demand and people seek USF first when they need care (Table 3). The evaluation of users indicates problems in scheduling medical appointments and, especially, in meeting spontaneous demand (Table 2), suggesting problems in the organization of first contact. Overall, the evaluations of professionals were more favorable regarding access conditions to USF, although the average waiting time for an appointment might exceed 15 days, according to 1/3 of professionals in Feira de Santana (Table 3). Although managers reported similar problems in consolidating PHC in the health region, strategies to face barriers and find solutions were still limited to the municipal level, being off the agenda of regional governance initiatives such as Comissão Intergestores Regional Table 2. Primary Health Care as first contact of preference according to users and families registered with Family Health Teams (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Dimension Indicator Organization of PHC as first contact of preference U: users/families; P: professionals (physicians and nurses) of family health teams; M: health managers; ESF: Family Health Strategy; USF: Family Health Unit; CIR: Comissão Intergestores Regional (Regional Inter-Managerial Commission); CHW: Community Health Worker; EqSF: Equipe de Saúde da Família (Family Health Team); HN: Healthcare Networks U: users/families; P: professionals (physicians and nurses) of family health teams; M: health managers; ESF: Family Health Strategy; USF: Family Health Unit; CIR: Comissão Intergestores Regional (Regional Inter-Managerial Commission); CHW: Community Health Worker; EqSF: Equipe de Saúde da Família (Family Health Team); HN: Healthcare Networks especially among teams with a high number of registered families. In the three municipalities, most users claimed to know where the USF facility was located and considered it easily accessible. The community health worker was known to respondents and a significant percentage reported receiving monthly home visits (Table 2). The results also indicated significant search for other service modalities as first contact, even at times and days in which USF was open and in contexts of low private health insurance coverage. In Feira de Santana, polyclinics – emergency services operating 24 hours a day with specialists – strongly competed for first contact (Table 2). According to managers, the operation of polyclinics was quite varied: some of them provided initial care and referred users to USF, while others didn’t. In Santo Antônio de Jesus, a regional hospital offering outpatient care by spontaneous demand, not linked to PHC, was routinely sought as first contact of preference, accounting for first contact services on weekdays, according to 33.0% of users (Table 2). For managers, hospital-centered culture was very strong, with spontaneous demand at the regional hospital consisting of users who firstly visited USF or sometimes didn’t even bother to go there, claiming there were no doctors or materials, either at the municipal center or other municipalities of the health region. In Vitória da Conquista, the excessive number of people under the care of a single EqSF limited the capacity of professionals to meet needs in a resolvable and timely manner. This situation was synthesized by the manager: (...) we see teams that are overcrowded with families. (...). We have been unable to ensure access even to those who seek it, so we have many accesses via emergency care of things that are not urgent.” (Municipal Manager/Vitória da Conquista). Indicator Feira de Santana Santo Antônio de Jesus Vitória da Conquista % % % Users aware of ESF (spontaneous + prompted) 97 (n = 588) 96 (n = 430) 73 (n = 572) Users aware of ESF facility location 99 (n = 572) 96 (n = 412) 95 (n = 420) Easy access to USF facility 97 (n = 565) 95 (n = 396) 97 (n = 399) Users who reported accessing ESF facility on foot 95 (n = 565) 93 (n = 396) 87 (n = 399) Users who are aware of CHW 98 (n = 572) 96 (n = 412) 96 (n = 420) Users who have been visited by CHW 98 (n = 560) 92 (n = 395) 95 (n = 404) Users visited by CHW at least once a month 68 (n = 546) 63 (n = 363) 68 (n = 386) Families with private health insurance 22 (n = 588) 16 (n = 430) 19 (n = 572) Family members covered by private health insurance n = 129 n = 69 n = 106 One family member 46 43 44 Two family members 18 25 21 Three or more family members 17 32 35 Assessment of users regarding scheduling medical appointments* n = 443 n = 308 n = 306 Very good/Good 68 58 54 Very poor/Poor 29 35 42 Assessment of users regarding obtaining medical appointments without previous scheduling – spontaneous demand* n = 443 n = 308 n = 306 Very good/Good 42 39 29 Very poor/Poor 42 50 52 Health service sought for illnesses on weekdays n = 588 n = 430 n = 572 Family Health Unit 43 49 39 Polyclinic 36 0 0 Public hospital outpatient/Emergency service 6 33 37 Clinic/Private hospital or emergency service 5 6 8 Private consultation 2 5 5 Other 8 7 11 Health service sought for illnesses on weekends/holidays/nighttime n = 588 n = 430 n = 572 Family Health Unit 0.3 0 4 Polyclinic 71 0 0 Public hospital outpatient/Emergency service 13 83 68 Clinic/Private hospital or emergency service 6 6 10 Private consultation 1.7 4 4 Other 8 7 14 ESF: Family Health Strategy; USF: Family Health Unit; CHW: Community Health Worker n = total number of respondents * Users who reported having received care at a Family Health Unit in the previous 12 months. Table 2. Primary Health Care as first contact of preference according to users and families registered with Family Health Teams (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Indicator Family Health Unit Polyclinic Public hospital outpatient/Emergency service Clinic/Private hospital or emergency service Private consultation Private consultation Health service sought for illnesses on weekends/holidays/nighttime Health service sought for illnesses on weekends/holidays/nighttime Family Health Unit Polyclinic Public hospital outpatient/Emergency service Clinic/Private hospital or emergency service ESF: Family Health Strategy; USF: Family Health Unit; CHW: Community Health Worker n = total number of respondents * Users who reported having received care at a Family Health Unit in the previous 12 months. ESF: Family Health Strategy; USF: Family Health Unit; CHW: Community Health Worker n = total number of respondents p * Users who reported having received care at a Family Health Unit in the previous 12 months Care coordination in regionalized networks Almeida PF and Santos AM Table 3. Primary Health Care (PHC) organization as first contact of preference, breadth and resolvability, provision of and access to specialized care, and communication between professionals according to physicians and nurses of Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Table 3. Primary Health Care (PHC) organization as first contact of preference, breadth and resolvability, provision of and access to specialized care, and communication between professionals according to physicians and nurses of Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Table 3. Primary Health Care (PHC) organization as first contact of preference, breadth and resolvability, provision of and access to specialized care, and communication between professionals according to physicians and nurses of Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Indicator Indicator Feira de Santanaa Santo Antônio de Jesus b Vitória da Conquista c % % % PHC organization as first contact of preference Completely agree/agree that the population first seeks USF for health care 92 77 91 Completely agree/agree that with ESF the population seeks fewer hospital and specialized services 81 74 84 Completely agree/agree that the population first seeks the hospital or emergency network for health care 19 10 19 Completely agree/agree that USF only provides care with previous scheduling 10 6 6 Completely agree/agree that USF provides care for spontaneous demand 88 97 98 Completely agree/agree that the average waiting time for a medical appointment at USF exceeds 15 days 33 16 19 PHC breadth and resolvability Evaluate as very satisfactory/Satisfactory the capacity of EqSF to solve 80% or more of cases treated at USF 89 77 78 Provision of sample collection for lab tests at USF 9 77 66 Evaluate as very satisfactory/Satisfactory the regular distribution of medicines by the team 69 84 75 Provision of and access to specialized care in Healthcare Network (HN) Most common pathway for patients from care at UBS to referral to specialized care The appointment is scheduled by UBS and the date is later informed to patient 60 48 94 The patient leaves UBS with a scheduled appointment 25 6 0 Patients are always/Most times able to schedule other services Appointment with specialists 75 61 63 Specialized test 76 55 75 Hospitalization 58 48 64 Communication between professionals Report completing medical record after each consultation 94 100 98 Always/Most times provide written information when patients are referred to other services 92 90 83 Always/Most times receive counter-referral following user’s consultation with a specialist 13 10 5 Completely agree/Agree that physicians are able to provide follow-up care for users of other health services 33 42 56 Completely agree/Agree that nurses are able to provide follow-up care for users of other health services 50 61 64 ESF: Family Health Strategy; USF: Family Health Unit; HN: Healthcare Networks; UBS: Unidade Básica de Saúde (Basic Health Unit) a 36 physicians and 70 nurses. b 10 h i i d 21 PHC organization as first contact of preference Managers mentioned they were unable Care coordination in regionalized networks Almeida PF and Santos AM Care coordination in regionalized networks Almeida PF and Santos AM to select professionals with the necessary profile to work in ESF, and frequently “turned a blind eye” to absences and carelessness, with physicians often being “auctioned” among the region’s cities. However, they reported no effort to seek regional solutions to the problem. to select professionals with the necessary profile to work in ESF, and frequently “turned a blind eye” to absences and carelessness, with physicians often being “auctioned” among the region’s cities. However, they reported no effort to seek regional solutions to the problem. Nonetheless, the results of the present study showed that physicians were responsible for the care of most users seeking USF. When asked about satisfaction concerning consultations with physicians, the majority (> 69%) reported being “very satisfied/satisfied” (Table 4). Users in the three municipalities reported positively on having their problem resolved at USF (> 70%) (Table 4), similar to the evaluation by professionals (Table 3). The survey with families indicated that even when they could not access a service or have their problems solved at ESF, in all three cases most people (50% to 62%) sought public services. However, an important percentage of the population sought care from the private sector (23% to 33%) (Table 4). In Feira de Santana, samples for lab tests were not collected at USF, which was confirmed by professionals (Table 3). In turn, polyclinics collected biological samples, which prompted demand for their services. In Santo Antônio de Jesus and Vitória da Conquista, according to managers, test samples were collected at USF, including rural areas. Among families that required lab tests requested by EqSF in the previous 12 months, 76%, 85% and 92%, respectively, in Santo Antônio de Jesus, Feira de Santana and Vitória da Conquista, reported being able to do them; however, between 52% and 67% of that total reported that all tests were done in the public network, with the best service offered in Vitória da Conquista. Between 14% and 26% of users reported doing tests in the private network, and about 1/4 were unable to do the prescribed test, especially in Feira de Santana (Table 4). Among users served by EqSF, 85% to 91% required medicines in the three cases studied. PHC organization as first contact of preference PHC organization as first contact of preference The appointment is scheduled by UBS and the date is later informed to patient The patient leaves UBS with a scheduled appointment Patients are always/Most times able to schedule other services Appointment with specialists Appointment with specialists Specialized test Hospitalization Communication between professionals Report completing medical record after each consultation Always/Most times receive counter-referral following user’s consultation with a specialist Completely agree/Agree that physicians are able to provide follow-up care for users of other health ser ESF: Family Health Strategy; USF: Family Health Unit; HN: Healthcare Networks; UBS: Unidade Básica de Saúde (Basic Health Unit) a 36 physicians and 70 nurses. b 10 physicians and 21 nurses. c 28 physicians and 36 nurses. b 10 physicians and 21 nurses. c 28 physicians and 36 nurses. (Regional Inter-Managerial Commission). This was confirmed in the interviews, in which managers explained they could only account for PHC dynamics in their own municipality. (Regional Inter-Managerial Commission). This was confirmed in the interviews, in which managers explained they could only account for PHC dynamics in their own municipality. PHC breadth and resolvability are important dimensions of care coordination. In Feira de Santana, according to managers, these elements were compromised by factors such as: insufficient number of Núcleos de Apoio à Saúde da Família (NASF – Centers for Family Health Support), lack of oral health teams in all USF, reduced or simplified offer of health actions, few community interventions, low involvement of doctors in collective and administrative actions, and poor logistics and communications support for integration. Interviews indicated scarce and irregular supply of inputs and inadequate physical infrastructure and facilities in many USF. In Vitória da Conquista, managers stressed the participation of professionals in offering technical and pedagogical support to EqSF as an alternative to the physician-centered view. It was emphasized that this is a recent experience in the health region, with insufficient numbers of NASF to meet the demands of EqSF, similar to the reality of Feira de Santana. In the three regions, according to managers, attraction and retention of physicians generated distortions, influencing care quality and resolvability. Care coordination in regionalized networks Almeida PF and Santos AM 7 DOI:10.1590/S1518-8787.2016050006602 PHC organization as first contact of preference Primary Health Care breadth and resolvability according to users and families registered with Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Table 4. Primary Health Care breadth and resolvability according to users and families registered with Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Table 4. Primary Health Care breadth and resolvability according to users and families registered with Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Indicator Feira de Santana Santo Antônio de Jesus Vitória da Conquista % % % EqSF professional responsible for carea n = 541 n = 391 n = 397 Physician 90 83 88 Nurse 8 13 8 Other/Doesn’t know/Did not answer 2 4 4 Satisfaction with care by physician n = 486 n = 323 n = 350 Very satisfied/Satisfied 81 69 83 Fair 15 21 13 Very dissatisfied/Dissatisfied 4 10 3 Doesn’t know/Did not answer 0 0 1 Health problem solved by EqSF carea n = 541 n = 391 n = 397 Yes 83 70 78 No 17 30 22 Action taken when health problem is not solved n = 91 n = 117 n = 87 Referral to public network specialist by EqSF 25 20 29 Direct search for care in public hospitals/emergency services 37 30 26 Direct search for care in private health network 23 33 24 No action taken/No search for health services 6 14 8 Other/Doesn’t know/Did not answer 9 3 13 Performance of clinical pathology tests requested by EqSF professionalb n = 280 n = 141 n = 207 All tests performed in the public network 58 52 67 Some tests performed in the public network 16 26 19 All tests performed in the private network 26 22 14 Access to medicines prescribed by EqSFc n = 461 n = 341 n = 360 Received all medicines 40 43 31 Received some medicines 53 50 64 Did not receive any medicines 7 7 5 n = Total number of respondents a Users who reported having received care at a Family Health Unit. b Families that reported having done clinical pathology tests requested by Family Health Team (EqSF) in the previous 12 months. PHC organization as first contact of preference Most (50% to 64%) reported receiving only a few (Table 4), but a significant percentage, around 40% in Santo Antônio de Jesus and Feira de Santana, received all medicines from SUS. Professionals, especially those working in Santo Antônio de Jesus, evaluated positively the regular distribution of medicines (Table 3). Regarding the provision of and access to specialized care in Healthcare Networks, an essential dimension of coordination, in the three municipal centers of the health regions, managers pointed out that provision of procedures for therapeutic support and financial resources for their expansion were insufficient. Therefore, they had to cope with the short supply, compromising health requirements and generating ethical conflicts between the EqSF, since they had to, among the numerous needs, choose which users had “priority.” Access to more complex technological services, in all three cases, mainly occurred via referral by PHC, according to managers. When the procedure was scheduled, the referral form usually returned to EqSF; therefore, users then returned to USF or were contacted by the community health worker (Tables 3 and 5). In Santo Antônio de Jesus, a significant percentage (29%) contacted the scheduling center directly (Table 5). In the three cases studied, most specialists were professionals with experience in the private sector. In Feira de Santana, there was no specialties center concentrating services, and therefore specialists hired by public notice could work in their private clinics, basic health units or polyclinics. In Santo Antônio de Jesus and Vitória da Conquista, specialized services were hired by public tender, with common non-attendance of providers. Besides the provision of care to the municipal center, services were procured to meet the needs of the health region’s Programação Pactuada e Integrada (Agreed and Integrated Program), which was not always possible based on regulated SUS prices. Often municipalities received a lower quota than agreed on, depending on what they managed to “negotiate” with private providers. For managers, the heavy reliance on the private sector represented an obstacle: “(...) we live in a capitalist world, and must work with SUS, which is a completely Care coordination in regionalized networks Almeida PF and Santos AM Care coordination in regionalized networks Almeida PF and Santos AM Table 4. Primary Health Care breadth and resolvability according to users and families registered with Family Health Team (EqSF). Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. Table 4. PHC organization as first contact of preference Part of users who were aware of ESF (31% to 43%) reported the need for specialized tests in the previous 12 months. Of those who did the tests, most users in Feira de Santana (72%) and Vitória da Conquista (85%) stated that they were scheduled by EqSF. The development of thematic networks (care lines) was seen as an advance in creating regionalized networks and overcoming fragmentation. Thematic networks would be a resource for creating a state plan for regional health care, according to state managers. Regarding communication between professionals in the network, the last dimension of coordination to be analyzed, managers said the system of counter-referral was deficient, further worsened when services were provided by the contracted network. Table 5. Provision of and access to specialized care in Healthcare Networks and communication between professionals according to registered users and families. Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. users and families. Feira de Santana, Santo Antônio de Jesus e Vitória da Conquista, BA, Northeastern Brazil, 2013. PHC organization as first contact of preference Indicator Feira de Santana Santo Antônio de Jesus Vitória da Conquista % % % Reported having seen a specialist in the previous 12 months 51 (n = 572) 51.5 (n = 412) 55 (n = 420) Service providing consultation with specialist n = 292 n = 212 n = 232 Polyclinic 17 17 18 Public hospital outpatient care 15 9 14 Public hospital 12 16 18 Private health service (doctor’s office/hospital) 45 51 44 Other 11 7 6 Specialized appointment referred by EqSF 47 (n = 292) 31 (n = 212) 40 (n = 232) Forms of access to specialized appointment referred by EqSF professional* n = 136 n = 65 n = 92 Scheduled by USF and date and time informed later 64 46 71 Scheduled by USF and date and time informed immediately 4 2 13 Referral and appointment information received directly at Scheduling Center 0 29 4 Referral received, but health service sought independently 7 6 1 Referral received, but health service sought and paid for independently 11 9 3 Other 14 8 8 Reported having needed to do a specialized test requested by EqSF in the previous 12 months 43 (n = 572) 31 (n = 412) 33 (n = 420) Performance of specialized test requested by EqSF 83 (n = 248) 59 (n = 126) 81 (n = 137) Forms of access to specialized test n = 206 n = 74 n = 111 Scheduled by USF and date and time informed later 69 34 75 Scheduled by USF and date and time informed immediately 3 5 10 Referral and appointment information received directly at Scheduling Center 0 24 1 Referral received, but health service sought and paid for independently 16 18 2 Other/Unable to inform 12 19 12 Families with hospitalization in the previous 12 months that received written post-discharge information for EqSF 24 (n = 78) 43 (n = 67) 17.5 (n = 80) Families with hospitalization in the previous 12 months who reported requesting information or visit by a EqSF professional during hospitalization 35 (n = 78) 40 (n = 67) 27.5 (n = 80) Professional responsible for seeking information or visiting patient during hospitalization – CHW 63 (n = 27) 67 (n = 27) 73 (n = 22) EqSF: Equipe de Saúde da Família (Family Health Team); USF: Family Health Unit; HN: Healthcare Networks; CHW: Community Health Worker n = total number of potential respondents. PHC organization as first contact of preference c Families that reported having needed medicines prescribed by Family Health Team (EqSF) in the previous 12 months. n = Total number of respondents a Users who reported having received care at a Family Health Unit. b Families that reported having done clinical pathology tests requested by Family Health Team (EqSF) in the previous 12 months. c Families that reported having needed medicines prescribed by Family Health Team (EqSF) in the previous 12 months. non-capitalist model; making this system work depending on the private sector, that generates conflict” (state manager). non-capitalist model; making this system work depending on the private sector, that generates conflict” (state manager). The purchase of specialized services within the logic of procurement procedures was also mentioned by managers as a factor that undermines the possibility of building an integrated network and care coordination via PHC: “It is fragmented from the start...” (municipal manager/Santo Antônio de Jesus). Queues and long waiting times for specialized care were common situations in all three cases. Requests for tests or procedures without protocols or well-defined criteria intensified problems in the regions. In Feira de Santana, private sector doctors could request “high complexity” tests, which were authorized by the scheduling center, competing with SUS users. Setting up task forces was a strategy to minimize waiting times in the three municipal centers. For managers, expanding services was not sufficient, since there were no specialists available in the inland areas of Bahia. They emphasized that care gaps varied widely in the state, although advances were mentioned with the building of regional hospitals. The difficulties in providing specialized care are reflected in the assessment of users. In all three cases, about half of those who were aware of ESF reported having needed to consult a Care coordination in regionalized networks Almeida PF and Santos AM specialist in the previous 12 months. Of those, 44% to 51% sought treatment in the private sector, commonly without referral by EqSF. In Feira de Santana and Santo Antônio de Jesus, 11% and 9% of users, respectively, sought private care, even with referral by EqSF (Table 5). About 2/3 of professionals reported they were able to schedule appointments with specialists “always or most times.” In all three cases, greater difficulties were observed in relation to scheduling hospitalization, according to professionals (Table 3). Only Vitória da Conquista had its own central service to regulate hospital beds and admissions. DISCUSSION The four dimensions investigated point to limits that endanger regional care coordinated by PHC. In this same perspective, Fausto et al.8 reaffirm that coordination by PHC depends on timely and resolvable first contact, attention to and care of spontaneous demand, comprehensive provision of services, and technical support. In the cases studied, the role of first contact of preference faces strong competition from hospital outpatient and emergency care services, outside the network, often devised to meet the demand for specialized care, such as regional hospitals. This situation is aggravated by problems of work organization in PHC, with disparity between programmed actions and spontaneous demand, barriers also identified by other studies2,8,9, and excessive number of users per EqSF. There is need to expand clinical nursing actions, Centers for Family Health Support, and oral health as measures that could help people recognize that EqSF is not restricted to physicians. The study also points to the need to strengthen PHC by the regular provision of medicines and other inputs. Despite the constraints, PHC, in the perception of users and professionals, seems to have good resolvability, which reaffirms the need for investments to enhance it. The indistinct search for public services as first contact, whether USF, polyclinics or hospitals, demonstrates the need to organize and coordinate services offered by the network, so that access via PHC is preferred. In all three cases, users expand their options as services are increasingly fragmented and coordination is dispersed. Cecilio et al.7 show that although users value USF, they multiply their possibilities by combining expected PHC resources with other services on the network. When coordination is extrapolated to health regions, the role of EqSF dissipates, for several reasons. Provision of specialized services is highly dependent on the private sector, whose public insufficiency and underfunding, combined with occasional inadequate use, can be identified as factors that hinder the establishment of regional networks. In addition, the selected health regions have historical care deficit, difficulty in attracting and offering certain specialties, physicians disinterested in the civil service (due to precarious labor relations in SUS), and high bargaining power of some specialties. Such problems go beyond the management capacity of isolated municipalities, requiring a shared stance15. Although access to specialties via PHC in the health region has been formally organized, the incorporation of specialists by procurement of procedures accentuates fragmentation and weakens the mechanisms of care coordination and regulation. PHC organization as first contact of preference * Users referred by EqSF who reported having seen a specialist. Care coordination in regionalized networks Almeida PF and Santos AM Counter-referral only occurred at the request or demand of the actual user. Referral was an instrument for scheduling procedures, and not of interprofessional communication. The survey with professionals showed that only 5% to 13% reported receiving counter-referrals frequently (Table 3). Computerization of USF was partial and, if any, limited to scheduling appointments, and never used as a communication flow instrument. In Feira de Santana, the existence of electronic medical records was reported in some units, although they were not shared with the other network services. According to managers, implementation of clinical protocols was incipient. In Santo Antônio de Jesus, about 43% of households reported receiving post-discharge information, and 40% reported requesting information or visits by EqSF professionals during hospitalization, especially community health workers (Table 5). For professionals, there were difficulties in follow-up care for users of services outside PHC (Table 3). 1 DOI:10.1590/S1518-8787.2016050006602 DISCUSSION While large municipalities prioritize the establishment of local specialized services1, in the context of health regions there seems to be a pressing need for other ways of incorporating therapeutic support, in a perspective of intercity networks. In this sense, initiatives such as Comissão Intergestores Regional (Regional Inter-Managerial Commission) need to be strengthened and recognized as 10 DOI:10.1590/S1518-8787.2016050006602 Care coordination in regionalized networks Almeida PF and Santos AM means of joint regional governance, able to overcome the municipal logic, which is ineffective to build regional networks for comprehensive care. To this end, Regional Intermanagers Commission (CIR) should bring together strategies for managers to establish joint health agreements to enhance the provision of public services, overcoming the fragile contract mechanisms with the private sector, since there is no proper monitoring of agreed goals among private providers, other than post factum auditing. In general terms, there is a relationship of mutual dependence between the public and private sectors; however, private contractors providing services to SUS have gained the upper hand, defining market prices for procedures and services to the detriment of public interest. The need for communication between professionals and providers to achieve better care coordination is a consensus5,13,a. Coordination actions will likely fail under the sole responsibility of physicians4. In this study, for example, the community health worker was responsible for seeking information and visiting patients during hospitalization, indicating potential ways to maximize action horizontal coordination. Unforeseen findings of this study, which nevertheless are relevant to the analyzed subject, include the absence of physicians at USF during data collection, carried out before and during the municipal elections. Physicians accounted for a greater percentage of losses, especially because of the absence or sporadic presence of these professionals at USF. Even though the election campaign may have generated instability and increased the precariousness of labor relations, especially in Feira de Santana and Santo Antônio de Jesus, turnover was higher among physicians, especially due to their greater employability and possibility of establishing new links in other municipalities. The losses are revealing of a concrete situation also experienced by users, influencing difficulty of access and timely care, leading to the search for emergency care services and, ultimately, thwarting care coordination by PHC. DISCUSSION The constraints of PHC coordination are varied, comprising a set of services that do not compose an integrated network with a view to coordinating subjects, knowledge and practices3, and lack of strong PHC essential attributes, requiring initiatives that go beyond municipal boundaries to organize the health care network within health regions. Nevertheless, it is clear that the success of the health regions studied requires the fulfillment of responsibilities among managers of different entities who, despite the normative expectations of regionalization, are unable to agree on a plan capable of taking the health territory beyond a bureaucratic and programmatic vision. The three regions indicate that, regarding care coordination, there is urgent need for expansion and qualification of first contact services via PHC due to persisting problems of fragmentation and disorderly search for services provided without adequate care regulation. The health regions indicate a shift from decentralization and a path to enable care integration16. However, this study shows that coordination, when it occurs, is limited to the municipal centers of the health regions, i.e., as the vast majority of municipalities need services offered by other entities, care coordination via PHC becomes unfeasible, losing itself in the bureaucratic flow of scheduling centers. Thus, even if a wide offer of services is achieved in a given territory, coordination is an essential attribute to enable continuous care and integrated services, requiring, more than ever, a strong PHC base. 1. Almeida PF, Giovanella L, Mendonça MHM, Escorel S. Desafios à coordenação dos cuidados em saúde: estratégias de integração entre níveis assistenciais em grandes centros urbanos. Cad Saude Publica. 2010;26(2):286-98. DOI:10.1590/S0102-311X2010000200008 2. Almeida PF, Giovanella L, Nunan BA. Coordenação dos cuidados em saúde pela atenção primária à saúde e suas implicações para a satisfação dos usuários. Saude Debate. 2012;36(94):375-91. DOI:10.1590/S0103-11042012000300010 2. Almeida PF, Giovanella L, Nunan BA. Coordenação dos cuidados em saúde pela atenção primária à saúde e suas implicações para a satisfação dos usuários. Saude Debate. 2012;36(94):375-91. DOI:10.1590/S0103-11042012000300010 REFERENCES 1. Almeida PF, Giovanella L, Mendonça MHM, Escorel S. Desafios à coordenação dos cuidados em saúde: estratégias de integração entre níveis assistenciais em grandes centros urbanos. Cad Saude Publica. 2010;26(2):286-98. DOI:10.1590/S0102-311X2010000200008 2. Almeida PF, Giovanella L, Nunan BA. Coordenação dos cuidados em saúde pela atenção primária à saúde e suas implicações para a satisfação dos usuários. Saude Debate. 2012;36(94):375-91. DOI:10.1590/S0103-11042012000300010 Care coordination in regionalized networks Almeida PF and Santos AM 3. Ayres JRCM. Cuidado: trabalho e interação nas práticas de saúde. Rio de Janeiro: Abrasco; 2009 [citado 2015 abr 9]. (Coleção Clássicos para Integralidade em Saúde). Disponível em: http://www.cepesc.org.br/wp-content/uploads/2013/08/miolo-livro-ricardo.pdf 3. Ayres JRCM. Cuidado: trabalho e interação nas práticas de saúde. Rio de Janeiro: Abrasco; 2009 [citado 2015 abr 9]. (Coleção Clássicos para Integralidade em Saúde). Disponível em: http://www.cepesc.org.br/wp-content/uploads/2013/08/miolo-livro-ricardo.pdf 4. Bodenheimer T. Coordinating care: a perilous journey through the health care system. N Engl J Med. 2008;358(10):1064-71. DOI:10.1056/NEJMhpr0706165 5. Boerma WGW. Coordenação e integração em atenção primária europeia. In: Saltman RB, Rico A, Boerma WGW, organizadores. Atenção Primária conduzindo as redes de atenção à saúde: reforma organizacional na atenção primária europeia. Berkshire: Open University Press; 2010. p.25-47. 6. Bynum JPW, Ross JS. A measure of care coordination? J Gen Intern Med. 2012;28(3):336-8. DOI:10.1007/s11606-012-2269-0 7. Cecilio LCO, Andreazza R, Carapinheiro G, Araújo EC, Oliveira LA, Andrade MGG, et al. A Atenção Básica à Saúde e a construção das redes temáticas de saúde: qual pode ser o seu papel? Cienc Saude Coletiva. 2012;17(11):2893-902. DOI:10.1590/S1413-81232012001100006 8. Fausto MCR, Giovanella L, Mendonça MHM, Seidl H, Gagno J. A posição da Estratégia Saúde da Família na rede de atenção à saúde na perspectiva das equipes e usuários participantes do PMAQ-AB. Saude Debate. 2014;38 N. Espec:13-33. DOI:10.5935/0103-1104.2014S003 9. Giovanella L, Mendonça MHM, Almeida PF, Escorel S, Senna MCM, Fausto MCR, et al. Saúde da família: limites e possibilidades para uma abordagem integral de atenção primária à saúde no Brasil. Cienc Saude Coletiva. 2009;14(3):783-94. DOI:10.1590/S1413-81232009000300014 10. Greene JC. Mixed methods in social inquiry. San Francisco: John Wiley; 2007. 11. Hofmarcher MM, Oxley H, Rusticelli E. Improved health system performance through better care coordination. Paris: OECD; 2007. (OECD Health Working Papers, No. 30). 12. Kringos DS, Boerma WGW, Hutchinson A, Zee J, Groenewegen PP. The breadth of primary care: a systematic literature review of its core dimensions. BMC Health Serv Res. 2010;10:65. DOI:10.1186/1472-6963-10-65 13. McDonald KM, Schultz E, Albin L, Pineda N, Lonhart J, Sundaram V, et al. Care coordination in regionalized networks Almeida PF and Santos AM REFERENCES Care Coordination Atlas Version 4. Rockville, MD: Agency for Healthcare Research and Quality; 2014. (AHRQ Publication, N.14-0037-EF). 14. Nolte E, McKee M, editors. Caring for people with chronic conditions: a health system perspective. Berkshire: Open University Press; 2008. Integration and chronic care: a review; p. 64-91. 15. Santos AM, Giovanella L. Governança regional: estratégias e disputas para gestão em saúde. Rev Saude Publica. 2014;48(4):622-31. DOI:10.1590/S0034-8910.2014048005045 16. Santos L, Campos GWS. SUS Brasil: a região de saúde como caminho. Saude Soc. 2015;24(2):438-46. DOI:10.1590/S0104-12902015000200004 17. Terraza Núñez R, Vargas Lorenzo I, Vásquez Navarrete ML. La coordinación entre niveles asistenciales: una sistematización de sus instrumentos y medidas. Gac Sanit. 2006;20(6):485-95. Funding: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Edital Universal 14/2011 (Process 479451/2011-7). Authors’ Contributions: PFA and AMS participated equally in design, manuscript draft, data interpretation, and critical review. Conflict of Interest: The authors declare no conflict of interest. Conflict of Interest: The authors declare no conflict of interest.
https://openalex.org/W4213215355	https://www.nature.com/articles/s43705-022-00098-4.pdf	English	null	Microbial communities across activated sludge plants show recurring species-level seasonal patterns	ISME communications	2,022	cc-by	46,364	ARTICLE OPEN Microbial communities across activated sludge plants show recurring species-level seasonal patterns Miriam Peces 1, Giulia Dottorini1, Marta Nierychlo1, Kasper Skytte Andersen 1, Morten Kam Dahl Dueholm 1 and Per Halkjær Nielsen 1✉ Miriam Peces 1, Giulia Dottorini1, Marta Nierychlo1, Kasper Skytte Andersen 1, Morten Kam Dahl Dueholm 1 and Per Halkjær Nielsen 1✉ © The Author(s) 2022 Microbial communities in activated sludge (AS) are the core of sanitation in wastewater treatment plants (WWTPs). Microbial communities in AS have shown seasonal changes, however, long-term experiments (>2 years) are rarely conducted, limiting our understanding of the true seasonal dynamics in WWTPs. In this study, we resolved the microbial seasonal dynamics at the species level in four municipal full-scale WWTPs, sampled every 7–10 days, during 3–5 consecutive years. By applying a new time-series analysis approach, we revealed that the seasonal pattern was species-speciﬁc, where species belonging to the same functional guild or genus may show different seasonal dynamics. Species could be grouped into cohorts according to their seasonal patterns, where seasonal cohorts showed repeatable annual dynamics across years and plants. Species were also grouped according to their net growth rate in the AS (i.e., growing species and disappearing species). Growing species were more prevailing in spring and autumn cohorts, while disappearing species, which were only present due to the continuous immigration from inﬂuent wastewater, were mostly associated with winter and spring cohorts. Most known process-critical species, such as nitriﬁers, polyphosphate accumulating organisms and ﬁlamentous organisms, showed distinct species-speciﬁc patterns. Overall, our study showed that overarching seasonal patterns affected microbial species in full-scale AS plants, with similar seasonal patterns across plants for many dominant species. These recurrent seasonal variations should be taken into account in the operation, understanding and management of the WWTPs. ISME Communications; https://doi.org/10.1038/s43705-022-00098-4 ISME Communications; https://doi.org/10.1038/s43705-022-00098-4 ISME Communications; https://doi.org/10.1038/s43705-022-00098-4 The degree to which the seasonal variations of the microbial communities in AS are cyclic is not well documented. Most longitudinal studies have been carried out only during one year, so it is unknown whether the seasonal variation is repeatable over the years, indicating a seasonal periodicity, or if the changes only occurred over a few months. In a 2-year survey, Flowers et al. [8] observed a repeatable seasonal pattern in microbial richness and diversity, whereas in a 5-year survey by Ju and Zhang [14], the overall community composition appeared seasonally indepen- dent. www.nature.com/ismecomms ARTICLE OPEN Microbial communities across activated sludge plants show recurring species-level seasonal patterns Although two studies are not enough to determine if seasonal periodicity is WWTP-speciﬁc, a key observation from these studies is the presence of different dynamic responses for individual taxa. For example, in Ju and Zhang [14] a Tetrasphaera- related operational taxonomic unit (OTU), clustered at 97% sequence identity, had a clear seasonal periodicity, while a Nitrosomonas-related OTU showed random ﬂuctuations. Addition- ally, 1-year surveys have also shown the presence, and absence, of temporal variations for process-critical taxa [9, 10, 15, 16]. How- ever, it is currently not known if some overarching factors are driving these variations and whether they are taxa-speciﬁc and/or WWTP-speciﬁc. Received: 3 September 2021 Revised: 26 January 2022 Accepted: 28 January 2022 Received: 3 September 2021 Revised: 26 January 2022 Accepted: 28 January 2022 1Department of Chemistry and Bioscience, Section of Biotechnology, Center for Microbial Communities, Aalborg University, Aalborg Ea Amplicon sequencing l d l Detailed sample preparation, DNA extraction and puriﬁcation protocol used can be found in the MiDAS ﬁeld guide (https://www.midasﬁeldguide. org/guide/protocols). Concisely, 160 μL of a homogeneous sample was used for DNA extraction with the FastDNA® spin kit for soil (MP biomedicals) and FastPrep-96 bead beater (MP Biomedicals) following the manufacturer’s protocol with minor modiﬁcations in the bead beating intensity and puriﬁcation. The V1–V3 region of 16S rRNA gene was ampliﬁed using the 27F (3′-AGAGTTTGATCCTGGCTCAG-5′) [24] and 534R (3′-ATTACCGCGGCTGCTGG-5′) [25] primers, as this primer set has shown to give the most representative community structure and the highest taxonomic resolution for bacteria in AS systems [17, 26]. Amplicon sequencing was conducted using the Illumina MiSeq platform (Illumina, USA) as described in Dottorini et al. [4]. Resolving species dynamics is important since many metabolic and functional traits are only conserved at the highest taxonomic resolution, which allows assigning known and putative functional roles to individual microorganisms [17–19]. A recent example to illustrate the importance of species-level resolution is the putative foam-forming genus Candidatus Microthrix [20]. In Danish WWTPs, Ca. M. parvicella and the novel Ca. M. subdominans (previous MiDAS 3 placeholder species name midas_s_2) are the two main coexisting species, and they show very different dynamics. Both species had substantial effects on the sludge settling properties, but Ca. M. parvicella showed a strong seasonality proliferating in the coldest months, while Ca. M. subdominans did not show any seasonal pattern [20]. Additionally, the combination of reprodu- cible species-level classiﬁcation and biomass mass-balances including immigration from inﬂuent wastewater, allows the grouping of species according to their net growth rate in the AS [4, 21]. Brieﬂy, ‘growing species’ are expected to grow in the AS and perform some critical process functions, while ‘disappearing species’ are expected to die-off in the AS and are only present because they are constantly transported with the inﬂuent waste- water [4]. The study of these growth groups separately can improve our understanding of the microbial community assembly since the growing species may be very dependent on temperature and process operation, while the disappearing species mainly depend on the immigration from the sewer system. Amplicon sequencing data were processed with AmpProc v.5.1.0 for downstream analyses (https://github.com/eyashiro/AmpProc). Brieﬂy, only forward sequencing reads were processed using usearch v.11.0.667 [27]. INTRODUCTION Microbial communities in activated sludge (AS) are the core of wastewater treatment plants (WWTPs) worldwide, where organic pollutants and nutrients are transformed by the action of diverse microbial groups to produce clean water, and in more advanced conﬁgurations, recover resources such as phosphorus. The under- standing of the microbial communities in WWTPs is being continuously resolved and improving [1], and a good understanding of the factors determining AS community assembly and dynamics is important for informed management of the WWTPs. The microbial community assembly is determined by a variety of factors, with the relative importance of different factors varying across WWTPs with different process design and operation. The controlling factors can be deterministic, such as environmental factors (e.g., wastewater temperature, the chemical composition of the inﬂuent wastewater), process design, operation (e.g., solid retention time (SRT), aeration time, chemicals dosage [2, 3]), and neutral, such as dispersal (i.e., microbial immigration from inﬂuent wastewater) [4, 5]. Given the multiple factors affecting the AS, it is clear that the composition and assembly of the microbial community are exposed to a variety of temporal responses that can alter its dynamics. Indeed, the few published longitudinal AS studies with good temporal resolution (e.g., ≥1 sample/month) have shown seasonal variations in community composition and abundance over the calendar year in different climate zones [6–13]. p A great challenge interpreting seasonal variation in previous studies was the lack of species-level taxonomic classiﬁcation. Previous studies aggregated taxa (usually resolved as 16S rRNA gene OTUs) in genera or families, or into functional guilds such as nce, Section of Biotechnology, Center for Microbial Communities, Aalborg University, Aalborg East 9220, Denmark. ✉email: phn@bio.aau. M. Peces et al. 2 nitriﬁers, polyphosphate accumulating organisms (PAOs) or ﬁlamentous organisms (hereafter referred to as ﬁlaments). This is problematic since not all species in the same guild may share the same ecophysiological traits. Consequently, grouping species could mask any species-level dynamics. This problem can be solved by using ecosystem-speciﬁc reference databases which can resolve the taxonomy to species level (>98.7% sequence identity) [17, 18]. For WWTPs and anaerobic digesters, the improved ecosystem-speciﬁc reference database MiDAS 4 provides repro- ducible species-level classiﬁcations based on a comprehensive set of amplicon sequence variant (ASV) resolved full-length 16S rRNA gene reference sequences. Moreover, it provides placeholder names for unclassiﬁed environmental taxa, providing a common taxonomy for all studies in the ﬁeld [1]. Data analysis D Downstream analyses and visualisation. Downstream statistical analyses and visualisation were performed in R v.4.0.3 [31] mainly using the following packages: tidyverse v.1.3.0 [32], ampvis2 v.2.6.1 [33] vegan v.2.5.6 [34], ComplexUpset v.1.1.0 [35], ggseas v.0.5.4 [36], and Harmonic Regression v.1.0 [37]. Prior to data analysis, samples with less than 10 000 reads were discarded and duplicate samples from the same sampling point were combined by averaging the relative abundance of each ASV. The total reads per sample ranged from 10 893 to 170 639 (Fig. S1). For alpha-diversity analyses, samples were rareﬁed at 10 000 reads and the number of unique ASVs and Simpson index were calculated using ampvis2. INTRODUCTION This approach has shown that many dominant species are shared among WWTPs with similar process conﬁguration [18]. Therefore, by comparing several WWTPs, it may be possible to ﬁnd species-speciﬁc recurrent patterns. nitriﬁers, polyphosphate accumulating organisms (PAOs) or ﬁlamentous organisms (hereafter referred to as ﬁlaments). This is problematic since not all species in the same guild may share the same ecophysiological traits. Consequently, grouping species could mask any species-level dynamics. This problem can be solved by using ecosystem-speciﬁc reference databases which can resolve the taxonomy to species level (>98.7% sequence identity) [17, 18]. For WWTPs and anaerobic digesters, the improved ecosystem-speciﬁc reference database MiDAS 4 provides repro- ducible species-level classiﬁcations based on a comprehensive set of amplicon sequence variant (ASV) resolved full-length 16S rRNA gene reference sequences. Moreover, it provides placeholder names for unclassiﬁed environmental taxa, providing a common taxonomy for all studies in the ﬁeld [1]. This approach has shown that many dominant species are shared among WWTPs with similar process conﬁguration [18]. Therefore, by comparing several WWTPs, it may be possible to ﬁnd species-speciﬁc recurrent patterns. 10–30 days and a yearly temperature range of 7–20 °C. The plants had minor differences in design conﬁguration, operational conditions, and inﬂuent composition (mostly municipal sewage but some discharge from nearby industries) that made each plant unique from an operational perspective (Supplementary Table S1). 10–30 days and a yearly temperature range of 7–20 °C. The plants had minor differences in design conﬁguration, operational conditions, and inﬂuent composition (mostly municipal sewage but some discharge from nearby industries) that made each plant unique from an operational perspective (Supplementary Table S1). AS samples were collected from the aeration tanks, or at the end of the aeration phase for plants with alternating operation, every 7–10 days between 2015 and 2020 (~1000 samples, Table S1). Routine monitoring at Damhusåen (line B) started in 2017. Brieﬂy, 500 mL of AS were collected, homogenised and subsampled in 2 mL cryotubes as detailed in the MiDAS ﬁeld guide (https://www.midasﬁeldguide.org/guide/protocols). The sam- ples were immediately stored at −18 °C at the WWTPs and shipped frozen to our lab in batches. Samples were stored at −18 °C until further processing. Amplicon sequencing l d l Raw fastq ﬁles were ﬁltered for phiX sequences using usearch -ﬁlter_- phix, trimmed to 250 bp using usearch -fastx_truncate –trunclen 250, and quality ﬁltered using usearch -fastq_ﬁlter with -fastq_- maxee 1.0. The sequences were dereplicated using usearch -fastx_uniques with -sizeout. ASVs were generated using -unoise3 [28] with standard settings. The ASVs were mapped to the full-length ASVs (FL-ASVs) of the MiDAS 4 wastewater ecosystem-speciﬁc reference database, allowing species resolution [1, 17] (available at https://www.midasﬁeldguide.org/guide/downloads). Taxon- omy was assigned using the SINTAX classiﬁer with a conﬁdence threshold of 0.8 [29]. In the MiDAS 4, taxonomic names are based on reproducible clustering with rank-speciﬁc identity thresholds and placeholder names [1]. Species-level classiﬁcation (>98.7% sequence identity as recommended by Yarza et al. [30]) was chosen to improve the read counts available for each investigated taxon and to provide a reproducible name recognisable across studies. For ASVs without species-level classiﬁcation, the taxonomy was assigned at the lowest available taxonomic level (e.g., genus), and these ASVs were treated as separate species. Microbial species were assigned to known functional guilds (e.g., nitriﬁers, PAO) based on the main in situ metabolism described to occur in AS (https://www. midasﬁeldguide.org/guide/search). In this study, we investigated the seasonal periodicity of all species in four full-scale nutrient removal WWTPs in temperate climate during a longitudinal survey of 3–5 consecutive years. The aims were to evaluate (i) if the microbial community structure could be seasonally described, (ii) which species showed signiﬁcant seasonal variations and if their dynamics were the same across different WWTPs, (iii) if similar seasonal patterns were observed for species within genera and functional guilds, and (iv) if growing species showed similar temporal dynamics to disappearing species, since the latter are present only due to transportation by the inﬂuent wastewater. compositional and sparse nature of the data as in Martino et al. [38]: In this study, we assigned the species according to their growth fate as identiﬁed by Dottorini et al. [4] since the WWTPs in this study were present among the WWTPs studied by Dottorini et al. [4], except for Damhusåen. However, Damhusåen WWTP has very similar taxa, process design and geographical location to our previous dataset, therefore it can be assumed that species will follow the same fate in AS. Species in WWTPs with different process design may not follow the same growth fate. The detailed mass- balance methodology can be found in Dottorini et al. [4]. Brieﬂy, paired samples from inﬂuent wastewater and AS were collected from 11 municipal WWTPs across Denmark every second week for months. The net growth rate of bacterial species in the AS process was estimated based on a mass- balance between paired inﬂuent and AS samples, knowing the ﬂow-rates of the plant and assuming: (i) steady-state process, (ii) that the apparent net growth rate (k) can be described as a ﬁrst-order process and (iii) that the biomass concentration of a species can be described by the relative abundance of that species multiplied by the total number of cells. According to the mass-balance, species can be assigned to three different groups: rclr xi ð Þ ¼ log g xi ð Þ geometric mean x ð Þ (1 rclr xi ð Þ ¼ log g xi ð Þ geometric mean x ð Þ (1) geometric mean x ð Þ ¼ YN i¼1 gðxiÞ 1 N (2) (1) geometric mean x ð Þ ¼ YN i¼1 gðxiÞ 1 N (2) (2) The rclr transformation is based on the log-ratio transformation introduced by Aitchison [39] where the rclr is the logarithm after dividing the number of reads of each species (g(xi)) by the geometric mean of the total sample reads (geometric mean(x)) for taxa (N) with a read abundance >0. Time-series decomposition and grouping in seasonal cohorts. For each WWTP, only species with a relative abundance >0.05% in at least one sample were retained to ﬁnd recurrent seasonal variations. To decrease the dimensionality inherently contained in time-series data [40], each species time-series was decomposed into trend, season and residual components based on local smoothing regression [41]. The time-series components were extracted using the function stl() implemented in the ggseas R package with an s.window = “periodic” and frequency = 52. Microbial composition and time-series dynamics of growth groups in full-scale WWTPs Species were classiﬁed according to their growth group (i.e., growing, disappearing, surviving or ambiguous) to evaluate: (i) the relative contribution of each growth group to the total read abundance, (ii) the identity and relative read abundance of species in each growth group, and (iii) their temporal dynamics in each WWTP. The growing fraction was dominant in all plants, representing about 60–70% of the total relative read abundance, and consisting of about 500 different species. The disappearing group contributed 8–15% of the relative read abundance in each WWTP, consisting of about 150 species (Fig. 1A). This group was only present due to the continuous immigration with the inﬂuent wastewater. Very few surviving species were sparsely observed in low abundance during the studied period. The ambiguous fraction was large, it harboured 1300–1600 species found in very low relative abundance, corre- sponding to a cumulative total read abundance of 20–25% in each WWTP (Fig. 1A). The growth groups showed a distinct dynamic response over time (Fig. 1B). The growing fraction showed signiﬁcant yearly seasonal dynamics only in one plant (Randers), whereas the disappearing and ambiguous fractions showed consistent yearly seasonal dynamics in all plants with maximum abundance around February–March and September–October, respectively. Species were assigned into seasonal cohorts depending on the temporal location of the maximum seasonal peak. The seasonal cohorts were based on the yearly process tank temperature variation (Fig. S3), using as a reference the deﬁnition of astronomical seasons for the northern hemisphere: ● Winter cohort: Species that have the maximum peak located between the 21st December and 20th March. Process tank temperature is about 10–12 °C. ● Spring cohort: Species that have the maximum peak located between the 21st March and the 21st June. Process tank temperature increases from about 11 to approximately 17 °C. pp y ● Summer cohort: Species that have the maximum peak located between the 22nd June and the 22nd September. Process tank temperature is about 17–19 °C. p ● Autumn cohort: Species that have the maximum peak located between the 23rd September and the 22nd December. Process tank temperature decreases from about 17 to approximately 12 °C. p pp y ● Non-signiﬁcant cohort: Species with a non-signiﬁcant harmonic model ﬁt (p > 0.01). The strength of the seasonal component was calculated based on the variance explained by the seasonal component over the residual component (Eq. (4)) [42]. compositional and sparse nature of the data as in Martino et al. [38]: Before time-series decomposi- tion, species reads were rclr transformed (section “Downstream analyses and visualisation”). When required, linear interpolation between dates was used to create an even weekly sampling distribution across years. The extracted seasonal component was used for further data processing and analyses. Each species seasonal component was ﬁtted to a simple harmonic model (Eq. (3)) to determine the statistical signiﬁcance of the seasonal response (1% signiﬁcance threshold, p < 0.01): ● Growing, where the apparent net growth rate is positive (k > 0). ● Disappearing, where the apparent net growth rate is negative (k < 0). ● Surviving, where the apparent net growth rate is close to zero (k ⋍0) and they may slowly grow or disappear depending on the process conditions. Species that could not be unequivocally assigned to any of the growth groups across WWTPs or species with a relative read abundance below 0.05% were assigned to the ambiguous group. It was considered that the low number of reads for species with a relative read abundance below 0.05% contains too much uncertainty to be thoroughly classiﬁed. The detailed methodology can be found in Dottorini et al. [4]. y ¼ m þ A cos ω t 2π φ (3) (3) where m is the mean value of the seasonal component, A is the amplitude of the oscillation, ω is the frequency of the oscillation, t is time (in days) and φ is the phase of the oscillation. A visual summary of this approach can be found in Supplementary Fig. S2. MATERIALS AND METHODS Wastewater treatment plant characterisation and sample collection This longitudinal survey was conducted between 2015 and 2020 in four full-scale municipal WWTPs operated as conventional AS with nitrogen and enhanced biological phosphorus removal (EBPR) conﬁguration. The four plants, Aalborg W, Aalborg E, Damhusåen and Randers, ran without major disturbances and operational changes during the sampling period. Only sporadic bulking events were reported by plant operators, with efﬂuent concentrations consistently complying with the Danish efﬂuent discharge limits (BOD5 < 15 mg/L, total nitrogen <8 mg/L, and total phosphorus < 1.5 mg/L) [22, 23]. The four plants are medium size municipal WWTPs (130 000 to 350 000 PE) with average SRTs of Differences in overall microbial community structure were explored by principal component analysis (PCA), where the ASV reads were Hellinger transformed prior to ordination using ampvis2. Statistical differences between PCA clusters were assessed by PERMANOVA using the adonis2 function in the vegan R package. The relative abundance of species was visualised with boxplots using the mean relative abundance for each plant. Prior to time-series analyses (described in “Time-series decomposition and grouping in seasonal cohorts”) sequencing reads were transformed using robust-centred log-ratio (rclr) transformation to account for the Differences in overall microbial community structure were explored by principal component analysis (PCA), where the ASV reads were Hellinger transformed prior to ordination using ampvis2. Statistical differences between PCA clusters were assessed by PERMANOVA using the adonis2 function in the vegan R package. The relative abundance of species was visualised with boxplots using the mean relative abundance for each plant. Prior to time-series analyses (described in “Time-series decomposition and grouping in seasonal cohorts”) sequencing reads were transformed using robust-centred log-ratio (rclr) transformation to account for the ISME Communications M. Peces et al. 3 compositional and sparse nature of the data as in Martino et al. [38]: compositional and sparse nature of the data as in Martino et al. [38]: Microbial composition and time-series dynamics of growth groups in full-scale WWTPs Within the most abundant bacteria in the growing fraction (Fig. 2), we found species and genera typical for Danish and global EBPR plants [1, 18] such as ﬁlamentous Ca. Microthrix and Ca. Amarolinea, the PAO Tetrasphaera and Dechloromonas, and other genera with unknown or poorly described in situ functions such as Rhodobacter, Rhodoferax, OLB8 and midas_g_17 (both family Saprospiraceae). Nitriﬁers were mostly represented by Nitrosomo- nas (midas_s_139 and midas_s_717), Nitrotoga (midas_s_181 and ASV223), and Nitrospira deﬂuvii, ranking within the top 100 most abundant growing species. Additionally, the most abundant growing bacteria were more evenly distributed than the disappearing species, which were dominated by Trichoccocus midas_s_4. The most abundant species of the ambiguous and surviving fraction are shown in Supplementary Fig. S5. varianceðresidual componentÞ The seasonal strength was independent of the growth group or functional guild (Fig. S9). The classiﬁcation of species into seasonal cohorts (i.e., species that peak in speciﬁc seasons) showed that the growing fraction contained a high number of species belonging to summer and autumn cohorts (Fig. S10). This is consistent with the dynamics of the estimated number of ASVs along the year. In contrast, the distribution of the cohorts in terms of relative read abundance was even between spring, summer, autumn, and non- signiﬁcant cohorts (Fig. 4A). This explains why the growing fraction rarely showed a signiﬁcant yearly variation, since aggregating species with different seasonality, but evenly distributed abundances can balance each other. The difference between WWTPs for the disappearing fraction (Fig. 4B) was related to the classiﬁcation of the most abundant disappearing species, i.e., Trichoccocus midas_s_4, where the estimated peak abundance was found near the winter and spring split (Fig. 5). The time-series of seasonal cohorts of ambiguous and surviving species are shown in Supplementary Fig. S11. The growing communities in the four WWTPs showed distinct clustering in PCA analysis (Fig. 3A) with a different community structure depending on the composition and abundance of each ASV, as also indicated at species level (Fig. 2). The disappearing communities also showed some plant-speciﬁc clustering, but much less pronounced than for the growing communities (Fig. 3C), as illustrated by the lower variance explained compared to the growing communities (R2 = 0.2290 and R2 = 0.4752, respectively). When each plant was analysed individually, both growth groups had similar variance explained by seasonal variations (i.e., winter, spring, summer and autumn) in all WWTPs (Figs. 3B, D). The seasonal patterns of functional guilds showed a mixed response depending on the functional guild and WWTP (Figs. 5 and S12). For some functional guilds, a similar pattern was observed in all WWTPs. For example, glycogen accumulating organisms (GAO) were more prevailing in summer and autumn, while for nitriﬁers and ﬁlaments, the seasonal response depended on the WWTP spanning from spring to autumn (Figs. S12 and S13). The PAOs lacked signiﬁcant seasonal patterns in most of the plants, although some PAO species such as Ca. Dechloromonas phosphoritropha (previous MiDAS 3 placeholder name varianceðresidual componentÞ S7). The detailed study of individual species dynamics showed that about 75% of all species had a signiﬁcant seasonal component. This was independent of whether they were high-abundant (maximum relative abundance ≥1%), low-abundant (maximum relative abundance < 1%), or unique (i.e. species observed in only one WWTP) (Fig. S8). The seasonal strength varied among species and WWTPs. On average, 30 species showed a very strong seasonal response, 296 strong, 838 moderate, 738 weak, and 14 very weak. The seasonal strength was independent of the growth group or functional guild (Fig. S9). The classiﬁcation of species into seasonal cohorts (i.e., species that peak in speciﬁc seasons) showed that the growing fraction contained a high number of species belonging to summer and autumn cohorts (Fig. S10). This is consistent with the dynamics of the estimated number of ASVs along the year. In contrast, the distribution of the cohorts in terms of relative read abundance was even between spring, summer, autumn, and non- signiﬁcant cohorts (Fig. 4A). This explains why the growing fraction rarely showed a signiﬁcant yearly variation, since aggregating species with different seasonality, but evenly distributed abundances can balance each other. The difference between WWTPs for the disappearing fraction (Fig. 4B) was related to the classiﬁcation of the most abundant disappearing species, i.e., Trichoccocus midas_s_4, where the estimated peak abundance was found near the winter and spring split (Fig. 5). The time-series of seasonal cohorts of ambiguous and surviving species are shown in Supplementary Fig. S11. Th l f f i l ild h d i d To evaluate the yearly dynamics of the community structure, alpha-diversity was calculated for each plant and ﬁtted to a simple harmonic model (Eq. 1) (Fig. S6). For most plants and diversity estimates, the model ﬁt was signiﬁcant (p < 0.01) indicating a seasonal response. Damhusåen was an exception, where the seasonality of the number of disappearing ASVs was not signiﬁcant (p = 0.5097) and for the growing species, Simpson diversity was above the 99% conﬁdence threshold (p = 0.0272). However, Damhusåen was the plant with the least sampling years. Overall, both diversity estimates showed a similar seasonal yearly pattern for all the plants, with a minimum diversity in late winter to early spring, and maximum richness in late summer to early autumn, for both growing and disappear- ing species groups. response, 296 strong, 838 moderate, 738 weak, and 14 very weak. varianceðresidual componentÞ Fs ¼ 1 variance residual component ð Þ þ varianceðseasonal componentÞ (4) Fs ¼ 1 variance residual component ð Þ þ varianceðseasonal componentÞ ¼ 1 variance residual component ð Þ þ varianceðseasonal componentÞ (4) Fs takes values from 0 to 1, 1 indicating that variation in species abundance is completely explained by the seasonal variation. For species with a signiﬁcant seasonal component (p < 0.01, harmonic model ﬁt), we deﬁned 5 categories for seasonal strength based on the distribution of Fs values obtained in this study (Fig. S4): Fs > 0.75, Very strong; 0.75 > Fs > 0.55, strong; 0.55 > Fs > 0.35, Moderate; 0.35 > Fs > 0.15, Weak; Fs < 0.15, Very weak. Classiﬁcation of activated sludge bacteria into growth groups. Most species in AS are detected in the inﬂuent wastewater, therefore it is possible to identify whether those species grow or die in AS based on mass balances [4]. ISME Communications M. Peces et al. 4 Fig. 1 Distribution of growth groups in four WWTPs over time. A Mean cumulative relative abundance of growth groups per sample. Values in the bars show the number of different species in each fraction. B Time-series of cumulative read abundance of each growth group, where dots show the experimental results and the solid lines show the corresponding harmonic model ﬁts. Only signiﬁcant (p < 0.01) harmonic ﬁts are shown. Fig. 1 Distribution of growth groups in four WWTPs over time. A Mean cumulative relative abundance of growth groups per sample. Values in the bars show the number of different species in each fraction. B Time-series of cumulative read abundance of each growth group, where dots show the experimental results and the solid lines show the corresponding harmonic model ﬁts. Only signiﬁcant (p < 0.01) harmonic ﬁts are shown. in all WWTPs. A total of 1254 species were only detected in one WWTP, constituting an average cumulative read abundance of 4.5 ± 1.0%. The rest of the species were observed in two or three WWTPs (Fig. S7). The detailed study of individual species dynamics showed that about 75% of all species had a signiﬁcant seasonal component. This was independent of whether they were high-abundant (maximum relative abundance ≥1%), low-abundant (maximum relative abundance < 1%), or unique (i.e. species observed in only one WWTP) (Fig. S8). The seasonal strength varied among species and WWTPs. varianceðresidual componentÞ On average, 30 species showed a very strong seasonal response, 296 strong, 838 moderate, 738 weak, and 14 very weak. The seasonal strength was independent of the growth group or functional guild (Fig. S9). The classiﬁcation of species into seasonal cohorts (i.e., species that peak in speciﬁc seasons) showed that the growing fraction contained a high number of species belonging to summer and autumn cohorts (Fig. S10). This is consistent with the dynamics of the estimated number of ASVs along the year. In contrast, the distribution of the cohorts in terms of relative read abundance was even between spring, summer, autumn, and non- signiﬁcant cohorts (Fig. 4A). This explains why the growing fraction rarely showed a signiﬁcant yearly variation, since aggregating species with different seasonality, but evenly distributed abundances can balance each other. The difference between WWTPs for the disappearing fraction (Fig. 4B) was related to the classiﬁcation of the most abundant disappearing species, i.e., Trichoccocus midas_s_4, where the estimated peak abundance was found near the winter and spring split (Fig. 5). The time-series of seasonal cohorts of ambiguous and surviving species are shown in Supplementary Fig. S11. Microbial community structure can be seasonally described The mean number of ASVs per sample was 774 for the growing fraction and 335 for the disappearing fraction, and the mean Simpson index (reported as 1-D) was slightly higher for the growing fraction (0.985) compared to the disappearing fraction (0.958), conﬁrming that growing communities were more diverse than the disappearing fraction (Fig. 2). in all WWTPs. A total of 1254 species were only detected in one WWTP, constituting an average cumulative read abundance of 4.5 ± 1.0%. The rest of the species were observed in two or three WWTPs (Fig. S7). The detailed study of individual species dynamics showed that about 75% of all species had a signiﬁcant seasonal component. This was independent of whether they were high-abundant (maximum relative abundance ≥1%), low-abundant (maximum relative abundance < 1%), or unique (i.e. species observed in only one WWTP) (Fig. S8). The seasonal strength varied among species and WWTPs. On average, 30 species showed a very strong seasonal response, 296 strong, 838 moderate, 738 weak, and 14 very weak. in all WWTPs. A total of 1254 species were only detected in one WWTP, constituting an average cumulative read abundance of 4.5 ± 1.0%. The rest of the species were observed in two or three WWTPs (Fig. The dynamics of seasonal cohorts vary between growth groups but are similar between WWTPs The species in the ambiguous and surviving group are shown in Fig. S5. Fig. 2 Relative read abundance of the top 25 most abundant bacterial species in the four WWTPs. A Species in the growing group, B species in the disappearing group. The species in the ambiguous and surviving group are shown in Fig. S5. B Dechloromonas midas_s_96) showed similar seasonal patterns in the surveyed WWTPs (Figs. 5 and S14). Therefore, the overall seasonal pattern of guilds, or the lack thereof, depended on the actual composition and diversity of individual species within the functional guilds, and detailed exploration at species level is warranted (section “Seasonal dynamics at species level”). Among nitrifying bacteria, Nitrosomonas was the only identiﬁed ammonia-oxidising genus, and various Nitrosomonas species coexisted in each WWTP. However, the dominant species differed between plants and different species were transiently abundant, and those transient species did not show systematic seasonal patterns (Figs. S12 and S15). Nitrosomonas midas_s_717 showed a weak to non-signiﬁcant seasonality, slightly peaking in spring, whereas Nitrosomonas midas_s_139 showed moderate to strong seasonality thriving during late autumn and winter depending on the WWTP (Fig. 5). In contrast, Nitrosomonas midas_s_723, that was only abundant in Damhusåen and Randers, showed similar seasonal patterns to Nitrosomonas midas_s_139 (Fig. S15, Supple- mentary ﬁle S1). Nitrotoga and Nitrospira represented the nitrite- oxidising bacteria (NOB). Nitrotoga midas_s_181 showed a weak to non-signiﬁcant seasonality, whereas a Nitrotoga-related ASV (ASV223) showed stronger seasonality peaking in late spring. N. deﬂuvii showed a seasonal response prevailing during late autumn to winter in the WWTPs where it coexisted with Nitrotoga species (Figs. 5 and S15) ISME Communications The dynamics of seasonal cohorts vary between growth groups but are similar between WWTPs To explore seasonal dynamics in WWTPs, a total of 2546 species with a relative read abundance higher than 0.05% in at least one sample were analysed. 502 species were shared across all WWTPs, constituting an average cumulative read abundance of 70.0 ± 1.3% ISME Communications ISME Communications M. Peces et al. 5 growing 0 5 10 15 20 Saprospiraceae; midas_s_6 Saccharimonadales; midas_s_8524 Run-SP154; midas_g_70_ASV32 Rhodobacteraceae; midas_s_57 Saprospiraceae; midas_s_20 RBG-13-54-9; midas_s_72 Sphingopyxis bauzanensis Rhodobacter; midas_s_63 Ca. Dechloromonas phosphoritropha PHOS-HE36; midas_s_179 Xanthobacteraceae_ASV22 Comamonadaceae; midas_s_50 Rhodobacter; midas_s_430 Ca. Villigracilis; midas_s_471 Rhodobacter; midas_s_24 OLB8; midas_s_29 Saprospiraceae; midas_g_17_ASV8 Acidovorax; midas_s_1484 Ca. Amarolinea; midas_s_1 Comamonadaceae_ASV28 Rhodoferax; midas_s_33 Rhodobacter_ASV6 Ca. Microthrix subdominans Tetrasphaera; midas_s_5 Ca. Microthrix parvicella Relative read abundance (%) A disappearing 0 5 10 15 Streptococcus parasuis Arcobacter cryaerophilus Simplicispira_ASV726 Proteiniclasticum; midas_s_140 Actinomycetaceae; midas_s_561 Comamonas denitrificans Lactococcus raffinolactis Actinomyces; midas_s_4035 Actinomycetaceae; midas_s_343 Ruminococcus faecis Tessaracoccus; midas_s_307 Ruminococcus bromii Enterococcus aquimarinus Intestinibacter bartlettii Comamonadaceae_ASV288 Acidovorax_ASV179 Thauera terpenica Fusicatenibacter saccharivorans Leptotrichia; midas_s_2907 Clostridium sensu stricto 1; midas_s_64 Clostridium sensu stricto 1; midas_s_101 Subdoligranulum; midas_s_348 Romboutsia; midas_s_34 Acidovorax; midas_s_2077 Trichococcus; midas_s_4 Relative read abundance (%) B Plant Aalborg E Aalborg W Damhusåen Randers disappearing 0 5 10 15 Streptococcus parasuis Arcobacter cryaerophilus Simplicispira_ASV726 Proteiniclasticum; midas_s_140 Actinomycetaceae; midas_s_561 Comamonas denitrificans Lactococcus raffinolactis Actinomyces; midas_s_4035 Actinomycetaceae; midas_s_343 Ruminococcus faecis Tessaracoccus; midas_s_307 Ruminococcus bromii Enterococcus aquimarinus Intestinibacter bartlettii Comamonadaceae_ASV288 Acidovorax_ASV179 Thauera terpenica Fusicatenibacter saccharivorans Leptotrichia; midas_s_2907 Clostridium sensu stricto 1; midas_s_64 Clostridium sensu stricto 1; midas_s_101 Subdoligranulum; midas_s_348 Romboutsia; midas_s_34 Acidovorax; midas_s_2077 Trichococcus; midas_s_4 Relative read abundance (%) B 5 growing 0 5 10 15 20 Saprospiraceae; midas_s_6 Saccharimonadales; midas_s_8524 Run-SP154; midas_g_70_ASV32 Rhodobacteraceae; midas_s_57 Saprospiraceae; midas_s_20 RBG-13-54-9; midas_s_72 Sphingopyxis bauzanensis Rhodobacter; midas_s_63 Ca. Dechloromonas phosphoritropha PHOS-HE36; midas_s_179 Xanthobacteraceae_ASV22 Comamonadaceae; midas_s_50 Rhodobacter; midas_s_430 Ca. Villigracilis; midas_s_471 Rhodobacter; midas_s_24 OLB8; midas_s_29 Saprospiraceae; midas_g_17_ASV8 Acidovorax; midas_s_1484 Ca. Amarolinea; midas_s_1 Comamonadaceae_ASV28 Rhodoferax; midas_s_33 Rhodobacter_ASV6 Ca. Microthrix subdominans Tetrasphaera; midas_s_5 Ca. Microthrix parvicella Relative read abundance (%) A B A B Plant Aalborg E Aalborg W Damhusåen Randers Fig. 2 Relative read abundance of the top 25 most abundant bacterial species in the four WWTPs. A Species in the growing group, B species in the disappearing group. The species in the ambiguous and surviving group are shown in Fig. S5. Fig. 2 Relative read abundance of the top 25 most abundant bacterial species in the four WWTPs. A Species in the growing group, B species in the disappearing group. Seasonal dynamics at species level Seasonal dynamics at species level The detailed study of seasonality at species level showed that species belonging to the same functional guild did not necessarily follow the same seasonal dynamics (Figs. 5 and S12). The same applies when aggregating species at higher taxonomic ranks (e.g., genus, family or higher), since different species in the same group showed diverse seasonal patterns. Signiﬁcant seasonal patterns could be found for higher taxonomic ranks (e.g., families), but they failed to represent all species within the rank since the overall group seasonal pattern was driven by few dominating species obscuring the dynamics of the less abundant ones. Different seasonal dynamics were also found for species in the same genus. For example, the two main species of Ca. Microthrix, i.e., Ca. M. parvicella and Ca. M. subdominans, showed very different patterns (Fig. 5). Ca. M. parvicella showed a strong seasonality increasing from early to late spring, while Ca. M. subdominans showed a weaker seasonality with maximum peaks varying from plant to plant. These were not isolated examples since for genera with more than one species (362 out of 967), only 28% had all species classiﬁed into the same seasonal cohort, and the rest was classiﬁed into two or more seasonal cohorts (Fig. S16, Supplementary ﬁle S1). On the contrary, the GAO Ca. Competibacter (Figs. 5 and S12) is an example where all ISME Communications M. Peces et al. Seasonal dynamics at species level midas_s_1 midas_s_5 f__Comamonadaceae_ASV28 g__midas_g_70_ASV32 g__Nitrotoga_ASV223 midas_s_6 midas_s_2 midas_s_49 midas_s_399 midas_s_5 R2 = 0.4752 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 PC1 [23.8%] PC2 [19.9%] Plant Aalborg E Aalborg W Damhusåen Randers A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A R2 = 0.1781 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [24%] PC2 [18%] W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A R2 = 0.1681 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [21.1%] PC2 [16.3%] n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D R2 = 0.2731 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [32.9%] PC2 [21.6%] s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R R2 = 0.1403 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [24%] PC2 [20.8%] Season Autumn Spring Summer Winter B midas_s_2077 g__Acidovorax_ASV179 midas_s_4 Streptococcus_parasuis g__Simplicispira_ASV726 Comamonas_denitrificans f__Comamonadaceae_ASV288 midas_s_4 Thauera_terpenica f__Comamonadaceae_ASV1505 R2 = 0.2290 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 PC1 [17.4%] PC2 [12.1%] Plant Aalborg E Aalborg W Damhusåen Randers C E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A R2 = 0.1874 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [22.7%] PC2 [12%] W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A R2 = 0.1093 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [14.2%] PC2 [11.3%] n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D R2 = 0.1460 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [17.7%] PC2 [6.1%] s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R R2 = 0.0799 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [13.8%] PC2 [10.6%] Season Autumn Spring Summer Winter D . Seasonal dynamics at species level 6 E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A R2 = 0.1781 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [24%] PC2 [18%] W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A R2 = 0.1681 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0 PC1 [21.1%] PC2 [16.3%] n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D R2 = 0.2731 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [32 9%] PC2 [21.6%] s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R R2 = 0.1403 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0 PC1 [24%] PC2 [20.8%] B midas_s_1 midas_s_5 f__Comamonadaceae_ASV28 g__midas_g_70_ASV32 g__Nitrotoga_ASV223 midas_s_6 midas_s_2 midas_s_49 midas_s_399 midas_s_5 R2 = 0.4752 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 PC1 [23.8%] PC2 [19.9%] A B E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A R2 = 0.1781 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [24%] PC2 [18%] B W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A R2 = 0.1681 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [21.1%] PC2 [16.3%] PC2 [16.3%] PC1 [21.1%] PC1 [21.1%] n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D R2 = 0.2731 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [32.9%] PC2 [21.6%] s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R R2 = 0.1403 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [24%] PC2 [20.8%] PC2 [20.8%] PC2 [21.6%] 2 0.1 0.0 0.1 0 PC1 [32.9%] PC1 [24%] PC1 [24%] Plant Aalborg E Aalborg W Damhusåen Randers Plant Aalborg E Aalborg W Damhusåen Randers Season Autumn Spring Summer Winter Season Autumn Spring Summer Winter Season Autumn Sp E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A E g r o b l a A R2 = 0.1874 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [22.7%] PC2 [12%] D midas_s_2077 g__Acidovorax_ASV179 midas_s_4 Streptococcus_parasuis g__Simplicispira_ASV726 Comamonas_denitrificans f__Comamonadaceae_ASV288 midas_s_4 Thauera_terpenica f__Comamonadaceae_ASV1505 R2 = 0.2290 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 PC1 [17.4%] PC2 [12.1%] C ring Summer Winter W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A W g r o b l a A R2 = 0.1093 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [14 2%] PC2 [11.3%] C D PC2 [11.3%] PC1 [14.2%] s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R s r e d n a R R2 = 0.0799 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [13.8%] PC2 [10.6%] i S Wi s n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D n e å s u h m a D R2 = 0.1460 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.1 0.0 0.1 0.2 0.3 0.4 PC1 [17.7%] PC2 [6.1%] Season Autumn Spr PC2 [10.6%] PC2 [6.1%] Plant Aalborg E Aalborg W Damhusåen Randers Fig. Seasonal dynamics at species level 3 PCA plots of the bacterial community structure. A Differences across plants for growing bacteria, B seasonal differences for growin cteria in each plant, C differences across plants for disappearing bacteria, D seasonal differences for disappearing bacteria in each plant. R lues show the explained variance by the grouping variable (A, C: WWTP; B, D: Season). Seasonal dynamics at species level 4 Time-series of seasonal cohorts in each WWTP. Overview of the cumulative relative read abundance of the seasonal cohorts (winter, spring, summer, autumn and non-signiﬁcant) in each WWTP. A Growing species, B disappearing species. Seasonal cohorts of ambiguous and surviving bacteria can be found in Supplementary Fig. S11. with a strong seasonal component, such as Ca. M. parvicella, Ca. Amarolinea midas_s_1, Ca. Competibacter midas_s_1775, Coma- monadaceae ASV28 or Trichococcus midas_s_4, among others, tended to peak during the same season or concomitant seasons in all WWTPs. The intersections where species were distributed within 2 or 3 cohorts plus the non-signiﬁcant cohort, capture the variability observed for some species (31.2% of the shared growing species and 17.1% of the shared disappearing species). These species tended to show a weak to moderate seasonal component and the estimated relative read abundance peak spanned over several months depending on the WWTP (e.g., Ca. D. phosphoritropha, N. deﬂuvii, or Clostridum sensu stricto 1 midas_s_101, Fig. 5). For the WWTPs studied, no species were classiﬁed into opposite cohorts. Finally, some species were exclusively found in the non-signiﬁcant cohort (eight growing species and three disappearing species) suggesting that their abundance in the WWTPs was not affected by any seasonally periodic component (Fig. 6, Supplementary ﬁle S1). with a strong seasonal component, such as Ca. M. parvicella, Ca. Amarolinea midas_s_1, Ca. Competibacter midas_s_1775, Coma- monadaceae ASV28 or Trichococcus midas_s_4, among others, tended to peak during the same season or concomitant seasons in all WWTPs. The intersections where species were distributed within 2 or 3 cohorts plus the non-signiﬁcant cohort, capture the variability observed for some species (31.2% of the shared growing species and 17.1% of the shared disappearing species). These species tended to show a weak to moderate seasonal component and the estimated relative read abundance peak spanned over several months depending on the WWTP (e.g., Ca. D. phosphoritropha, N. deﬂuvii, or Clostridum sensu stricto 1 midas_s_101, Fig. 5). For the WWTPs studied, no species were classiﬁed into opposite cohorts. Finally, some species were exclusively found in the non-signiﬁcant cohort (eight growing species and three disappearing species) suggesting that their abundance in the WWTPs was not affected by any seasonally periodic component (Fig. 6, Supplementary ﬁle S1). four WWTPs and they were similar to other Danish plants with nutrient removal [18]. Seasonal dynamics at species level 3 PCA plots of the bacterial community structure. A Differences across plants for growing bacteria, B seasonal differences for growing bacteria in each plant, C differences across plants for disappearing bacteria, D seasonal differences for disappearing bacteria in each plant. R2 values show the explained variance by the grouping variable (A, C: WWTP; B, D: Season). species in the same genus showed a very similar pattern, which was also similar across WWTPs. Damhusåen WWTP harboured the highest diversity and abundance of Ca. Competibacter species with estimated peaks yearly recurring between August and November, depending on the species. Additionally, species seasonality was highly reproducible among WWTPs, yet signiﬁcant variations in maximum peak estimates could be found for some species (see Ca. Competibacter mid- as_s_1820 in Fig. S13). Detailed seasonal estimates for all species in each WWTP are shown in Supplementary ﬁle S1. species in the same genus showed a very similar pattern, which was also similar across WWTPs. Damhusåen WWTP harboured the highest diversity and abundance of Ca. Competibacter species with estimated peaks yearly recurring between August and November, depending on the species. Additionally, species seasonality was highly reproducible among WWTPs, yet signiﬁcant variations in maximum peak estimates could be found for some species (see Ca. Competibacter mid- as_s_1820 in Fig. S13). Detailed seasonal estimates for all species in each WWTP are shown in Supplementary ﬁle S1. patterns in different plants. The intersection plots (Figs. 6 and S17) show the co-occurrence of species in the seasonal cohorts among the four plants, where the dotted chart represents the intersection between seasonal cohorts and the bar chart shows the number of species found in each intersection. For both growth groups (growing fraction and disappearing fraction), most species with a very strong to moderate seasonality were assigned to the same seasonal cohort, or concomitant cohorts, in the four WWTP, corresponding to 32.5% of the shared growing species and 34.2% of the shared disappearing species (Fig. 6). This phenomenon summarises the visualisation dynamics in Fig. 5, where species The shared species among the four WWTPs (502 out of 909) were used to evaluate the repeatability of species seasonal ISME Communications M. Peces et al. Seasonal dynamics at species level 7 Non-significant Autumn cohort Summer cohort Spring cohort Winter cohort Jan 2015 Jul 2015 Jan 2016 Jul 2016 Jan 2017 Jul 2017 Jan 2018 Jul 2018 Jan 2019 Jul 2019 Jan 2020 Jul 2020 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Date Relative abundance (%) A Non-significant Autumn cohort Summer cohort Spring cohort Winter cohort Jan 2015 Jul 2015 Jan 2016 Jul 2016 Jan 2017 Jul 2017 Jan 2018 Jul 2018 Jan 2019 Jul 2019 Jan 2020 Jul 2020 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Date Relative abundance (%) A Non-significant Autumn cohort Summer cohort Spring cohort Winter cohort Jan 2015 Jul 2015 Jan 2016 Jul 2016 Jan 2017 Jul 2017 Jan 2018 Jul 2018 Jan 2019 Jul 2019 Jan 2020 Jul 2020 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Date Relative abundance (%) B Plant Aalborg E Aalborg W Randers Damhusåen Fig. 4 Time-series of seasonal cohorts in each WWTP. Overview of the cumulative relative read abundance of the seasonal cohorts (winter, spring, summer, autumn and non-signiﬁcant) in each WWTP. A Growing species, B disappearing species. Seasonal cohorts of ambiguous and surviving bacteria can be found in Supplementary Fig. S11. Non-significant Autumn cohort Summer cohort Spring cohort Winter cohort Jan 2015 Jul 2015 Jan 2016 Jul 2016 Jan 2017 Jul 2017 Jan 2018 Jul 2018 Jan 2019 Jul 2019 Jan 2020 Jul 2020 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Date Relative abundance (%) B Plant Aalborg E Aalborg W Randers Damhusåen A B Non-significant Autumn cohort Summer cohort Spring cohort Winter cohort Jan 2015 Jul 2015 Jan 2016 Jul 2016 Jan 2017 Jul 2017 Jan 2018 Jul 2018 Jan 2019 Jul 2019 Jan 2020 Jul 2020 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 0 10 20 30 40 Date Relative abundance (%) B Plant Aalborg E Aalborg W Randers Damhusåen Relative abundance (%) Date Date Date Fig. Seasonal dynamics at species level This suggests that the clustering was determined by differences in their relative read abundances as well as differences in low-abundant species. Both growth groups showed a seasonal pattern, but they showed a different seasonal response. The growing group was characterised by a higher number of low-abundant species peaking during the summer and autumn cohorts. However, the cumulative relative read abun- dance explained by species within each seasonal cohort showed an even distribution. In contrast, the species in the disappearing group were more prevalent during winter/spring, indicating the inﬂuence of different seasonal drivers affecting the species in the two growth groups. Since the disappearing group consists of only immigrating bacteria that do not thrive in the process tanks, their abundance must be controlled by upstream factors in the sewer system and by factors affecting their degradation rate in the AS. system and by factors affecting their degradation rate in the AS. Growing bacteria are assumed to be active in the AS system and perform process-critical functions [21, 44], so a common approach is to group microorganisms according to their potential functionality (e.g., nitriﬁers or PAO) or similar morphology (e.g., ﬁlamentous organisms) to evaluate WWTP performance. However, these guilds may encompass species from different phylogenetic groups and have diverse metabolic potentials. Some guilds showed an overall seasonal pattern, but the results should be extrapolated with great caution since the dynamics depended on the actual composition, diversity, and abundance of species within the deﬁned guild. For example, in Damhusåen WWTP, the two main ﬁlamentous species Ca. M. parvicella (spring cohort) and Ca. Amarolinea midas_s_1 (autumn cohort) showed a similar mean relative read abundance over the years (~5.7% and 4.0%, respectively), hence the ﬁlamentous guild did not show any overall seasonality. Distinguishing the individual taxa dynamics beyond the overall guild has great practical implications since these two ﬁlamentous bacteria are the main responsible taxa for poor sludge settling and ﬁlamentous bulking, and the different physiologies may require distinct control measures [20, 45]. However, ﬁlamentous bacteria share a morphological trait rather than a metabolic trait, therefore, the different species might not occupy the same niche and compete for the same resources. Therefore, the identity of species in these groups may vary depending on the geographical location of the WWTP and environmental conditions [1, 46]. ISME Communications DISCUSSION O l d Our longitudinal survey of four full-scale WWTP with nutrient removal during 3–5 consecutive years with a new approach, including time-series decomposition and ecosystem-speciﬁc spe- cies-level classiﬁcation, showed signiﬁcant seasonal dynamics for about 75% of the identiﬁed species. The seasonal dynamic was cyclic (i.e., repeating each year) and comparable across WWTPs with similar process conﬁguration. Moreover, our approach allowed us to evaluate if distinct seasonal patterns could be observed for bacterial species depending on their functional guild, level of taxonomic aggregation, and net growth rate in the AS bioreactor. Interestingly, the growing and disappearing growth groups exhibited different alpha- and beta-diversity patterns. In particular, the beta-diversity of disappearing microbial communities resembled the beta-diversity commonly observed in inﬂuent wastewater communities [4, 43]. This can be explained by the fact that disappearing species in AS are only present as a consequence of mass immigration of bacteria with the inﬂuent wastewater [4]. The beta-diversity of growing and disappearing groups in each WWTP showed distinct clustering patterns (PCA, Fig. 3). This indicated that each WWTP had its own signature microbial community, where differences were stronger for the growing bacteria than for the disappearing species. However, the top abundant species in each growth group were the same across all Nitriﬁers form a well-described and deﬁned guild with few genera abundant in WWTPs worldwide, i.e., Nitrosomonas, Nitrotoga and Nitrospira [1]. In our study, all WWTPs reported a ISME Communications M. Peces et al. Fig. 5 Fitted seasonal component of the top ﬁve species in the main functional guilds. Each WWTP is denoted by its initials (AAE Aalborg E, AAW Aalborg W, DAM Damhusåen, RND Randers). Colour intensity of the ﬁtted trends represents the seasonal strength of each species in the WWTP (darker intensity represents a stronger seasonal component, grey colour represents non-signiﬁcant seasonality). Background colours show the astronomical seasons for the northern hemisphere (winter, spring, summer and autumn). The text box shows the week at maximum read abundance with the 95% conﬁdence interval. 8 Fig. 5 Fitted seasonal component of the top ﬁve species in the main functional guilds. Each WWTP is denoted by its initials (AAE Aalborg E, AAW Aalborg W, DAM Damhusåen, RND Randers). Colour intensity of the ﬁtted trends represents the seasonal strength of each species in the WWTP (darker intensity represents a stronger seasonal component, grey colour represents non-signiﬁcant seasonality). DISCUSSION O l d Temperature affects growth and decay rates, degradation rates and biochemical transforma- tions, or liquid-gas solubility among others, hence having a profound impact on species ecophysiology. Some ecophysiologi- cal characteristics can be found for pure cultures, but they are rarely described in situ, making it very difﬁcult to associate the impact of temperature over individual species. Nevertheless, temperature alone cannot explain all the different seasonal dynamics observed, nor the variability for some species between WWTPs. For example, the maximum relative abundance peak estimates could vary about 2 months between WWTPs, even for strongly seasonal species (e.g., Ca. M. parvicella, spring cohort), while for weakly seasonal species even greater variability could be observed (e.g., Ca. M. subdominans). Substrate availability and composition is another important factor that can affect seasonal dynamics [11, 49, 50]. For example, in combination with temperature, a seasonal inﬂuent lipid loading was found to increase the abundance of a foam-forming microorganism related good nitrogen removal performance [24, 25, Fig. S18, Supple- mentary ﬁles S2 and S3] and showed a comparable relative read abundance of nitriﬁers (0.8–2.0%) with some ﬂuctuations during the years (Fig. S15). Different species of Nitrosomonas coexisted in the WWTPs and some showed seasonal patterns. However, both the seasonal response and the dominant Nitrosomonas species varied between plants, making it difﬁcult to elucidate the drivers for each species. The dominant NOB was also WWTP dependent. In the Randers plant only N. deﬂuvii was above the quantiﬁcation limit (but Nitrotoga could be detected with a read abundances below 0.01%), while in the other WWTPs two Nitrotoga species coexisted with seasonal increases in N. deﬂuvii. Intriguingly, within the plants where different NOB coexisted, N. deﬂuvii showed a similar seasonality increasing from late autumn to late winter in all WWTPs concomitant with lower read abundances of Nitrotoga species. Based on genomic studies, Nitrotoga and Nitrospira have diverse metabolic potential as well as different membrane-bound orientations of the nitrite oxidoreductase enzyme, indicating different afﬁnities in nitrite uptake [47, 48], which suggests the possibility of coexistence. The long-term dynamic pattern observed between Nitrotoga and Nitrospira species in different years and WWTPs, suggests that for a given WWTP nitriﬁcation capacity, some degree of competition occurs between these NOB genera. However, as discussed below, immigration may partly be responsible for these dynamics, so further studies are necessary to determine the factors inﬂuencing species seasonality and the practical implications for the WWTP operation. DISCUSSION O l d Background colours show the astronomical seasons for the northern hemisphere (winter, spring, summer and autumn). The text box shows the week at maximum read abundance with the 95% conﬁdence interval. good nitrogen removal performance [24, 25, Fig. S18, Supple- mentary ﬁles S2 and S3] and showed a comparable relative read abundance of nitriﬁers (0.8–2.0%) with some ﬂuctuations during the years (Fig. S15). Different species of Nitrosomonas coexisted in the WWTPs and some showed seasonal patterns. However, both the seasonal response and the dominant Nitrosomonas species varied between plants, making it difﬁcult to elucidate the drivers for each species. The dominant NOB was also WWTP dependent. In the Randers plant only N. deﬂuvii was above the quantiﬁcation limit (but Nitrotoga could be detected with a read abundances below 0.01%), while in the other WWTPs two Nitrotoga species coexisted with seasonal increases in N. deﬂuvii. Intriguingly, within the plants where different NOB coexisted, N. deﬂuvii showed a similar seasonality increasing from late autumn to late winter in all WWTPs concomitant with lower read abundances of Nitrotoga species. Based on genomic studies, Nitrotoga and Nitrospira have diverse metabolic potential as well as different membrane-bound orientations of the nitrite oxidoreductase enzyme, indicating different afﬁnities in nitrite uptake [47, 48], which suggests the possibility of coexistence. The long-term dynamic pattern observed between Nitrotoga and Nitrospira species in different years and WWTPs, suggests that for a given WWTP nitriﬁcation capacity, some degree of competition occurs between these NOB genera. However, as discussed below, immigration may partly be responsible for these dynamics, so further studies are necessary to determine the factors inﬂuencing species seasonality and the practical implications for the WWTP operation. (Figs. 5 and 6). Variations in process parameters could explain some of the species’ observed variance (Supplementary Note 1, Figs. S19 and S20). However, most of the measured process parameters such as inﬂuent chemical oxygen demand, ammonia or phosphate, among others, ﬂuctuated randomly during the years suggesting a minor impact on the observed seasonal dynamics (Supplementary Note 1, Supplementary ﬁles S2 and S3). Temperature was the only measured parameter that showed a recurrent seasonal pattern across the WWTPs (Supplementary ﬁle S4, Supplementary Note 1). Process tank temperature has been suggested as the most important factor to explain seasonal dynamics in WWTPs [6, 8, 9, 11–13]. REFERENCES 1. Dueholm MS, Nierychlo M, Andersen KS, Joergensen VR, Knutsson S, Consortium the MG, et al. MiDAS 4: a global catalogue of full-length 16S rRNA gene sequences and taxonomy for studies of bacterial communities in wastewater treatment plants. 2021. https://www.biorxiv.org/content/10.1101/2021.07.06.451231v1. 2021. https://www.biorxiv.org/content/10.1101/2021.07.06.451231v1 2. Xia Y, Wen X, Zhang B, Yang Y. Diversity and assembly patterns of activated sludge microbial communities: a review. Biotechnol Adv. 2018;36:1038–47. to Gordonia, causing seasonal bulking [49]. Other factors, such as SRT, are typically considered in WWTP operation since SRT is inversely proportional to net microbial growth rates [51], but it is unclear if typical SRT ﬂuctuations in WWTP can apply enough selective pressure to inﬂuence strong seasonal microbial dynamics. Indeed, full-scale experiments have shown minor differences in bacterial classes between an SRT of 12 and 30 days [52], which is a common operational range for many municipal AS bioreactors in temperate climates. 3. Yuan H, Mei R, Liao J, Liu WT. Nexus of stochastic and deterministic processes on microbial community assembly in biological systems. Front Microbiol. 2019:1536. 4. Dottorini G, Michaelsen TY, Kucheryavskiy S, Andersen KS, Kristensen JM, Peces M, et al. Mass-immigration determines the assembly of activated sludge micro- bial communities. PNAS. 2021;118:e2021589118. 5. Oﬁţeru ID, Lunn M, Curtis TP, Wells GF, Criddle CS, Francis CA, et al. Combined niche and neutral effects in a microbial wastewater treatment community. PNAS. 2010;107:15345–50. 6. Jiang X-T, Ye L, Ju F, Wang Y-L, Zhang T. Toward an intensive longitudinal understanding of activated sludge bacterial assembly and dynamics. Environ Sci Technol. 2018;52:8224–32. Immigrating bacteria with inﬂuent wastewater could also affect the observed seasonal dynamics, although their contribution to seasonality in AS has not yet been studied in detail. Bacteria from the source communities (e.g., sewer systems) are continuously added to the AS, and recent longitudinal studies have shown some seasonal patterns in sewer and AS inﬂuent microbial communities [12, 53, 54]. The inﬂuence of mass immigration is clear for the disappearing group, which dies in the AS tank and it is only present due to the wastewater inﬂow. A good example is Trichococcus. Trichococcus has been reported to be more abundant in the sewer systems in colder climates and during colder months [53, 54], which may explain our observations of the strong dynamics of Trichococcus showing a maximum seasonal peak late winter and early spring. DISCUSSION O l d Many species showed similar seasonal patterns across the four WWTPs suggesting the inﬂuence of some overarching factors ISME Communications ISME Communications M. Peces et al. 9 0 5 10 15 20 25 30 35 Species counts Seasonal strength Strong Moderate Weak NS Non-significant Autumn_cohort Summer_cohort Spring_cohort Winter_cohort A 0 2 4 6 8 10 12 14 16 Species counts Seasonal strength Strong Moderate Weak NS Non-significant Autumn_cohort Summer_cohort Spring_cohort Winter_cohort B Fig. 6 Comparison of species assigned into seasonal cohorts. Distribution of shared species between WWTPs across seasonal cohorts by growth group. A Growing fraction, B Disappearing fraction. The dotted chart represents all the possible intersections among the seasonal cohorts where the bar chart plot shows the number of species found in each intersection coloured by their seasonal strength. Colour intensity in the top bars represents the seasonal strength of the shared species in each cohort intersection, grey colour represents non-signiﬁcant seasonal species. Ambiguous and surviving bacteria can be found in Supplementary Fig. S17. 0 5 10 15 20 25 30 35 Species counts Seasonal strengt Strong Moderate Weak NS Non-significant Autumn_cohort Summer_cohort Spring_cohort Winter_cohort A seasonal dynamic. However, to date, this is undescribed. Proving a mechanistic causation of speciﬁc drivers for each species is challenging. However, the combination of targeted experiments of the microbial immigration, the chemistry of inﬂuent waste- water, the application of deterministic and stochastic models, or deep learning may in the near future allow a better understanding of the factors that drive the species dynamics. A p y The observed recurrent long-term seasonal dynamics of many species has several practical implications for WWTP understand- ing, operation and performance. Importantly, we have shown that seasonal dynamics are species-speciﬁc and some variations exist among WWTPs without clear correlations to the process parameters. Therefore, WWTPs should analyse their community pattern at species level by standardised taxonomy (e.g., MiDAS 4) for at least 2 years to establish a “normal” baseline for the plant, concomitantly with a detailed surveillance of the process and operational parameters and inﬂuent wastewater. This will allow improving full-scale experiments design, results interpretation and comparability among studies. Currently, this is, in general, far beyond normal practice. Alternatively, when carrying out experi- ments in full-scale WWTPs, replicate bioreactors (i.e., independent parallel AS lines) could be used as seasonal control, although this option may not be feasible for most WWTPs. DATA AVAILABILITY Amplicon sequencing data is deposited in NCBI project PRJNA757616. Amplicon sequencing data is deposited in NCBI project PRJNA757616. DISCUSSION O l d Looking beyond, where on-site or “online” sequencing is implemented with the control system of full-scale AS bioreactors, seasonal dynamics will need to be considered to develop robust and effective process control loops. B 0 2 4 6 8 10 12 14 16 Species counts Seasonal strength Strong Moderate Weak NS Non-significant Autumn_cohort Summer_cohort Spring_cohort Winter_cohort B Fig. 6 Comparison of species assigned into seasonal cohorts. Distribution of shared species between WWTPs across seasonal cohorts by growth group. A Growing fraction, B Disappearing fraction. The dotted chart represents all the possible intersections among the seasonal cohorts where the bar chart plot shows the number of species found in each intersection coloured by their seasonal strength. Colour intensity in the top bars represents the seasonal strength of the shared species in each cohort intersection, grey colour represents non-signiﬁcant seasonal species. Ambiguous and surviving bacteria can be found in Supplementary Fig. S17. ISME Communications REFERENCES This supports the ﬁndings that abundant species in AS are inﬂuenced by their abundance in the inﬂuent wastewater. The inﬂuence of mass immigration can be extended to the growing fraction, where most species are expected to perform process-critical functions in the AS. The AS growing species are generally present in low abundance in the inﬂuent wastewater [4], and their abundance may also be seasonal in the inﬂuent wastewater and thereby determine, at least partly, the AS 7. Johnson JS, Spakowicz DJ, Hong B-Y, Petersen LM, Demkowicz P, Chen L, et al. Evaluation of 16S rRNA gene sequencing for species and strain-level microbiome analysis. Nat Commun. 2019;10:5029. 8. Flowers JJ, Cadkin TA, McMahon KD. Seasonal bacterial community dynamics in a full- scale enhanced biological phosphorus removal plant. Water Res. 2013;47:7019–31. 9. Grifﬁn JS, Wells GF. Regional synchrony in full-scale activated sludge bioreactors due to deterministic microbial community assembly. ISME J. 2017;11:500–11. 10. Johnston J, Behrens S. Seasonal dynamics of the activated sludge microbiome in sequencing batch reactors, assessed using 16S rRNA transcript amplicon sequencing. Appl Environ Microbiol. 2020;86:e00597–20. 11. Petrovski S, Rice DTF, Batinovic S, Nittami T, Seviour RJ. The community com- positions of three nitrogen removal wastewater treatment plants of different conﬁgurations in Victoria, Australia, over a 12-month operational period. Appl Microbiol Biotechnol. 2020;104:9839–52. 12. Sun C, Zhang B, Ning D, Zhang Y, Dai T, Wu L, et al. Seasonal dynamics of the microbial community in two full-scale wastewater treatment plants: diversity, composition, phylogenetic group based assembly and co-occurrence pattern. Water Res. 2021;200:117295. 13. Wells GF, Park H-D, Eggleston B, Francis CA, Criddle CS. Fine-scale bacterial community dynamics and the taxa–time relationship within a full-scale activated sludge bioreactor. Water Res. 2011;45:5476–88. ISME Communications M. Peces et al. 10 42. Hyndman R, Athanasopoulos G. Forecasting: principles and practice, 2nd ed. 2018. Australia: OTexts, Monash University. 14. Ju F, Zhang T. Bacterial assembly and temporal dynamics in activated sludge of a full-scale municipal wastewater treatment plant. ISME J. 2015;9:683–95. 15. Wang P, Yu Z, Qi R, Zhang H. Detailed comparison of bacterial communities during seasonal sludge bulking in a municipal wastewater treatment plant. Water Res. 2016;105:157–66. 43. Lee S-H, Kang H-J, Park H-D. Inﬂuence of inﬂuent wastewater communities on temporal variation of activated sludge communities. Water Res. 2015;73:132–44. 44. Mei R, Liu W-T. Quantifying the contribution of microbial immigration in engi- neered water systems. Microbiome. 2019;7:144. 16. REFERENCES Wang Y, Qin W, Jiang X, Ju F, Mao Y, Zhang A, et al. Seasonal prevalence of ammonia-oxidizing archaea in a full-scale municipal wastewater treatment plant treating saline wastewater revealed by a 6-year time-series analysis. Environ Sci Technol. 2021;55:2662–73. 45. Nierychlo M, McIlroy SJ, Kucheryavskiy S, Jiang C, Ziegler AS, Kondrotaite Z, et al. Candidatus Amarolinea and Candidatus Microthrix are mainly responsible for ﬁlamentous bulking in Danish municipal wastewater treatment plants. Front Microbiol. 2020;11:1214 17. Dueholm MS, Andersen KS, McIlroy SJ, Kristensen JM, Yashiro E, Karst SM, et al. Generation of comprehensive ecosystem-speciﬁc reference databases with species-level resolution by high-throughput full-length 16S rRNA gene sequen- cing and automated taxonomy assignment (AutoTax). mBio. 2020;11:e01557–20. 46. Wu L, Ning D, Zhang B, Li Y, Zhang P, Shan X, et al. Global diversity and bio- geography of bacterial communities in wastewater treatment plants. Nat Microbiol. 2019;4:1183–95. p y g g p g g q cing and automated taxonomy assignment (AutoTax). mBio. 2020;11:e01557–20. nd automated taxonomy assignment (AutoTax). mBio. 2020;11:e0155 18. Nierychlo M, Andersen KS, Xu Y, Green N, Jiang C, Albertsen M, et al. MiDAS 3: an ecosystem-speciﬁc reference database, taxonomy and knowledge platform for activated sludge and anaerobic digesters reveals species-level microbiome composition of activated sludge. Water Res. 2020;182:115955. 47. Lücker S, Wagner M, Maixner F, Pelletier E, Koch H, Vacherie B, et al. A nitrospira metagenome illuminates the physiology and evolution of globally important nitrite-oxidizing bacteria. PNAS. 2010;107:13479–84. 48. Boddicker AM, Mosier AC. Genomic proﬁling of four cultivated Candidatus Nitrotoga spp. predicts broad metabolic potential and environmental distribu- tion. ISME J. 2018;12:2864–82. 19. Martiny JBH, Jones SE, Lennon JT, Martiny AC. Microbiomes in light of traits: a phylogenetic perspective. Science. 2015;350:aac9323. 20. Nierychlo M, Singleton CM, Petriglieri F, Thomsen L, Petersen JF, Peces M, et al. 20. Nierychlo M, Singleton CM, Petriglieri F, Thomsen L, Petersen JF, Peces M, et al. Low global diversity of Candidatus Microthrix, a troublesome ﬁlamentous organism in full-scale WWTPs. Front Microbiol. 2021;12:1593. 49. Frigon D, Michael Guthrie R, Timothy Bachman G, Royer J, Bailey B, Raskin L. Long-term analysis of a full-scale activated sludge wastewater treatment system exhibiting seasonal biological foaming. Water Res. 2006;40:990–1008. Low global diversity of Candidatus Microthrix, a troublesome ﬁlamentous organism in full-scale WWTPs. Front Microbiol. 2021;12:1593. 21. Saunders AM, Albertsen M, Vollertsen J, Nielsen PH. The activated sludge ecosystem contains a core community of abundant organisms. ISME J. 2016;10:11–20. 50. ACKNOWLEDGEMENTS 28. Edgar RC. UNOISE2: improved error-correction for Illumina 16S and ITS amplicon sequencing. https://www.biorxiv.org/content/10.1101/081257v1. 2016. sequencing. https://www.biorxiv.org/content/10.1101/081257v1. 20 We would like to thank all the wastewater treatment plant operators for the collection of samples. The project was funded by Aalborg University and the Villum Foundation (Grant 13351) as part of the MiDAS activities. We would like to thank all the wastewater treatment plant operators for the collection of samples. The project was funded by Aalborg University and the Villum Foundation (Grant 13351) as part of the MiDAS activities. 29. Edgar RC. Accuracy of taxonomy prediction for 16S rRNA and fungal ITS sequences. PeerJ. 2018;6:e4652. 30. Yarza P, Yilmaz P, Pruesse E, Glöckner FO, Ludwig W, Schleifer K-H, et al. Uniting the classiﬁcation of cultured and uncultured bacteria and archaea using 16S rRNA gene sequences. Nat Rev Microbiol. 2014;12:635–45. 31. R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2020. ADDITIONAL INFORMATION 37. Lück S, Thurley K, Thaben PF, Westermark PO. Rhythmic degradation explains and uniﬁes circadian transcriptome and proteome data. Cell Reports. 2014;9:741–51. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s43705-022-00098-4. 38. Martino C, Shenhav L, Marotz CA, Armstrong G, McDonald D, Vázquez-Baeza Y, et al. Context-aware dimensionality reduction deconvolutes gut microbial com- munity dynamics. Nat Biotechnol. 2020;39(2):165–8. Correspondence and requests for materials should be addressed to Per Halkjær. Nielsen. 39. Aitchison J. The statistical analysis of compositional data. J R Stat Soc Ser B (Methodol). 1982;44:139–60. Reprints and permission information is available at http://www.nature.com/ reprints REFERENCES Zhang B, Ning D, Yang Y, Van Nostrand JD, Zhou J, Wen X. Biodegradability of wastewater determines microbial assembly mechanisms in full-scale wastewater treatment plants. Water Res. 2020;169:115276. 22. Forsyning, A. Aalborg Forsyning og miljø og energiforvaltningen miljøredegørelse 2019 for vanddivisionen. Aalborg Forsyning og Miljø og Energiforvaltningen, Denmark; 2020. p. 3–33. 51. Metcalf & Eddy, Tchobanoglous G, Stensel HD, Tsuchihashi R, Burton F. Waste- water engineering: treatment and resource recovery, 5th ed. New York, USA: McGraw-Hill Education; 2014. 23. Kommune, K. Udledningstilladelse renseanlæg Damhusåen. Københavns Kom- mune, Denmark; 2020. p. 1–71. 24. Lane DJ. 16S/23S rRNA sequencing. Nucleic acid techniques in bacterial sys- tematics. New York, NY: John Wiley and Sons Ltd; 1991. p. 115–75. 52. Vuono DC, Munakata-Marr J, Spear JR, Drewes JE. Disturbance opens recruitment sites for bacterial colonization in activated sludge. Environ Microbiol. 2016;18:87–99. 25. Muyzer G, de Waal EC, Uitterlinden AG. Proﬁling of complex microbial popula- tions by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-ampliﬁed genes coding for 16S rRNA. Appl Environ Microbiol. 1993;59:695–700. 53. McLellan SL, Roguet A. The unexpected habitat in sewer pipes for the propa- gation of microbial communities and their imprint on urban waters. Curr Opin Biotechnol. 2019;57:34–41. 26. Albertsen M, Karst SM, Ziegler AS, Kirkegaard RH, Nielsen PH. Back to basics – the inﬂuence of DNA extraction and primer choice on phylogenetic analysis of activated sludge communities. PLoS ONE. 2015;10:e0132783. 54. LaMartina EL, Mohaimani AA, Newton RJ. Urban wastewater bacterial commu- nities assemble into seasonal steady states. Microbiome. 2021;9:116. 27. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioin- formatics. 2010;26:2460–1. COMPETING INTERESTS 35. Krassowski M. krassowski/complex-upset version 1.1.0. 2021. https://github.com/ krassowski/complex-upset The authors declare no competing interests. 36. Ellis P. ggseas: ‘stats’ for Seasonal adjustment on the ﬂy with ‘ggplot2’. R package version 0.5.4. 2018. https://github.com/ellisp/ggseas AUTHOR CONTRIBUTIONS MP, GD and PHN devised the study and its main conceptual ideas. MP performed data analyses and data visualisation with contributions from GD, MN and KSA. MP, GD and PHN wrote the manuscript. MN coordinated sampling and plant data collection, and lab-procedures. KSA and MKDD provided bioinformatic support. All authors contributed to the manuscript and approved the submitted version. 32. Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, et al. Wel- come to the Tidyverse. J Open Source Softw. 2019;4:1686. 33. Andersen KS, Kirkegaard RH, Karst SM, Albertsen M. ampvis2: an R package to analyse and visualise 16S rRNA amplicon data. https://www.biorxiv.org/content/ 10.1101/299537v1.full. 2018. 34. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: Community ecology package version 2.5.6. 2020. https://CRAN.R-project.org/ package=vegan Reprints and permission information is available at http://www.nature.com/ reprints 40. Aghabozorgi S, Seyed Shirkhorshidi A, Ying Wah T. Time-series clustering – a decade review. Inf Syst. 2015;53:16–38. 41. Cleveland RB, Cleveland WS, McRae JE, Terpenning I. STL: a seasonal-trend decomposition. J Off Stat. 1990;6:3–73. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. ISME Communications ISME Communications M. Peces et al. M. Peces et al. 11 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. © The Author(s) 2022 ISME Communications
W2883894412.txt	https://andjournal.sgu.ru/sites/andjournal.sgu.ru/files/text-pdf/2019/03/bastrakov.pdf	ru		ПОДАВЛЕНИЕ ВОЗБУЖДЕНИЙ В АКТИВНОЙ СРЕДЕ С ПОМОЩЬЮ СЛАБОГО ВНЕШНЕГО ВОЗДЕЙСТВИЯ	Izvestiâ vysših učebnyh zavedenij. Prikladnaâ nelinejnaâ dinamika	2,014	cc-by	5,427	61 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 61 УДК 517.977 ПОДАВЛЕНИЕ ВОЗБУЖДЕНИЙ В АКТИВНОЙ СРЕДЕ С ПОМОЩЬЮ СЛАБОГО ВНЕШНЕГО ВОЗДЕЙСТВИЯ И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов Нижегородский государственный университет имени Н. И. Лобачевского В данной работе представлены два новых метода подавления импульса в одномерной и двумерной возбудимых средах с помощью внешнего воздействия. В предложенных методах использовалось кратковременное импульсное воздействие, приводящее к изменению скоростей распространения фронтов, что, в свою очередь, привело к дестабилизации распространяющегося импульса и переходу среды в невозбужденное состояние. Исследования проводились на модели Зыкова, которая при некотором наборе параметров является моделью возбудимой среды. Были определены условия на амплитуду и длительность внешних воздействий, необходимые для подавления возбуждений. Ключевые слова: Нелинейная динамика, активные среды, спиральные волны, модель Зыкова, волны возбуждения. Введение Модель возбудимой среды – одна из базовых моделей активных сред. В таких средах возможно существование устойчивых самоподдерживающихся волн. Возбудимые среды широко распространены в автокаталитических химических реакциях [1, 2] и в биологии – нейронные структуры, нейронная и мышечная ткани [3–6]. Такие среды интересны тем, что в них могут распространяться волны возбуждения, которые можно рассматривать как один из механизмов связи между различными частями сред. В некоторых случаях существование распространяющихся волн является нежелательным эффектом. Например, самопроизвольное разрушение спиральной волны на несколько волн и их последовательное дробление из-за неоднородности могут приводить к хаотическому пространственно-временному поведению. Такая динамика рассматривается как возможный механизм для начала желудочковой фибрилляции в сердечной мышце [7]. Таким образом, существует необходимость в разработке эффективных методов для инициации, управления и уничтожения волн возбуждения. Разработке стратегий по борьбе с сердечными аритмиями на основе математических моделей посвящено большое число исследований. В [8] описано развитие 62 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 систем, моделирующих поведение клеток сердечной ткани, механизмы возникновения аритмий сердца и способы борьбы с ними. Способ стимуляции сердца для снижения вероятности сердечной недостаточности, использующий управление с обратной связью (вызываются малые возмущения постоянной амплитуды с чередующимися частотами), предложен в [9]. Этот способ более устойчив к шумам, чем многие аналогичные методы, и поэтому более эффективен, с точки зрения реализации экспериментов и применения для лечения аритмии. В [10] предложен способ подачи внешнего воздействия для подавления спиральной волны в одномерной кусочно-линейной модели ФитцХью–Нагумо – периодическое импульсное воздействие малой амплитуды непрерывное во времени. Управление волнами с помощью обратной связи обсуждалось в [11–14]. В [15] предложено для подавления фибрилляции желудочков подавать последовательность из нескольких (5 или 7) электрических импульсов малой амплитуды, вместо одного импульса с большой амплитудой. При этом снижение суммарной энергии импульсов составило в среднем 84%. Для удержания клеток миокарда в «заблокированном» состоянии в течении заданного промежутка времени в [16] предложено использование синусоидального высокочастотного переменного тока. В работе предлагается метод, который использует кратковременное импульсное воздействие на среду для уничтожения нестатических состояний в возбудимых средах, таких как распространяющиеся импульсы, одиночная спиральная волна и хаос спиральных волн. 1. Модель Рассмотрим достаточно общую модель, которая описывает возбудимую среду через двумерное уравнение реакции-диффузии  ∂u   = F (u, v) + Du ∆u,  ∂t  ∂v   = εG(u, v) + Dv ∆v, ∂t (1) где ε ≪ 1 малый параметр, который управляет пространственно-временными масштабами в системе; u и v – быстрые и медленные переменные, соответственно; ∆ – оператор Лапласа в пространственных координатах; Du,v – коэффициенты диффузии для быстрой и медленной переменных. В нервной и мышечных тканях диффузия медленной переменной отсутствует, то есть Dv = 0 в (1). Функции F (u, v) и G(u, v) определяют локальную кинетику системы. Динамика изолированного элемента среды (в (1) Du = Dv = 0) описывается следующей системой:  du   = F (u, v),  dt (2)  dv   = εG(u, v). dt Функции F (u, v) и G(u, v), используемые при моделировании возбудимых сред, качественно представлены на рис. 1. Изоклина F (u, v) = 0 имеет N -образную c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 63 форму. Обозначим левую убывающую ветвь через h− (v), правую убывающую ветвь как h+ (v) и среднюю возрастающую ветвь как h0 (v). Изоклина G(u, v) = 0 монотонная функция и пересекает левую ветвь h− (v) изоклины F (u, v) = 0 в одной точке. Обозначим vmax и vmin значения в стационарных точках на изоклине F (u, v) = 0, где достигаются локальные максимальное и минимальное значения переРис. 1. Фазовый портрет возбудимой системы после подачи внешнего стимула. Изоклины F (u, v) = 0 менной v, соответствующие расположеи G(u, v) = 0. Маршрут ABCD соответствует нию A и C на рис. 1. Единичный элеодиночному кратковременному импульсу. Интервал мент среды имеет на пересечении изоAB соответствует переднему фронту импульса, ин- клин F (u, v) = 0 и G(u, v) = 0 сотервал BC – возбужденной части импульса, интервал CD – заднему фронту импульса и интервал DA стояние равновесия (u0 , v0 ). Это состояние устойчиво по отношению к ма– невозбужденной части импульса лым возмущениям. В пространственнораспределенной системе однородное состояние с координатами u = u0 и v = v0 линейно устойчиво. Однако оно может быть неустойчиво к большим возмущениям, которые могут приводить к появлению неоднородного движения. В частности, в зависимости от возмущения можно получить различные возбудимые движения [17]. Будем рассматривать одиночные импульсы в однородной среде, одиночную спираль и спирально-волновой беспорядок. Для распространяющегося импульса обозначим время возбуждения через TAB , время гашения через TCD , длину переднего фронта волны через lAB , длину заднего фронта волны через lCD , длину возбужденной части через lBC и длину невозбужденной части импульса lDA . Сначала рассмотрим процессы подавления возбуждения в одномерной модели, а затем продемонстрируем их для двумерной. 2. Подавление возбуждения в одномерной среде Рассмотрим систему (1) в одномерном случае без диффузии медленной переменной. Одиночный элемент описывается моделью Зыкова [18]  ∂u   = F (u, v) + Du ∆u,  ∂t (3)  ∂v   = εG(u, v), ∂t где x ∈ [0, L], L – параметр, определяющий размер среды. F и G следующие кусочно-линейные функции:  −k1 u − v, u ≤ σ,     kf (u − a) − v, σ < u < b − σ, F (u, v) = (4)     k2 (b − u) − v, b − σ ≤ u, 64 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 { G(u, v) = kg u − v, kg u − v ≥ 0, kε (kg u − v), kg u − v < 0. (5) Здесь k1,2 – определим из непрерывности функции f (u, v) = F (u, v) + v, при u = σ и u = b − σ, то есть k1 = −kf (σ − a)/σ, k2 = kf (b − σ − a)/σ. Рассмотрим систему (3) c функциями (4) и (5) при следующих значениях параметров: a = 5, b = 20, kf = 0.4, kg = 1.7, σ = 0.2, kε = 6, ε ≪ 1. Выберем Du = 1 и L = 35, граничные условия считаем периодиче- Рис. 2. Фазовый портрет единичного элемента срескими u(x + L, t) = u(x, t). Кроме про- ды при ε = 0.001. Маршрут ABCD соответствует странственно однородного устойчивого одиночному кратковременному импульсу возбуждения состояния u(x, t) = u0 и v(x, t) = v 0 , при определенных условиях существует решение в виде распространяющегося импульса. Его форма определяется контуром ABCD на рис. 2. Маршрут ABCD состоит из четырех различных интервалов: • два участка быстрых движений: передний фронт импульса – AB и задний фронт импульса – CD; • два участка медленных движений: BC и DA, соответствующие возбужденному и не возбужденному состоянию среды, соответственно. Найдем условия, при которых одиночный импульс может быть подавлен относительно коротким импульсным воздействием. Это может быть достигнуто за счет введения внешней импульсной силы e(x, t) во второе уравнение системы (3).  ∂u   = F (u, v) + Du ∆u,  ∂t (6)  ∂v   = εG(u, v) + e(x, t), ∂t где   E0 , t ∈ ∆t, e(x, t) =  0, t ̸∈ ∆t. (7) Здесь ∆t – продолжительность внешнего импульса, E0 – его амплитуда. Существование малого параметра ε в системе (3) позволяет разделить все движения на быстрые и медленные. Медленная переменная v(x, t) имеет характерное время эволюции τv = 1/ε, а быстрая переменная u(x, t) может значительно измениться в течение намного более короткого промежутка времени. Поэтому для переднего фронта AB и заднего CD фронта импульса значения медленной переменной v в нулевой аппроксимации могут приниматься как константы: vf – соответствует переднему фронту и vb – соответствует заднему фронту. c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 65 Если время действия внешнего импульса ∆t короче, чем время возбуждения TAB или время релаксации TСD , то эти предположения могут быть применены к системе (6). Тогда, после подачи импульса, значение медленной переменной для переднего и заднего фронтов импульса, в нулевой аппроксимации для ε, тоже являются imp константами. Из второго уравнения (6) находим vf,b – значения медленной переменной на переднем и заднем фронтах, соответственно: imp vf,b = vf,b + E, (8) где E = E0 ∆t. В данном случае E0 ∆t – энергия внешнего воздействия. Модель, описывающую эволюцию переднего и заднего фронтов волны, можно представить как ∂u ∂2u imp (9) = F (u, vf,b ) + Du 2 . ∂t ∂x В зависимости от величины внешнего импульса будем различать три типа воздействий: слабое, среднее и сильное. В качестве слабого воздействия рассмотрим такой импульс, после подачи которого значение медленной переменной не выходит imp из интервала (vmin , vmax ). При постоянном значении vf,b уравнение (9) может иметь три пространственно однородных статических состояния с абсциссами imp f,b imp f,b imp uf,b 1 = −vf,b /k1 ∈ h− (v), u2 = a + vf,b /kf ∈ h0 (v), u3 = b − vf,b /k2 ∈ h+ (v), (10) imp определяемыми уравнением Fb ≡ F (u, vf,b ) = 0, где величина Vfimp определяет состояние равновесия для переднего фронта волны, а величина Vbimp определяет состояние равновесия для заднего фронта волны. При среднем внешнем воздействии может существовать только одно пространственно однородное устойчивое статическое состояние, координата которого u ∈ h− (v) или u ∈ h+ (v). При сильном внешнем воздействии устойчивых статических состояний нет. 2.1. Подавление возбуждения с помощью слабого внешнего воздействия. Пусть имеем слабое внешнее воздействие. Будем рассматривать случай когда f,b f,b f,b f,b uf,b 1 < u2 < u3 . Тогда два из этих установившихся состояний u1 и u3 устойчиf,b вы, а u2 неустойчиво. Поэтому (9) описывает две стационарные волны переключения: • волна от uf1 до uf3 , соответствующая переднему фронту; • волна от ub3 до ub1 , соответствующая заднему фронту. Профили этих волн переключения, распространяющиеся со скоростями cf,b , являются решениями уравнения −cf,b du d2 u imp = F (u, vf,b , ) + Du dξf,b dξf,b 2 (11) где ξf,b = x − cf,b t, cf – скорость переднего фронта, cb – скорость заднего фронта. Граничные условия du du =0 и =0 (12) dξf ξf =0 dξf ξf =T 66 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 для переднего фронта волны и du dξb ξb =0 =0 и du dξb ξb =T =0 (13) для заднего фронта волны. Решение для переднего и заднего фронтов волны может быть найдено при малых значениях σ. Тогда в (11) остается только ветвь h0 (v) imp imp F (u, vf,b ) = kf (u − a) − vf,b . (14) Система (11) с одним из граничных условий (12) или (13) является задачей на собственные числа для единственных значений скорости распространения cf,b . imp При этом значения vf,b определяют скорости и направления движения переднего и заднего фронтов распространяющегося импульса. Поэтому кратковременное импульсное воздействие может вызывать изменение значения медленной переменной на переднем и заднем фронтах бегущего импульса, что приводит к изменению скоростей распространения фронтов. Распространение переднего и заднего фронтов с различными скоростями может привести к дестабилизации распространяющегося импульса и переходу в состояние равновесия. Собственные значения задач (11), √ (12) и (11), (13) являются комплексно-сопряf,b женными и равны p1,2 = −cf,b /2 ± i 4kf − c2f,b /2. Тогда, для переднего и заднего фронтов волны получим выражения [ ] u(ξf,b ) = eγξf,b A1f,b cos δξf,b + A2f,b sin δξf,b + uf,b 2 , где γ = −cf,b /2, δ = √ (15) 4kf − c2f,b /2, A1f = uf1 − uf2 , A1b = ub3 − ub2 , A2f = = ((uf3 − uf2 )e−γT − A1f cos δT )/ sin δT , A2b = ((ub1 − ub2 )e−γT − A1b cos δT )/ sin δT . Скорости переднего и заднего фронтов c2f,b b − 1) af,b = , b 2 2 π + ln ( − 1) af,b 4kf ln2 ( (16) с продолжительностью √ Tf,b 1 =√ kf π2 + ln2 ( b − 1), af,b (17) где af = b − a − vfimp /kf , ab = a − vbimp /kf – имеют смысл порогового значения медленной переменной. Таким образом, подавая внешнее воздействие, можно изменять значения медленной переменной на переднем и заднем фронтах импульса и, как следствие, регулировать скорости распространения обоих фронтов (рис. 3). В зависимости от знака E0 можно выделить два случая подавления распространяющегося импульса: c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 67 • При E0 < 0 происходит дестабилизация импульса, распространяющегося в среде за счет увеличения возбужденной части BC и уменьшения невозбужденной части AD (см. рис. 2). Довозбуждение импульса обусловлено замедлением движения заднего фронта и ускорением движения переднего фронта (см. рис. 3). • При E0 > 0 происходит дестабилизация импульса, распространяющегося в среде за счет уменьшения возбужденной части BC и увеличения невозбужденной части AD (см. рис. 2). Гашение импульса обусловлено ускорением движения заднего фронта и замедлением движения переднего фронта (см. рис. 3). Подавление возбуждения возможно, только если амплитуда и продолжительность внешних импульсов достаточно велики. При этом одной из важнейших характеристик является ширина возбужденной части импульса lBC . Значение lBC может быть найдено из уравнения для медленных движений следующей системы:    F (u, v) = h+ (v) = 0, (18) dv   −c∗ = εG(u, v), dξf,b где c∗ – скорость распространения невозмущенного импульса. kg ) ε( lBC = ∗ 1 + . c k2 (19) Для невозмущенного распространяющегося импульса (E = 0) скорости переднего и заднего фронтов должны быть равны (cf = cb ). Тогда при фиксированных значениях параметров получаем следующее соотношение медленной переменной для переднего и заднего фронта импульса vb = (b − 2a)kf − vf . Воспользовавшись этим соотношением, получим выражение для скорости невозмущенного распространяющегося импульса ( a + v /k ) f f 4kf ln2 b − a − v /k f f (20) c2f,b = ( a + v /k ) . f f π2 + ln2 b − a − vf /kf Зная значения ширины возбужденной части импульса и обеих скоростей переднего и заднего фронтов, можно получить оценки продолжительности и амплитуды внешнего импульса, необходимого для подавления всех возбуждений в среде. T такое, что Найдем значение Ecr после подачи импульса в систему передний и задний фронты с течением некоторого времени T достигнут друг друга. В течение этого времени после прекращения подачи импульса в нулевой аппроксимации можно рассматривать новые распространяющиеся значения переднего cf и заднего cb фронтов как Рис. 3. Скорости переднего cf и заднего cb фронтов константы. Таким образом, можно найT , такое что передний и импульса в зависимости от E ти значение Ecr 68 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 задний фронты смогут пройти интервал, равный ширине исходного невозмущенного T . Для этого импульса lBC в течение времени T . Оценим критическую величину Ecr рассмотрим уравнение cf T + lBC = cb T, (21) где T ≤ TAB = lAB /c∗ и lBC могут быть найдены из (21). Время, в течение которого возбужденная часть импульса уменьшится до нуля, T = lBC . cb − cf (22) Приближенные значения времени подавления импульса могут быть найдены в зависимости от значения E, так как величины cb и cf зависят от E. Предполагая, что новые скорости возмущенного импульса остаются постоянными в течение T после этого распространявремени T ≤ TAB , оценка критической величины Ecr ющего импульса, может быть получена из (22) при T = TAB . Аналогичная оценка времени подавления импульса и критической величины амплитуды внешнего воздействия может быть получена для случая довозбуждения импульса. При этом в качестве необходимого условия для подавления импульса, можно рассматривать условие расширения возбужденной части импульса на всю среду. Общее время подавления импульса состоит из • времени, в течение которого вся среда окажется возбужденной, T1 = L − lBC , cf − cb (23) где L – размер среды, • времени перехода от возбужденного состояния среды к невозбужденному T2 = TCD + TDA . (24) Таким образом, общее время возбуждения может быть выражено как: T Tобщ = T1 + T2 , что является неявным заданием оценки критической величины Ecr при Tобщ = TAB . 2.2. Подавление возбуждения через параметрическое управление. Ранее мы использовали системное управление, то есть изменяли позиции в фазовом пространстве системы импульсным воздействием при постоянных значениях параметров среды. Теперь продемонстрируем подавление возбуждения через параметрическое управление, то есть будем изменять значение одного из системных параметров. Возьмем в качестве F (u, v) и G(u, v) следующие функции, описывающие модель ФитцХью–Нагумо: u3 F (u, v) = u − − v, (25) 3 G(u, v) = u − γv + β, (26) где β определяет асимметрию между возбуждением и восстановлением для каждого элемента, γ характеризует диссипацию медленной переменной v. В качестве управляющих параметров рассмотрим релаксационный параметр ε и параметр β. c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 69 Аналогично предыдущим случаям будем рассматривать только кратковременное импульсное воздействие. Тогда для параметрического ε-управления можно записать   E0 , t ∈ ∆t, ε ≡ ε(x, t) = (27)  0, t ̸∈ ∆t, и для β-управления β ≡ β(x, t) =   β0 + E0 t, t ∈ ∆t,  β , t ̸∈ ∆t. 0 (28) В (27) и (28) E0 и ∆t имеют тот же смысл, что и ранее. Импульсное изменение параметра β эквивалентно добавлению внешнего импульса ε во второе уравнение системы (3). Поэтому теория, рассмотренная ранее, может быть применена для случая параметрического управления. Изменение параметра β приводит к изменению значений медленных переменных в переднем и заднем фронтах распространяющихся волн. 2.3. Локализованное в одномерной среде импульсное воздействие. Изменение скоростей переднего и заднего фронтов может быть достигнуто за счет воздействия только на небольшой участок среды, содержащий передний или задний фронт волны. В этом случае импульсное воздействие может быть записано как   E0 , t ∈ ∆t, x ∈ ∆x, (29) e ≡ e(x, t) =  0, t ̸∈ ∆t, x ̸∈ ∆x, где ∆t – время воздействия, ∆x – область воздействия. Так как положения переднего и заднего фронтов соответствуют, например, положениям максимальных значений производных быстрой и медленной переменных, то время и область воздействия могут быть однозначно определены. 3. Численные эксперименты Проведены численные эксперименты для различных значений параметров E0 и ∆t. На рис. 4 представлены численные результаты для различных значений амплитуды E0 при фиксированной длительности ∆t = 10 приложенного внешнего импульса. Внешний импульс приложен в момент времени t = 0. В случае успешного довозбуждения импульса происходит расширение возбужденной части исходного импульса на всю среду, что приводит к его уничтожению и переходу среды в невозбужденное состояние. В случае успешного гашения импульса происходит сужение возбужденной части исходного импульса, что приводит к его уничтожению и переходу среды в невозбужденное состояние. В случае неуспешного подавления значение импульсного воздействия E = E0 ∆t недостаточно для дестабилизации начального импульса. Расширение возбужденной части исходного импульса, продолжающееся и 70 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 Рис. 4. Пространственно-временные диаграммы эволюции начального импульса (по горизонтали отложена пространственная координата, по вертикали – временная; темным цветом обозначены возбужденные области, светлым – невозбужденные области). Успешное довозбуждение импульса (а–в) при значениях амплитуды E0 : −5.5 (а), −3.5 (б), −2.5 (в). Успешное гашение импульса (ж–и) при значениях амплитудым E0 : 3.5 (ж), 3.59 (з), 3.6 (и). Неуспешное подавление (г–е): неуспешное довозбуждение при E0 = −0.5 (г); невозмущенная модель Зыкова при E0 = 0 (д); неуспешное гашение при E0 = 3.0 (е) после прекращения подачи внешнего воздействии, через некоторое время заканчивается, область возбуждения начинает сужаться и в итоге начальный импульс восстанавливается. Таким образом, подавление распространяющегося импульса возможно, T. если значение E больше некоторой критической величины Ecr 4. Подавление возбуждения в двумерной среде 4.1. Подавление одиночной спиральной волны. Рассмотрим двумерную модель с импульсным внешним воздействием  ( 2 ) ∂u ∂ u ∂2u   = F (u, v) + Du + 2 ,  c ∂t ∂x2 ∂y (30)  ∂v   = εG(u, v) + e(x, t), ∂t со cвободными граничными условиями ∂u ∂x и ∂u ∂y = 0, (31) c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 71 x=0, x=L =0 y=0, y=L где e(x, t) определено системой (7). Спиральная волна имеет ширину возбужденной части, которая является относительно постоянной вдали от ядра спиральной волны и от границ среды [19]. Предположим, что ширина возбужденной части волны совпадает с шириной импульса, распространяющегося в одномерной среде с тем же набором параметров. Заметим, что эволюция спиральной волны может рассматриваться как движение свободного конца спирали и движение переднего и заднего фронтов спиральной волны. Поэтому внешний импульс, применяемый к одиночной спиральной волне или пространственно-временному беспорядку, приведет к результатам, аналогичным полученным ранее для одномерной среды. Внешнее воздействие может привести к изменению скорости распространения переднего и заднего фронтов спиральной волны. Таким образом, изменится ширина возбужденной части, что приведет к дестабилизации и уничтожению спиральной волны. Проведены численные эксперименты для различных значений параметров E0 и ∆t. На рис. 5, 6 представлены результаты расчетов для различных значений амплитуды E0 при фиксированной длительности приложенного внешнего импульса ∆t = 10. Внешний импульс приложен в момент времени t = 0. В случае успешного довозбуждения происходит расширение возбужденной части импульса, что приводит к переходу среды в возбужденное состояние, дестабилизации и уничтожению спиральной волны и последующему переходу среды в невозбужденное состояние. В случае успешного гашения происходит сужение возбужденной части импульса, что приводит к дестабилизации и уничтожению спиральной волны и переходу среды в невозбужденное состояние. Рис. 5. Мгновенные распределения эволюции начального импульса (по горизонтали отложена пространственная координата x, по вертикали – пространственная координата y; темным цветом обозначены возбужденные области, светлым – невозбужденные области). Довозбуждение импульса при амплитуде E0 = −2.5 и временах t: а – 0, б – 4.0, в – 10.0, г – 17.0 Рис. 6. Мгновенные распределения эволюции начального импульса (по горизонтали отложена пространственная координата x, по вертикали – пространственная координата y; темным цветом обозначены возбужденные области, светлым – невозбужденные области). Гашение импульса при амплитуде E0 = 5.0 и временах t: а – 0, б – 2.0, в – 4.0, г – 7.0 72 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 4.2. Подавление спирально-волнового беспорядка. По схеме, изложенной в предыдущем пункте, были проведены численные эксперименты для различных значений параметров E0 и ∆t. На рис. 7, 8 приведены мгновенные распределения эволюции начального импульса u(x, y, 0). В случае успешного довозбуждения происходит расширение возбужденной части импульса, что приводит к переходу среды в возбужденное состояние, дестабилизации и уничтожению спирально-волнового беспорядка и последующему переходу среды в невозбужденное состояние. В случае успешного гашения происходит сужение возбужденной части импульса, что приводит к дестабилизации и уничтожению спирально-волнового беспорядка и переходу среды в невозбужденное состояние. Рис. 7. Мгновенные распределения эволюции начального импульса (по горизонтали отложена пространственная координата x, по вертикали – пространственная координата y; темным цветом обозначены возбужденные области, светлым – невозбужденные области). Довозбуждение импульса при амплитуде E0 = −3.0 и временах t: а – 0, б – 3.0, в – 11.0, г – 16.0 Рис. 8. Мгновенные распределения эволюции начального импульса (по горизонтали отложена пространственная координата x, по вертикали – пространственная координата y; темным цветом обозначены возбужденные области, светлым – невозбужденные области). Гашение импульса при амплитуде E0 = 4.5 и временах t: а – 0, б – 3.0, в – 5.0, г – 10.0 Заключение В работе предложены способы подавления нестатических состояний в возбудимых средах, таких как распространяющиеся импульсы, одиночная спиральная волна, спирально-волновой беспорядок. Было показано, что короткое импульсное воздействие может вызывать изменение значений медленной переменной переднего и заднего фронтов волн и соответствующее изменение скоростей фронтов. Распространение фронтов с различными скоростями ведет к дестабилизации распространяющегося импульса и переходу среды в невозбужденное состояние. Приведенные методы не требуют постоянного во времени применения воздействия. И в случае c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 73 довозбуждения и в случае гашения внешние воздействие не приводит явно к возбуждению или невозбуждению всей среды, а влияет только на передний и задний фронты. Теоретические исследования поддержаны грантом РНФ № 14-2-00811. Частично работа поддержана грантом (Соглашение от 27 августа 2013г. № 02.В.49.21.0003 между МОН РФ и ННГУ.) Библиографический список 1. Merkin J.H., Petrov V., Scott S.K., Showalter K. Wave-induced chemical chaos // Phys. Rev. Lett. 1996. Vol. 76, № 3. P. 546. 2. Zimmermann M.G., Firle S.O., Natiello M.A. et al. Pulse bifurcation and transition to spatio-temporal chaos in an exitable reaction-diﬀusion model // Physica D. 1997. Vol. 110. P. 92. 3. Резниченко Г.Ю. Математические модели в биофизике и экологии, Ижевск: Институт компьют. исследований, 2006. 184 с. 4. FitzHugh R.A. Impulses and physeological states in theoretical model of nerve membrane // Biophyse. J. 1961. P. 445. 5. Hodgikin A.L., Huxley A.F. A quantitative description of membrane current and its application conduction and excitation in nerve // J. Physiol. 1952. P. 500. 6. Zeeman E.C. Diﬀerential equations for the heartbeat and nerve impulses. Mathematical institute, Univer. of Warvick, Coventry, 1972. 7. Востриков В.А., Горбунов Б.Б. Сравнение биполярных импульсов, генерируемых внешними дефибрилляторами // Клиническая анестезиология и реаниматология. 2006. Т. 3, № 6. 8. Елькин Ю.Е., Москаленко А.В. Базовые механизмы аритмий сердца // Клиническая аритмология. Под ред. проф. А.В. Ардашева. М.: ИД Медпрактика-М, 1200:, 2009. C. 45. 9. Sridhar S., Duy-Manh Le, Yun-ChiehMi, Sinha S., Pik-Yin Lai, Chan C.K. Suppression of cardiac alternans by alternating-period-feedback stimulations // Physical Review. Vol. 2013. Vol. 87, № 4. 042712. 10. Konishi K., Takeuchi M., Shimizu T. Design of external forces for eliminating traveling wave in a piecewise linear FitzHugh–Nagumo model // Chaos: An Interdisciplinary Journal of Nonlinear Science. 2011. Vol. 21, № 2. 023101. 11. Sakurai T., Mihaliuk E., Chirila F., Showalter K. Design and control of wave propagation patterns in excitable media // Science. 2009(2002). Vol. 296, № 5575; Vilas C., Garcia M.R., Banga J.R., Alonso A.A. Robust feed-back control of distributed chemical reaction systems // Chemical engineering science. 2007. Vol. 62, № 11. P. 2941. 12. Yoneshima H., Konishi K., Kokame H. Symposium on nonlinear theory and its applications // Chaos: Proceedings of the International. 2008. Vol. 21, № 2. 023101. 13. Guo W., Qiao C., Zhang Z., Ouyang Q., Wang H. Spontaneous suppression of spiral turbulence based on feedback strategy // Physical Review. 2010. Vol. 81. 056214. 14. Sakaguchi H., Nakamura Y. Sample entropy of GPi neurons dependence on the level of alertness in 6OHDA rats // Journal of the Physical Society of Japan. 2010. 74 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 15. 16. 17. 18. 19. Vol. 79. 074802. Luther S., Fenton F.H., Kornreich B.G., Squires A., Bittihn P. Low-energy control of electrical turbulence in the heart // Nature. 2011. Vol. 475, № 7355. 235-9. Tandri H., Weinberg S.H., Chang K.C., Zhu R., Trayanova N.A., Tung L., Berger R.D. Reversible cardiac conduction block and deﬁbrillation with high-frequency electric ﬁeld // Science Translational Medicine. 2011. Vol. 3, № 102. 102ra96. Лоскутов А.Ю., Михайлов А.С. Введение в синергетику. М.: Наука, 1990. 272 с. Zykov V.S., Mikhailov A.S., Muller S.C. Controlling spiral waves in conﬁned geometries by global feedback // Phys. Rev. Lett. 1997. Vol. 78. Гулько А.Б., Петров А.А. Механизм образования замкнутых путей проведения в возбудимой среде // Биофизика. 1972. Т. 17, вып. 2. 270 с. Поступила в редакцию После доработки 30.01.2014 7.04.2014 THE SUPPRESSION OF THE EXCITATION OF THE ACTIVE MEDIUM WITH A WEAK EXTERNAL ACTION I. I. Bastrakov, K. A. Gavrilova, S. A. Grigorieva, G. V. Osipov Lobachevsky State University of Nizhni Novgorod This paper presents two new methods of suppressing an impulse in one-dimensional and two-dimensional excitable media using an external inﬂuence. In the proposed methods, we used short-impulse inﬂuence, leading to a change in velocity of the front , which in turn led to the destabilization of the propagating impulse and transition medium unexcited state. The studies were conducted on the Zykov model that a certain set of parameters is a model of an excitable medium. The conditions were determined for the amplitude and duration of the external inﬂuences required for suppressing excitation. Keywords: Nonlinear dynamics, active medium, spiral waves, Zykov model, excitation waves. Бастраков Илья Иванович – родился в 1992 году в Омутнинске. Получил степень бакалавра прикладной математики и информатики в 2013 году в Нижегородском государственном университете им. Н.И. Лобачевского. С 2013 года проходит обучение в магистратуре по специальности «Математическое моделирование» факультета ВМК ННГУ. Выполнял исследования по проектам ФЦП «Исследования и разработки по приоритетным направлениям развития научнотехнологического комплекса России на 2007–2013 годы». Им опубликовано 3 тезиса научных конференций. 603950 Нижний Новгород, пр-т Гагарина, д. 23 Нижегородский государственный университет имени Н. И. Лобачевского E-mail: ilya.bastrakov@gmail.com c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014 75 Гаврилова Ксения Андреевна – родилась в 1992 году в Дзержинске. Получила степень бакалавра прикладной математики и информатики в 2013 году в Нижегородском государственном университете им. Н.И. Лобачевского. С 2013 года проходит обучение в магистратуре по специальности «Математическое моделирование» факультета ВМК ННГУ. Выполняла исследования по проектам ФЦП «Исследования и разработки по приоритетным направлениям развития научно-технологического комплекса России на 2007–2013 годы». Ею опубликовано 2 тезиса научных конференций. 603950 Нижний Новгород, пр-т Гагарина, д. 23 Нижегородский государственный университет имени Н. И. Лобачевского E-mail: gavrilovakseniaan@gmail.com Григорьева Светлана Анатольевна – окончила механико-математический факультет Нижегородского государственного университета им. Н.И. Лобачевского с дипломами бакалавра (2006) и магистра (2008) математики. В 2008–2011 проходила обучение в аспирантуре на факультете Вычислительной математики и кибернетики Нижегородского государственного университета им. Н.И. Лобачевского по специальности «Дифференциальные уравнения, динамические системы и оптимальное управление». С 2008 года ассистент кафедры Теории управления и динамики машин факультета ВМК ННГУ. Участвует в разработке курса Methods of optimization для иностранных студентов 3-го курса специализации Information Technologies. Выполняет исследования по российским и международным проектам: ФЦП «Научные и научно-педагогические кадры инновационной России» на 2009–2013 годы, ФЦП «Исследования и разработки по приоритетным направлениям развития научно-технологического комплекса России на 2007–2013 годы». Принимала участие в работе российских и международных научных коллективов под руководством: проф. J. Suykens из Католического университета г. Левена, Бельгия; проф. C.K. Chan из Тайваньской Академии наук Sinika, Тайвань; проф. И.Р. Ефимова из Вашингтонского университета, США; проф. J. Kurths из Потсдамского института исследования климата, Германия. Научные исследования С.А. Григорьевой проводятся по математическому моделированию и оптимальному управлению динамики сложных систем. Ею опубликовано 7 научных трудов из них: 5 тезисов научных конференций, 1 методическая работа, также 1 статья в журнале Прикладной нелинейной динамики (ВАК). 603950 Нижний Новгород, пр-т Гагарина, д. 23 Нижегородский государственный университет имени Н. И. Лобачевского E-mail: sv.grigoryeva@gmail.com Осипов Григорий Владимирович – родился в 1960 году. Является высококвалифицированным преподавателем и выполняет все виды педагогической работы: читал лекции по общим курсам «Классическая механика» и «Теория колебания», ведет практические и лабораторные занятия, руководит курсовыми и дипломными работами студентов. Разработал специальные курсы «Синхронизация, структуры и хаос в нелинейных дискретных средах», «Численные методы в исследовании нелинейных систем» и «Возбудимые среды: динамика и управление». Опубликовал более 120 научных работ, а за последние 5 лет – 30 работ в ведущих отечественных и зарубежных журналах и в сборниках материалов нескольких международных конференций, на которых выступал с докладами в качестве приглашенного лектора. В 2004 году защитил докторскую диссертацию по теме «Синхронизация в неоднородных ансамблях локально диффузионно связанных регулярных и хаотических осцилляторов». За последние пять лет участвовал в выполнении двух международных (INTAS) и пяти российских (РФФИ) научных проектов (в трех из них – в качестве руководителя). Работал по приглашениям, в том числе как приглашенный профессор, зарубежных университетов в Дармштадте, Леувене, Тайбее, Потсдаме, Вене, Гонконге, Ланкастере и Бостоне. 603950 Нижний Новгород, пр-т Гагарина, д. 23 Нижегородский государственный университет имени Н. И. Лобачевского E-mail: grosipov@gmail.com 76 c И. И. Бастраков, К. А. Гаврилова, С. А. Григорьева, Г. В. Осипов ⃝ Изв. вузов «ПНД», т. 22, № 2, 2014
https://openalex.org/W2789733605	https://europepmc.org/articles/pmc5870158?pdf=render	English	null	Correction: Computing the T-matrix of a scattering object with multiple plane wave illuminations	Beilstein journal of nanotechnology	2,018	cc-by	244	Correction Address: 1Institute of Theoretical Solid State Physics, Karlsruhe Institute of Technology, Wolfgang-Gaede-Strasse 1, 76131 Karlsruhe, Germany, 2Institute of Nanotechnology, Karlsruhe Institute of Technology, P.O. Box 3640, 76021 Karlsruhe, Germany and 3Department of Physics, National Technical University of Athens, Heroon Polytechniou 9, 15780 Zografou, Greece This correction refers to Beilstein J. Nanotechnol. 2017, 8, 614–626. doi:10.3762/bjnano.8.66 This correction refers to Beilstein J. Nanotechnol. 2017, 8, 614–626. doi:10.3762/bjnano.8.66 This correction refers to Beilstein J. Nanotechnol. 2017, 8, 614–626. doi:10.3762/bjnano.8.66 In the original publication, the unnumbered equation that appears in the top right corner on page 624 before Equation 23 contains an error. The order of the svd vectors should be as follows: In the original publication, the unnumbered equation that appears in the top right corner on page 624 before Equation 23 contains an error. The order of the svd vectors should be as follows: Martin Fruhnert*1, Ivan Fernandez-Corbaton2, Vassilios Yannopapas3 and Carsten Rockstuhl1,2 Open Access Beilstein J. Nanotechnol. 2018, 9, 953. doi:10.3762/bjnano.9.88 Received: 27 February 2018 Accepted: 09 March 2018 Published: 22 March 2018 Guest Editor: H. Hahn © 2018 Fruhnert et al.; licensee Beilstein-Institut. License and terms: see end of document. License and Terms This is an Open Access article under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The authors would like to acknowledge Mr. Philipp Gutsche for pointing out this error. 953 953
https://openalex.org/W4392778144	https://jazindia.com/index.php/jaz/article/download/4348/3814	English	null	Design And Evaluation Of Self Emulsifying Mouth Dissolving Film Of Ranolazine By Solvent Casting Method	null	2,023	cc-by	8,677	Design And Evaluation Of Self Emulsifying Mouth Dissolving Film Of Ranolazine By Solvent Casting Method Corresponding Author: Deepak Kumar Jain Principal, Chetana College of Pharmacy, Infront of New Police Station, RLM Campus, Rithore, Khurai, Dist-Sagar, MP, 470117, Mail id: jaindeepak2022@gmail.com Keywords: Ranolazine, Self-emulsifying, Mouth dissolving film, Pseudo ternary phase Factorial design, Solvent casting method Journal of Advanced Zoology ISSN: 0253-7214 Volume 44 Issue -5 Year 2023 Page 1433-1445 When a dosage form must have a quicker onset of action and be suitable for administration, ranolazine self-emulsifying mouth dissolving film (SEMDF) might be viewed as an anti-anginal formulation. Keywords: Ranolazine, Self-emulsifying, Mouth dissolving film, Pseudo ternary phase Factorial design, Solvent casting method Journal of Advanced Zoology ISSN: 0253-7214 Volume 44 Issue -5 Year 2023 Page 1433-1445 Journal of Advanced Zoology ISSN: 0253-7214 Volume 44 Issue -5 Year 2023 Page 1433-1445 Design And Evaluation Of Self Emulsifying Mouth Dissolving Film Of Ranolazine By Solvent Casting Method Neha Dipak Desai1, Dr Shweta Mishra2, Rahul Chaurasia3, Dikhsha Jat4, Manoj Kumari More4, Namrata Soni4, Deepak Kumar Jain5 1Ashokrao Mane College of Pharmacy, Front of KTM, Peth Vadgaon, Maharashtra 416112 2Professor and Head of Department (Pharmacology), Guru Ramdas Khalsa Institute of Science and Technology (Pharmacy), Barela, Kukrikheda, Jabalpur, MP, 483001 3Bhagyoday Tirth Pharmacy College, Khurai Road, Achrya Vidhya Marg, Sagar, Mp, 470002 4Oriental College of Pharmacy, Raisen Rd, Patel Nagar, Bhopal, MP, 462022 5Chetana College of Pharmacy, Infront of New Police Station, RLM Campus, Rithore, Khurai, Dist-Sagar, MP, 470117 Corresponding Author: Deepak Kumar Jain Principal, Chetana College of Pharmacy, Infront of New Police Station, RLM Campus, Rithore, Khurai, Dist-Sagar, MP, 470117, Mail id: jaindeepak2022@gmail.com CC License CC-BY-NC-SA 4.0 ABSTRACT A novel self-emulsifying mouth dissolving film (SEMDF) containing ranolazine (RZ) is being produced in the current study with the aid of a mouth dissolving film (MDF) mixed with self-emulsifying components. Using a solvent casting process, the films for ranolazine were made from the water-soluble polymer HPMC K15M. Ethyl oleate was used as the oil phase, Tween 80 as the surfactant, PEG 400 as the co-surfactant, and distilled water as the solvent to create the pseudoternary phase diagram (aqueous phase). There are ten different possible surfactant mixture to oil combinations with different Km values for the phase diagram investigation of RZ SEDDS (1, 2, 3, and 4 were employed). The phase diagram at Km value 3 shows better microemulsion existence zones when compared to Km values 1, 2, and 4. In this study, a 32-factorial design was used to evaluate two factors at each of three levels, and experimental batches were conducted in all conceivable combinations. In testing of their physical characteristics, such as uniformity of weight, thickness, folding durability, drug content uniformity, surface pH, and tensile strength, the developed mouth- dissolving films functioned satisfactorily. The formulations underwent disintegration, in-vitro drug release testing, and stability studies. The FTIR and DSC analyses showed no physicochemical interaction between the excipients and the medicine. F5 showed a maximal drug release of 93.85% at 5 minutes. Studies on stability demonstrated the dependability of the modified formulation. INTRODUCTION More than 40% of recently developed chemical entities are essentially insoluble in water. To be absorbed, the medication needs to be present in solution form at the absorption site. Solubility, then, represents a considerable challenge for formulation scientists1. Poor solubility and ineffective medication absorption contribute to low bioavailability, which also jeopardises the product's efficacy and safety2. The complex structure, size, high molecular weight, high lipophilicity, compound H-bonding to solvent, intramolecular H- bonding, intermolecular H-bonding (crystal packing), crystallinity, polymorphic forms, ionic charge status, pH, and salt form of drugs are a few physicochemical factors that contribute to their poor solubility3. To improve their solubility, dissolution rate, and absorption, various methods have been used, including particle size reduction, nanonization, Co-solvency, Hydrotropy, pH adjustment, so no-crystallization, the supercritical fluid (SCF) process, solid dispersion, inclusion complexation, self-emulsifying or self-micro emulsifying systems, liquid-solid methods, etc.4. Self-emulsifying drug delivery systems are a popular and economically sound formulation choice for tackling these problems. SMEDDs can dramatically improve the oral bioavailability and solubility of medications that are only marginally water soluble5. The primary elements of SMEDDS are drug, oil, surfactant, co-surfactant, and co-solvents. This dosage form's basic concept is to interfere with drug absorption to increase bioavailability by forming microscopic oil-in-water (o/w) microemulsions with light agitation after aqueous phase dilution6, 7. Mouth dissolving film (MDF) is a convenient dosage form that dissolves in the mouth without chewing in a few minutes, enhancing the onset of therapeutic action and increasing treatment efficacy. The objective of this work was to develop a novel self- emulsifying mouth dissolving film (SEMDF) based on an MDF with integrated self-emulsifying components. Quick drug release, great potential for improving oral dissolution and bioavailability of poorly water-soluble drugs, no need for water during administration, the potential for taste masking, the absence of choking risk, high patient compliance, flexibility and portability for ease of handling, and the avoidance of first past metabolism are all advantages of SEMDF8. An anti-anginal medication called ranolazine (RZ) is used to treat a number of cardiovascular conditions. The biochemical classification system places it in class II, which is characterised by low solubility and high permeability. Due to its low solubility in biological fluids, the drug's poor oral bioavailability is one of its key problems. Ranolazine poor dissolution and limited solubility reduce its bioavailability. Therefore, it is critical to improve ranolazine solubility and water dissolution for medicinal purposes. INTRODUCTION Ranolazine aqueous solubility and dissolution can be improved by formulating it in SEDDS9. ABSTRACT A novel self-emulsifying mouth dissolving film (SEMDF) containing ranolazine (RZ) is being produced in the current study with the aid of a mouth dissolving film (MDF) mixed with self-emulsifying components. Using a solvent casting process, the films for ranolazine were made from the water-soluble polymer HPMC K15M. Ethyl oleate was used as the oil phase, Tween 80 as the surfactant, PEG 400 as the co-surfactant, and distilled water as the solvent to create the pseudoternary phase diagram (aqueous phase). There are ten different possible surfactant mixture to oil combinations with different Km values for the phase diagram investigation of RZ SEDDS (1, 2, 3, and 4 were employed). The phase diagram at Km value 3 shows better microemulsion existence zones when compared to Km values 1, 2, and 4. In this study, a 32-factorial design was used to evaluate two factors at each of three levels, and experimental batches were conducted in all conceivable combinations. In testing of their physical characteristics, such as uniformity of weight, thickness, folding durability, drug content uniformity, surface pH, and tensile strength, the developed mouth- dissolving films functioned satisfactorily. The formulations underwent disintegration, in-vitro drug release testing, and stability studies. The FTIR and DSC analyses showed no physicochemical interaction between the excipients and the medicine. F5 showed a maximal drug release of 93.85% at 5 minutes. Studies on stability demonstrated the dependability of the modified formulation. When a dosage form must have a quicker onset of action and be suitable for administration, ranolazine self-emulsifying mouth dissolving film (SEMDF) might be viewed as an anti-anginal formulation. Keywords: Ranolazine, Self-emulsifying, Mouth dissolving film, Pseudo ternary phase Factorial design, Solvent casting method 1433 Available online at: https://jazindia.com Journal Of Advance Zoology Methods Determination of saturation solubility of ranolazine in oils, surfactants and co-surfactants Oils such as oleic acid, ethyl oleate, isopropyl myristate, and castor oil, as well as surfactants like Tween 80, Tween 20, and Span 20, and cosurfactants like PEG 200, PEG 400, and PEG 600, were all used to test ranolazine saturation solubility. Excess RZ was added to a vial holding 2mL of each selected solvent. Following sealing, the liquid was vortexed using a cyclomixer for 10min to make sure that the vehicles and ranolazine were properly mixed. Mixtures were kept at room temperature for 72hours to attain equilibrium in an orbital shaking incubator before being centrifuged at 2000 rpm for 15 minutes. Mixtures were kept at room temperature for 72 hours to attain equilibrium in an orbital shaking incubator before being centrifuged at 2000 rpm for 15 minutes and filtered through a membrane filter (0.45 mm). After aliquots of supernatant were diluted with methanol, the drug concentration was assessed using a UV-visible double beam spectrophotometer (Jasco V-630) against methanol as a blank solution at max 272 nm. Three measurements were taken for each measurement10, 11. Materials We received a complimentary sample of ranolazine from Unichem Laboratories Ltd. in Goa, India. The following ingredients were purchased from Loba Chemie Pvt. Ltd., Mumbai: ethyl oleate, PEG 400, PEG 600, isopropyl myristate, and tween 80. The supplier of the hydroxy propyl methyl cellulose (K 15M) was Colorcorn Asia Pvt. Ltd. in Verna, Goa. From Molychem Pvt. Ltd., Mumbai, we acquired citric acid and propylene glycol. From Mumbai's Hi Media Pvt. Ltd., aspartame was purchased. Preparation of liquid self-emulsifying drug delivery system (SEDDS) p q y g g y y The phase diagrams were created using a variety of Km values, and the Km value that produced the largest microemulsion area was selected for further study. Four formulations were selected from this microemulsion zone and used in additional evaluation experiments. Calculations were made to determine the proportions of water, oil, and surfactant/co-surfactant concentration in each formulation. Weighing out the proper proportions of oil, surfactant, and co-surfactant, they were then combined with gentle stirring. According to its solubility in the formulation amount, ranolazine was dissolved into the mixture of oil and surfactants. The ranolazine was then thoroughly dissolved by combining the components at 37oC were using a combination of gentle stirring and vortex mixing. After sealing the glass vial, the solution was added there and stored until required13. Thermodynamic stability study of SEDDS y y y Using freeze-thaw, the formulations' stability was evaluated. Three to four cycles of freezing and thawing, including a 24-hour period of freezing at - 4°C and a 24-hour period of thawing at 40°C, were performed on the formulations. At 3000 rpm, 5 minutes of centrifugation were completed. The formulations were then examined for evidence of phase separation. For subsequent research, only formulations that could tolerate phase separation were chosen15. Globule size determination p p Using the lesser light scattering method, the Malvern zetasizer was utilised to gauge the microemulsions average droplet size, size distribution, and polydispersity index16 (Nano-ZS, Malvern Instruments, and U. K). Evaluation of liquid self-emulsifying drug delivery system (SEDDS) Dilution and self-emulsification study y In this study, selected formulations were diluted in distilled water at several ratios (1: 10, 1:50, and 1:100), and the formulations were then rated visually. The SEDDS's emulsification time was calculated using the USP Dissolution Test Apparatus Type II. 300 mg of each formulation were added drop wise to 500 ml of pure water that had been boiled to 37°C. A simple dissolving paddle made of stainless steel that rotated at 50 rpm offered light agitation. A visual measurement of emulsification time was made14. Pseudo-ternary phase diagram Available online at: https://jazindia.com 1434 Using the water titration method at room temperature, the homogenous liquid mixture of oil, surfactant, and co-surfactant was added drop by drop to produce the pseudo-ternary phase diagrams. Based on the outcomes 1434 Journal Of Advance Zoology of solubility tests and excipients screening, ethyl oleate, tween 80, and PEG 400 were selected as the oily phase, surfactant, and co-surfactant, respectively. The appropriate mixture (Smix), corresponding to the chosen surfactant to co-surfactant ratio (Km), was formed. At the necessary Km values (1:1, 2:1, 3:1, and 4:1), Smix and oil were blended in a test tube at the following ratios: 0.5:1, 1:1, 1.5:1, 2:1, 2.5:1, 3:1, 3.5:1, 4:1, 4.5:1, and 5:1. (Table V and Fig.1 a-d). Double distilled water was gradually added to the resulting mixtures until the first sign of turbidity appeared in order to ascertain the endpoint and equilibrium. The water addition was then restarted if the system became clear. Visual inspection of the combinations' ability to flow and exhibit distinct phases was performed once perfect equilibrium had been reached12. Zeta potential analysis By using electrophoretic mobility in a Malvern Zetasizer device (Malvern Instruments, UK) outfitted with the appropriate software and calibrated with the provided standard, the zeta potential of the Microemulsion droplet surface was ascertained. At 25ºC three measurements are made back-to-back with a constant cell drive of 150 mV. Using the dielectric constants and viscosity of the dispersion medium, the Smoluchowsky equation transforms the electrophoretic mobility into zeta potential values16. Available online at: https://jazindia.com Preparation of self emulsifying mouth dissolving film The solvent casting method was used to create the ranolazine mouth dissolving film. Cold water was used to dissolve water-soluble polymers, such as HPMC K15M, and create a homogeneous viscous solution while being simultaneously stirred at 1000 rpm. The viscous mixture is maintained at room temperature. Following this, an emulsion comprising the API ranolazine, the plasticizer glycerol, and other chemicals such as aspartame was added. It was combined with a flavoring agent (orange flavor) and citric acid. For defoaming, the final film solution was cast on a typical petri dish. For three hours, it was dried in the hot air oven at 40° C. The film was carefully taken out of the petri dish, examined for flaws, and sliced to the proper size (2x2cm2) per strip to administer the prescribed dose. The samples were placed in a desiccator until further analysis17 Table III. Available online at: https://jazindia.com 1435 Journal Of Advance Zoology y g g The physical parameters of the prepared self emulsifying mouth dissolving films such as weight variation, thickness, tensile strength, folding endurance and surface pH of the film were calculated and reported19,20. Y=b0+b1X1+b2X2+b12X1X2+b11X1X1+b22X2X2+є Characterization of self-emulsifying mouth dissolving films Table II: Amount of variables in a 32 factorial design Coded Level -1 0 +1 HPMC K15M (X1) 175 225 275 Glycerol (X2) 1 1.25 1.50 Table III: Formulation of self emulsifying mouth dissolving film Components F1 F2 F3 F4 F5 F6 F7 F8 F9 RZ Emulsion (mL) 5 5 5 5 5 5 5 5 5 HPMC K 15 (mg) 175 175 175 225 225 225 275 275 275 Glycerol (mL) 1 1.25 1.50 1 1.25 1.50 1 1.25 1.50 Citric acid (mg) 50 50 50 50 50 50 50 50 50 Aspartame (mg) 10 10 10 10 10 10 10 10 10 Water (mL) 10 10 10 10 10 10 10 10 10 Orange colour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Orange flavour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Characterization of self-emulsifying mouth dissolving films Table II: Amount of variables in a 32 factorial design Coded Level -1 0 +1 HPMC K15M (X1) 175 225 275 Glycerol (X2) 1 1.25 1.50 Table II: Amount of variables in a 32 factorial design Table III: Formulation of self emulsifying mouth dissolving film Components F1 F2 F3 F4 F5 F6 F7 F8 F9 RZ Emulsion (mL) 5 5 5 5 5 5 5 5 5 HPMC K 15 (mg) 175 175 175 225 225 225 275 275 275 Glycerol (mL) 1 1.25 1.50 1 1.25 1.50 1 1.25 1.50 Citric acid (mg) 50 50 50 50 50 50 50 50 50 Aspartame (mg) 10 10 10 10 10 10 10 10 10 Water (mL) 10 10 10 10 10 10 10 10 10 Orange colour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Orange flavour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Y=b0+b1X1+b2X2+b12X1X2+b11X1X1+b22X2X2+є Y=b0+b1X1+b2X2+b12X1X2+b11X1X1+b22X2X2+є In the equation above, Y is the dependent variable, and b0 is the arithmetic mean response of the nine trials. Bi (b1, b2, b12, b11, and b22) is the estimated coefficient for the corresponding factor Xi (X1, X2, X1X2, X12, and X22), which represents the average result of changing one factor at a time from its low to high value. The interaction term demonstrates how the answer changes when two factors are changed at once (X1X2). The polynomial terms (X1X1 and X2X2) are provided to analyse the nonlinearity. The symbol є denotes random error 18. Table I: Full factorial experimental design layout Trials Variable levels in coded form X1 X2 1 -1 -1 2 -1 0 3 -1 +1 4 0 -1 5 0 0 6 0 +1 7 1 -1 8 1 0 9 1 +1 Table I: Full factorial experimental design layout Trials Variable levels in coded form X1 X2 1 -1 -1 2 -1 0 3 -1 +1 4 0 -1 5 0 0 6 0 +1 7 1 -1 8 1 0 9 1 +1 1 -1 -1 2 -1 0 3 -1 +1 4 0 -1 5 0 0 6 0 +1 7 1 -1 8 1 0 9 1 +1 Table II: Amount of variables in a 32 factorial design Coded Level -1 0 +1 HPMC K15M (X1) 175 225 275 Glycerol (X2) 1 1.25 1.50 Table III: Formulation of self emulsifying mouth dissolving film Components F1 F2 F3 F4 F5 F6 F7 F8 F9 RZ Emulsion (mL) 5 5 5 5 5 5 5 5 5 HPMC K 15 (mg) 175 175 175 225 225 225 275 275 275 Glycerol (mL) 1 1.25 1.50 1 1.25 1.50 1 1.25 1.50 Citric acid (mg) 50 50 50 50 50 50 50 50 50 Aspartame (mg) 10 10 10 10 10 10 10 10 10 Water (mL) 10 10 10 10 10 10 10 10 10 Orange colour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Orange flavour (mg) q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. q.s. Factorial design g In this work, a 32-factorial design was employed to examine two factors at each of three levels, and experimental batches were run in all possible combinations. Tensile strength and cumulative% drug release were chosen as dependent variables, whilst HPMC K4M (X1) and glycerol (X2) amounts were chosen as independent factors. In PCP Disso 2.08, 3-D response surface methodology was applied to the data to ascertain the impact of the types and concentrations of polymers on the various dependent variables Table I displays the whole factorial experimental design arrangement. Table II shows the values of the variables in a 32Factorial Design. The responses were computed using a statistical model with interactive and polynomial terms. Stability studies In the present study, stability studies were carried out for a specific time period up to 30 days for selected formulations, ambient temperature and humidity 40°C ± 2°C/75% ± 5%RH in stability chamber for 30 days. After 30 days sample were removed and characterized for tensile strength, % drug content and cumulative % drug release of optimized formulation9,19,20. Drug content uniformity The four corners and the centre of the moulded film (n = 3) were sliced into three film strips (2 × 2 cm2). A separate conical flask containing 100 mL of distilled water was used for each film strip. For two hours, the flasks were shaken in a mechanical shaker. In a UV-Visible spectrophotometer, all of the solutions were filtered and examined at a wavelength of 272 nm19, 20. In-vitro dissolution studies Utilizing a USP dissolving apparatus I (basket device) in 300mL of simulated saliva fluid (pH 6.8) held at 37±0.5°C and agitated at 50 rpm, the dissolution investigation was conducted. The film was divided into patches measuring (2 × 2 cm2). At intervals of 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 minutes, 2 mL samples were taken, filtered, and subjected to spectrophotometric analysis at 272 nm in a UV Spectrophotometer19,20. Differential scanning calorimetry (DSC) Thermograms of the physical combination and the Optimized formulation were acquired using an intracooler-equipped DSC (Pyris Diamond TG/DTA, Make-Perkin Elmer). Alpha alumina powder and a platinum crucible are used as a standard for calibrating the DSC temperature and enthalpy scale. The 2-10 mg powder samples were hermetically sealed in an aluminum pan and heated steadily to 10°C19, 20. Characterization of self-emulsifying mouth dissolving films Characterization of self-emulsifying mouth dissolving films The physical parameters of the prepared self emulsifying mouth dissolving films such as weight variation, thickness, tensile strength, folding endurance and surface pH of the film were calculated and reported19,20. Available online at: https://jazindia.com 1436 Journal Of Advance Zoology Journal Of Advance Zoology Journal Of Advance Zoology FTIR (Fourier transform infrared spectroscopy) In order to identify any potential interactions between the API and the excipients used, FTIR experiments were conducted. Using the KBr dispersion method, a Fourier transform infrared spectrophotometer (Jasco V- 530 model) was used to obtain the IR absorption spectra of ranolazine. In a nutshell, 2 mg of the sample was fully ground with previously dried KBr at 120°C for 30 minutes. The drug sample was then uniformly mixed with the ground sample and maintained in the sample holder, and spectra were recorded over the wave number 400-4000 cm-1. The infrared Spectra of the Optimised batch, physical mixture of formulation, and API were recorded19, 20. Scanning electron microscopy (SEM) Using a 20 kV SEM from JEOL in Japan, the exterior macroscopic structure (morphology) of the Film was examined. The sample was glued to a SEM stub before a small layer of gold or platinum was applied19, 20. In-vitro disintegration test A petri plate disintegration test was conducted. Each batch's film sample (2×2 cm2) was put in 10mL of simulated saliva. A film begins to shatter or dissolve at the disintegration time (n=3) 19, 20. RESULTS AND DISCUSSION Selection and screening of potential emulsion components (surfactant, co surfactant and oil), from solubility data (Table IV), Ethyl oleate shows good solubilizing power for ranolazine (46.96 mg/mL) amongst other oils investigated. In case of surfactant, Tween 80 has more solubilizing capacity (48.094 mg/mL) of drug followed by PEG 400 (36.65 mg/mL). Available online at: https://jazindia.com 1437 Table IV: Solubility of RZ in oils, surfactants and co-surfactants Type of oil Solubility (mg/mL)* Oleic acid 42.653  0.025 Ethyl oleate 46.960  0.007 Isopropyl myristate 16.830  0.019 Castor oil 42.248  0.017 Type of surfactants Tween 80 48.094  0.023 Table IV: Solubility of RZ in oils, surfactants and co-surfactants 1437 Available online at: https://jazindia.com Journal Of Advance Zoology Tween 20 34.804  0.034 Span 20 36.344  0.019 Type of co-surfactants PEG 400 36.65  0.022 PEG 200 25.286  0.017 PEG 600 23.376  0.013 Propylene glycol 24.58  0.007 Indicates average triplicates ±SD (n=3) Construction of pseudoternary phase diagram Construction of pseudoternary phase diagram Oil, a surfactant, and a co-surfactant were chosen for microemulsion formulation based on the findings of solubility investigations. The phase diagram analysis of RZ SEDDS utilised ten alternative combinations of surfactant mixture to oil at various Km values (1, 2, 3, and 4) (Table V). Oil-in-water (o/w) and water-in-oil (w/o) did not clearly convert into one another. Each phase diagram revealed the o/w microemulsions boundary layer (Fig.1 a-d). Table V: Composition of ethyl oleate/tween 80 /peg 400/water at Km=1,2,3,4 Sr. No. Smix (mL) Oil (mL) Water(mL) Km=1 Km=2 Km=3 Km=4 1 0.5 1 0.4 0.2 0.1 0.3 2 1 1 0.9 0.9 0.7 0.9 3 1.5 1 1.8 2.5 1.9 1.2 4 2 1 2.1 3.7 3.1 1.4 5 2.5 1 2.5 4.9 3.7 1.8 6 3 1 2.8 5.8 5.3 2.1 7 3.5 1 3.5 6.7 6.5 2.6 8 4 1 2.6 7.5 7.1 3.5 9 4.5 1 4.1 9.1 7.5 3.8 10 5 1 5.6 10.5 9.5 3.9 1.a) Pseudoternary phase diagram at Km=1 1.b)Pseudoternary phase diagram at Km=2 Table V: Composition of ethyl oleate/tween 80 /peg 400/water at Km=1,2,3,4 1.a) Pseudoternary phase diagram at Km=1 1.b)Pseudoternary phase diagram at Km=2 1.a) Pseudoternary phase diagram at Km=1 1.b)Pseudoternary phase diagram at Km=2 1438 Available online at: https://jazindia.com Journal Of Advance Zoology 1.c) Pseudoternary phase diagram at Km=31.d) Pseudoternary phase diagram at Km=4 Fig.1: Pseudoternary phase diagram All the combinations of km i.e. surfactant and co-surfactant ratio in certain different concentrations were taken and constructed as pseudo ternary phase diagram. But the diagram shows highest water absorption i.e. wider self micro emulsifying region was considered to be a better combination (Km) in terms of self-micro emulsification efficiency. The phase diagram at Km value 3 show better microemulsion existence regions than 1, 2 and 4 and value didn’t show further increase in microemulsion existence region which shown in Fig. 2. So, Km 3 system was selected as final formulation of SEDDS. Journal Of Advance Zoology Journal Of Advance Zoology 1.c) Pseudoternary phase diagram at Km=31.d) Pseudoternary phase diagram at Km=4 Fig.1: Pseudoternary phase diagram 1.c) Pseudoternary phase diagram at Km=31.d) Pseudoternary phase diagram at Km=4 Fig.1: Pseudoternary phase diagram All the combinations of km i.e. surfactant and co-surfactant ratio in certain different concentrations were taken and constructed as pseudo ternary phase diagram. But the diagram shows highest water absorption i.e. Construction of pseudoternary phase diagram wider self micro emulsifying region was considered to be a better combination (Km) in terms of self-micro emulsification efficiency. The phase diagram at Km value 3 show better microemulsion existence regions than 1, 2 and 4 and value didn’t show further increase in microemulsion existence region which shown in Fig. 2. So, Km 3 system was selected as final formulation of SEDDS. Fig. 2: Selected composition of formulation RZ 1 to RZ 4 Fig. 2: Selected composition of formulation RZ 1 to RZ 4 Globule size determination It was discovered that the optimised batch RZ 3 illustrated in Fig. 3 had dimensions of 14.72 nm and 0.277, respectively. All of the reconstituted SEDDS had extremely small average droplet sizes, all of which were in the nanometer range (<100 nm). Globules with a size in the nano- or micron range are more transparent and have a larger surface area, which is important for dividing drugs between oil and water. Results for droplet size are shown in Table VIII. All formulations have polydispersibility indices that are less than 1, which indicates that globules are distributed uniformly throughout the formulation. Evaluation of liquid self emulsifying drug delivery system (SEDDS) Dilution & self emulsification study The self-emulsification of SEDDS in 100mL double distilled water at 37oC with gentle agitation was evaluated visually. After two hours of storage, as indicated in Table VI, all solid SEDDS batches displayed spontaneous emulsification and showed no indication of phase separation or phase inversion of microemulsion. The grade of the optimized batch was A, indicating quick formation of a clear microemulsion with a 20-second self-microemulsion formation time. Table VI: Dilution & self-emulsification time study of solid SEDDS batches Sr. No. Batch code Dilution study Emulsification time (S) 1:10 1:50 1:500 1 RZ 1 B A A 50 2 RZ 2 B A A 35 3 RZ 3 A A A 20 4 RZ 4 B A A 22 Available online at: https://jazindia.com Available online at: https://jazindia.com Available online at: https://jazindia.com 1439 Journal Of Advance Zoology Journal Of Advance Zoology Thermodynamic stability study of SEDDS Thermodynamic stability study of SEDDS RZ 1, RZ 2, RZ 3 showed better results but in case of RZ 4 there was slightly precipitation occurred after centrifugation. Liquid SEDDS formulation showed good storage stability at different temperature shown in Table VII (+ Phase separation, + + Drug precipitation, - No phase separation, -- No precipitation) Thermodynamic stability study of SEDDS RZ 1, RZ 2, RZ 3 showed better results but in case of RZ 4 there was slightly precipitation occurred after centrifugation. Liquid SEDDS formulation showed good storage stability at different temperature shown in Table VII (+ Phase separation, + + Drug precipitation, - No phase separation, -- No precipitation) Table VII: Thermodynamic stability study of SEDDS Sr. No. Batch code Thermodynamic Stability at 4°C at 40°C After centrifugation 1 RZ 1 -,-- -,-- -,-- 2 RZ 2 -,-- -,-- -,-- 3 RZ 3 -,-- -,-- -,-- 4 RZ 4 -,-- -,+ -,++ Table VII: Thermodynamic stability study of SEDDS Globule size determination Globule size determination Because it affects both drug release and absorption, the droplet size of the emulsion is a key component in how well it performs self-emulsification. Additionally, it has been suggested that the emulsion droplets' lower particle size may promote quicker absorption and increase bioavailability (Table VIII). Table VIII: Globule size of SEDDS formulation Sr. No. Batch Code Mean particle size (nm) Polydispersibility index (PDI) 1 RZ 1 21.50 0.177 2 RZ 2 16.79 0.315 3 RZ 3 14.72 0.277 4 RZ 4 18.76 0.278 Fig. 3: Particle size distribution of optimized formulation RZ 3 The globule size determination study was used to calculate the mean particle size and polydispersity index. It was discovered that the optimised batch RZ 3 illustrated in Fig. 3 had dimensions of 14.72 nm and 0.277, respectively. All of the reconstituted SEDDS had extremely small average droplet sizes, all of which were in the nanometer range (<100 nm). Globules with a size in the nano- or micron range are more transparent and have a larger surface area, which is important for dividing drugs between oil and water. Results for droplet size are shown in Table VIII. All formulations have polydispersibility indices that are less than 1, which indicates that globules are distributed uniformly throughout the formulation. Table VIII: Globule size of SEDDS formulation Sr. No. Batch Code Mean particle size (nm) Polydispersibility index (PDI) 1 RZ 1 21.50 0.177 2 RZ 2 16.79 0.315 3 RZ 3 14.72 0.277 4 RZ 4 18.76 0.278 Table VIII: Globule size of SEDDS formulation Fig. 3: Particle size distribution of optimized formulation RZ 3 The globule size determination study was used to calculate the mean particle size and polydispersity index. It was discovered that the optimised batch RZ 3 illustrated in Fig. 3 had dimensions of 14.72 nm and 0.277, respectively. All of the reconstituted SEDDS had extremely small average droplet sizes, all of which were in the nanometer range (<100 nm). Globules with a size in the nano- or micron range are more transparent and have a larger surface area, which is important for dividing drugs between oil and water. Results for droplet size are shown in Table VIII. All formulations have polydispersibility indices that are less than 1, which indicates that globules are distributed uniformly throughout the formulation. The globule size determination study was used to calculate the mean particle size and polydispersity index. Zeta potential analysis Table IX displays the zeta potential values of the diluted SEDDS formulations. The examined SEDDS formulations zeta potential showed a large, substantial variance. All SEDDS batches, with the exception of RZ 3, had zeta potential values that ranged from -0.3 to -0.4 mV. RZ 3 was the only batch to have a mean zeta potential of -38.21 mV, indicating that it was more stable than the other batches. 1440 Journal Of Advance Zoology Table IX: Zeta potential of SEDDS formulation Sr. No. Batch code Mean zeta potential (mV) 1 RZ 1 -0.411 2 RZ 2 -0.365 3 RZ 3 -38.21 4 RZ 4 -0.441 Table IX: Zeta potential of SEDDS formulation Mean zeta potential Table IX: Zeta potential of SEDDS formulation 2 RZ 2 -0.365 3 RZ 3 -38.21 4 RZ 4 -0.441 Fig. 4: Zeta potential of optimized batch RZ 3 Characterization of self emulsifying mouth dissolving films The weight of the film increases together with the weight of the polymer. The films from F1 through F9 ranged in weight from 162mg to 277mg. The F3 formulation had the lightest weight of the film, and the F7 formulation had the heaviest weight of the film, as stated in Table X. The homogeneous weight of the films is indicated by the low standard deviation figures. The film's thickness was discovered in ascending sequence. As polymer concentration rises, the film's thickness rises as well, as illustrated in Table X. The range of 0.405-0.926 mm was obtained for the film thickness of formulations F1-F9. It was discovered that formulation F5 was 0.688 mm thick. The film's physical consistency is indicated by the low standard deviation numbers. The F5 formulation's folding resistance was measured at 305. The films' folding endurance values were discovered to be at their maximum, and as a result, they displayed good physical and mechanical qualities. All of the films' surfaces had pH values that fell within the range of salivary pH. There was no discernible variation in the surface pH of any films. All of the formulations measured surface pH values were found to be near to neutral, reducing their propensity to irritate the buccal mucosa and indicating that they should be tolerable. The film's tensile strength was determined to be between 11.90 and 50.00 N/mm2. It was discovered that formulation F5 had a tensile strength of 50.00 N/mm2. Zeta potential analysis It is evident from the surface response plot that the tensile strength increases when the amount of HPMC K15M is increased; this could be the result of hydrogen bonds between the medication and the polymer. The drug content of ranged between 91.39 and 98.00%, as reported in Table X. 98.00% of the drug was discovered in formulation F5. As can be observed, the fact that it is significantly closer to 100% indicates that there was no drug loss during the production of the film. The disintegration time was discovered to be between 20 to 30 seconds, as stated in Table X. The disintegration time for formulation F9 was determined to be the slowest, at 30 seconds, and the fastest, at 20 seconds for F5. Characterization of self emulsifying mouth dissolving films Characterization of self emulsifying mouth dissolving films The weight of the film increases together with the weight of the polymer. The films from F1 through F9 ranged in weight from 162mg to 277mg. The F3 formulation had the lightest weight of the film, and the F7 formulation had the heaviest weight of the film, as stated in Table X. The homogeneous weight of the films is indicated by the low standard deviation figures. The film's thickness was discovered in ascending sequence. As polymer concentration rises, the film's thickness rises as well, as illustrated in Table X. The range of 0.405-0.926 mm was obtained for the film thickness of formulations F1-F9. It was discovered that formulation F5 was 0.688 mm thick. The film's physical consistency is indicated by the low standard deviation numbers. The F5 formulation's folding resistance was measured at 305. The films' folding endurance values were discovered to be at their maximum, and as a result, they displayed good physical and mechanical qualities. All of the films' surfaces had pH values that fell within the range of salivary pH. There was no discernible variation in the surface pH of any films. All of the formulations measured surface pH values were found to be near to neutral, reducing their propensity to irritate the buccal mucosa and indicating that they should be tolerable. The film's tensile strength was determined to be between 11.90 and 50.00 N/mm2. It was discovered that formulation F5 had a tensile strength of 50.00 N/mm2. Zeta potential analysis It is evident from the surface response plot that the tensile strength increases when the amount of HPMC K15M is increased; this could be the result of hydrogen bonds between the medication and the polymer. The drug content of ranged between 91.39 and 98.00%, as reported in Table X. 98.00% of the drug was discovered in formulation F5. As can be observed, the fact that it is significantly closer to 100% indicates that there was no drug loss during the production of the film. The disintegration time was discovered to be between 20 to 30 seconds, as stated in Table X. The disintegration time for formulation F9 was determined to be the slowest, at 30 seconds, and the fastest, at 20 seconds for F5. Fig. 4: Zeta potential of optimized batch RZ 3 Characterization of self emulsifying mouth dissolving films The weight of the film increases together with the weight of the polymer. The films from F1 through F9 ranged in weight from 162mg to 277mg. The F3 formulation had the lightest weight of the film, and the F7 formulation had the heaviest weight of the film, as stated in Table X. The homogeneous weight of the films is indicated by the low standard deviation figures. The film's thickness was discovered in ascending sequence. As polymer concentration rises, the film's thickness rises as well, as illustrated in Table X. The range of 0.405-0.926 mm was obtained for the film thickness of formulations F1-F9. It was discovered that formulation F5 was 0.688 mm thick. The film's physical consistency is indicated by the low standard deviation numbers. The F5 formulation's folding resistance was measured at 305. The films' folding endurance values were discovered to be at their maximum, and as a result, they displayed good physical and mechanical qualities. All of the films' surfaces had pH values that fell within the range of salivary pH. There was no discernible variation in the surface pH of any films. All of the formulations measured surface pH values were found to be near to neutral, reducing their propensity to irritate the buccal mucosa and indicating that they should be tolerable. The film's tensile strength was determined to be between 11.90 and 50.00 N/mm2. It was discovered that formulation F5 had a tensile strength of 50.00 N/mm2. In-vitro dissolution studies To know the release at varied polymer concentrations, the drug release at various time intervals was computed and determined. The collected values were translated to % drug release. In Table XI & Fig. 5, the percentage cumulative release was displayed. Fig. 5: Dissolution profile of formulations 0 20 40 60 80 100 0 1 2 3 4 5 6 Cumulative drug release (%) Time (min) F1 F2 F3 F4 F5 F6 F7 F8 Fig. 5: Dissolution profile of formulations 0 20 40 60 80 100 0 1 2 3 4 5 6 Cumulative drug release (%) Time (min) F1 F2 F3 F4 F5 F6 F7 F8 Fig. 5: Dissolution profile of formulations Table XI: Percent cumulative release of formulations Time (min) F1 F2 F3 F4 F5 F6 F7 F8 F9 0 0 0 0 0 0 0 0 0 0 0.5 33.45 32.48 36 28 36.82 32.45 28.34 29.3 25.45 1 43.54 40.01 50.4 45.18 48.34 40.01 33.45 32.79 28.81 1.5 51.87 45.05 54.76 51.87 52.76 47.4 46.8 42.68 37.45 2 56.37 52.04 60.24 55.24 56.75 52.04 52.24 48.45 43.31 2.5 59.74 54.79 65.75 58.47 60.98 56.48 56.47 50.64 48.86 3 62.14 62.03 70.94 62.14 74.25 65.34 60.26 52.31 50.38 3.5 66.04 68.47 72.29 66.04 76.85 70.98 65.93 56.98 56.02 4 68.25 70.15 76.35 69.48 85.49 72.84 68.45 62.87 60.68 4.5 78.12 80.45 79.45 85.45 92.2 78.65 75.49 70.24 68.1 5 79.25 81.6 80.36 86.59 93.85 79.36 76.2 72.69 68.25 Differential scanning calorimetry (DSC) The physical condition of the medication in the formulation was investigated using a DSC investigation. W assessed the physical mixture, formulation F5, and pure medication. Fig. 6 displays the thermograms ranolazine, HPMC K15M, the physical combination, and the Optimized formulation (F5). 0 20 40 60 80 100 0 1 2 3 4 5 6 Cumulative drug release (%) Time (min) F1 F2 F3 F4 F5 F6 F7 F8 Fig. Zeta potential analysis It is evident from the surface response plot that the tensile strength increases when the amount of HPMC K15M is increased; this could be the result of hydrogen bonds between the medication and the polymer. The drug content of ranged between 91.39 and 98.00%, as reported in Table X. 98.00% of the drug was discovered in formulation F5. As can be observed, the fact that it is significantly closer to 100% indicates that there was no drug loss during the production of the film. The disintegration time was discovered to be between 20 to 30 seconds, as stated in Table X. The disintegration time for formulation F9 was determined to be the slowest, at 30 seconds, and the fastest, at 20 seconds for F5. Available online at: https://jazindia com 1441 Table X: evaluation of mouth dissolving film formulation batches F1 to F9 B. code Weight variation (mg) Thickness (mm)* Folding endurance Surface pH Tensile strength (N/mm2)* Drug content uniformity (%) Disintegr ation time (sec)* F1 163  0.02 0.405  0.02 260  0.03 7.1  0.05 11.90 ± 0.2 96.52  2.32 22  0.4 F2 167  0.05 0.524  0.05 270  0.02 7.0  0.03 22.91 ± 0.05 95.21  1.30 22  0.6 F3 162  0.03 0.574  0.07 273  0.05 6.9  0.05 17.50 ± 0.2 95.41  1.32 23  0.4 F4 189  0.07 0.678  0.03 276  0.02 7.0  0.02 16.06 ± 0.3 96.41  2.30 24  0.4 1441 Journal Of Advance Zoology Journal Of Advance Zoology F5 199  0.06 0.688  0.02 305  0.05 6.8  0.05 50.00 ± 0.2 98.00  1.34 20  0.2 F6 206  0.02 0.690  0.08 293  0.07 7.1  0.07 26.31 ± 0.05 96.01  0.75 25  0.4 F7 277  0.03 0.711  0.07 280  0.02 7.0  0.03 34.21 ± 0.1 93.56  0.25 26  0.1 F8 247  0.05 0.734  0.05 281  0.09 6.9  0.09 33.3 ± 0.10 92.85  1.85 28  0.2 F9 251  0.05 0.926  0.03 287  1.27 7.2  0.02 36.36 ± 0.2 91.39  0.39 30  0.2 *All readings are average ± SD (n=3) In-vitro dissolution studies 5: Dissolution profile of formulations Table XI: Percent cumulative release of formulations Time (min) F1 F2 F3 F4 F5 F6 F7 F8 F9 0 0 0 0 0 0 0 0 0 0 0.5 33.45 32.48 36 28 36.82 32.45 28.34 29.3 25.45 1 43.54 40.01 50.4 45.18 48.34 40.01 33.45 32.79 28.81 1.5 51.87 45.05 54.76 51.87 52.76 47.4 46.8 42.68 37.45 2 56.37 52.04 60.24 55.24 56.75 52.04 52.24 48.45 43.31 2.5 59.74 54.79 65.75 58.47 60.98 56.48 56.47 50.64 48.86 3 62.14 62.03 70.94 62.14 74.25 65.34 60.26 52.31 50.38 3.5 66.04 68.47 72.29 66.04 76.85 70.98 65.93 56.98 56.02 4 68.25 70.15 76.35 69.48 85.49 72.84 68.45 62.87 60.68 4.5 78.12 80.45 79.45 85.45 92.2 78.65 75.49 70.24 68.1 5 79.25 81.6 80.36 86.59 93.85 79.36 76.2 72.69 68.25 g y ( ) The physical condition of the medication in the formulation was investigated using a DSC investigation. We assessed the physical mixture, formulation F5, and pure medication. Fig. 6 displays the thermograms for ranolazine, HPMC K15M, the physical combination, and the Optimized formulation (F5). Differential scanning calorimetry (DSC) The physical condition of the medication in the formulation was investigated using a DSC investigation. We assessed the physical mixture, formulation F5, and pure medication. Fig. 6 displays the thermograms for ranolazine, HPMC K15M, the physical combination, and the Optimized formulation (F5). 1442 Available online at: https://jazindia.com Journal Of Advance Zoology Fig. 6: Overlain DSC of RZ, HPMC K 15M, physical mixture and optimized formulation F5 Ranolazine, when taken in its purest form, produces an endothermic peak that corresponds to melting at 123°C, demonstrating its crystalline nature. The endothermic peak in the DSC of the physical combination does not significantly change. The DSC data showed that there was no specific alteration in the melting endothermic peak, indicating that there was no interaction between the medication and the employed polymers. 123.67°C 81.48°C 56.87°C 123.70°C 123.67°C 81.48°C 56.87°C 123.70°C Available online at: https://jazindia.com 1443 Fig. 6: Overlain DSC of RZ, HPMC K 15M, physical mixture and optimized formulation F5 Ranolazine, when taken in its purest form, produces an endothermic peak that corresponds to melting at 123°C, demonstrating its crystalline nature. The endothermic peak in the DSC of the physical combination does not significantly change. The DSC data showed that there was no specific alteration in the melting endothermic peak, indicating that there was no interaction between the medication and the employed polymers. Powder X-ray diffraction analysis (PXRD) The physical mixture and formulation batch diffractograms showed complete amorphism of the drug and excipients. No sharp peak, as seen in Fig. 7 was noticed. Fig. 7: Overlain of PXRD of pure RZ, HPMC K 15M and physical mixture 123.70°C Physical Mixture HPMC K15M RZ ig. 6: Overlain DSC of RZ, HPMC K 15M, physical mixture and optimized formulation F5 Ranolazine, when taken in its purest form, produces an endothermic peak that corresponds to melting at 123°C, demonstrating its crystalline nature. The endothermic peak in the DSC of the physical combination does not significantly change. The DSC data showed that there was no specific alteration in the melting endothermic peak, indicating that there was no interaction between the medication and the employed polymers. Ranolazine, when taken in its purest form, produces an endothermic peak that corresponds to melting at 123°C, demonstrating its crystalline nature. The endothermic peak in the DSC of the physical combination does not significantly change. Differential scanning calorimetry (DSC) The DSC data showed that there was no specific alteration in the melting endothermic peak, indicating that there was no interaction between the medication and the employed polymers. Powder X-ray diffraction analysis (PXRD) The physical mixture and formulation batch diffractograms showed complete amorphism of the drug and excipients. No sharp peak, as seen in Fig. 7 was noticed. Powder X-ray diffraction analysis (PXRD) Available online at: https://jazindia.com 1 Fig. 7: Overlain of PXRD of pure RZ, HPMC K 15M and physical mixture Physical Mixture HPMC K15M RZ Physical Mixture HPMC K15M RZ Physical Mixture HPMC K15M RZ Available online at: https://jazindia.com Fig. 7: Overlain of PXRD of pure RZ, HPMC K 15M and physical mixture 1443 Journal Of Advance Zoology Journal Of Advance Zoology Journal Of Advance Zoology Scanning electron microscopy (SEM) Scanning electron microscopy (SEM) Investigations using a SEM revealed that the surface structures of the SEMDF varied. Fig.8 depicts typical characteristics of the MDF surface with visible cellulose fibres. Self-emulsifying components in SEMDF that contain oils, surfactants, and co-surfactants tightened up on the surface and were entirely contained by the cellulose fibre surface. Fig. 8: SEM of optimized formulation F5 The self-emulsifying elements on the surface of SEMDF may quickly self-emulsify in water and contribute significantly to the rapid commencement of action of ranolazine. Fig. 8: SEM of optimized formulation F5 Fig. 8: SEM of optimized formulation F5 The self-emulsifying elements on the surface of SEMDF may quickly self-emulsify in water and contribute significantly to the rapid commencement of action of ranolazine. The self-emulsifying elements on the surface of SEMDF may quickly self-emulsify in water significantly to the rapid commencement of action of ranolazine. Stability studies The formulations underwent a stability assessment for 30 days at 40°C with a 1°C temperature fluctuation and 75% relative humidity. The samples drug content was determined using different time intervals, and it is clear that there were minor variations, as indicated in Table XII. Table XII: Evaluation of optimized formulation F5 after stability period Parameters Time period Before After 30 days Tensile strength (N/mm2) 50 ± 0.2 50 ± 0.1 % Drug release 93.85 93.16 Drug content (%) 98.0 ± 01.34 98.23 ± 0.02 Table XII: Evaluation of optimized formulation F5 after stability period CONCLUSION In order to achieve greater therapeutic effectiveness with enhanced bioavailability and better patient compliance, the present study demonstrated that the SEMDF of ranolazine could be successfully manufactured by solvent casting process. For the phase diagram analysis of RZ SEDDS, ten distinct possible surfactant mixtures to oil combinations at various Km values (1, 2, 3, and 4) were employed. Compared to Km values 1, 2, and 4, the phase diagram at Km value 3 displays better microemulsion existence zones. Km 3 system was ultimately chosen as the SEDDS formulation. Additionally, it was determined that formulation F5 had the best physicochemical and mechanical characteristics out of all the other formulations. Additionally, the new formulation's stability analysis validated SEMDF's increased shelf life. The current study thus illustrates the enormous potential of SEMDF for improving patient comfort and compliance by expediting the commencement of action and avoiding hepatic first-pass metabolism, especially in pediatric and geriatric patients. Available online at: https://jazindia.com CONFLICTS OF INTEREST The authors declare that there is no conflict of interest. REFERENCES REFERENCES 1. Savjani KT, Gajjar AK, Savjani JK. Drug solubility: importance and enhancement techniques, ISRN. Pharm. 2012; 1-10 2. Sharma D, Soni M, Kumar S, Gupta GD. Solubility enhancement- eminent role in poorly soluble drugs. Res. J. Pharm. Technol. 2009; 2(2):220-24. 3. Lipinski CA. Drug-like properties and the causes of poor solubility and poor permeability. J. Pharmacol. Toxicol. Methods 2000; 44(1): 235-249. 4. Kumar S, Singh P. Various techniques for solubility enhancement: An overview. J. Pharm. Innov. 2016; 5(1): 23-28 5. Suares D, Meghani N. Self micro-emulsifying drug delivery system (SMEDDS): A promising tool to improve bioavailability. J. Pharm. Phytother. 2013;2(1):17-21 6. Shah NH, Carvajal MT, Patel CI, Infeld MH, Malick AW. Self emulsifying drug delivery systems (SEDDS) with polyglycolysed glycerides for improving in vitro dissolution and oral absorption of lipophilic drugs. Int. J. Pharm. 1994; 106:15- 23. p p g 7. Maurya SD, Rajeshwar KK, Rajpal AG, Dhakar RC. Self-micro emulsifying drug delivery systems (SMEDDS): a review on physico-chemical and biopharmaceutical aspects. J. Drug Deliv. Ther. 2017; 7(3):55-65. m H, Na S-J, Shin D, Jo K, Lee J. Thin films as an emerging platform for drug delivery. Asian ci. 2016; 11:559-74. 8. Karki S, Kim H, Na S-J, Shin D, Jo K, Lee J. Thin films as an emerging platform for drug J. Pharm. Sci. 2016; 11:559-74. 9. Kumar CB. Aparna, CA, Srinivas, DP. Formulation and evaluation of solid self-emulsifying drug delivery system of Ranolazine. J. Glob. Trends Pharm. Sci. 2014;5(4):2238-2247 10. Reddy MS, Fazal S, Apte SS. Solubility enhancement of fenofibrate: A BCS class 2 drug, by self emulsifying drug delivery system. Int. Res. J. Pharm. 2011; 2(11):173-177. y g g y y 11. Patel J, Kevin G, Patel A, Raval M, Sheth N. Design and development of a self-nanoemulsifying drug delivery system for telmisartan for oral drug delivery. Int. J. Pharm. Investig. 2011; 1(2): 112-118. 12. Kirankumar V, Aruna M, Bikshapathi DVRN. Development of solid self emulsifying drug delivery system containing efavirenz: in-vitro and in-vivo evaluation. Int. J. Pharm. Bio. Sci. 2013; 4(1): 869- 882. 13. Kokare CR, Kumbhar SA, Patil A. Formulation and evaluation of self emulsifying drug delivery system 12. Kirankumar V, Aruna M, Bikshapathi DVRN. Development of solid self emulsifying drug delivery system containing efavirenz: in-vitro and in-vivo evaluation. Int. J. Pharm. Bio. Sci. 2013; 4(1): 869- 882. 13. Kokare CR, Kumbhar SA, Patil A. CONFLICTS OF INTEREST Available online at: https://jazindia.com CONFLICTS OF INTEREST The authors declare that there is no conflict of interest. 1444 Journal Of Advance Zoology REFERENCES Formulation and evaluation of self emulsifying drug delivery system of carbamazepine. Indian J. Pharm. Educ. Res. 2013; 47(2):172-17. 13. Kokare CR, Kumbhar SA, Patil A. Formulation and evaluation of self emulsifying drug delivery system of carbamazepine. Indian J. Pharm. Educ. Res. 2013; 47(2):172-17. 14. Tayal A, Jamil F, Sharma R. Self emulsifying drug delivery systems: a review. Int. Res. J. Pharm. 2012; 3(5):32-36. 15. Sapraa K, Sapra A, Singha SK, Kakkarb S. Self emulsifying drug delivery system: a tool in solubility enhancement of poorly soluble drugs. Indo-Glob. J. Pharm. Sci. 2012, 2(3): 313-332. 16. Bhikshapathi D, Madhukar P, Kumar BD, Kumar GA. Formulation and characterization of pioglitazone HCl self emulsifying drug delivery system. Der Pharm. Lett. 2013; 5(2):292-305. 17. Koland M, Sandeep VP, Charyulu NR. Fast dissolving sublingual films of ondansetron hydrochloride: effect of additives on in vitro drug release and mucosal permeation. J. Young Pharm. 2010; 2(3):216-22. 18. Mishra R, Amin A. Formulation development of taste-masked rapidly dissolving films of cetirizine 17. Koland M, Sandeep VP, Charyulu NR. Fast dissolving sublingual films of ondansetron hydrochloride: effect of additives on in vitro drug release and mucosal permeation. J. Young Pharm. 2010; 2(3):216-22. 18. Mishra R, Amin A. Formulation development of taste-masked rapidly dissolving film hydrochloride. Pharm. Technol. 2009; 33(2):48-56. y ; ( ) 19. Panchal MS, Patel H, Bagada A, Vadalia KR. Formulation and evaluation of mouth dissolving film of ropinirole hydrochloride by using pullulan polymers. Int. J. Pharm. Res. Allied Sci. 2012; 1(3):60-72. 20. Raza SN, Kar AH, Wani TU, Khan NA. Formulation and evaluation of mouth dissolving films of losartan potassium using 32 Factorial design. Int. J. Pharm. Sci. Res. 2019; 10(3):1402-11. 1445 Available online at: https://jazindia.com
https://openalex.org/W2799443297	https://europepmc.org/articles/pmc5932356?pdf=render	English	null	Enhancing EPA Content in an Arctic Diatom: A Factorial Design Study to Evaluate Interactive Effects of Growth Factors	Frontiers in plant science	2,018	cc-by	9,644	ORIGINAL RESEARCH published: 17 April 2018 doi: 10.3389/fpls.2018.00491 Enhancing EPA Content in an Arctic Diatom: A Factorial Design Study to Evaluate Interactive Effects of Growth Factors Microalgae with a high content of the omega-3 polyunsaturated fatty acids (PUFAs), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) are of great demand for microalgae-based technologies. An Arctic strain of the diatom Attheya septentrionalis was shown in previous experiments to increase its EPA content from 3.0 to 4.6% of dry weight (DW) in the nutrient-replete exponential phase and nutrient-depleted stationary phase, respectively. In the present study, a factorial-design experiment was used, to investigate this effect in more detail and in combination with varying salinities and irradiances. A mathematical model revealed that both growth phase and salinity, alone and in combination, inﬂuenced the EPA content signiﬁcantly. Maximum EPA values of 7.1% DW were obtained at a salinity of 22 and after 5 days in stationary phase, and might be related to a decreased silica content, an accumulation of storage lipids containing EPA, or both. However, growth rates were lower for low salinity (0.54 and 0.57 d−1) than high salinity (0.77 and 0.98 d−1) cultures. Edited by: Jianhua Fan, East China University of Science and Technology, China Reviewed by: Adolfo Rivero-Muller, Turku Centre for Biotechnology, Finland Zhi-Gang Zhou, Shanghai Ocean University, China Keywords: eicosapentaenoic acid (EPA), arctic diatom, factorial design, salinity, growth phase, interactive effects, microalgal biotechnology Zhi Gang Zhou, Shanghai Ocean University, China *Correspondence: Pia Steinrücken pia.steinrucken@uib.no INTRODUCTION The omega-3 polyunsaturated fatty acids (PUFAs), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) are known to contribute signiﬁcantly to human health (Martins et al., 2013). The current major source for the two fatty acids (FA) is ﬁsh oil from marine wild ﬁsh. The ﬁsh obtain and accumulate these PUFAs themselves predominantly via the marine food chain from EPA- and DHA-synthesizing microalgae (Spolaore et al., 2006). As EPA and DHA are also essential for farmed ﬁsh (Khozin-Goldberg et al., 2016), ﬁsh oil is an important additive in aquaculture feed. The increasing demand for EPA and DHA, particularly from the growing aquaculture industries, but also increasingly from the health and food sectors, necessitates other sustainable production sources. Many marine microalgae naturally produce EPA and DHA, and are therefore considered a promising alternative (Patil et al., 2005; Chauton et al., 2015). Although there has been intensive research in this ﬁeld, the costs associated with microalgae large-scale cultivation and processing for FAs are still greater than in ﬁsh oil production. Improvements at the diﬀerent parts of the production chain are therefore essential in order to reduce production costs. The selection of suitable species that are superior to existing cultures in terms of growth and EPA and DHA content, and the optimization of cultivation conditions are important contributions to the ongoing improvement of microalgae-based technologies (Adarme-Vega et al., 2012). Specialty section: This article was submitted to Plant Biotechnology, a section of the journal Frontiers in Plant Science Received: 24 January 2018 Accepted: 03 April 2018 Published: 17 April 2018 Keywords: eicosapentaenoic acid (EPA), arctic diatom, factorial design, salinity, growth phase, interactive effects, microalgal biotechnology Experimental Design p g Factorial design was used to investigate the eﬀects of salinity, irradiance, and growth phase and their interactions on the EPA and total fatty acid (TFA) content, and the FA composition in the diatom A. septentrionalis, by growth of batch cultures at diﬀerent conditions. The eﬀects of salinity and irradiance were assessed at two levels (22 and 35, and 50 and 200 µmol photons m−2 s−1, respectively), and the growth phase at three levels; exponential phase (e), Day 3 of stationary phase (ﬁrst stationary phase, s1), and Day 5 of stationary phase (second stationary phase, s2), resulting in 12 treatment groups. Salinity and irradiance levels were selected as being representative of those occurring under natural conditions in the Arctic and additionally, so as to induce diﬀerent responses but not to impair the cultures. Before the start of the batch experiment, pre-cultures (one biological replicate for each condition) were grown semi-continuously for 14 d at either low or high salinity and low or high irradiance (LSLI, HSLI, LSHI, and HSHI), to acclimate the cultures to their respective conditions. Two sterile and nutrient enriched media (Walne, 1970) were prepared with aged SW (salinity of 35) and respective dilutions with distilled water (salinity of 22). Pre-cultures were prepared in glass tubes (300 mL, 3.5 cm inner diameter), by adding 1 mL inoculum to 60 mL fresh medium of the respective salinity, and placed into temperature-controlled water tanks (10◦C). Continuous illumination with 50 or 200 µmol photons m−2 s−1 (measured with 4π quantum scalar irradiance sensor [QSL-100, Biospherical Instruments, San Diego, CA, USA], inside the empty glass cylinder) was provided by banks of six white ﬂuorescent tubes (Philips MASTER, TL-D 90 Graphica, 58W/95) in the back of the water tanks, running perpendicular to the glass tubes. To ensure adequate mixing and carbon supply, 0.2 µm-ﬁltered and 1% CO2-enriched air was bubbled through glass capillaries into the bottom of each glass tube. Pre-cultures were kept in exponential phase by maintaining an optical density (OD750) between 0.15 and 0.50, by the addition of fresh medium, successively added until volumes reached 260 mL. Thereafter, half of the culture volume was replaced with fresh medium daily or every other day. After the 2-week acclimation period, the batch experiment was started. Experimental Design Biomass from each pre-culture was distributed into two sterile glass tubes to obtain two biological g It is well-known that microalgae modify their biochemical composition and FA content in response to environmental factors, including nutrient availability, irradiance, temperature, and salinity (Dunstan et al., 1993; Renaud and Parry, 1994; Tatsuzawa and Takizawa, 1995; Xu and Beardall, 1997; Van Wagenen et al., 2012; Boelen et al., 2013; Cepák et al., 2013). To investigate the impact of the diﬀerent factors, traditional methods vary one condition at a time, while keeping all other factors constant. However, FA composition and content in microalgae are dependent on synergistic and antagonistic interactions of cultivation conditions. Factorial designs are based on a multivariate approach, in which the variation in diﬀerent factors are tested simultaneously (Duarte et al., 2001). These yield a predictive model which provides information on the magnitude of the eﬀects of both individual factors and combinations of factors, and on their statistical signiﬁcance (Chen et al., 2012). In this study, we aimed to elucidate additional information on the dynamics of the EPA content and relative FA composition in A. septentrionalis, and to determine the experimental conditions that might lead to a further increase in the EPA content. A factorial-design experiment was used to investigate the impact of nutrient starvation in greater detail, together with the eﬀects of salinity and irradiance, and their respective interactions. Both salinity and irradiance are known to aﬀect the FA composition in microalgae (Xu and Beardall, 1997; Lu et al., 2001; Chen et al., 2008), and are highly variable in Arctic environments due to melting and freezing sea ice, and strong variations in photoperiod. Hence, microalgae from these environments are expected to be promising in this context, and possess the necessary adaptations to cope with these changing conditions. METHODS Stock Cultures and Inoculum Citation: Steinrücken P, Mjøs SA, Prestegard SK and Erga SR (2018) Enhancing EPA Content in an Arctic Diatom: A Factorial Design Study to Evaluate Interactive Effects of Growth Factors. Front. Plant Sci. 9:491. doi: 10.3389/fpls.2018.00491 April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 1 Increasing EPA in a Diatom Steinrücken et al. Spitzbergen (N 79◦25.14′ E 08◦18.84′) in May 2014 (3.5◦C water temperature, salinity of 35, and 24 h daylight). For strain characterization, a region of the 28S ribosomal RNA (rRNA) gene was sequenced and compared with previously published sequences from diatoms at GenBank (Steinrücken et al., 2017). Partial sequence of the 28S rRNA gene have been deposited in GenBank (http://www.ncbi.nlm.nih.gov/) with accession number MH020639. Stock cultures were maintained in 50 mL Erlenmeyer ﬂasks at 10◦C and 50 µmol photons m−2 s−1, in nutrient- enriched (Walne, 1970) 80% seawater (SW) by monthly dilution. Eighty percent SW was obtained by dilution of ﬁltered and autoclaved fjord-SW (from 90 m depth, salinity of 35) with distilled water (80:20, v:v), giving a salinity of 29. For the inoculum, biomass was harvested from exponential phase stock cultures by centrifugation (2,264 × g, 5 min), washed twice with fresh medium and re-inoculated into 10 mL fresh medium. In previous research, we searched for new and fast growing microalgal strains from North Atlantic habitats with high growth rates and high EPA and DHA contents (Steinrücken et al., 2017). A strain of the diatom Attheya septentrionalis, isolated from Arctic Waters, demonstrated rapid growth at temperatures of 10◦C (0.7 d−1) and a high EPA content, which increased from 3.0 to 4.6% of the dry weight (DW) from exponential phase to Day 3 of the stationary phase. This EPA content under nutrient-depleted conditions was higher than typically found in industrially applied microalgae and this strain was therefore suggested to be a potential candidate for future large- scale cultivation and EPA production. A high growth rate and EPA content in stationary phase in A. septentrionalis were also found by Knuckey et al. (2002) who suggested this diatom to be a promising feed source for oyster hatcheries. However, only few studies on Attheya species exist (Aizdaicher and Markina, 2011; Stonik et al., 2017), and more explicit investigations are needed in order to assess the potential of this diatom for microalgae-based technologies. Regression Models The model for growth rate as a function of salinity, irradiance and their interactions, and models for TFA and EPA content as functions of salinity, irradiance, growth phase, and their interactions were calculated by multiple least squares regression. The models reported in the paper are based on the coded factor levels, where the low values are assigned −1 and the high values are assigned +1. For the growth phase, there are three levels; exponential phase, and ﬁrst and second stationary phase. Exponential and second stationary phase were assigned the levels −1 and +1, respectively. The level for ﬁrst stationary phase was set to 0.74. This level was found by iteratively testing values from −1 to 1 with increments of 0.01, and selecting the value that minimized the sum of squared residuals of the model. Models and model statistics were calculated by the ﬁtlm function in the Statistics and Machine Learning Toolbox running under Matlab R2017a (Mathworks, Natick, MA, USA). Optical density measurements were performed using a spectrophotometer (UV1800, Shimadzu Corporation, Kyoto, Japan) at 750 nm and if required, samples were diluted to give an attenuation between 0.2 and 0.8. The QY was measured with AquaPen (AquaPen-C, AP-C 100, Photon System Instruments, Brno, Czech Republic) after initial dark incubation between 10 and 60 min. DWs (expressed as weight of the dried biomass [g] per volume [L]) were determined in triplicates as described by Zhu and Lee (1997), with 0.5 M ammonium formate as a washing buﬀer. Exponential and second stationary phase were assigned the levels −1 and +1, respectively. The level for ﬁrst stationary phase was set to 0.74. This level was found by iteratively testing values from −1 to 1 with increments of 0.01, and selecting the value that minimized the sum of squared residuals of the model. Models and model statistics were calculated by the ﬁtlm function in the Statistics and Machine Learning Toolbox running under Matlab R2017a (Mathworks, Natick, MA, USA). For FA analysis, triplicate 10 mL microalgal cultures were harvested by centrifugation (6 min at 2,264 × g) into glass tubes (PYREX), the supernatant discarded, and the pellet covered in nitrogen atmosphere and stored at −20◦C until analysis. FAs were extracted and derivatized to fatty acid methyl esters (FAME) by direct esteriﬁcation (Meier et al., 2006). Stock Cultures and Inoculum Attheya septentrionalis is a single celled diatom with four long setae, and is broadly distributed in Arctic and Temperate Waters (Rampen et al., 2009; Stonik et al., 2017). The strain used in this experiment was isolated from Arctic Waters north-west of April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 2 Increasing EPA in a Diatom Steinrücken et al. replicates, and each diluted with fresh medium to yield a starting OD and volume of ∼0.15 and 260 mL, respectively. Cultures were then grown until Day 5 of the stationary phase, and all cultures were sampled daily for OD and maximum quantum yield (QY), and on Day 1 in exponential phase (e) and Days 3 and 5 in stationary phase (s1 and s2, respectively) for DW and FA analysis. Additionally, cultures were sampled for nutrient analysis of the media at the start of the experiment, and on the DW and FA sampling days. 30 cm s−1 was used as carrier gas in constant ﬂow mode. Injector and detector temperatures were 250 and 300◦C, respectively (Prestegard et al., 2015). The FAMEs were identiﬁed by analysis on gas chromatography coupled to mass spectrometry (GC-MS) as described in Wasta and Mjøs (2013), and by using libraries of mass spectra and retention indices available at www.chrombox. org/data. For nutrient measurements of the media, 20 mL GF/F ﬁltrates were collected in white plastic vials, 100 µl chloroform added and stored at 4◦C before analysis. Dissolved inorganic nitrate, nitrite, orthophosphate, and silicate were analyzed at the Institute of Marine Research, Bergen, which oﬀers accredited and standardized service for nutrient analyses, using colorimetric absorption measurements on an Alpkem- Lab analyzer (Alpkem Corporation, Oregon USA) according to Parsons et al. (1992). Analytical Procedures Relative growth rates between repeated dilutions during pre- cultivation were calculated according to the changes in attenuation with Equation (1). Nx0 and Nx are OD750 after dilution (tx0) and before the subsequent dilution (tx), respectively. µ d−1 = ln (Nx) −ln(Nx0) (tx −tx0) (1) (1) Regression Models The pellet was dried in the 10 mL tube by evaporating the remaining water under a nitrogen stream, and 18 or 34 µg internal standard (23:0 FAME dissolved in isooctane) was added to exponential or stationary phase samples, respectively. The solvent was evaporated under nitrogen stream and 0.5 mL methylation reagent (2M HCl in methanol) was added. Samples were covered with nitrogen gas, sealed and incubated at 90◦C for 2 h. After cooling to room temperature, half of the methylation reagent was evaporated, 0.5 mL water was added, and the samples were extracted twice with 1 mL isooctane. The combined extracts of stationary phase samples were diluted with isooctane (1:1, v:v) to yield a ﬁnal internal standard concentration of approximately 18 µg mL−1 (Steinrücken et al., 2017). FAMEs were analyzed by GC (7890 gas chromatograph, Agilent, Santa Clara, CA, USA) equipped with an autosampler, split-splitless injector, ﬂame ionization detector (FID), and a 60 m BPX70 capillary column (SGE, Ringwood, Australia) with internal diameter of 0.25 mm and ﬁlm thickness of 0.25 µm. One microliters of sample volumes were injected splitless at 60◦C. This temperature was maintained for 3 min before raised by 40◦C min−1 to 150◦C and by 1.5◦C min−1 to 230◦C. Helium, with an estimated average velocity of Batch Experiment—Dry Weights (DW), Total Fatty Acids (TFA), and EPA y ( ) DW together with TFA and EPA contents (% DW) for the 12 diﬀerent treatment groups, each with two biological replicates are shown in Figure 3. The DW (average of the biological replicates) increased from the exponential phase to the ﬁrst stationary phase by 0.13, 0.12, 0.13, and 0.11 g L−1 and further by 0.04, 0.03, 0.01, 0.02 g L−1 to the second stationary phase, giving a total increase of 0.17, 0.15, 0.14, and 0.13, g L−1 for low salinity- low irradiance (LSLI), high salinity-low irradiance (HSLI), low salinity-high irradiance (LSHI), and high salinity-high irradiance (HSHI) conditions, respectively (Figure 3A). p y g The TFA and EPA contents (Figures 3B,C respectively) increased from the exponential to the stationary phase at all growth conditions, but to diﬀerent extents. In exponential phase, EPA contents (average of replicates) were 2.8, 3.2, 2.4, and 3.1% DW for LSLI, HSLI, LSHI, and HSHI conditions, respectively, and increased to 6.8, 5.7, 5.8, and 5.5% DW, respectively, in the ﬁrst stationary phase. For LSLI, LSHI, and HSHI cultures, EPA content increased further to 7.2, 7.1, and 5.8% DW, respectively, in the second stationary phase, while it decreased in the HSLI conditions to 5.2% DW. A similar pattern with increase from the exponential phase to the ﬁrst stationary phase in all cultures and further increase to the second stationary phase in LSLI, LSHI and HSHI cultures, and decrease from the ﬁrst to the second stationary phase in HSLI cultures was found for the TFA content. Corresponding values with standard deviation can be found in the Supplementary Material (Table S2). In exponential phase, TFA and EPA contents were higher in HS cultures, while after 5 days of stationary phase, levels were higher for LS cultures. EPA and TFA contents (% DW), estimated by the model, are shown in Figure S1 in the Supplementary Material. µ (d−1)est = 0.715+0.163∗X1+0.062∗X2 +0.043∗X1X2 (3) Statistics The batch experiment was performed with two individual cultures (biological replicates) for each treatment, which is suﬃcient for solid statistics when using regression analysis and factorial design. One measurement replicate was taken for OD and QY measurements, whereas FA and DW were analyzed in triplicates for each biological replicate. The FA content and the DW were analyzed from individual subsamples, and the standard deviation (SD) for FA content relative to the DW was calculated using Equation (2). SDFA/DW = p %SDFA2 + %SDDW2 100 × FA DW (2) (2) Euclidean dendrograms and Principal Component Analysis (PCA) of treatment groups and their FA composition were calculated using Sirius 10.0 (Pattern Recognition Systems AS, Bergen, Norway) and edited in GraphPad Prism 6. April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 3 Increasing EPA in a Diatom Steinrücken et al. RESULTS and by 23, 45, and 38% for high salinity cultures, respectively (Figures 2C,D). At the ﬁrst stationary phase (Day 3), all nutrients had been consumed in all cultures (nitrate 99%, silicate 99%, and phosphate 97%). Batch Experiment—Growth and Maximum Quantum Yield (QY) ( ) Growth curves and QY during the batch experiment were very similar for the biological replicates, but there were diﬀerences between the cultivation conditions, especially between low and high salinities (Figures 2A,B). Cultures grown at LS, revealed a declined slope of the growth curves and reached stationary phase 1 day later (Day 4) than cultures grown at HS (Day 3). During exponential growth, the QY was slightly lower for the LS cultures (replicate averages of 0.64 at Day 1, both for low and high irradiances) than for the HS cultures (replicate averages of 0.69 and 0.68 at Day 1, for low and high irradiance, respectively). During stationary phase, the QY decreased in all cultures. In HS cultures, QY dropped by 10% (LI) and 30% (HI) after entering stationary phase, and by 23 and 32%, respectively, at the end of cultivation period. A lower reduction was observed for the LS cultures, where QY was reduced by 6 and 8% (low and high irradiance, respectively) after entering stationary phase, and by 14 and 17% at the end of cultivation. The mathematical models, representing the EPA and the TFA content as a function of salinity (X1), irradiance (X2), growth phase (X3), and their combinations (X1X2, X1X3, X2X3) are expressed by Equations (4, 5), respectively. A positive term for the combinations in the equations indicates a synergistic eﬀect (increasing values increase the content), whereas negative terms indicate an antagonistic eﬀect (increasing values decrease the content; Chen et al., 2012). Coeﬃcients in bold indicate statistical signiﬁcance for the corresponding coeﬃcients (black: p < 0.05 and red p < 0.01). For the EPA content, coeﬃcients for growth phase, salinity, and the combination of salinity and growth phase were statistically signiﬁcant whilst the other factors were not. The most signiﬁcant variable with highest impact on EPA content was the growth phase with a positive estimated eﬀect of 1.74 (p = 2.0∗10−13). The combined eﬀect of salinity and growth phase was lower (−0.48, p = 0.00003) and negative, while salinity had a less negative eﬀect on EPA content (−0.19, p = 0.0241). Frontiers in Plant Science \| www.frontiersin.org Pre-cultivation—Growth Rates Pre-cultivation—Growth Rates Pre-cultures were grown for 2 weeks to allow for cells to acclimate to the respective salinities and irradiances. After three dilutions, growth rates for each condition became more constant although they still varied slightly between the repeated dilutions. Only growth rates from the ﬁnal seven dilutions prior to the batch experiment were used for analyses (Table S1). Average values, together with the estimates provided by the mathematical model, are shown in Figure 1A. Lower growth rates (0.54 d−1 for low irradiance [LI] and 0.57 d−1 for high irradiance [HI]) were observed for low salinity (LS) cultures, compared to high salinity (HS) cultures (0.77 d−1 for LI and 0.98 d−1 for HI), together with increased values for HI cultures. High irradiance had a stronger positive eﬀect on the growth rate at HS. The mathematical model representing the growth rate as a function of salinity (X1), irradiance (X2), and their combination (X1X2) in the experimental setup is expressed by Equation (3). Positive coeﬃcients in the equations indicate that increasing values increase the growth rate and bold numbers indicate a high statistical signiﬁcance (black: p < 0.05 and red: p < 0.01). According to the model, salinity had the strongest positive inﬂuence on the growth rate, with high signiﬁcance (0.163, p- value 1.4∗10−09), while irradiance (0.062, p = 0.001) and the combination of salinity and irradiance (0.043, p = 0.016) had lower, but still signiﬁcant, impacts. A strong and signiﬁcant correlation (R2 = 0.8484) between the estimated values of the model and the measured values indicates a good ﬁt between the model and the experimental data (Figure 1B). Details on the measured and estimated growth rates can be found in the Supplementary Material (Table S1). Batch Experiment—Dry Weights (DW), Total Fatty Acids (TFA), and EPA Batch Experiment—Growth and Maximum Quantum Yield (QY) The TFA content was also signiﬁcantly inﬂuenced by growth phase (8.34, p = 1.1∗10−15), salinity (−0.77, p = 0.0116), The nitrate, silicate, and phosphate concentrations of the media decreased from the start of the experiment to the ﬁrst day of exponential phase by 14, 39, and 27% for low salinity cultures April 2018 \| Volume 9 \| Article 491 4 Steinrücken et al. Increasing EPA in a Diatom FIGURE 1 \| Growth rates (µ) during pre-cultivation. (A) Average growth rates of repeated dilutions (bars, n = 7) and estimated growth rates by the model (dots) of the diatom Attheya septentrionalis grown at four different cultivation conditions (combinations of low or high salinity [22 and 35] and irradiance [50 and 200 µmol photons m−2 s−1]. (B) Average and standard deviation (n = 7) of measured growth rates, plotted against the estimated growth rates by the model, with linear regression. FIGURE 1 \| Growth rates (µ) during pre-cultivation. (A) Average growth rates of repeated dilutions (bars, n = 7) and estimated growth rates by the model (dots) of the diatom Attheya septentrionalis grown at four different cultivation conditions (combinations of low or high salinity [22 and 35] and irradiance [50 and 200 µmol photons m−2 s−1]. (B) Average and standard deviation (n = 7) of measured growth rates, plotted against the estimated growth rates by the model, with linear regression. FIGURE 2 \| Growth during batch experiment. Batch cultures of the diatom Attheya septentrionalis grown at four conditions (combinations of low and high salinity [22 and 35] and irradiance [50 and 200 µmol photons m−2 s−1] with two biological replicates for each condition). (A,B) Optical density (OD750) based growth curves and maximum quantum yields (QY, 8max) for low and high salinities, respectively. Circles indicate sampling time points (e, exponential phase; s1, ﬁrst stationary phase; s2, second stationary phase). LI, low irradiance; HI, high irradiance. (C,D) Superimposed nutrient concentration in the media at the start of the experiment and during exponential (e) and stationary (s1 and s2) sampling time points for low and high salinities, respectively. Values are the averages and standard deviations from samples of the respective salinity. FIGURE 2 \| Growth during batch experiment. Batch cultures of the diatom Attheya septentrionalis grown at four conditions (combinations of low and high salinity [22 and 35] and irradiance [50 and 200 µmol photons m−2 s−1] with two biological replicates for each condition). Batch Experiment—Growth and Maximum Quantum Yield (QY) (A,B) Optical density (OD750) based growth curves and maximum quantum yields (QY, 8max) for low and high salinities, respectively. Circles indicate sampling time points (e, exponential phase; s1, ﬁrst stationary phase; s2, second stationary phase). LI, low irradiance; HI, high irradiance. (C,D) Superimposed nutrient concentration in the media at the start of the experiment and during exponential (e) and stationary (s1 and s2) sampling time points for low and high salinities, respectively. Values are the averages and standard deviations from samples of the respective salinity. and 0.9802 for EPA and TFA content, respectively) of estimated values by the model and measured values indicates a good ﬁt between the model and the experimental data (Figure S2). More and the combination of salinity and growth phase (−1.43, p = 0.0002), and additionally, by the combination of salinity and irradiance (0.73, p = 0.0126). A high correlation (R2 = 0.9651 April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org Frontiers in Plant Science \| www.frontiersin.org Increasing EPA in a Diatom Steinrücken et al. FIGURE 3 \| Dry weights (DW) and fatty acid content during factorial-design batch experiment. Average DW (A), and total fatty acid (TFA) and EPA contents (B,C respectively) of Attheya septentrionalis cultures at 12 different treatments (two biological replicates per treatment, a and b). Values are average of three measurement replicates. Detailed values with standard deviations can be found in Table S2. e, exponential phase; s1, ﬁrst stationary phase (Day3); s2, second stationary phase (Day 5). abundant, whereas C18-FA were present in only low amounts. In all treatments, palmitoleic acid (16:1 n−7), and EPA (20:5 n−3) were the two major FA, together accounting for between 45 and 52% TFA, followed by myristic acid (14:0) with 10–18% TFA, palmitic acid (16:0) with 6–16% TFA, and DHA (22:6 n−3) with 3–6% TFA. However, small variations in the relative FA content between the diﬀerent treatment groups were apparent. These diﬀerences became more distinct by means of a principal component analysis (PCA, Figure 5A). The distribution of the 12 diﬀerent treatment groups (objects) represents their similarities and diﬀerences in the relative FA composition (% TFA), and the distribution of the FAs (vectors) indicate their contribution to the grouping of the objects. Objects were arranged along component 1 and component 2 according to the growth phases and salinities, respectively. DISCUSSION Impact of Culture Conditions on Growth Growth rates during the pre-cultivation were strongly dependent on the salinity, followed by irradiance and the interaction of both factors. Growth rates were 43% (LI) and 72% higher (HI) at HS compared to LS, and 5% (LS) and 27% (HS) higher when grown at HI compared to LI. The negligible eﬀect of HI on the growth rate at LS, and the much stronger eﬀect it caused at HS conditions, reveals the combined eﬀect of salinity and irradiance and emphasizes the importance of investigating combinatory eﬀects of diﬀerent growth factors. EPA% DWest = 4.61 −0.19∗X1 −0.12∗X2 + 1.74∗X3 + 0.14∗X1X2 −0.48∗X1X3 + 0.02∗X2X3 (4) TFA% DWest = 18.74 −0.77∗X1 −0.25∗X2 + 8.34∗X3 + 0.73∗X1X2 −1.43∗X1X3 −0.26∗X2X3 (5) Batch Experiment—Growth and Maximum Quantum Yield (QY) The exponential phase objects, grouping on the left side of component 1, were clearly separated from the stationary phase objects, which clustered on the right side, and objects were shifted further to the right along component 1 with increasing nutrient starvation (stationary phases). Palmitic and palmitoleic acids were correlated positively with stationary phase samples while the PUFAs were correlated with exponential phase samples. Except for the treatment HSLI_e (high-salinity, low-irradiance, and exponential phase), treatments of the same salinity were grouped together, with LS samples on the upper region along component 2, and HS samples arranged on the lower part of component 2. Myristic acid, and to a lesser extent hexadecatetraenoic acid (16:4 n-1) were correlated positively with HS treatments, while hexadecadienoic acid (16:2 n-4) was correlated positively with LS samples. Irradiance had only small eﬀect on the FA composition. The Euclidean Dendrogram illustrates the distinct grouping of the treatments (Figure 5B). Exponential phase samples were separated from stationary phase samples and both groups were then further divided according to the salinity treatment. For stationary phase samples, all LS samples grouped separately from the HS samples, followed by grouping according to their growth phase (s1 and s2) and ﬁnally by irradiance. Within exponential phase samples, HSHI samples grouped between HSLI and the LS samples, and separation was followed by irradiance, where replicates of LI were separated from the HI replicates. The two biological replicates of each treatment group were very similar. FIGURE 3 \| Dry weights (DW) and fatty acid content during factorial-design batch experiment. Average DW (A), and total fatty acid (TFA) and EPA contents (B,C respectively) of Attheya septentrionalis cultures at 12 different treatments (two biological replicates per treatment, a and b). Values are average of three measurement replicates. Detailed values with standard deviations can be found in Table S2. e, exponential phase; s1, ﬁrst stationary phase (Day3); s2, second stationary phase (Day 5). details on measured and estimated values can be found in the Supplementary Material (Table S2). Batch Experiment—Relative Fatty Acid (FA) Composition p In total 36 FAs were detected in the GC for A. septentrionalis, from which 11 (14:0, 16:0, 16:1 n−7, 16:2 n−4, 16:3 n−4, 16:4 n−1, 18:1 n−7, 18:4 n−3, 20:4 n−3, 20:5 n−3, and 22:6 n−3) constituted more than 1% TFA (Figure 4). C16-FA were the most These growth characteristics during the pre-cultivation also became evident when considering the growth curves during the batch experiment. The transition from exponential to stationary April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 6 Increasing EPA in a Diatom Steinrücken et al. GURE 4 \| Relative fatty acid (FA) composition during factorial-design batch experiment. Effect of the 12 different treatments on the relative FA composition (% of al fatty acids [TFA]) of the diatom Attheya septentrionalis (each with two biological replicates, a and b). Treatments were altering combinations of the three factors inity (22 and 35), irradiance (50 and 200 µmol photons m−2 s−1) and growth phase (exponential [e], 3 days stationary phase [s1], and 5 days stationary phase ]). Values are average of three measurement replicates. GURE 5 \| Similarities and differences in relative fatty acids (FA) composition (% of total fatty acids [TFA]) of Attheya septentrionalis cultures at 12 treatments during e factorial-design batch experiment (each with two biological replicates). Treatments were altering combinations of the three factors salinity (22 and 35), irradiance 0 and 200 µmol photons m−2 s−1) and growth phase (e, s1, s2). (A) Principal component analysis (PCA). Twenty-four objects representing the 12 treatment oups, and eight variables, representing the FAs with highest impact on the distributions. Values are average of three measurement replicates. e, exponential phase; ﬁrst stationary phase (Day 3); s2, second stationary phase (Day 5). (B) Euclidean Dendrogram showing dissimilarities between the treatment groups. LSLI, low inity and low irradiance; HSLI, high salinity and low irradiance; LSHI, low salinity and high irradiance; HSHI, high salinity and high irradiance. FIGURE 4 \| Relative fatty acid (FA) composition during factorial-design batch experiment. Effect of the 12 different treatments on the relative FA composition (% of total fatty acids [TFA]) of the diatom Attheya septentrionalis (each with two biological replicates, a and b). Treatments were altering combinations of the three factors salinity (22 and 35), irradiance (50 and 200 µmol photons m−2 s−1) and growth phase (exponential [e], 3 days stationary phase [s1], and 5 days stationary phase [s2]). Batch Experiment—Relative Fatty Acid (FA) Composition Values are average of three measurement replicates. FIGURE 4 \| Relative fatty acid (FA) composition during factorial-design batch experiment. Effect of the 12 different treatments on the relative FA composition (% of total fatty acids [TFA]) of the diatom Attheya septentrionalis (each with two biological replicates, a and b). Treatments were altering combinations of the three factors salinity (22 and 35), irradiance (50 and 200 µmol photons m−2 s−1) and growth phase (exponential [e], 3 days stationary phase [s1], and 5 days stationary phase [s2]). Values are average of three measurement replicates. FIGURE 5 \| Similarities and differences in relative fatty acids (FA) composition (% of total fatty acids [TFA]) of Attheya septentrionalis cultures at 12 treatments during the factorial-design batch experiment (each with two biological replicates). Treatments were altering combinations of the three factors salinity (22 and 35), irradiance (50 and 200 µmol photons m−2 s−1) and growth phase (e, s1, s2). (A) Principal component analysis (PCA). Twenty-four objects representing the 12 treatment groups, and eight variables, representing the FAs with highest impact on the distributions. Values are average of three measurement replicates. e, exponential phase; s1, ﬁrst stationary phase (Day 3); s2, second stationary phase (Day 5). (B) Euclidean Dendrogram showing dissimilarities between the treatment groups. LSLI, low salinity and low irradiance; HSLI, high salinity and low irradiance; LSHI, low salinity and high irradiance; HSHI, high salinity and high irradiance. FIGURE 5 \| Similarities and differences in relative fatty acids (FA) composition (% of total fatty acids [TFA]) of Attheya septentrionalis cultures at 12 treatments during the factorial-design batch experiment (each with two biological replicates). Treatments were altering combinations of the three factors salinity (22 and 35), irradiance (50 and 200 µmol photons m−2 s−1) and growth phase (e, s1, s2). (A) Principal component analysis (PCA). Twenty-four objects representing the 12 treatment groups, and eight variables, representing the FAs with highest impact on the distributions. Values are average of three measurement replicates. e, exponential phase; s1, ﬁrst stationary phase (Day 3); s2, second stationary phase (Day 5). (B) Euclidean Dendrogram showing dissimilarities between the treatment groups. LSLI, low salinity and low irradiance; HSLI, high salinity and low irradiance; LSHI, low salinity and high irradiance; HSHI, high salinity and high irradiance. phase was deﬁned by the decline of the growth curve with a concomitant decrease in the QY and was one day earlier for HS than for LS cultures. Batch Experiment—Relative Fatty Acid (FA) Composition In contrast to other elements that are essential for survival, diatoms rarely take up more silicon than is required for cell division (Reynolds, 2006). When silicon becomes scarce, its uptake depends on special silicic transport proteins (SITs); however, when silicon is abundant, its uptake is by diﬀusion (Shrestha and Hildebrand, 2015). Another reason for the increase of the FA fraction in stationary phase might be a decrease of silica in the cells. As a result of the siliciﬁed cell walls of diatoms, silicate availability in the medium is a key factor regulating their growth, as cells can only divide when new valves can be synthesized (Martin- Jézéquel et al., 2000). Studies have shown that in silicate-limited diatom cultures, uptake is restricted to the SITs (Shrestha and Hildebrand, 2015), and siliciﬁcation is reduced, resulting in thinner cell walls and a decreased silica content per cell (Martin- Jézéquel et al., 2000; Javaheri et al., 2015). Knuckey et al. (2002) found comparable results to our ﬁndings in an A. septentrionalis isolate from coastal waters in Tasmania, with an EPA content increasing from 1.3 to 4.2% DW from exponential to stationary phase. Concomitantly, the ash content fell sharply from 26.1 to 8.8% DW resulting in a corresponding increase of the other major organic fractions; proteins, carbohydrates, and lipids. The decreased ash content was most likely related to a diminished silica content of the DW, due to silicate limitation in the stationary phase. The same eﬀect might have contributed to the observed increase in TFA and EPA contents relative to DW in the stationary phase in our study. The typical ash content of microalgae contributes between 5 and 12% DW, but these values are higher in siliciﬁed diatoms, between 20 and 55% DW (Nalewajko, 1966; Renaud and Parry, 1994), where much of it is attributable to the extent of silicon in the cell walls. Hildebrand et al. (2012) stated that expressing the FA content as percentage of DW might underestimate the actual amount of FA in diatoms in terms of a per cell carbon basis, when compared with other microalgae, due to their high silica content. Expressing the FA content relative to the ash-free DW might preclude such an underestimation and furthermore might give a better understanding of the FA content and dynamics during silicate-replete and silicate-depleted conditions. Batch Experiment—Relative Fatty Acid (FA) Composition The QY reﬂects the photosynthetic performance of photosystem II and is used as a vitality indicator for cultures, as decreasing values are associated with stressful growth conditions (Maxwell and Johnson, 2000; Kräbs and Büchel, 2011). During exponential growth, QYs were slightly lower for LS cultures, but decreased to a lesser extent in stationary phase than they did for HS cultures. The strongest decrease in QY was observed for cultures grown at HS and HI. Hence, while April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 7 Increasing EPA in a Diatom Steinrücken et al. lipids consist predominantly of saturated and monounsaturated FAs, while PUFAs are generally present in polar membrane lipids (Olofsson et al., 2012). Therefore, TAG accumulation is typically accompanied by a noticeable increase of both palmitic (16:0) and palmitoleic acids (16:1 n-7), which often constitute the predominant FA in the TAG. Yet, PUFAs have also been reported to accumulate in TAG in diﬀerent microalgae species (Tonon et al., 2002; Sharma et al., 2012). The relative FA compositions (% TFA) observed during the experiment revealed a slight increase of the palmitic and palmitoleic acid fractions, together with a weak decrease of PUFAs after cultures progressed from exponential to stationary phase in all conditions. However, these shifts toward palmitic and palmitoleic acids were lower than typically observed during TAG accumulation, and revealed a concurrent increase of all major FA. This might indicate that TAG accumulated in the cells during stationary phase, containing PUFAs such as EPA, alongside palmitic, and palmitoleic acids. the combination HSHI appeared most advantageous during nutrient replete conditions as it caused the highest growth rates in exponential phase, it was also most stressful for the cultures during nutrient starvation. After the transition from exponential to stationary phase, cell division was assumed to have stopped due to nutrient depletion. Although all nutrients had been consumed on Day 3 in stationary phase, silicate most likely became the major limiting nutrient, as almost half of the silicate was consumed after one day in exponential phase, whereas nitrate and phosphate were both consumed to a lesser extent. When one element becomes limiting, other elements, that are more abundant, may be accumulated in the cell (Reynolds, 2006). Therefore, nitrate and phosphate, whilst not necessarily limiting, might have been taken up by the microalgae cells after silicate was depleted. Frontiers in Plant Science \| www.frontiersin.org Batch Experiment—Relative Fatty Acid (FA) Composition EPA Content During Batch Experiment Within the range of experimental variables considered, three factors were identiﬁed by the model as having aﬀected the EPA content in the present A. septentrionalis strain signiﬁcantly; the growth phase, salinity, and the interaction of both. Growth phase had the greatest impact, with increasing nutrient starvation leading to a higher EPA content relative to DW. The eﬀect of salinity and the combined eﬀects of both salinity and growth phase were lower and negative. Irradiances used in this experimental set-up did not aﬀect the EPA content signiﬁcantly. The mathematical model reﬂected the measured EPA values very accurately, with one exception: during the experiment, the EPA content decreased slightly from ﬁrst to second stationary phase under HSLI conditions, while the model predicted a further increase, similar to under the other conditions. Both the measured data and the model emphasized, that combining LS with 5 days’ nutrient starvation yielded a maximum EPA content of 7.1% DW on average. This DW-based EPA content is, to our knowledge, higher than previously reported for microalgae (Sukenik et al., 1991; Lu et al., 2001; Jiang and Gao, 2004; Hu and Gao, 2006; Patil et al., 2006). The EPA dynamics revealed the same pattern as for the TFA, suggesting that they were triggered by the same processes. Following exponential growth, the DW continued to increase between 5 and 20% from the ﬁrst to the second stationary phase, although nutrients were depleted, and accordingly cell division inhibited. When cell division is hampered due to an insuﬃcient nutrient supply, microalgae often produce carbonaceous storage compounds like carbohydrates and lipids. Many diatoms accumulate neutral storage lipids in the form of triacylglycerol (TAG), causing the lipid content to increase up to 50% of DW (Hu et al., 2008). Total FA content increased in all cultures from on average 10% in exponential phase, to 25 and 27% DW in the two stationary phases. Storage Salinity also aﬀected the TFA and EPA content, especially in combination with the growth phase. Interestingly, TFA and EPA contents were higher for HS cultures in exponential phase, but in contrast were higher for LS cultures in stationary phase. Hence, a greater increase in TFA and EPA from exponential to stationary phase occurred in the LS cultures. These observations April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 8 Increasing EPA in a Diatom Steinrücken et al. CONCLUSION The eﬀect of growth phase, salinity and irradiance, and their interactions on the EPA content in an Arctic A. septentrionalis strain was investigated by means of a factorial design experiment. The highest EPA values of 7.1% DW were achieved at a salinity of 22 and Day 5 of the stationary phase. However, at the same time, growth rates during exponential phase were reduced considerably at low salinities. Mathematical models revealed interactive eﬀects of salinity and irradiance on growth and of salinity and growth phase on the EPA content, emphasizing the importance of investigating the additive eﬀects of diﬀerent growth factors. Batch Experiment—Relative Fatty Acid (FA) Composition would be necessary in order to achieve higher cell densities and productivities before cultures reach the stationary phase. Furthermore, other growth conditions such as irradiance, temperature and pH might change considerably when moving from small-scale to large-scale systems, and can thereby aﬀect the EPA content of the cells. The strain used in the current study was isolated from an Arctic habitat and adapted to low temperatures, and therefore temperatures in the experiment were maintained at 10◦C. Several studies have shown that low temperatures can increase the PUFA content to maintain membrane ﬂuidity (Boelen et al., 2013). EPA values in the present experiments were higher than values recorded for the A. septentrionalis strain by Knuckey et al. (2002) grown at 20◦C. However, at the same time, growth rates of their strain were twice as high as the ones observed in this study. Hence, changing temperatures could additionally aﬀect both the EPA content and growth rates. This should also be evaluated with further work. might also be linked to diﬀerences in the FA accumulation and silica content of cells grown at diﬀerent salinities, but might also be related to the changing FA composition of membrane lipids, as an adaptation to variable salt concentrations and the resultant osmotic stress, as has been reported in several studies (Chen et al., 2008; Kumari et al., 2013). Microalgae grown at higher salinities might also reveal an increased ash content, due to an increased ion concentration (Renaud and Parry, 1994). Determining the ash and silicon contents of the cells and diﬀerentiating between polar (membrane) and neutral (storage) lipids, and their FA compositions in future experiments, might contribute to a better understanding of the reasons we observe diﬀerent EPA and TFA contents relative to DW under diﬀerent salinities and growth phases, and furthermore, might reveal in which lipid fraction the increased EPA levels are located. Relative Fatty Acid (FA) Composition During Batch Experiment Knuckey et al. (2002) suggested A. septentrionalis to be an excellent feed species for juvenile bivalve molluscs and other ﬁlter feeders. Its cell size is within the range that is suitable for ingestion by ﬁlter feeders and its protein level (32% DW) remained stable from exponential to stationary phase, while carbohydrate and lipid fractions increased. As shown in our study, EPA contents can be further increased in stationary phase, by changing growth conditions. This could make this diatom strain a promising EPA source for the North Atlantic ﬁsh aquaculture industry or for other application areas, such as the health and food sectors. Interestingly, irradiance did not signiﬁcantly aﬀect the EPA content relative to DW, although irradiances have been shown to aﬀect photosynthetic membranes. Generally, photosynthetic membranes increase at low irradiance and are reduced at high irradiances and hence, increasing irradiance has been reported to result in a decrease in EPA and other PUFAs in diﬀerent microalgae species (Adlerstein et al., 1997; Fábregas et al., 2004). However, irradiance did aﬀect the FA proﬁle in this study, although only to a minor degree. The relative amounts of the main FAs (% TFA) were for the most part aﬀected by nutrient availability primarily, followed by salinity, time of nutrient starvation (days in stationary phase) and irradiance. The diﬀerences between the growth phases were mainly due to palmitic and palmitoleic acids and therefore might be related to an accumulation of TAG in stationary phase. The eﬀects of the growth conditions (salinity and irradiance) were less distinct and are more diﬃcult to explain, but might be related to reconstructions of cellular membranes as an adaptation to the cultivation conditions. Potential for Microalgae-Based Technologies g The low salinity of 22 was more eﬀective in increasing EPA content in the stationary phase in the prevalent A. septentrionalis strain, but at the same time, it considerably decreased growth rates compared to a HS of 35. In future large-scale cultivations, EPA productivity would be dependent on both the growth rates and the EPA content in the cells. Calculating the EPA productivities from exponential phase until Day 5 of the stationary phase revealed 0.97 mg L−1 d−1 for LS cultures and 0.72 mg L−1 d−1 for HS cultures. Thus, under the prevailing conditions, higher productivities were obtained for the LS cultures, although these productivities are much lower than those seen in commercial production due to the much lower biomass concentrations used in this experimental setup. Whether our results can be successfully repeated in up-scaled systems needs to be evaluated further. Higher nutrient concentrations AUTHOR CONTRIBUTIONS PS, SP, SM, and SE: designed research; PS: performed research; PS and SM: analyzed data; PS: wrote the paper; SE and SM: revised the paper. All authors read and approved the ﬁnal manuscript. Frontiers in Plant Science \| www.frontiersin.org REFERENCES doi: 10.1007/s10811-013-0119-z Lu, C., Rao, K., Hall, D., and Vonshak, A. (2001). Production of eicosapentaenoic acid (EPA) in Monodus subterraneus grown in a helical tubular photobioreactor as aﬀected by cell density and light intensity. J. Appl. Phycol. 13, 517–522. doi: 10.1023/A:1012515500651 Chauton, M. S., Reitan, K. I., Norsker, N. H., Tveterås, R., and Kleivdal, H. T. (2015). A techno-economic analysis of industrial production of marine microalgae as a source of EPA and DHA-rich raw material for aquafeed: research challenges and possibilities. Aquaculture 436, 95–103. doi: 10.1016/j.aquaculture.2014.10.038 Martin-Jézéquel, V., Hildebrand, M., and Brzezinski, M. A. (2000). Silicon metabolism in diatoms: Implications for growth. J. Phycol. 36, 821–840. doi: 10.1046/j.1529-8817.2000.00019.x Martins, D. A., Custódio, L., Barreira, L., Pereira, H., Ben-Hamadou, R., Varela, J., et al. (2013). Alternative sources of n-3 long-chain polyunsaturated fatty acids in marine microalgae. Mar. Drugs 11, 2259–2281. doi: 10.3390/md11072259 Chen, G. Q., Jiang, Y., and Chen, F. (2008). Salt-induced alterations in lipid composition of diatom Nitzschia laevis (Bacillariophyceae) under heterotrophic culture condition. J. Phycol. 44, 1309–1314. doi: 10.1111/j.1529-8817.2008.00565.x well, K., and Johnson, G. N. (2000). Chlorophyll ﬂuorescenc Maxwell, K., and Johnson, G. N. (2000). Chlorophyll ﬂuorescence - a practic guide. J. Exp. Bot. 51, 659–668. doi: 10.1093/jexbot/51.345.659 j Chen, J., Li, Y., Xie, M., Chiu, C., Liao, S., and Lai, W. (2012). Factorial design of experiment for biofuel production by Isochrysis galbana. Int. Conf. Environ. Energy Biotechnol. 33, 91–95. Meier, S., Mjøs, S. A., Joensen, H., and Grahl-Nielsen, O. (2006). Validation of a one-step extraction/methylation method for determination of fatty acids and cholesterol in marine tissues. J. Chromatogr. A 1104, 291–298. doi: 10.1016/j.chroma.2005.11.045 Duarte, M. M. M. B., Da Silva, J. E., Passavante, J. Z. D. O., Fernanda Pimentel, M., De Barros Neto, B., and Da Silva, V. L. (2001). Macroalgae as lead trapping agents in industrial eﬄuents - a factorial design analysis. J. Braz. Chem. Soc. 12, 499–506. doi: 10.1590/S0103-505320010004 00010 Nalewajko, C. (1966). Dry weight, ash, and volume data for some freshwater planktonic algae. J. Fish. Res. Board Canada 23, 1285–1288. doi: 10.1139/f66-119 Olofsson, M., Lamela, T., Nilsson, E., Bergé, J. P., del Pino, V., Uronen, P., et al. (2012). Seasonal variation of lipids and fatty acids of the microalgae Nannochloropsis oculata grown in outdoor large-scale photobioreactors. Energies 5, 1577–1592. doi: 10.3390/en5051577 Dunstan, G. A., Volkman, J. K., Barrett, S. M., and Garland, C. D. (1993). REFERENCES Javaheri, N., Dries, R., Burson, A., Stal, L. J., Sloot, P. M. A., and Kaandorp, J. A. (2015). Temperature aﬀects the silicate morphology in a diatom. Sci. Rep. 5:11652. doi: 10.1038/srep11652 Adarme-Vega, T. C., Lim, D. K. Y., Timmins, M., Vernen, F., Li, Y., and Schenk, P. M. (2012). Microalgal biofactories: a promising approach towards sustainable omega-3 fatty acid production. Microb. Cell Fact. 11:96. doi: 10.1186/1475-2859-11-96 Jiang, H., and Gao, K. (2004). Eﬀects of lowering temperature during culture on the production of polyunsaturated fatty acids in the marine diatom Phaeodactylum tricornutum (Bacillariophyceae). J. Phycol. 40, 651–654. doi: 10.1111/j.1529-8817.2004.03112.x Adlerstein, D., Bigogno, C., Khozin, I., and Cohen, Z. (1997). The eﬀect of growth temperature and culture density on the molecular species composition of the galactolipids in the red microalga Porpyridium cruentum (Rhodophyta). J. Phycol. 33, 975–979. doi: 10.1111/j.0022-3646.1997. 00975.x Khozin-Goldberg, I., Leu, S., and Boussiba, S. (2016). “Microalgae as a source for VLC-PUFA production,” in Lipids in Plant and Algae Development, eds Y. Nakamura and Y. Li-Beisson (Berlin: Springer), 471–509. (Rhodophyta). J. Phycol. 33, 975–979. doi: 10.1111/j.0022-3646.1997. 00975.x Knuckey, R. M., Brown, M. R., Barrett, S. M., and Hallegraeﬀ, G. M. (2002). Isolation of new nanoplanktonic diatom strains and their evaluation as diets for juvenile Paciﬁc oysters (Crassostrea gigas). Aquaculture 211, 253–274. doi: 10.1016/S0044-8486(02)00010-8 Aizdaicher, N. A., and Markina, Z. V. (2011). Inﬂuence of changes in sea water salinity on the growth, photosynthetic pigment content, and cell size of the benthic alga Attheya ussurensis Stonik, Orlova et Crawford, 2006 (Bacillariophyta). Russ. J. Mar. Biol. 37, 472–477. doi: 10.1134/S1063074011060034 Kräbs, G., and Büchel, C. (2011). Temperature and salinity tolerances of geographically separated Phaeodactylum tricornutum Böhlin strains: maximum quantum yield of primary photochemistry, pigmentation, proline content and growth. Bot. Mar. 54, 231–241. doi: 10.1515/bot.2011.037 Boelen, P., van Dijk, R., Sinninghe Damsté, J. S., Rijpstra, W. I. C., and Buma, A. G. J. (2013). On the potential application of polar and temperate marine microalgae for EPA and DHA production. AMB Express 3:26. doi: 10.1186/2191-0855-3-26 Kumari, P., Kumar, M., Reddy, C. R., and Jha, B. (2013). “Algal lipids, fatty acids and sterols,” in Functional Ingredients From Algae for Foods and Nutraceuticals, ed H. Domínguez (Cambridge: Woodhead Publishing Limited), 87–134. Cepák, V., Pribyl, P., Kohoutková, J., and Kaštánek, P. (2013). Optimization of cultivation conditions for fatty acid composition and EPA production in the eustigmatophycean microalga Trachydiscus minutus. J. Appl. Phycol. 26, 181–190. FUNDING This work was supported by EU MIRACLES project and has received funding from the European Union’s Seventh Framework Programme for research, technological development, and demonstration under grant agreement No 613588. April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 9 Steinrücken et al. Increasing EPA in a Diatom SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2018. 00491/full#supplementary-material Big thanks to Dorinde Kleinegris for the helpful input, and to Bryan Wilson for valiantly giving up 1 week of lunch hours to read about algae and proofread the English!! REFERENCES Changes in the lipid composition and maximisation of the polyunsaturated fatty acid content of three microalgae grown in mass culture. J. Appl. Phycol. 5, 71–83. doi: 10.1007/BF02182424 Parsons, T. R., Maita, Y., and Lalli, C. M. (1992). A Manual of Chemical and Biological Methods for Seawater Analysis. Oxford: Pergamon Press. Fábregas, J., Maseda, A., Domínguez, A., and Otero, A. (2004). The cell composition of Nannochloropsis sp. changes under diﬀerent irradiances in semicontinuous culture. World J. Microbiol. Biotechnol. 20, 31–35. doi: 10.1023/B:WIBI.0000013288.67536.ed Patil, V., Källqvist, T., Olsen, E., Vogt, G., and Gislerød, H. R. (2006). Fatty acid composition of 12 microalgae for possible use in aquaculture feed. Aquac. Int. 15, 1–9. doi: 10.1007/s10499-006-9060-3 Hildebrand, M., Davis, A. K., Smith, S. R., Traller, J. C., and Abbriano, R. (2012). The place of diatoms in the biofuels industry. Biofuels 3, 221–240. doi: 10.4155/bfs.11.157 Patil, V., Reitan, K. I., Knutsen, G., Mortensen, L. M., Källqvist, T., Olsen, E., et al. (2005). Microalgae as source of polyunsaturated fatty acids for aquaculture. Curr. Top. Plant Biol. 6, 57–65. Hu, H., and Gao, K. (2006). Response of growth and fatty acid compositions of Nannochloropsis sp. to environmental factors under elevated CO2 concentration. Biotechnol. Lett. 28, 987–992. doi: 10.1007/s10529-006- 9026-6 Prestegard, S. K., Erga, S. R., Steinrücken, P., Mjøs, S. A., Knutsen, G., and Rohloﬀ, J. (2015). Speciﬁc metabolites in a Phaeodactylum tricornutum strain isolated from western Norwegian fjord water. Mar. Drugs 14:9. doi: 10.3390/md14010009 Hu, Q., Sommerfeld, M., Jarvis, E., Ghirardi, M., Posewitz, M., Seibert, M., et al. (2008). Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances. Plant J. 54, 621–639. doi: 10.1111/j.1365-313X.2008.03492.x Rampen, S. W., Schouten, S., Elda Panoto, F., Brink, M., Andersen, R. A., Muyzer, G., et al. (2009). Phylogenetic position of Attheya longicornis and Attheya septentrionalis (bacillariophyta). J. Phycol. 45, 444–453. doi: 10.1111/j.1529-8817.2009.00657.x Frontiers in Plant Science \| www.frontiersin.org April 2018 \| Volume 9 \| Article 491 10 Steinrücken et al. Increasing EPA in a Diatom Renaud, S. M., and Parry, D. L. (1994). Microalgae for use in tropical aquaculture II: Eﬀect of salinity on growth, gross chemical composition and fatty acid composition of three species of marine microalgae. J. Appl. Phycol. 6, 347–356. doi: 10.1007/BF02181949 to triacylglycerols in four microalgae. Phytochemistry 61, 15–24. doi: 10.1016/S0031-9422(02)00201-7 Van Wagenen, J., Miller, T. W., Hobbs, S., Hook, P., Crowe, B., and Huesemann, M. (2012). Frontiers in Plant Science \| www.frontiersin.org REFERENCES Eﬀects of light and temperature on fatty acid production in Nannochloropsis salina. Energies 5, 731–740. doi: 10.3390/en50 30731 Reynolds, C. S. (2006). Ecology of Phytoplankton. Cambridge: Cambridge University Press. Sharma, K. K., Schuhmann, H., and Schenk, P. M. (2012). High lipid induction in microalgae for biodiesel production. Energies 5, 1532–1553. doi: 10.3390/en5051532 Walne, P. R. (1970). Studies on the food value of nineteen genera of algae to juvenile bivalves of the genera Ostrea, Crassostrea, Mercenaria and Mytilus. Fish. Invest. Ser. 2, 1–62. Shrestha, R. P., and Hildebrand, M. (2015). Evidence for a regulatory role of diatom silicon transporters in cellular silicon responses. Eukaryot. Cell 14, 29-40. doi: 10.1128/EC.00209-14 Wasta, Z., and Mjøs, S. A. (2013). A database of chromatographic properties and mass spectra of fatty acid methyl esters from omega-3 products. J. Chromatogr. A 1299, 94–102. doi: 10.1016/j.chroma.2013. 05.056 Spolaore, P., Joannis-Cassan, C., Duran, E., and Isambert, A. (2006). Commercial applications of microalgae. J. Biosci. Bioeng. 101, 87–96. doi: 10.1263/jbb.101.87 Xu, X. Q., and Beardall, J. (1997). Eﬀect of salinity on fatty acid composition of a green microalga from an antarctic hypersaline lake. Phytochemistry 45, 655–658. doi: 10.1016/S0031-9422(96)00868-0 Steinrücken, P., Erga, S. R., Mjøs, S. A., Kleivdal, H., and Prestegard, S. K. (2017). Bioprospecting North Atlantic microalgae with fast growth and high polyunsaturated fatty acid (PUFA) content for microalgae-based technologies. Algal Res. 26, 392–401. doi: 10.1016/j.algal.2017.07.030 Zhu, C. J., and Lee, Y. K. (1997). Determination of biomass dry weight of marine microalgae. J. Appl. Phycol. 9, 189–194. doi: 10.1023/A:1007914806640 Stonik, I. V., Kapustina, I. I., Aizdaicher, N. A., and Svetashev, V. I. (2017). Sterols and fatty acids from Attheya planktonic diatoms. Chem. Nat. Compd. 53, 422–425. doi: 10.1007/s10600-017-2013-4 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Sukenik, A., Zmora, O., and Carmeli, Y. (1991). Biochemical quality of marine unicellular algae with special emphasis on lipid composition. II. Nannochloropsis sp. Aquaculture 117, 313–326. doi: 10.1016/0044-8486(93)90328-V Copyright © 2018 Steinrücken, Mjøs, Prestegard and Erga. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. REFERENCES No use, distribution or reproduction is permitted which does not comply with these terms. Copyright © 2018 Steinrücken, Mjøs, Prestegard and Erga. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Tatsuzawa, H., and Takizawa, E. (1995). Changes in lipid and fatty acid composition of Pavlova lutheri. Phytochemistry 40, 397–400. doi: 10.1016/0031-9422(95)00327-4 Tonon, T., Harvey, D., Larson, T. R., and Graham, I. A. (2002). Long chain polyunsaturated fatty acid production and partitioning April 2018 \| Volume 9 \| Article 491 Frontiers in Plant Science \| www.frontiersin.org 11
https://openalex.org/W2750931391	https://miun.diva-portal.org/smash/get/diva2:1137962/FULLTEXT01	English	null	Tightrope walking: external impact factors on workplace health management in small-scale enterprises	Society, health and vulnerability	2,017	cc-by	12,762	Society, Health & Vulnerability Society, Health & Vulnerability ISSN: (Print) 2002-1518 (Online) Journal homepage: http://www.tandfonline.com/loi/zvgi20 Date: 30 August 2017, At: 02:18 Download by: [Mittuniversitetet] KEYWORDS small-scale enterprises; workplace health management; qualitative methods; management; Norway; Sweden Marianne Hedlund, Bodil J Landstad & Stig Vinberg To cite this article: Marianne Hedlund, Bodil J Landstad & Stig Vinberg (2017) Tightrope walking: external impact factors on workplace health management in small-scale enterprises, Society, Health & Vulnerability, 8:sup1, 1350551, DOI: 10.1080/20021518.2017.1350551 To link to this article: http://dx.doi.org/10.1080/20021518.2017.1350551 © 2017 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published online: 29 Aug 2017. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=zvgi20 Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=zvgi20 Download by: [Mittuniversitetet] SOCIETY, HEALTH & VULNERABILITY, 2017 VOL. 8, 1350551 https://doi.org/10.1080/20021518.2017.1350551 SOCIETY, HEALTH & VULNERABILITY, 2017 VOL. 8, 1350551 https://doi.org/10.1080/20021518.2017.1350551 KEYWORDS ABSTRACT Small-scale enterprises (SSEs) are important for ensuring growth, innovation, job creation, and social integration in working life. Research shows that SSEs pay little attention to and have insufficient competence in workplace health management. From the perspective of managers, this study explores how external factors influence the development of this management. The article refers to a case study among eight Norwegian and ten Swedish managers of SSEs in the middle part of Norway and Sweden. We used a stepwise qualitative approach to analyse data, using an interpretive indexing of main categories. Two main categories were found to have an influence on the development of workplace health management: (1) restricted leeway and (2) commitments. Concerning the first main category, areas that managers highlight as important comprise the legal framework and regulations; workforce and market situation, production, economy; and occupational safety and health issues. Areas related to the second main category were advice from the board, guidance from mentors, work-related networks, and family and friends as buffers. One conclusion is that despite limited scope for developing workplace health management, managers find supportive guidance and inspiration from environments that are committed to helping them and their enterprise. Downloaded by [Mittuniversitetet] at 02:18 30 August 2017 MacEachen et al., 2010). Studies have shown that most SSEs pay little attention to OHS issues (Andersson & Hägg, 2006; Breucker, 2001; Frick, Langaa Jensen, Quinlan, & Wilthagen, 2000: Hasle & Limborg, 2006) and that specific strategies are needed to implement solutions in SSEs (Breslin et al., 2010; European Agency for Safety and Health at Work, 2013b). Similarly, research also suggests that SSEs have limited competence in creating health-promoting workplaces (Landstad, Hedlund, & Vinberg, 2017; Moser & Karlqvist, 2004; Torp & Moen, 2006). Nevertheless, the European Network for Workplace Health Promotion (2001) states that SSEs have a unique ability to affect employee health positively because of factors such as the family atmosphere and SSE managers’ immediate control of working condi- tions. In addition, Meggeneder (2007) argues that small enterprises have organizational characteristics that are suitable for introducing and implementing workplace health promotion. Thus, the research has arrived at contradictory results regarding SSEs’ ability to develop WHM. Tightrope walking: external impact factors on workplace health management in small-scale enterprises Marianne Hedlunda,b, Bodil J Landstad c,d and Stig Vinbergc aFaculty of Nursing and Health Science, Nord University, Levanger, Norway; bDepartment of Social work, NTNU, Trondheim, Norway; cDepartment of Health Sciences, Mid Sweden University, Östersund, Sweden; dLevanger Hospital, Nord-Trøndelag Hospital Trust, Levanger, Norway © 2017 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. lund marianne.hedlund@nord.no Faculty of Nursing and Health Science, Nord University, Høgskolevegen 27, Previous research: WHM Employers now generally consider a reasonable working environment a prerequisite for legitimacy among the organization’s stakeholders (Almqvist & Henningsson, 2009; Frick & Zwetsloot, 2007). Research shows that companies must pay close attention to WHM to prevent sickness and create healthy workplaces (Department of Health, 2004; Holt & Powell, 2015; Lindström, Schrey, Ahonen, & Kaleva, 2000). A literature review concludes that active management, the involvement of all staff, and comprehensiveness in measures are success factors for implementing health-promoting practices (Chu et al., 2000; Sparling, 2010). Currently, there is also increased attention on the development of SSEs and their investments in health and the working environment (Abrahamsson, 2006; Hasle, Limborg, Kallehave, Klitgaard, & Rye Andersen, 2012; Witt, Olsen, & Ablah, 2013). However, several studies indicate that health-promo- tion practices are less developed in SSEs, and there RQ1: How do these external factors affect WHM? RQ1: How do these external factors affect WHM? Key concepts The Ottawa Charter defines health promotion as ‘the process of enabling people to increase control over, and to improve, their health’ (WHO, 1986). This definition approaches health not only as an absence of disease, but also as a resource in everyday life that includes physical, mental, and social well-being and capacity (Eriksson, 2011). There is no consensus on the definition of WHM. Jiménez, Winkler, and Dunkl (2016) assert that WHM consists of a set of leadership behaviours that is in continuous interaction with the working environment, with the goal of designing that environment to enhance employee health. As stated, we regard WHM as the actions taken by managers in the workplace to promote OHS, a good working environment, and employee health. External factors refer to factors that affect WHM in SSEs from the outside, that is, the market, financials, business sector, legislation, or the personal circumstances of managers. SSEs are defined as enter- prises employing fewer than 20 people (EU, 2003). p g Research indicates that the SSE manager (who is often the company owner) is a key person because his or her opinions and values influence the company’s approach to health and safety improvements (Hasle & Limborg, 2006). However, holding a managerial posi- tion in an SSE often involves long and irregular work- ing hours (Gunnarsson, Vingård, & Josephson, 2007: Nordenmark, Vinberg, & Strandh, 2012), along with high and conflicting work demands (Bornberger- Dankvardt, Ohlson, & Westerholm, 2003; Stephan & Roesler, 2010). Several of these conditions may hin- der the implementation of health-promotion prac- tices in the enterprise. In addition, SSE managers may consider these practices and working-environ- ment regulations and demands as a financial burden that is too heavy for a small enterprise to bear (Hasle & Limborg, 2006). Based on the above-mentioned characteristics in SSEs, it is important to gain more knowledge about how external factors influence the development of WHM. By external factors, we mean legal regulations, the market situation, or personal circumstances. OHS consists of strategies to reduce ill health at work. This can be achieved by promoting the use of systematic managerial processes to detect and abate workplace hazards and actively manage the quality of the overall work environment (Frick et al., 2000). Introduction Small-scale enterprises (SSEs) are important for Europe’s economy, and the European Commission considers them a key factor to ensuring growth, innovation, job creation, and social integration (European Agency for Safety and Health at Work, 2013a). In Sweden, approximately 900,000 indivi- duals – more than one-fifth of the population working in the private sector – are employed in this enterprise group (Statistics Sweden, 2011). The corresponding figure for Norway is 550,000 indivi- duals, which is one-fifth of the population working in the private sector (Statistics Norway, 2014). Because of this group’s importance to working life and SSEs’ less developed workplace health manage- ment (WHM), there is a need for more knowledge about how external factors affect WHM in this enterprise group. By WHM, in this article, we refer to actions taken by managers in the workplace to promote occupational health and safety (OHS), a good working environment, and employee health. According to various studies, SSEs represent a par- ticular challenge in terms of working with OHS issues (Eakin, Lamm, & Limborg, 2000; European Agency for Safety and Health at Work, 2013b; According to Sverke (2009), Scandinavian work- environment regulations emphasize a consensus F CONTACT Marianne Hedlund Levanger 7600, Norway M. HEDLUND ET AL. 54 54 ability to develop WHM in SSEs. We focus on the following research questions: ability to develop WHM in SSEs. We focus on the following research questions: model according to which a motivated and harmonized workforce will result in long-term organizational effec- tiveness. Trade unions and employers are therefore the most important stakeholders, operationalizing public policy regulations and transforming them into practice (Hasle, Limborg, & Nielsen, 2014). Legislation indicates that both employers and employees have a general responsibility to create a sound working environment as part of maintaining a general ‘license to operate’ (Hasle et al., 2014, p. 74). State regulation of the work- ing environment in both Norway and Sweden requires a working life and working environment that provide the basis for a healthy and meaningful working situa- tion (The Norwegian Working Environment Act, 2015; The Swedish Working Environment Act, 2014). Managers in SSEs may have low competence (Hasle & Limborg, 2006) but good organizational conditions for developing healthy workplaces (Meggeneder, 2007). This makes it important to gain knowledge about how structural and external factors affect managers’ internal conditions and possibilities for WHM. The same applies to the manner of managers’ reasoning and prio- rities. Introduction Managers of SSEs usually have limited human- resource management, economic resources, and elbow room for manoeuvring. Consequently, they may have to balance on a fine line – like a tightrope walker – to meet different requirements of the enterprise, such as creating a foundation for a healthy workplace, keeping the budget within its limit, maintaining a good market position, and being oriented towards customers. model according to which a motivated and harmonized workforce will result in long-term organizational effec- tiveness. Trade unions and employers are therefore the most important stakeholders, operationalizing public policy regulations and transforming them into practice (Hasle, Limborg, & Nielsen, 2014). Legislation indicates that both employers and employees have a general responsibility to create a sound working environment as part of maintaining a general ‘license to operate’ (Hasle et al., 2014, p. 74). State regulation of the work- ing environment in both Norway and Sweden requires a working life and working environment that provide the basis for a healthy and meaningful working situa- tion (The Norwegian Working Environment Act, 2015; The Swedish Working Environment Act, 2014). Managers in SSEs may have low competence (Hasle & Limborg, 2006) but good organizational conditions for developing healthy workplaces (Meggeneder, 2007). This makes it important to gain knowledge about how structural and external factors affect managers’ internal conditions and possibilities for WHM. The same applies to the manner of managers’ reasoning and prio- rities. Managers of SSEs usually have limited human- resource management, economic resources, and elbow room for manoeuvring. Consequently, they may have to balance on a fine line – like a tightrope walker – to meet different requirements of the enterprise, such as creating a foundation for a healthy workplace, keeping the budget within its limit, maintaining a good market position, and being oriented towards customers. RQ1: Which external factors have an impact on WHM? RQ1: Which external factors have an impact on WHM? Aim and research questions Managers who work with broader intervention strategies exert a greater influence on outcomes related to employee health than managers do who work with more one- dimensional strategies (Dellve, Skagert, & Vilhelmsson, 2007; Grawitch, Gottschalk, & Munz, 2006). However, a review of studies in the Nordic countries revealed that most studies had an individual focus on changing workers’ lifestyles or behaviour by using a top-down approach that does not focus on settings-related factors (Torp, Eklund, & Thorpenberg, 2011). In addition, the SSE workplace is a challenging context for managers. They must consider various fac- tors when they work with health-promotion issues at the workplace, such as the number of employees, busi- ness age, structure, workforce, manager centricity, and culture (Cunningham, Sinclair, & Schulte, 2014, p. 148). According to a qualitative study of SSEs, managers try to adapt the workplace for sick employees (The Swedish Social Insurance Inspectorate, 2012). However, their experience is that the role of the Social Insurance Agency and their own coordinating role are unclear (The Swedish Social Insurance Inspectorate [Inspektionen för Socialförsäkring], 2012). In addition, OHS research shows that only 10–55% of Swedish employees in SSEs have access to occupational health services (OH) Gunnarsson, Andersson, & Josephson, 2011. The Norwegian workforce displays a similar ten- dency (Moen, Hanoa, Lie, & Larsen, 2015). One con- sequence of these facts is that many SSEs engage in only a limited use of these resources. Studies of health and safety practices in the workplace identify several factors that either hinder or facilitate implementation (Whysall, Haslam, & Haslam, 2006). Hindering factors include management commitment, managers’ general attitudes towards health, insufficient resources, and prioritization of production. Facilitating factors include supportive managers, local control over budget spend- ing for health, and good communication among man- agers and co-workers (Whysall et al., 2006). Improved safety and product quality were the origi- nal goals of WHM (Rootman et al., 2001). Later, this concept was developed into various approaches to improve employees’ health. According to Gjerstad and Lysberg (2012), in recent years, WHM has received increased attention in the Nordic countries. Managers influence the interaction of individual and organiza- tional aspects. Important concepts include health awareness, workload, control, reward, community, fair- ness, and values (Jiménez et al., 2016). Aim and research questions Managers who work with broader intervention strategies exert a greater influence on outcomes related to employee health than managers do who work with more one- dimensional strategies (Dellve, Skagert, & Vilhelmsson, 2007; Grawitch, Gottschalk, & Munz, 2006). However, a review of studies in the Nordic countries revealed that most studies had an individual focus on changing workers’ lifestyles or behaviour by using a top-down approach that does not focus on settings-related factors (Torp, Eklund, & Thorpenberg, 2011). In addition, the SSE workplace is a challenging context for managers. They must consider various fac- tors when they work with health-promotion issues at the workplace, such as the number of employees, busi- ness age, structure, workforce, manager centricity, and culture (Cunningham, Sinclair, & Schulte, 2014, p. 148). The position of this study is to explore and con- tribute to knowledge about how mentioned external factors contribute to development of WHM in SSEs. Research about WHM and OHS in SSEs has more been focused on management culture (Meggeneder, 2007) and internal workplace-related strategies (Frick et al., 2000; Landstad et al., 2017). Aim and research questions This study’s overall research aim was to explore which and how external factors affects managers’ 55 SOCIETY, HEALTH & VULNERABILITY 55 Many managers in small-scale enterprises coop- erate locally in professional networks (Gunnarsson, 2010). Regional or local professional networks may improve health and safety in small enterprises (Vinberg, 2006). The characteristics of long-lasting networks are trust, good relations, and usefulness to entrepreneurs (Antonsson, Birgersdotter, & Bornberger-Dankvardt, 2002). The results from a study of three Danish networks on OHS issues (Limborg & Grøn, 2014) indicate that small enter- prises are more affected by the actions and attitudes of their competitors and collaborators within their industry than by general campaigns, regulations, and visits from the Labour Inspectorate. The authors conclude that this result suggests the need for recon- sidering or supplementing the previous strategy towards SSEs, which in general has closely matched the strategy towards large enterprises. A new strategy should include support for the establishment and should facilitate networking between similar compa- nies that can support the companies’ joint effort to achieve a common commitment to satisfy health and safety standards (Limborg & Grøn, 2014). are several reasons for the low involvement of small enterprises in health-promotion issues (Griffin, Hall, & Watson, 2005; Moser & Karlqvist, 2004). They lack the motivation and resources to work with health issues, there are few organizational mechanisms for communication, and they have limited resources to devote to occupational health issues (Breucker, 2001; Hasle & Limborg, 2006). are several reasons for the low involvement of small enterprises in health-promotion issues (Griffin, Hall, & Watson, 2005; Moser & Karlqvist, 2004). They lack the motivation and resources to work with health issues, there are few organizational mechanisms for communication, and they have limited resources to devote to occupational health issues (Breucker, 2001; Hasle & Limborg, 2006). Hasle & Limborg, 2006). Improved safety and product quality were the origi- nal goals of WHM (Rootman et al., 2001). Later, this concept was developed into various approaches to improve employees’ health. According to Gjerstad and Lysberg (2012), in recent years, WHM has received increased attention in the Nordic countries. Managers influence the interaction of individual and organiza- tional aspects. Important concepts include health awareness, workload, control, reward, community, fair- ness, and values (Jiménez et al., 2016). Method This study analysed interview data from managers in 18 SSEs in central regions of Norway and Sweden. The methodology used to study conditions for creat- ing WHM was based on a stepwise inductive method (Patton, 2002; Tjora, 2012). This means that analyti- cal categories are not stipulated in advance (Patton, 2002; Tjora, 2012) but rather through a stepwise process. The researchers did not use predefined themes, but instead identified and extracted data across the empirical material based on their purpose- fulness and relevance to answering the research ques- tions (Patton, 2002; Tjora, 2012). We searched for patterns and concepts that echo patterns found to answer the research questions. Eventually, we rela- belled these patterns and concepts into categories linked to adequate theories and reanalysed them. The goal was to generate and derive subtopics that reflected patterns found in the data analysis and then to relate them to pertinent theories and research so main categories could be constructed. In the analysis section, we give further details on this process. ( g , , , , p ) According to a qualitative study of SSEs, managers try to adapt the workplace for sick employees (The Swedish Social Insurance Inspectorate, 2012). However, their experience is that the role of the Social Insurance Agency and their own coordinating role are unclear (The Swedish Social Insurance Inspectorate [Inspektionen för Socialförsäkring], 2012). In addition, OHS research shows that only 10–55% of Swedish employees in SSEs have access to occupational health services (OH) Gunnarsson, Andersson, & Josephson, 2011. The Norwegian workforce displays a similar ten- dency (Moen, Hanoa, Lie, & Larsen, 2015). One con- sequence of these facts is that many SSEs engage in only a limited use of these resources. Studies of health and safety practices in the workplace identify several factors that either hinder or facilitate implementation (Whysall, Haslam, & Haslam, 2006). Hindering factors include management commitment, managers’ general attitudes towards health, insufficient resources, and prioritization of production. Facilitating factors include supportive managers, local control over budget spend- ing for health, and good communication among man- agers and co-workers (Whysall et al., 2006). The foundation for the analysis was the similarity of verbal references among the participants’ or managers’ viewpoints, but representing empirical contours that 56 M. HEDLUND ET AL. 56 56 The interviews lasted from 90 to 120 minutes, and they were conducted at a location convenient for the participants (Patton, 2002, p. 341). Method are more typical or general for the strategic sample. All of the enterprises investigated participated in a WHM project intended to give managers improved skills and competence in health, safety, and work-environment issues. One Norwegian and one Swedish OH led the project, and the focus was on management issues, psy- chosocial working conditions, and employees’ health. Components of the project were investigations of work- ing conditions and employee health, network meetings, and leadership support. We used an interview guide to collect data. The guide involved asking for managers’ experiences and reflec- tions on external factors that influenced their WHM and how those factors either affected their opportunities or presented obstacles to the creation of a health-pro- moting workplace. Immediately following the inter- views, the tape-recorded interviews were transcribed. This article does not present data from the inter- vention study. Instead, the focus is on what managers identified as their possibilities and obstacles for WHM, including knowledge gain from the intervention. Analysis We used a stepwise method when analysing the data (Charmaz, 2000; Mason, 2002; Tjora, 2012). Stepwise analysis provides a flexible, heuristic strategy (Charmaz, 2000, p. 510) for analysing meaning and interpretation in data material. We used this strategy because we continuously compared utterances and the expressed experiences in the data and searched for patterns of meaning about the research questions (Patton, 2002). Below, we introduce the analytical steps in the order in which they were performed. Recruitment criteria To ensure a wider range of SSE manager types in the strategic sample, we recruited managers from differ- ent branches of the private sector. We recruited infor- mants from SSEs in Norway and Sweden who agreed to participate in an intervention project on WHM in SSEs. One selection criterion was that the enterprises had no more than 20 employees. Further criteria were that the enterprises employed both sexes, that they were located in the middle of Norway and Sweden, and that they represented different types of services in the private sector (see Table 1). The sampling is qualitative and is not intended to be representative. The first step was to conduct a naive reading of the data to determine distinct patterns or displayed com- monalities. The next step was to read these distinct patterns or displayed commonalities thoroughly and then search for condensations of meaning and differ- ences in condensation to describe and compare the data. In this analytical step, we analysed the distinct patterns that seemed to form main categories and sub-topics of a main category. The first and second authors individually evaluated the credibility of their understandings of the analytical categories, critically challenging them and searching for alternative patterns that could likewise be applied (Marshall & Rossman, 2006; Tjora, 2012). These researchers analysed data through a creative and inter- pretative process. Together, they then constructed the main categories, categories, and their sub-topics (Charmaz, 2000). The first and second author compared notes for analysing data to glean a more nuanced outline of the core descriptions and categories (initial coding). Then all three authors discussed and agreed on the codes and focused coding. This process was repeated and modified until saturation was reached and the cate- gories were validated (Corbin & Strauss, 2008; Mason, 2002; Patton, 2002). Figure 1 illustrates the coding and stepwise forming of categories, as described in more detail in the following section. Data collection We collected data between March and May 2015 from eight managers in Norway and ten managers in Sweden. The data-collection method was focused informant interviews (Denzin, 2001; Tjora, 2012). Table 1. Descriptions of sample participants’ criteria Country Norway Sweden Managers in total 8 10 Sex Male 4 6 Female 4 4 Age (years) <40 3 2 41–50 4 5 51–60 1 2 >61 0 1 Education High school 0 2 Vocational training school 2 2 Upper secondary school 1 1 University 5 5 Civil status Married/cohabiting 5 8 Single 3 2 Years in the enterprise <5 3 1 6–10 3 5 >11 2 4 Branches Building and construction/industy 1 3 Service delivery 7 7 Table 1. Descriptions of sample participants’ criteria Findings Managers experienced anxiety about not following regulations to ensure a safe working place and the rules in the Working Environment Act. These rules and requirements may seem overwhelming, given the reality of SSEs. The analysis showed that two main categories were relevant to the research questions: (1) restricted lee- way and (2) commitments. Below, we present each category in relation to its focus area in the main category. We also describe the categories and their sub-topics at varying lengths. This is, however, an expression of variation in characteristics and their links to the main category. The length of the descrip- tions should not be seen as a sign of difference or less significance. It is just that when you are a small business . . . all these rules about the work environment. . . . They can be a huge burden because they are the same rules as for a big company. In a way, there should be a work environment . . . a “light” version of the work envir- onment for the small company. Do not get me wrong, I do not mean that the employees should be any worse off. (IP 3) Downloaded by [Mittuniversitetet] at 02:18 30 August 2017 Figure 1. The steps of the analysis: coding and the emerging category exemplified. Figure 1. The steps of the analysis: coding and the emerging category exemplified. public regulations and rules. These regulations appear to create a “Gordian knot” for some managers. without giving any reason. We immediately anon- ymized identifying data in the transcriptions of the interviews. All of the data were properly stored accord- ing to the Swedish Act on Ethical Review of Research Involving Humans (SFS 2003:460). Then there are a lot of . . . how should I put it? Demands both from the municipality, even though we are not a municipal [organization], they also have their policies. [Customers] have opinions, employees . . . Yes. Politicians making cutbacks are also a signal. (IP 11) Then there are a lot of . . . how should I put it? Demands both from the municipality, even though we are not a municipal [organization], they also have their policies. [Customers] have opinions, employees . . . Yes. Politicians making cutbacks are also a signal. (IP 11) Downloaded by [Mittuniversitetet] Category 1: restricted leeway Managers want to ensure a safe working environment for their employees and develop a healthy workplace. However, they experienced that it was difficult to follow rules and the various legal frameworks that governed them. This main category addresses external factors that affected the managers’ leeway to model WHM. It refers to factors that managers considered important but out of their control. These factors structured the managers’ room to manoeuvre. The only thing I worry about . . . the work environ- ment requirements and that kind of thing . . . that you are in the grey area of the law some situations. Of course, I am not on top of all that. Then, you are going to be rapped over the knuckles eventually, when you have not met some requirement or you Ethics Sweden’s Regional Ethical Committee, Department of Medical Research approved the method design of the study (Dnr 2014-28-31M). The informants gave written consent to participate in the study. The informants were informed about their right to withdraw from the study SOCIETY, HEALTH & VULNERABILITY 57 Figure 1. The steps of the analysis: coding and the emerging category exemplified. economy Many people ask me, “Yes, but what, as the CEO – what do you do then?” “I buy coffee, run the swee- per, and fetch the laundry” . . . Yes, I am actually a – I am really more of a facilitator, trying to make sure that everything around this group works as well as possible. If I create the right situation for them, they will do a fantastic job. Like, I make sure that as little as possible gets in their way. (IP 10) Managers communicated how they need to man- oeuvre in a market situation in which seasonal work is more common and sector-specific fluctuations influence employment. These external factors inter- fere with managers’ freedom and flexibility to address WHM in a way they find valuable. Therefore, the advantages of a small company, when you own it . . . you can shape it . . . quite a lot yourself. You have this great power to make deci- sions . . . the opportunity . . . at the same time, you have the disadvantages of the small company in that you do not have the same . . . support functions in the company. (IP 1) Managers were constantly present and active in their efforts to develop the enterprise, keep it going, adapt to the market, and be efficient. In the current eco- nomic situation, although managers had a heavy work- load, they tried to create a psychosocial working environment that was flexible and adaptable to the needs of both managers and employees. Several man- agers demonstrated an almost entrepreneurial willing- ness to make their enterprise a flexible workplace – a place characterized by solidarity and tolerance. Managers experienced that they had insufficient sup- port functions related to workforce and production, for example human resources, administrative sup- port, and economy systems. Because of the size of their enterprises and sector-specific conditions, man- agers cannot afford to lose employees during certain periods significant to production. It was a challenge for such managers to reorganize workloads due to sickness or to obtain new qualified employees for reasons related to SSEs’ lower employee numbers. Managers relied on employees and their qualifica- tions to maintain a favourable market situation. Employees’ adaptation to multiple roles and their interdependent place in a team’s productivity ren- dered managers dependent on their work capacity. economy This has just strengthened my interest in carrying on, whatever I do for the rest of my life, that you see that you are creating a flexible environment in which you are open to changes . . . to be able to adapt to them, that is very important . . . it is a bit . . . what you read about, many Silicon Valley companies. There is a lot of flexibility. In that respect, that is what I am driven by, creating an effective and flex- ible environment. (IP 2) It is expensive for SSEs to have employees on sick leave. Therefore, several managers worried about how they could prevent illness and adapt the work so that sick leave did not occur. The diverse responsibilities for managers made them feel they were working in the interest of the entire enterprise when working proactively to prevent prolonged sick leave among employees. However, it is not easy in this line of business. You are, like, trying to get the person into production as soon as you can . . . Training and checking the qual- ity, that they know how to do the work . . . and . . . the workers have to manage the job. Workers also need to have a safety mind-set and do the right thing. Moreover, it obviously works in some way then, when no one has hurt himself or herself. (IP 4) I depend on people coming to me and letting me know . . . when people do not do that, we have experienced several times that people push them- selves too far, and then they are on full-time sick leave for a long period. Then, I think that if they had spoken up a bit earlier, perhaps they would have had part-time sick leave and perhaps for a shorter period as well, if they had spoken up earlier. (IP 11) Managers stated that the financial situation of the enterprise could affect how they addressed absence among employees and if they could hire contract workers. Then, we have a financial situation today that allows us to do that. It was not like that when I started as the manager – it was a hard slog for the first few years to get everything together. Now, we have an upward trend where we are still managing – and a bit more. Legal framework and regulations When managers searched for appropriate tools to prevent illness and form a health-promoting work- place, they needed simultaneously to comply with M. HEDLUND ET AL. 58 58 do not have some document that is supposed to be in place. (IP 4) do not have some document that is supposed to be in place. (IP 4) Managers do not have sufficient time to address safety and healthy activities in the workplace because they are busy in the “production line”. They were often the ones who were most qualified to make judgements about the market and what to do to achieve a favourable position compared to competi- tors. Managers also stepped into the core business in case of sickness or other types of absence. Managers do not have sufficient time to address safety and healthy activities in the workplace because they are busy in the “production line”. They were often the ones who were most qualified to make judgements about the market and what to do to achieve a favourable position compared to competi- tors. Managers also stepped into the core business in case of sickness or other types of absence. This subcategory shows how managers experienced limitations in relation to complying with working- environment legislation related to other regulations and the company’s regular business. Occupational safety and health issues Most managers have previously worked preventively by buying tailored health insurance. This health insurance would pay for prompt treatment when an employee became ill or had an accident. The managers stated that OH represented an unused opportunity for preventing illness and promoting health in the workplace. The managers were aware that this was unused potential. They had good experi- ences with this service when participating in the intervention program. As mentioned above, these managers participated in an intervention program intended to improve their skills and competence in health and work environment issues. Through this participation, managers gained access to and signed agreements with qualified persons from OH provi- ders. Managers received help and support to promote health in the workplace and obtained a different perspective on OH issues. I tried to tackle it the easy way. So we have this sickness . . . Wait . . . but heavens, what is it called now . . . I was sitting with all these insurance schemes for this “go to the front of the queue” insurance, as it is popularly called, allowing you to get treatment quickly . . . It is a kind of insurance, illness insur- ance. . . (IP 4) In Norway, some managers expressed satisfaction with their experience of being part of a public coop- eration agreement for a more inclusive working life (i.e. with being part of an IW enterprise). The main goals of such an agreement are to improve the work- ing environment, prevent and reduce sick leave, and prevent exclusion and withdrawal from working life. The contact with the OHS – I think that perhaps helps me to be a little clearer as a manager . . . The three-day courses, they have helped me a lot as a manager in relation to employees . . . I mean, the perspective . . . not that I had a bad perspective before, but it has influenced the way I see employees. How to work with employees in wise and sensible ways. (IP 11) It was the [name OHS] who began saying that we should have a meeting with the working life centre at NAV and go on to become an IA business [IW enter- prise]. There is not, at least as I saw it, very much to consider, because it does not cost anything. SOCIETY, HEALTH & VULNERABILITY SOCIETY, HEALTH & VULNERABILITY 59 5 59 facilitator and ensuring good working conditions for their employees, such as a trustful relationship to managers, and to provide adjusted work tasks to employees with reduced capacity, thus creating a health-promoting workplace. Some people who are ill . . . with sick notes and several challenges and we have some employees who have some health-related challenges that we must address. After we started this collaboration, we became an IA business [“IW” enterprise1]. Therefore, we now have a few more tools to help us through NAV and things like that. (IP 16) A small company always receives a failing grade when it is compared with a big company that has a HR department, a finance department, and various special functions. We are the same person with a hundred functions instead . . . we must satisfy the Occupational safety and health issues There are only benefits from being included in the system . . . After all, our aim must be to be a business where people can work until they are pensioners. It should be possible to create a basis for them to have a good workday and a secure future until they retire. (IP 16) In addition to new awareness and insights, managers experienced the benefit of using the OH for preventive purposes. Employees could obtain external expertise if needed. This project has been good . . . in the same way that I can buy health examinations, I can buy help from a psychologist . . . if one of my staff members is having a hard time with something, needs to talk, and I cannot take that on. I mean, one has cropped up . . . it is important to be preventive . . . to work more before a problem arises to create a better framework. (IP 1) Sometimes, managers found it difficult to prioritize OHS issues because so many tasks needed their atten- tion. Another difficulty was that they did not have time to obtain specialized knowledge in this field and sufficient administrative qualifications. . . .In a bigger company, there you usually have a Human Relations [HR] function where you can get support, you have a finance department . . . and you might also have executive colleagues if you are the CEO. Therefore, there are . . . more organized support functions that you can use in a larger company. (IP 1) Managers considered it important to have access to OH because in doing so, they gained partners to consult before problems could occur, a consulting body from which they could obtain advice and dis- cuss appropriate measures. Working with health and safety issues in the work- place was a job that fell to the managers, although they did not necessarily feel they were trained or qualified to do so. The result was a great deal of firefighting and spur-of-the-moment work when managers addressed these issues. . . .if one has a black day, I can phone one of [name staff members at a local OHS] . . . I can talk a bit about the project we are running as a team to obtain some new ideas and get motivated. That is quite fun. . . [laughter]. (IP 16) . . economy Therefore, there is the opportunity . . . then we can hire extra staff, and we do not have to wear ourselves out every day. (IP 6) In this subcategory, the managers described how they have limited resources to work with health and safety issues and sick leave. Managers often have to work in the production line to maintain market position and ensure adequate quality and financial solidity. The leaders are therefore interested in being a flexible In this subcategory, the managers described how they have limited resources to work with health and safety issues and sick leave. Managers often have to work in the production line to maintain market position and ensure adequate quality and financial solidity. The leaders are therefore interested in being a flexible Occupational safety and health issues .if one has a black day, I can phone one of [name staff members at a local OHS] . . . I can talk a bit about the project we are running as a team to obtain some new ideas and get motivated. That is quite fun. . . [laughter]. (IP 16) For managers, it was particularly important to get support from OH when employees had recurring health problems and needed partial sick leave. In these cases, managers needed advice for helping these employees could return to full-time work when possible. A small company always receives a failing grade when it is compared with a big company that has a HR department, a finance department, and various special functions. We are the same person with a hundred functions instead . . . we must satisfy the 60 M. HEDLUND ET AL. 60 . . .This is an issue that many small business owners have, that you are pretty much on your own on some issues . . . there is an advantage of having a board . . . in that forum; it is the advantage of bouncing off ideas and thoughts with somebody who knows the company well. Speaking for myself, it is important to create some structure there, where you can find support. (IP 1) tax authorities, we must satisfy the work environ- ment authority, and there is the chemicals agency, and there is . . . [laughs] the customer. The customer is the most important, after all, because if we did not have that, the company would not exist . . . no work- environment work would exist either. (IP 3) This subcategory shows how the managers experienced restricted leeway that made them “tightrope walking”, especially when dealing with health and safety issues at the workplace. This is partly due to managers having limited possibilities to gain knowledge of and qualifica- tions for working preventively with such issues and partly because they did not have sufficient knowledge about how to use the public social insurance systems and OH providers . Some board members were especially supportive and important to managers. The board chair appeared to play such a role, and the use of this person to discuss difficult personnel matters or difficult cases at the work- place helped managers focus. Category 2: commitments This category addresses managers’ external engagements that were not directly related to the enterprise but influ- enced the managers’ method of developing WHM. When managers developed WHM, personal engagement out- side the enterprise had an impact on how they prioritized and focused. Managers who owned their enterprise expressed the view that the entire workplace relied on their efforts. For this reason, managers were eager to invest in affordable health and safety equipment if it would benefit the entire enterprise. The impact of com- mitments outside the enterprise involved personal rela- tionships. The managers received advice and correction from family members and friends; this information was useful because it came from people who knew them and the business well and wanted them to succeed. In addition to continuity and familiarity with the board members, it was important for the managers that the board members did not change too often. Most have stayed; some have been replaced, but the backbone has always stayed. That is a bit reassuring because then you know that the board will not come in and turn everything upside down. (IP 6) This subcategory related to external boards and board members and how these affect both managers and matters of importance to WHM. The board, particu- larly the chair, gives managers important input and advice about what is essential or insignificant. Occupational safety and health issues The board chair appeared to be an important adviser because he or she was well acquainted with the enterprise, could give qualified advice, and was therefore a person to whom the man- agers easily turned. Continuity and a good understand- ing of the enterprise were important qualities of board members if managers were to use them for advice and guidance. This was particularly important to managers with respect to obtaining advice about how to handle conflicts and challenges in the working environment. Work-related networks Managers could participate in various types of busi- ness networks. These could be industry-specific net- works or networks targeted at female managers. It was important to develop work-related networks, particu- larly in relation to personnel management and other conditions that could affect the work environment. Friends or family had in-depth knowledge about the manager from their personal relationships. Some managers found that this helped them respond more reflectively to the issues they discussed. I do not want to sit and work every evening. Then I want to concentrate on my family and my personal life . . . I have some important support around me, my partner or friends . . . important supportive peo- ple in life that I can discuss things with and reflect on things together with. (IP 11) We have a networking group with 12 [managers]. It is fantastically useful. We meet about every sixth week . . . for a period of a few years, we have worked with professional development, professional focus and a professional boost. However, we can also bring up cases involving problems at our own places. For example, I need some guidance about follow-up with staff on sick leave and how do you do it . . . we can discuss that. (IP 11) Managers found it fruitful to have someone outside the enterprise to talk to about their everyday work and found that friends gave them new ideas and energy because they draw attention to the positive aspects of working in a stressful environment. Some managers participated in formal networks, whereas others used informal networks. In both cases, networks were used to discuss general matters regarding WHM, not individual cases or staff matters. Managers found they could usually benefit from net- works in the same industry, but that did not apply to advice related to the psychosocial working environ- ment. It was equally valuable to discuss these questions with networks and managers from other industries. Some managers participated in formal networks, whereas others used informal networks. In both cases, networks were used to discuss general matters regarding WHM, not individual cases or staff matters. It is wonderful to have [female] friends who know me well. They see me in a different way from how my staff sees me, so I have found my support there. Advice from boards Managers mentioned mentors as significant in help- ing them develop the enterprise and gain support in their daily business. Although mentors needed to be at a distance, it was also necessary for them to be sufficiently close to dilemmas and issues that might arise for a manager in this business. Boards were something all enterprises had. They played an important role in enabling managers to obtain advice on any issue. Some boards provided managers with a regular dose of corrections and a reality check, with a “kick in the pants” for managers to behave appropriately and within the board’s fra- mework and directions. Several managers found that the board provided important guidance and advice both in relation to how to develop the business and in how they should invest in human relations. I have found them through [name of Swedish business association] . . . it has been fantastic. Incredible support: I ask for it, and then I get help. Yes, things happen that I have to tackle in different ways, so that they . . . oh, it has been good. Here, you have to work everything out yourself. It is very important to know so much in all areas. Therefore, mentors are good. (IP 9) I have an active and resourceful board at the company . . . it should be a board with people who complement me. I talk to these people every day about negotiations in relation to contracts and tenders and such things. In addition, I use them as sparring partners . . . I also use them in relation to . . . staff and that kind of thing, if there are special challenges. (IP 16) Mentors may have an outsider’s perspective on man- agerial challenges. In this way, managers can obtain an overview and strive for a role model. . . .He was my greatest mentor until he died. There was no need to be in touch all the time, but when something cropped up, then he . . . always had time somehow. It is important to have someone you can confide in. (IP 8) Because their jobs were lonely and difficult, managers usually took the board’s advice. They needed some- one to team up with who was familiar with the enterprise and its type of business. Family and friends as buffers Managers noted how family, friends, and partners gave valuable advice and guidance for how to engage in the enterprise, how to commit in a man- ner that balanced their personal engagement with engagement in the enterprise, and how to correct their behaviour in a way that would promote better health and work–life balance for themselves and their employees. This subcategory linked to managers who sought advice and direction about WHM from external men- tors. It was important for the mentors to have perso- nal involvement in the challenges faced by the managers and to be able to provide qualified advice and comprehensive analysis, thus providing the man- agers with direction. Someone at home puts on the brakes. My husband might say that “Yes, it would be nice if you were at home some evenings too”. Nevertheless, every now and then, he comes here, and we help each other. (IP 9) Advice from boards SOCIETY, HEALTH & VULNERABILITY 6 61 61 It was important for mentors to have management experience, insight, and maturity. address in the different networks. In some cases, it was advantageous to receive advice from someone who was not a competitor or working in the same industry. . . .You have a contact network outside your own things, so you aren’t snowed under in your own . . . without having to have . . . well, mentors then. I have always had mentors, always had. Both when times have been really tough . . . then you really need hon- est mentors. I have had that. (IP 8) Discussion From a managerial perspective, this study explores which external factors influence the development of WHM and how these factors have an impact. From earlier research, we know that SSEs pay little attention to OH issues (e.g. Frick et al., 2000; Hasle & Limborg, 2006) and haveinsufficient competence to createhealth- promoting workplaces (Moser & Karlqvist, 2004; Torp & Moen, 2006). This study’s findings explain that work- ing environment regulations, market situation with sec- tor-specific fluctuations, insufficient time and limited resources to address health and safety activities, and insufficient knowledge about how to get support from occupational health services and social insurance system restrict SSE managers’ leeway to engage in WHM. Managers want the best for their enterprise and its employees and want to provide a healthy workplace, even if this aim can be difficult to fulfil. According to the European Agency for Safety and Health at Work (2013a), SSEs have a beneficial position in working life. This study confirms that managers partially agree with that statement. Nevertheless, as this study confirms, SSE managers can experience the demands of OHS regula- tions both as a financial burden and as too bureaucratic (Hasle & Limborg, 2006; Hasle et al., 2012, 2014). The managers studied here experienced difficulties in con- sidering all types of legal requirements and in imple- menting OHS regulations. This is partially attributable to their lack of flexibility and ability to migrate these requests and regulations into the reality and conditions of small enterprises. The managers lack both the time and appropriate methods to implement these regula- tions. Managers note that it is difficult to prioritize OHS regulations while focusing on important external factors such as market changes and sector-specific fluc- tuations in the requirements for the enterprise. This finding is in accordance with a study by Tappura, Syvänen, and Saarela (2014), who found that high eco- nomic pressure and a lack of resources were the most significant factors that affected managerial ability to design workplaces and promote employee health. The managers of the SSEs in this study, especially Swedish managers, referred to their restricted ability to use methods provided by the social security system to prevent sickness and ill health. According to Ahlberg et al. (2008), the absence of rehabilitation procedures is an important aspect of the difficulty of reducing sick leave. Work-related networks Because they may have seen that “but how are things?” or “now we must do something fun” . . . you can pour out everything on your mind. They are good. Everything feels much better afterwards [laughs]. (IP 9) g g Managers found they could usually benefit from net- works in the same industry, but that did not apply to advice related to the psychosocial working environ- ment. It was equally valuable to discuss these questions with networks and managers from other industries. Family ownership and close relationships with family members can influence managers’ behaviour and management practice. Although family members can relieve managers in their work tasks, this could have been problematic if managers had a strained relationship with family members who had previously led the enterprise. It helps you to develop, to meet colleagues, especially those who are in the same situation. In completely different lines of business – like now – it gives you quite a lot, I think. The problems, they are the same everywhere it seems. As far as the psychosocial part is concerned – how to manage staff and so on. (IP 8) This subcategory includes formal or informal external networks that affect managers in developing the work- place in a health-promoting direction. Commitment and dedication were characteristic of managers and those involved in the networks, and this made it easier to utilize the knowledge represented by these net- works. The networks did not have to be in the same industry for managers to benefit from them. The managers knew which questions were best suited to This subcategory was associated with external buf- fers that assist the managers in the development of WHM by guiding them to balance between their different skills and qualifications, and work–life bal- ance for themselves and their employees. The buffers came from private relationships, friends, or family who made managers aware of their own need to care for their family lives, not only the workplace’s M. HEDLUND ET AL. 62 needs. In that way, the managers received important correctives about how to organize their daily lives and focus on universal measures for WHM. so. This finding is in accordance with other studies underlining how managers must be supportive, hands-on, and inclusive to create a health-promoting workplace (Jiménez et al., 2016; Skarholt, Blix, Sandsund, & Andersen, 2015). Work-related networks In this study, OH services were an important exter- nal factor that the managers could utilize better to create a good work environment and a healthy workplace. Other research studies show great heterogeneity among SSE managers regarding the priority of work- environment issues (Hasle et al., 2012). In our study, managers explain why this can be the case: they lack appropriate regulations, tools, and resources to improve the work environment. In addition, they did not have sufficient information and knowledge about how to use the OH. SSEs in Norway and Sweden are affiliated with OH only to a slight degree (Gunnarsson, 2010; Moen et al., 2015). However, as shown in this study, that situation can easily change. Managers pointed to the experience they had gained from being exposed to a workplace health intervention led by OHS and indi- cated that this was a door opener for obtaining assis- tance in difficult WHM issues. Participation in a health intervention project provided them with new experi- ences and increased awareness and knowledge about how to benefit from OH services. Therefore, there is a potential for OH to develop adjusted services for SSEs. Research shows that efficient collaboration between enterprises and OH is dependent on a continuous dia- logue, where the services had to be to be flexible and adjusted to enterprises’ needs (Schmidt, Sjöström, & Antonsson, 2011). Conclusions and implications Our conclusions are related to the research questions about which external factors have an impact on WHM and how these impacts work. One conclusion is that rigid working-environment regulation and laws, a lack of tools and methods, and limited use of OH and the social security system are hindrances to WHM. Nevertheless, the managers in this study expressed high awareness and willingness to develop skills and knowledge about WHM. Another conclusion is that external commitments from the board, mentors, net- works, and family and friends are crucial and a resource for how managers develop a health-promot- ing workplace. Additionally, the societal support sys- tem for WHM does not seem to recognize SSEs’ special characteristics. Managers must therefore learn to walk as equilibrists, such as developing competence and skills in WHM without necessary support functions, having restricted leeway for prioritizing WHM and personal commitments they have outside the enter- prise. They must take into account the limitations in their room for manoeuvre while also taking advantage of the external resources that have been committed to help them and the enterprise. The managers in this study considered WHM to be an area in which they strived to balance indepen- dence, requests from individual employees, and the requirements of official regulations and the market situation. They emphasized the need to consider both individual-directed and organization-directed tools, which is in line with Jiménez et al.’s (2016) finding that managers influence the interaction of individual and organizational aspects. Therefore, it is difficult for SSE managers to cope with WHM. It is possible to reflect on this study’s findings in relation to theoretical aspects of organizational health. According to Lindström et al. (2000), organizational health implies that an organization can not only opti- mize its effectiveness and the well-being of its employ- ees but also cope effectively with both internal and external changes. In this study, the managers seem to realize the connection between employee well-being and organizational outcomes and the importance of employee health and working capacity. However, they point to several obstructing factors such as the lack of flexibility in working-environment regulations, market fluctuations, the firm’s financial situation, the lack of service from external resources, and managers’ demanding working conditions. Discussion However, an interview study among SSEs showed both that they were unsure of how to utilize the resources offered by the social security system and that they did not know what these public authorities could offer SSEs (The Swedish Social Insurance Inspectorate [Inspektionen för Socialförsäkring], 2012). Norwegian managers expressed a need to sign a formal agreement with the social security system so they could better benefit from it. Engaging in a long-term relationship with the social security system and working with that system to prevent or avoid long-term health issues was an overlooked aspect of managers’ efficacy. Some man- agers buy private health insurance for their employ- ees instead of relying on the public system. Managers explained that they lacked the compe- tence, but not the willingness to develop skills in WHM because of a work overload, the need to accom- modate other priorities, and a lack of administrative and management resources. However, they were eager both to learn more and to share knowledge with man- agers from other companies about psychosocial work- ing conditions. Managers expressed a need to pay more attention to developing a health-promoting workplace and were highly motivated but lacked the capacity to do The managers in this study emphasized commit- ments based on external factors, including the engagement of board members, mentors, business networks, and family and friends. These external SOCIETY, HEALTH & VULNERABILITY 6 63 factors and commitments support managers in WHM. Research indicates that networks can be a way of improving health and working environment in SSEs (Vinberg, 2006). Trust and close relations are important in these networks if they are to be long lasting (Antonsson et al., 2002; Street & Cameron, 2007). According to Limborg and Grøn (2014), it can be an advantage to have networks of enterprises in the same sector when seeking solutions to sector- specific problems in the working environment. In our study, managers valued the exchange of experi- ence with their counterparts from other sectors when they discussed issues relating to the psychosocial working environment. It is noteworthy that the man- agers in this study referred to friends and families who caused them to be reflective in their manage- ment and to balance their obligations to the enter- prise, employee issues, and their own personalities. This managerial reflectiveness is important to the development of WHM (Larsson & Vinberg, 2010). Discussion For example, managers reflect about leadership beha- viour and how they foster a positive culture in the enterprise. health-promoting workplace. However, another con- tribution of this study is that models for WHM and theoretical aspects of leadership theory must consider the special nature of the culture of SSEs and their managers’ special challenges. In many SSEs, the man- ager is also the owner, and his or her beliefs and cultural values provide the guidelines for developing the enterprise (Hasle & Limborg, 2006). Being an owner-manager often means a large amount of responsibility and a high workload that can lead to stress and ill health (Gunnarsson et al., 2007; Nordenmark et al., 2012). Therefore, models for WHM in SSEs also must focus on measures for improving the working conditions, lifestyle habits, and health of managers, not just those of employees. Note 1. An IW enterprise or an “IA business” is a Norwegian term for an enterprise that has entered into a colla- boration agreement with the NAV (The Norwegian Labour and Welfare Administration) for Inclusive Workplace Support. g Breslin, F. C., Kyle, N., Bigelow, P., Irvin, E., Morassaei, S., MacEachen, E., . . . Amick III, B. C. (2010). Effectiveness of health and safety in small enterprises: A systematic review of quantitative evaluations of interventions. Journal of Occupational Rehabilitation, 20, 163–179. doi:10.1007/s10926-009-9212-1 Bodil J Landstad http://orcid.org/0000-0001-6558-3129 Bodil J Landstad http://orcid.org/0000-0001-6558-3129 Bodil J Landstad http://orcid.org/0000-0001-6558-3129 References Abrahamsson, L. (2006). Småföretag – hjältestory, familjeaffär eller ren business. In H. Ylinenpää, B. Johansson, & J. Johansson (Eds.), Ledning i småföretag [Swedish] [Management in small enterprises]. Lund: Studentlitteratur. Acknowledgements The authors want to thank Linda Näsström and Bente Rømo Søreng for their assistance in data collection and/ or transcribing interviews. We owe special gratitude to the managers of SSEs in Norway and Sweden who generously shared their experiences with us. Breucker, G. (2001). Small, healthy and competitive. New strategies for improved health in small and medium-sized enterprises (Report on the Current Status of Workplace Health Promotion in Small and Medium-Sized Enterprises (SMEs)). Essen: Federal Association of Company Health Insurance Funds. Charmaz, K. (2000). Grounded theory, objectivist and con- structivist methods. In N. K. Denzin & Y. S. Lincoln (Eds.), The Sage handbook of qualitative research (pp. 509-535)). Thousand Oaks, CA: Sage. Disclosure statement No potential conflict of interest was reported by the authors. Chu, C., Breucker, G., Harris, N., Stitzel, A., Gan, X., Gu, X., & Dwyer, S. (2000). Health promoting workplaces – interna- tional settings development. Health Promotion International, 15(2), 155–167. doi:10.1093/heapro/15.2.155 Trustworthiness and limitation The findings should be interpreted with caution both because of the sample size in a single geographical context in Sweden and Norway and because the enterprises participated in a workplace health-promo- tion project. One limitation might be that the man- agers had positive attitudes about WHM because they were participating in a project. However, since we asked for managers’ previous experiences before they participated in the project, the intention was to capture their past experiences, not what they experi- enced because they were positive about participating in the workplace health-promotion project. Ahlberg, G., Bergman, P., Ekenvall, L., Parmsund, M., Stoetzer, U., Waldenström, M., & Svartengren, M. (2008). Hälsa och framtid. [Swedish] [Health and future]. Stockholm: Karolinska Institutet. Almqvist, R., & Henningsson, J. (2009). When capital mar- ket actors reduce the complexity of corporate personnel and work environment information. Journal of Human Resource Costing & Accounting, 13(1), 46–60. doi:10.1108/14013380910948072 Downloaded by [Mittuniversitetet] at 02:18 30 August 2017 Andersson, I.-M., & Hägg, G. M. (2006). Arbetsmiljöarbete i Sverige 2004. En kunskapssammanställning över strategier, metoder och arbetssätt för arbetsmiljöarbete [Swedish] (Vol. 6) [Working environment issues in Sweden 2004. A review of strategies, methods and approach to working environment issues]. Arbete och Hälsa: Arbetslivsinstitutet. Nevertheless, the aim of qualitative research is not to extend findings derived from selected samples to the world at large, but rather to transform and apply them to similar situations in other contexts (Polit & Beck, 2004). One strength of this study is its focus on SSEs in different sectors, along with the relatively extensive interviews. Downloaded by [Mittuniversitetet] at 02:18 Antonsson, A.-B., Birgersdotter, L., & Bornberger-Dankvardt, S. (2002). Small enterprises in Sweden. Health and safety and the significance of intermediaries in preventive health and safety. Arbete och Hälsa. (1), (Work and Health 2002:1). Stockholm: National Institute for Working Life. g Bornberger-Dankvardt, S., Ohlson, C.-G., & Westerholm, P. (2003). Arbetsmiljö- och hälsoarbete i småföretag – försök till helhetsbild. [Swedish] [Working environment and healthy work in small enterprises – an attempt to an overall picture]. Stockholm: Arbetslivsinstitutet (Arbetsliv i omvandling 2003:1). Conclusions and implications One interesting result is that the managers are more likely to obtain external support from the board, mentors, networks, family, and friends than from, for example, OH, which was created to assist enterprises with their health and safety issues at the workplace. In future models for WHM in SSEs, it will be important to find ways to cope with the obstructing factors found in this study and to consider SSE managers’ external support systems so that they can be assisted in developing both to WHM and a This study has several implications. One implica- tion is that there is a need to develop models and tools for implementing occupational and health reg- ulations that are adapted to SSEs. The second impli- cation is that it is important for SSE managers to obtain more knowledge about WHM and workplace health issues. This can be accomplished by develop- ing local networks dedicated to these issues; research indicates that such networks can be successful if there is trust among the network members (Antonsson et al., 2002). Given that there is insufficient coopera- tion between SSEs and OH, a third implication is that OH consultants can coordinate the networks dis- cussed in this study and support managers in WHM and workplace health processes. It is also important that managers receive support for improving their M. HEDLUND ET AL. 64 64 0615), and the Swedish Work Environment Authority (Working Life Research 2016-2018), Sweden, for financing this study. 0615), and the Swedish Work Environment Authority (Working Life Research 2016-2018), Sweden, for financing this study. own working conditions and work–life balance. The final implication is that future research should focus more on tools for WHM and the significance of external factors for SSEs using both qualitative and quantitative methods. In particular, there is a need to determine more about how external factors such as OH, the Social Insurance Agency, boards, and men- tors can support SSE managers in WHM and work- place health processes. 1. An IW enterprise or an “IA business” is a Norwegian term for an enterprise that has entered into a colla- boration agreement with the NAV (The Norwegian Labour and Welfare Administration) for Inclusive Workplace Support. Funding Creating a healthy working environment with leadership: The con- cept of health-promoting leadership. The International Journal of Human Resource Management, 1–19. doi:10.1080/09585192.2015.1137609 g EU. (2003). OJL (Official Journal of the European Union). L124/37-41. Retrieved from http://europa.eu.int/eur-lex/ en/oj/ Landstad, B. J., Hedlund, M., & Vinberg, S. (2017). How managers of small-scale enterprises can create a health promoting corporate culture. International Journal of Workplace Health Management, 10(3), 228–248. doi:10.1108/IJWHM-07-2016-0047 European Agency for Safety and Health at Work. (2013a). Promoting health and safety in European Small and Medium-sized Enterprises. Luxembourg: Office for Official Publications of the European Communities. Larsson, J., & Vinberg, S. (2010). Leadership behaviour in successful organisations: Universal or situation-depen- dent? Total Quality Management & Business Excellence, 21, 317–334. doi:10.1080/14783360903561779 European Agency for Safety and Health at Work. (2013b). Safety and health in micro and small enterprises. Retrieved from https://osha.europa.eu/en/themes/safety- and-health-micro-and-small-enterprises. Limborg, H. J., & Grøn, S. (2014). Networks as a policy instru- ment for smaller companies. Nordic Journal of Working Life Studies, 4(3), 53–55. doi:10.19154/njwls.v4i3.4179 European Network for Workplace Health Promotion. (2001, June). The Lisbon statement on workplace health in small and medium-sized enterprises (SMEs). Adopted in Lisbon, Portugal. j Lindström, K., Schrey, K., Ahonen, G., & Kaleva, S. (2000). The effects of promoting Organizational health on worker well-being and organizational effectiveness in small and medium-sized enterprises. In L. R. Murphy & C. L. Cooper (Eds.), Healthy and Productive work – an international Perspective. London: Taylor & Francis. g Frick, K., Langaa Jensen, P., Quinlan, M., & Wilthagen, T. (Eds.). (2000). Systematic occupational health and safety management. Oxford: Elsevier Science Ltd. g Frick, K., & Zwetsloot, G. I. J. M. (2007). From safety management to corporate citizenship: An overview of approaches to managing health. In U. Johansson, G. Ahonen, & R. Roslender (Eds.), Work health and manage- ment control (pp. 9–134). Stockholm: Thomson Fakta. MacEachen, E., Kosny, A., Scott-Dixon, K., Facey, M., Chambers, L., Breslin, C., . . . Mahood, Q. (2010). Workplace health understandings and processes in small businesses: A systematic review of the qualitative literature. J Occup Rehabil, 20, 180–198. doi:10.1007/ s10926-009-9227-7 Gjerstad, P., & Lysberg, F. (2012). Leadership and health promotion workplaces. In S. T. Innstrand (Ed.), Health promotion – theory and practice, research centre for health promotion and resources. Trondheim: HiST/NTNU. Marshall, C., & Rossman, G. B. (2006). Designing Qualitative Research. 4th edition. London: Sage. Grawitch, M. J., Gottschalk, M., & Munz, D. C. (2006). Funding We wish to express our gratitude to AFA Försäkring Insurance (Dnr: 130190), The Swedish Council for Health, Working Life and Welfare (Grant number 2012- Corbin, J. M., & Strauss, A. L. (2008). Basics of qualitative research: Techniques and procedures for developing grounded theory. Thousand Oaks, CA: Sage. SOCIETY, HEALTH & VULNERABILITY 65 Cunningham, T. R., Sinclair, R., & Schulte, P. (2014). Better understanding the small business construct to advance research on delivering workplace health and safety. Small Enterprise Research, 21(2), 148–160. doi:10.1080/ 13215906.2014.11082084 Gunnarsson, K., Andersson, I.-M., & Josephson, M. (2011). Swedish entrepreneurs’ use of occupational health ser- vices. Workplace Health & Safety, 59, 437–445. p f y Gunnarsson, K., Vingård, E., & Josephson, M. (2007). Self rated health and working conditions of small-scale enterprisers in Sweden. Industrial Health, 45, 775–780. Dellve, L., Skagert, K., & Vilhelmsson, R. (2007). Leadership in workplace health promotion projects: 1- and 2-year effects on long-term work attendance. European Journal of Public Health, 17(5), 471–476. doi:10.1093/eurpub/ckm004 Hasle, P., & Limborg, H. J. (2006). A review of the litera- ture on preventive occupational health and safety activ- ities in small enterprises. Industrial Health, 44, 6–12. Hasle, P., Limborg, H. J., Kallehave, T., Klitgaard, C., & Rye Andersen, T. (2012). The working environment in small firms: Responses from owner-managers. International Small Business Journal, 30(6), 622–639. doi:10.1177/ 0266242610391323 p Denzin, N. K. (2001). The reflexive interview and a perfor- mative social science. Qualitative Research, 1(1), 23–46. doi:10.1177/146879410100100102 Department of Health. (2004). Choosing health: Making healthier choices easier (The Wanless Report, 16). London: TSO. Hasle, P., Limborg, H. J., & Nielsen, K. T. (2014). Working environment interventions ? Bridging the gap between policy instruments and practice. Safety Science, 68, 73– 80. doi:10.1016/j.ssci.2014.02.014 Eakin, J. M., Lamm, F., & Limborg, H. J. (2000). International perspective on the promotion of health and safety in small workplaces. In K. Frick, P. L. Jensen, M. Quinlan, & T. Wilhagen (Eds.), Systematic occupational health and safety management (pp. 227– 247). Oxford: Elsevier. j Holt, M., & Powell, S. (2015). Health and well-being in small and medium-sized enterprises (SMEs). What public health support do SMEs really need? Perspectives in Public Health, 135(1), 49–55. doi:10.1177/1757913914521157 Eriksson, A. (2011). Health-Promoting Leadership: A study of the concept and critical conditions for implementation and evaluation (Thesis). Nordic School of Public Health, Gothenburg. Jiménez, P., Winkler, B., & Dunkl, A. (2016). Funding The path to a healthy workplace: A critical review linking healthy workplace practices, employee well-being, and organizational improvements. Consulting Psychology Journal: Practice and Research, 58(3), 129–147. doi:10.1037/1065-9293.58.3.129 g Mason, J. (2002). Qualitative researching. London: Sage. Meggeneder, O. (2007). Style of management and the rele- vance for workplace health promotion in small and medium sized enterprises. Journal of Public Health, 15, 101–107. doi:10.1007/s10389-006-0088-7 Griffin, B. L., Hall, N., & Watson, N. (2005). Health at work in small and medium sized enterprises. Issues of engagement. Health Education, 105(2), 126–141. doi:10.1108/09654280510584571 Moen, B. E., Hanoa, R. O., Lie, A., & Larsen, Ø. (2015). Duties performed by occupational physicians in Norway. Occupational Medicine (Oxford, England), 65, 139–142. doi:10.1093/occmed/kqu184 Gunnarsson, K. (2010). Entrepreneurs and small-scale enterprises. Self-reported health, work conditions, work environment management and occupational health ser- vices (Thesis 574). Uppsala: Uppsala University. q Moser, M., & Karlqvist, L. (2004). Small and medium sized enterprises. A literature review of workplace health pro- motion (Arbetslivsrapport 17 [Working life report]). Stockholm: National Institute for Working Life. 66 M. HEDLUND ET AL. 66 Street, C., & Cameron, A.-F. (2007). External relationships and the small business: A review of small business alli- ance and network research. Journal of Small Business Management, 45(2), 239–266. doi:10.1111/jsbm.2007.45. issue-2 Nordenmark, M., Vinberg, S., & Strandh, M. (2012). Job control and demands, work life balance and wellbeing among self-employed men and women in Europe. Vulnerable Groups & Inclusion, 3, doi:10.3402/vgi. v3i0.18896. Sverke, M. (2009). The importance of the psychosocial work environment for employee well-being and work motivation. Scandinavian Journal of Work, Environment & Health, 35, 241–243. doi:10.5271/sjweh.1336 The Norwegian Working Environment Act. (2015). Lov om arbeidsmiljø, arbeidstid og stillingsvern mv. [Norwegian] [Act of working environment, working hours and job protection etc.], Oslo: Norway]. (Arbeidsmiljøloven), LOV-2015-12-18-104. The Swedish Social Insurance Inspectorate [Inspektionen för Socialförsäkring]. (2012). Arbetsgivare i små företag. En intervjustudie om deras erfarenheter av sjukskrivning- sprocessen. [Swedish] [Managers in small enterprises. An interview study of their experiences of the sick leave process]. Stockholm: Inspektionen för Socialförsäkring. Patton, M. Q. (2002). Qualitative research & evaluation methods (3rd ed.). Thousand Oak, CA: Sage Publication. Polit, D. F., & Beck, C. T. (2004). Nursing research : principles and methods. 7th edition. Philadelphia: Lippincott Williams & Wilkins. Rootman, M., Goodstadt, B., Hyndman, D., McQueen, L., Potvin, J., & Springett, E. Z. (2001). Evaluation in health promotion: Principles and perspectives. Copenhagen: WHO Regional Publications. European Series, No 92. Funding The Swedish Working Environment Act. (2014). Arbetsmiljölag [Working environment law], SFS nr [Swedish] (pp. 659). Stockholm: The Swedish Work Environment Authority. g p Schmidt, L., Sjöström, J., & Antonsson, A.-B. (2011). Vägar till framgångsrikt samarbete med företagshäslovård [Swedish] [Break fresh ground to successful cooperation with occupational health services]. Stockholm: IVL Svenska Miljöinstitutet AB. y Tappura, S., Syvänen, S., & Saarela, K.-L. (2014). Challenges and needs for support in managing occupa- tional health and safety from managers’ viewpoints. Nordic Journal of Working Life Studies, 4(3), 31–51. doi:10.19154/njwls.v4i3.4178 j SFS 2003:460. (2005). Lag om etikprövning av forskning som avser människor. [Swedish] [Swedish Law in Force]. Retrieved from http://www.riksdagen.se/sv/Dokument- Lagar/Lagar/Svenskforfattningssamling/Lag-2003460- om_etikprovning_sfs-2003-460/. Tjora, A. (2012). Kvalitative forskningsmetoder i praksis [Norwegian] [Practising Qualitative methods] (2nd ed.). Oslo: Gyldendal Akademiske. Torp, S., Eklund, L., & Thorpenberg, S. (2011). Research on workplace health promotion in the Nordic countries: A literature review, 1986–2008. Global Health Promotion, 18(3), 15–22. doi:10.1177/1757975911412401 p g Skarholt, K., Blix, E. H., Sandsund, M., & Andersen, T. K. (2015). Health promoting leadership practices in four Norwegian industries. Health Promotion International, 31(4), 936–945. Torp, S., & Moen, B. E. (2006). The effects of occupational health and safety management on work environment and health: A prospective study. Applied Ergonomics, 37(6), 775–783. doi:10.1016/j.apergo.2005.11.005 Sparling, P. B. (2010). Worksite health promotion: Principles, resources, and challenges. Preventing Chronic Disease, 7(1). Retrieved from http://www.cdc. gov/pcd/issues/2010/jan/09_0048.htm j p g Vinberg, S. (2006). Health and performance in small enter- prises: Studies of organizational determinants and change strategy (Thesis). Luleå: Luleå tekniska universitet. Statistics Norway. (2014). Antall virksomheter etter størrelse [Norwegian] [Number of establishments by size]. Retrieved from https://www.ssb.no/virksomheter-foretak-og- regnskap/statistikker/bedrifter/aar/. gy WHO (World Health Organization). (1986). Ottawa charter for health promotion. Geneva: World Health Organization. p g Whysall, Z., Haslam, C., & Haslam, R. (2006). Implementing health and safety interventions in the workplace: An exploratory study. International Journal of Industrial Ergonomics, 36(9), 809–818. doi:10.1016/j. ergon.2006.06.007 Statistics Sweden. (2011). Aktuell statistik ur företagsregistret [Swedish] [Current statistics from the establishment register]. Retrieved from http://www.scb. se/sv_/Vara-tjanster/Foretagsregistret/Aktuell-statistik- ur-Foretagsregistret/. g Witt, L. B., Olsen, D., & Ablah, E. (2013). Motivating factors for small and midsized businesses to implement worksite health promotion. Health Promotion Practice, 14(6), 876–884. doi:10.1177/1524839912472504 g g Stephan, U., & Roesler, U. (2010). Health of entrepreneurs versus employees in a national representative sample. Journal of Occupational and Organizational Psychology, 83, 717–738. doi:10.1348/096317909X472067
https://openalex.org/W2024616134	https://europepmc.org/articles/pmc3792528?pdf=render	English	null	Protocol for Evaluation of Robotic Technology in Orthopedic Surgery	Advances in orthopedics	2,013	cc-by	4,190	Milad Masjedi, Zahra Jaffry, Simon Harris, and Justin Cobb MSk Lab, Charing Cross Hospital, Imperial College London, London W6 8RF, UK Correspondence should be addressed to Milad Masjedi; m.masjedi@imperial.ac.uk Received 25 October 2012; Accepted 25 August 2013 Correspondence should be addressed to Milad Masjedi; m.masjedi@imperial.ac.uk Correspondence should be addressed to Milad Masjedi; m.masjedi@imperial.ac.uk Received 25 October 2012; Accepted 25 August 2013 Academic Editor: Jess H. Lonner Copyright © 2013 Milad Masjedi et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. In recent years, robots have become commonplace in surgical procedures due to their high accuracy and repeatability. The Acrobot Sculptor is an example of such a robot that can assist with unicompartmental knee replacement. In this study, we aim to evaluate the accuracy of the robot (software and hardware) in a clinical setting. We looked at (1) segmentation by comparing the segmented data from Sculptor software to other commercial software, (2) registration by checking the inter- and intraobserver repeatability of selecting set points, and finally (3) sculpting (𝑛= 9 cases) by evaluating the achieved implant position and orientation relative to that planned. The results from segmentation and registration were found to be accurate. The highest error was observed in flexion extension orientation of femoral implant (0.4±3.7∘). Mean compound rotational and translational errors for both components were 2.1±0.6 mm and 3±0.8∘for tibia and 2.4±1.2 mm and 4.3±1.4∘for the femur. The results from all processes used in Acrobot were small. Validation of robot in clinical settings is highly vital to ensure a good outcome for patients. It is therefore recommended to follow the protocol used here on other available similar products. Hindawi Publishing Corporation Advances in Orthopedics Volume 2013, Article ID 194683, 5 pages http://dx.doi.org/10.1155/2013/194683 Hindawi Publishing Corporation Advances in Orthopedics Volume 2013, Article ID 194683, 5 pages http://dx.doi.org/10.1155/2013/194683 Hindawi Publishing Corporation Advances in Orthopedics Volume 2013, Article ID 194683, 5 pages http://dx.doi.org/10.1155/2013/194683 2. Materials and Methods The root mean squared (RMS) error of the registration process was recorded for each observer. In order to check the effect of different tracking and sculpting arm positions, the same procedure was repeated by changing the fixation of the femur. The positions mimicked those found in surgical operations for various patients’ size or surgeons’ preference.h g The last step is to measure the accuracy of constraints set at the planning stage during bone resection. A senior surgeon (JPC) was recruited to plan the operation using the Planner Software. Uniglide implants (Corin, Cirencester, UK) were chosen (a size four tibial component and size three femoral component) to restore the natural joint line, incorporating a seven-degree posterior slope in the tibial component. Nine UKAs were implanted on identical dry bone knee models (Imperial knee, Medical Models Company, Bristol, UK) by three experienced users of the Sculptor (three each). The models used were CT-based replicas of a patient’s arthritic knee consisting of a capsule, replica ligaments, and muscle tissues. Following implantation, the knee joint was separated from femur and tibia and each bone was individually scanned using the NextEngine Desktop 3D scanner (NextEngine, Santa Monica, CA, USA). Prior to implantation, the implant was painted in white enamel paint to improve pick-up of the laser spot from the scanner on the metal surface.h Other than the validity of the software, potential sources of error which can influence the outcome of the surgery include (1) the inaccuracy in position of sculpting arm or tracking arm (poor calibration), (2) inaccuracy in the registration algorithms to match the CT data to the bone, and finally (3) the robotic control system that constrains the surgeon to resect only on the safe zone area. Additionally, there may be other errors arising from surgeons in charge such as poor fixture of bones to tracking arm or inaccuracies in the use of tools [16]. The accuracy of registration, specifically, is an aspect that remains to be determined. There are a number of methods through which registration can take place, such as use of X- ray or ultrasound [17]. Some systems use fiducial markers in order to register the bone and some use landmarks on the bone such as discrete identifiable points or the ridge line [17]. Each is subject to a certain type of error including fiducial localization and registration error and target registration error [18]. 2. Materials and Methods as computed tomography (CT) to identify the pathology and plan the surgery virtually. During the surgery, a registration process takes place that matches the preoperative plan and imaging data to the patient [15]. This means the transfor- mation between the virtual environment and the patient is known and any points in the plan can be located during the surgery. Results can be affected by both the software and hardware used by the robot [16]. In order to evaluate the accuracy of the Acrobot Sculptor, the following steps were taken.h The initial step was determining the accuracy of the segmentation procedure. We compared the segmentation result of Modeller (Stanmore Implants, London, UK) from a single femur using various software used to convert CD data to 3D models. These are Mimics (Materialise, Leuven, Bel- gium) and Robin 3D (Cavendish Medical, London UK) [14]. These surfaces were then matched together using 3-matic (Materialise, Leuven, Belgium) software and the differences in size were analysed.h The patient’s CT scan is often segmented using available commercial software (e.g., Acrobot Modeller for Acrobot Sculptor). This software can generate the surface structure of the specified bones. It is possible to use these three- dimensional (3D) images to diagnose the pathology even though the surface geometry is not accurate or in scale. How- ever, in robotic procedures, the accuracy of these surfaces has a direct influence on the outcome of surgery. This surface model is then loaded onto Acrobot Planner software to carry out preoperative planning. The Sculptor has a cutting burr attached to its three degrees of freedom (DoF) arm which can sculpt the bone based on a predefined plan. A tracking arm is pinned to the bone so that the system is aware of the 3D position of that bone relative to the robot at all times. Following attachment of the bone to the tracking arm, the intraoperative procedure also requires registration of points on the bone surfaces [15]. The second step in determining the reliability and repro- ducibility of the Sculptor was by placing a set of points (using a marker pen) on the dry bone femur that was CT scanned. A total of 45 points were selected randomly on the distal part of the femur, focussed on for medial UKA procedures and four observers used the Sculptor to register these points. 1. Introduction the patient’s morphology. The main problem with the latter is that these models are often created based on a normal anatomy dataset, and using them for pathological subjects can be problematic [12]. Audenaert et al. described the estimated accuracy of imageless surgery as poor because of the significant difference between the actual location of the probe during surgery and what is displayed on the navigation platform screen [13].h In recent years, robots have become commonplace in indus- try due to their high accuracy and repeatability especially during procedures that require movement that is beyond the human control [1, 2]. As imaging and robotic technology has advanced, there is real potential to use these capabilities in the field of surgery, from planning to performing the procedure. This is especially useful in operations such as unicompartmental knee arthroplasty (UKA) where previous studies have shown the substantial effect of implant position inaccuracy [3–5].h The Acrobot Sculptor (Stanmore Implants Worldwide Ltd.) is a semiactive robot, uses the computer tomography (CT) data as input, and could assist with bone resection for UKA surgery in a consistent manner to minimise variability [14]; however, the repeatability and accuracy of this robot in clinical settings are yet to be determined. In this study, we have set up various steps to determine the accuracy and repeatability of the Stanmore Sculptor which we believe will also be applicable to a wide range of other available similar products. The robotic procedures can be fully controlled (active) [6], can be shared as control or semiactive, where the robot monitors surgeon performance and provides stability and support through active constraint [7], they can be tele- surgical where the surgeon performs the operation from a console distant to operating table [8]. The input to the robot can vary from the actual imaging data of the patient to statistical shape models (SSM) [9] or active shape models (ASM) [10, 11] that are based on a few point estimates of There are a number of processes involved in the use of the Sculptor, as with most robotic systems, with potential for error. Surgeons often use the imaging technology such 2 2 Advances in Orthopedics 2. Materials and Methods The Acrobot Sculptor uses a mechanical digitizer (a secondary use of the robotic arm) to register the surface, where the tip of the cutter (ball point) is used as a probe which has a 2 mm diameter. As a result of inaccuracy in calibration or radius of the ball point, the captured data can be displaced from the true surface. These scans were exported as Stereolithography (STL) files to 3-matic software. The positions of the tibial and femoral components were then compared to those of the ideal plan by recording the coordinates of four points on the planned implants versus the achieved implants (Figure 1). Using MATLAB, a local frame of reference was created using these four points for the achieved implant and was compared to that of the planned in all six DoF. These coordinates were created so that they follow the anatomical frame of reference such that the 𝑥-, 𝑦- and 𝑧-axes correspond to mediolateral, anteroposterior, and superoinferior directions accordingly. The magnitude of translational (a combination of the medial- lateral, anterior-posterior, and superior-inferior directions errors) and rotational (a combination of the axial, flexion- extension, and coronal alignment errors) errors were calcu- lated for each case for both tibial and femoral components. Validation of robot is vital to ensure a good outcome and highlight their value in use with patients [19]. Although there are several technical papers that have talked in detail about the accuracy of registration algorithms and robotic manipulations [20], there is no real simple method to test the accuracy of the robot in a clinical environment, and the main reference point simply remains the manufacturer’s information. In this study, we aim to evaluate the above possible cause of errors in a clinical setting. 3 Advances in Orthopedics 3 Advances in Orthopedics 3 (a) (b) (c) (d) Figure 1: Four points selected on the (a) femoral and (b) tibial implants to construct the local frame of reference. Comparison of the planned versus achieved rotational and translational errors based on the local frame of reference for (c) femoral and (d) tibial implants. (a) (b) (a) (b) (b) ( ) (c) (d) (d) (c) Figure 1: Four points selected on the (a) femoral and (b) tibial implants to construct the local frame of reference. Comparison of the planned versus achieved rotational and translational errors based on the local frame of reference for (c) femoral and (d) tibial implants. 2. Materials and Methods Table 1: Translational and rotational error values in UKA implant placement (𝑛= 9). Tibia Femoral Translational error (mm) Rotational error (∘) Translational error (mm) Rotational error (∘) Lateral medial Anterior posterior Distal proximal Flexion extention Varus valgus Axial rotation Lateral medial Anterior posterior Distal proximal Flexion extention Varus valgus Axial rotation Mean 0.0 −0.9 0.8 2.1 −0.8 0.4 0.6 −1.5 −1.1 0.4 −0.5 0.9 SD 1.5 0.5 1.1 0.6 1.6 1.4 0.9 1.4 1.0 3.7 0.9 2.6 Max 1.8 −0.2 2.1 2.2 2.2 3.1 1.9 0.4 −0.1 5.3 0.5 5.3 Min −2.8 −1.6 −1.7 −3.3 −1.3 −2.4 −1.2 −3.2 −3.5 −4.7 −1.9 −2.1 Table 1: Translational and rotational error values in UKA implant placement (𝑛= 9). the maximum error among all subjects was found to be 1.8 mm with mean maximum being 1.53 ± 0.2 mm.h the maximum error among all subjects was found to be 1.8 mm with mean maximum being 1.53 ± 0.2 mm. The results for implantation are shown in Table 1. Place- ment of the femoral component in general was more prone to error with a maximum error of 5.3∘around the 𝑥- and 𝑧-axes. Mean compound rotational and translational errors for tibia component were 2.1 ± 0.6 mm and 3 ± 0.8∘and for femoral component were 2.4 ± 1.2 mm and 4.3 ± 1.4∘ 4. Discussion We acknowledge that during the planning phase, land- marks used to define reference frames are located manually by the surgeon. Srivastava et al. [24] describe the effects of landmark placement variability on kinematic descriptions of the knee. The positions of these landmarks may be open to placement inaccuracy and variability between surgeons. In addition to the accuracy measurements described above, a sensitivity analysis should be performed to determine the likely variability in frame of reference orientations and implant position relative to these introduced by the human operator during planning. In this study, we set protocols that make it possible to evaluate robot’s accuracy in house, which include testing both software and hardware and are applicable to a variety of similar products on the market. Robotics technology can improve surgical outcomes by providing the surgeon with the greatest amount of accuracy and precision regardless of long surgical training [21]. The robot gives the surgeon more control in terms of the position and alignment of the tools. The accompanied software can also assist in the planning of the surgery and also during the operation by supplying information on direction and amount of cut by enabling the surgeon to visualise these on the screen. These robots are prone to errors both systematic and those due to the operator. Validating the accuracy of image guided surgery is therefore an important issue that needs to be addressed.h In this study, we set protocols that make it possible to evaluate robot’s accuracy in house, which include testing both software and hardware and are applicable to a variety of similar products on the market. Robotics technology can improve surgical outcomes by providing the surgeon with the greatest amount of accuracy and precision regardless of long surgical training [21]. The robot gives the surgeon more control in terms of the position and alignment of the tools.ht The accompanied software can also assist in the planning of the surgery and also during the operation by supplying information on direction and amount of cut by enabling the surgeon to visualise these on the screen. These robots are prone to errors both systematic and those due to the operator. Validating the accuracy of image guided surgery is therefore an important issue that needs to be addressed.h Often in the literature, errors are based on the transla- tional or angular location of the implant and cuts; however, Simon et al. 4. Discussion [17] argued that there are ambiguities associated with these data due to a dependence upon the selected coordinate system. It is therefore anticipated in the future for the implant manufacturer to provide a standard protocol for evaluation of location of the implant. The use of dry bones meant that soft tissue balancing could not be recreated and the tibio-femoral angle could not be measured. Although this is an important measure of functional outcome following a UKA, it is widely accepted that component alignment is a major influence on the limb’s tibiofemoral angle [24]. In this study, we used a laser scanner instead of CT to find the position of the implant postoperatively since a metallic implant will create artefact in the CT scan and inaccuracy in segmentation. There could also be inaccuracies during segmentation and in CT data itself; however, this is not part of the system and would be operator error, not that of the software. The initial step was determining the accuracy of the segmentation procedure. There are numerous techniques available to create the bone surface from CT images. In lack of any available straight forward method, in this study we used comparative validation and found almost identical data using different software. For the second part we evaluated the landmarks that create the frame of reference. In this study, we found that placement of landmarks can be inaccurate. It is therefore important for surgeons to rely on their own experience for planning the procedure as well as the suggested values given by software for the implant position. t In this study, rotational error was found to be the highest source of error in both femoral and tibial components. This we believe is actually not only depends on the cut made by robot but also, when the implant is hammered into the plastic bone. This is on especially important with use of plastic bone, as it is possible to deform this under higher loads which may increase the error seen here. Nevertheless, these errors were accepted, far superior to what has been reported for conventional surgeries [14, 22], and similar to other robots available. For example, Dunbar et al. [22] found that 3. Results The results from segmentation using different software were almost identical. The measurements were performed in 3- matic software, and the difference was far less than 0.5 mm at all points. The results for implantation are shown in Table 1. Place- ment of the femoral component in general was more prone to error with a maximum error of 5.3∘around the 𝑥- and 𝑧-axes. Mean compound rotational and translational errors for tibia component were 2.1 ± 0.6 mm and 3 ± 0.8∘and for femoral component were 2.4 ± 1.2 mm and 4.3 ± 1.4∘ The results for the registration repeatability (second step) showed a consistent mean RMS error were of 0.5 mm while 4 Advances in Orthopedics 0 1 2 3 4 5 6 Tibial translational error (mm) Femoral translational error (mm) Tibial rotational error (∘) Femoral rotational error (∘) Figure 2: Magnitude of resultant rotational and translation error for tibial and femoral components when compared to planned posi- tions. the MAKO robot’s mean RMS errors for the tibia were 1.4 mm and 2.6∘and for the femur 1.2 mm and 2.1∘. Furthermore, the ranges found are within the safe range reported by Biomet [23]. We recognise the inherent limitations of our study, one of which is the use of dry replica bones rather than patients. To compensate, the dry bones were replicas of a patient’s arthritic tibia and femur, with replica ligaments as well as a surrounding capsule attached and hence were as realistic to a real patient as possible. Fixation of the tracking arm to the bone may also cause inaccuracy if fixation pins bend or loosen due to stress on the fixation point. It is therefore important to design fixtures that are robust and rigid and not loosened (i.e., no movement between the tracking device and the anatomy should be allowed). Furthermore, if after screwing the fixtures, the anatomy of the subject deforms due to overloading of the segment, the possible error as a result of this needs to be evaluated for each robot. In the Sculptor, the tracking arm is quite light, and therefore we assumed stress on the fixation point because the weight of the arm would be minimal. Figure 2: Magnitude of resultant rotational and translation error for tibial and femoral components when compared to planned posi- tions. (Figure 2). The highest error was found in rotational elements for both components. (Figure 2). The highest error was found in rotational elements for both components. References [1] A. D. Pearle, D. Kendoff, and V. Musahl, “Perspectives on computer-assisted orthopaedic surgery: movement toward quantitative orthopaedic surgery,” The Journal of Bone & Joint Surgery A, vol. 91, no. supplement 1, pp. 7–12, 2009. [18] J. M. Fitzpatrick, “The role of registration in accurate surgical guidance,” Journal of Engineering in Medicine H, vol. 224, no. 5, pp. 607–622, 2010. [2] W. L. Bargar, “Robots in orthopaedic surgery: past, present, and future,” Clinical Orthopaedics and Related Research, no. 463, pp. 31–36, 2007. [19] C. A. Willis-Owen, K. Brust, H. Alsop, M. Miraldo, and J. P. Cobb, “Unicondylar knee arthroplasty in the UK National Health Service: an analysis of candidacy, outcome and cost efficacy,” Knee, vol. 16, no. 6, pp. 473–478, 2009. [3] P. Hernigou and G. Deschamps, “Alignment influences wear in the knee after medial unicompartmental arthroplasty,” Clinical Orthopaedics and Related Research, no. 423, pp. 161–165, 2004. [20] P. L. Yen and B. L. Davies, “Active constraint control for image-guided robotic surgery,” Proceedings of the Institution of Mechanical Engineers H, vol. 224, no. 5, pp. 623–631, 2010. [4] P. Hernigou and G. Deschamps, “Posterior slope of the tibial implant and the outcome of unicompartmental knee arthro- plasty,” The Journal of Bone & Joint Surgery A, vol. 86, no. 3, pp. 506–511, 2004. [21] M. Karia, M. Masjedi, B. Andrews et al., “Robotic assistance enables inexperienced surgeons to perform unicompartmental knee arthroplasties on dry bone models with accuracy superior to conventional methods,” Advances in Orthopedics, vol. 2013, Article ID 481039, 7 pages, 2013. [5] E. M. Mariani, M. H. Bourne, R. T. Jackson, S. T. Jackson, and P. Jones, “Early failure of unicompartmental knee arthroplasty,” The Journal of Arthroplasty, vol. 22, supplement 2, no. 6, pp. 81– 84, 2007. [22] N. J. Dunbar, M. W. Roche, B. H. Park, S. H. Branch, M. A. Conditt, and S. A. Banks, “Accuracy of dynamic tactile- guided unicompartmental knee arthroplasty,” The Journal of Arthroplasty, vol. 27, no. 5, pp. 803.e1–808.e1, 2012. [6] T. Hananouchi, N. Nakamura, A. Kakimoto, H. Yohsikawa, and N. Sugano, “CT-based planning of a single-radius femoral component in total knee arthroplasty using the ROBODOC system,” Computer Aided Surgery, vol. 13, no. 1, pp. 23–29, 2008. [23] Biomet. Oxford partial knee-manual of the surgical technique, 2012. [7] J. E. Lang, S. Mannava, A. J. Floyd et al., “Robotic systems in orthopaedic surgery,” The Journal of Bone & Joint Surgery B, vol. 5. Conclusions Overall our results of segmentation, registration, and cuts made by robot were satisfactory for both components using the Acrobot Sculptor. It is possible to apply the full or part of this protocol in this study in a variety of other 5 Advances in Orthopedics products available on the market for better understanding and validation of robotic technology. products available on the market for better understanding and validation of robotic technology. controlled study of the acrobot system,” The Journal of Bone & Joint Surgery B, vol. 88, no. 2, pp. 188–197, 2006. [15] B. L. Davies, F. M. Rodriguez y Baena, A. R. W. Barrett et al., “Robotic control in knee joint replacement surgery,” Journal of Engineering in Medicine H, vol. 221, no. 1, pp. 71–80, 2007. Acknowledgments The authors would like to thank Wellcome trust and EPSRC for funding part of this study. The authors would like to thank Wellcome trust and EPSRC for funding part of this study. [16] M. Masjedi, K. Davda, S. Harris et al., “Evaluate your robot accuracy,” in Proceedings of the Hamlyn Symposium on Medical Robotics, London, UK, 2011. [17] D. Simon, R. V. O’Toole, M. Blackwell et al., “Accuracy val- idation in image-guided orthopaedic surgery,” in Proceedings of the 2nd International Symposium on Medical Robotics and Computer Assisted Surgery, pp. 185–192, Baltimore, Md, USA, 1995. References 93, no. 10, pp. 1296–1299, 2011. [24] A. Srivastava, G. Y. Lee, N. Steklov, C. W. Colwell Jr., K. A. Ezzet, and D. D. D’Lima, “Effect of tibial component varus on wear in total knee arthroplasty,” The Knee, vol. 19, no. 5, pp. 560–563, 2011. [8] R. Agganval, J. Hance, and A. Darzi, “Robotics and surgery: a long-term relationship?” International Journal of Surgery, vol. 2, no. 2, pp. 106–109, 2004. [9] S. Schumann, M. Tannast, L.-P. Nolte, and G. Zheng, “Vali- dation of statistical shape model based reconstruction of the proximal femur—a morphology study,” Medical Engineering and Physics, vol. 32, no. 6, pp. 638–644, 2010. [10] J. C. Baker-LePain, K. R. Luker, J. A. Lynch, N. Parimi, M. C. Nevitt, and N. E. Lane, “Active shape modeling of the hip in the prediction of incident hip fracture,” Journal of Bone and Mineral Research, vol. 26, no. 3, pp. 468–474, 2011. [11] G. Zheng and S. Schumann, “3D reconstruction of a patient- specific surface model of the proximal femur from calibrated x-ray radiographs: a validation study,” Medical Physics, vol. 36, no. 4, pp. 1155–1166, 2009. [12] J. Schmid and N. Magnenat-Thalmann, “MRI bone segmenta- tion using deformable models and shape priors,” in Proceedings of the 11th International Conference on Medical Image Computing and Computer-Assisted Intervention (Miccai ’08), vol. 5241 of Lecture Notes in Computer Science, part 1, pp. 119–126, 2008. [13] E. Audenaert, B. Smet, C. Pattyn et al., “Imageless versus image-based registration in navigated arthroscopy of the hip: a cadaver-based assessment,” The Journal of Bone & Joint Surgery B, vol. 94, no. 5, pp. 624–629, 2012. [14] J. Cobb, J. Henckel, P. Gomes et al., “Hands-on robotic uni- compartmental knee replacement: a prospective, randomised
https://openalex.org/W2315230561	https://zenodo.org/record/4751099/files/5116ijbiss01.pdf	English	null	ASSESSING THE IMPACT OF RELATIONSHIP QUALITY ON ONLINE ADOPTION	Deleted Journal	2,016	cc-by	6,334	ternational Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 ternational Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 Abstract Relationships are strongly embedded in Indian culture due to its high-context nature. Because of this, the role of relationship marketing has been advanced in a variety of services viz., banking, retailing, telecom, etc and hence the competition amongst various service providers has increased. Traditionally, relationship management used to be personal interaction based phenomenon but with IT as an enabler for online channels, the need for creating and growing relationships have transpired in a big way. With the advent of technological changes in the business environment, customers are seeking better alternatives to reduce the frequency of visiting the bank branch physically and are relying more on the virtual mode for transacting. Hence, banks and other service organizations are strongly focussing on the digital (online) channels for facilitating monetary transactions and hence gaining competitive advantage. In the virtual mode, the concept of relationships appears to be unrealistic due to the absence of human interface. However, significant studies have been conducted in this regard which focus on the influence of relationship quality factors viz., trust, commitment, satisfaction, etc. towards adopting the online channel for carrying out financial transactions. Relationship quality focuses on evaluating the strength of relationships, which affects customer loyalty. There exists a dearth of such significant studies in the Indian context. The present research will focus on bridging this gap in the literature. This paper will follow a causal research design for empirically analysing the impact of RQ factors towards adopting online mode for banking transactions. It will also investigate the future intentions of the customers toward using the online channels for carrying out financial transactions. Dr. Renu Aggarwal YMCA, Faridabad and Monica Bhardwaj Fortune Institute of International Business, Plot no 5, Rao Tula Ram Marg,,VasantVihar, Delhi, Delhi, India- 110057 Dr. Renu Aggarwal YMCA, Faridabad and Monica Bhardwaj Fortune Institute of International Business, Plot no 5, Rao Tula Ram Marg,,VasantVihar, Delhi, Delhi, India- 110057 Keywords Relationship Quality, Online, Adoption, India 1.INTRODUCTION Relationship marketing in the context of banking aims at establishing and maintaining long-term relationships with the customers (Ritter, 1993). The task of relationship becomes more challenging for online transactions (Mukherjee and Nath, 2003). In order to understand this challenge in a more sophisticated manner an assessment of the relationship strength is required. This can be judged using Relationship Quality (RQ) measures. DOI : 10.14810/ijbiss.2016.5101 1 Banking industry has been strongly enabled by the internet, which makes it convenient for the consumers to transact without physically visiting the bank.Commonly used banking operations by the customers via online medium include access to account information, funds transfer and bill payment. E-banking involves using internet technologies viz., internet, ATMs, mobile phones,etc. 1 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 for carrying out banking transactions (Abbasnejad, 2006 cited by Esmaeili et al, 2013). E-banking systems communicate customers about the banks’ efforts at three levels viz., notification, communication, and transaction (Ghane et al., 2010). Consumers are using internet banking for online shopping, mobile recharges, travel booking, movie ticket booking, etc.Consumers are benefitted in terms of easy access to their account, which can be controlled by them without any intervention/help from the bank employees and hence the banking consumers can make prompt financial decisions. In other words, online banking system is a self-service technology interface Meuteret al (2000). An important and challenging issue for the banks is the perceived security, which is perhaps due to the absence of bank employees and hence the banks are required to look at initiatives for developing online banking relationships and hence shall look at strengthening the online relationship quality (Warrington et al, 2000). Banks are expected to inculcate trust in the relationships such that the online banking is promoted which further looks at building customer loyalty. The importance of trust is at the paramount level due to the complexities arising from the virtual environment (Mukherjee and Nath,2003). Interestingly, trust is understood as the key factor, which is one of the components of RQ. The other important factors include commitment and satisfaction. These are discussed in detail under the review of literature section. The popularity of online banking has propelled since mid 1990s across the globe especially among the developed countries (Pikkarainenet al., 2004). The rate of penetration for online banking has risen immensely in the last few years. 1.INTRODUCTION In 2012, the total number of internet users were 137 million which is expected to rise to somewhere in the range of 330 million to 370 million (Internet World Stats, 2010; McKinsey Digital Consumer Research, 2013). In addition, the digital maturity index suggests that banking industry in India is at the starting stage of digital maturity (Mckinsey Analysis, 2013) while industries like personal computing and travel have reached the digital acceleration levels in India. According to the McKinsey Digital Consumer Survey 2012, the penetration for digital high-value consumers is 24% and online banking is used 0.19 times per week, while for overall consumers it is 7%, who use online banking 0.02 times per week. It is worth noting that the purchases made by the digital high-value consumers is just 1% while researching for the same and making final decision for purchase score 64% and 42% respectively thereby suggesting the lack of trust in the virtual medium. According to Mckinsey’s (2012) article on the impact of internet on India, the online payment system needs to be strengthened for which the Indian banking consumers exhibit low levels of trust and confidence. However, on the positive note, India still has a good scope for improvement in online payment systems and internet readiness, which is also mentioned in the same report. Therefore, it is important to evaluate the factors, which significantly affect the consumer’s intention to adopt the online medium for banking and payments and their plan to continue with the online banking system in future.In addition to this, the online technology interface provided by the bank has been evolving over the years and hence suggest the need for research towards the adoption of online banking which is aimed at enhancing the customer intentions (Loureiro et al., 2014). This study attempts to integrate RQ and online banking adoption thereby extending the current state of knowledge. 2 ternational Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 To address this, following research questions are put forward- To address this, following research questions are put forward- RQ1: Which relationship quality factors are the significant influencers of online banking adoption? RQ2: How do relationship quality factors influence the consumer’s adoption process towards online banking? 2.REVIEW OF LITERATURE It is well understood in the literature that much of the research pertaining to online adoption refers to the Technology Acceptance Model (TAM) and Theory of Reasoned Action (TRA). However, there have been many variations in the conceptualization and application of the model, which has resulted into various modified versions. In addition, the elusive nature of the online environment calls for more focus on the human factors, which ponder on relationships and hence the conceptualization of RQ towards its influence on adoption intentions holds importance. The consideration of RQ has been different across various studies in a very creative manner thereby carving out different dimensions. Trust and Satisfaction have been widely accepted however, several researches also consider Commitment as the key factor comprising RQ (Johnson et al., 2004). Initially, the research by Crosby et al (1990) discussed trust and satisfaction as the antecedents of RQ while Wong and Sohal(2002) contributed commitment as the other dimensions but still as the antecedent to RQ. However, there are several research works, which consider these factors as dimensions of RQ and not as antecedents. These include the works of Storbacka et al (1994), Henning-Thurau and Klee (1997), Dorsch et al (1998), Smith(1998), Baker et al (1999), Garbarino and Johnson (1999) , Lang and Colgate (2003), Walter et al (2003), Van Bruggen et al (2005), Ulaga and Eggert ( 2006), Leonidou et al (2006), Carr (2006). However, only the work of Lang and Colgate (2003) focus on banks and more specifically the online banking. The need for a fresh research perspective is expected which considers these three dimensions, which have been among the more popular RQ dimensions in the extant literature. These three main RQ factors are discussed in the following paragraphs along with a focus towards the online adoption. 1.INTRODUCTION The next section of the paper discusses the review of literature in which the factors pertaining to RQ viz., trust, commitment and satisfaction are reviewed. This is followed by the methodology section wherein the data collection method and the survey instrument along with scale development are discussed. Next, analysis of data is presented along with the structural model. This is followed by the discussion and managerial implications. 2.1.Trust Trust is the confidence exhibited by the buyer in the seller’s future performance. Trust is a key construct of RQ and a favourably perceived future performance results in high level of RQ and hence positively influences customer loyalty (Zhang et al., 2011).Trust holds greater importance in the online environment in relation to the offline-banking environment (Ratnasingham, 1998). Trust is related with sub-factors viz., benevolence, competence, integrity, predictability, etc. which is based on customer’s beliefs about the quality of services provided by the service 3 3 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 provider in online contexts (Gefen et al.,2003). Yee and Faziharudean (2010) discussed, through review of literature, how trust influences customer loyalty. Cyr et al (2007) found a direct and positive association between trust and loyalty in the context of e-services while the same was supported by Lin and Wang (2006) in the m-commerce context. In addition, Floh and Treiblmaier (2006) discussed trust and satisfaction apart from website quality as antecedents to customer loyalty in the online banking context.Yee and Faziharudean (2010) empirically investigated the impact of trust along with other factors on customer loyalty and found trust significantly influencing the customer loyalty in Malaysian online banking environment. Kim et al (2013) studied the mediating role of trust on the usage intentions in online shopping context. Yousafzai et al (2009)found that trust is an antecedent of behavioural intention and holds a multidimensional position in adoption to internet banking. Moreover, online banking is conditional to evoking of trust in the consumer’s mind, which depends on the ability of the bank. Trust also had significant and positive effect on internet banking adoption in the Vietnamese context (Chong et al, 2010). Some research works, which focus on TAM, also included trust in the structural model while evaluating the adoption intentions. Trust was considered as an antecedent to PU and PEOU dimensions of TAM and found that the trust has positive influence on both these dimensions (Erikssonet al, 2005). Also, PU and PEOU in turn influence the attitude towards online banking (Chiou and Shen, 2012). Suh and Han (2002) found that trust significantly influences customer’s attitude towards adoption to online banking. Suh and Han (2002) found trust as a key factor in elucidating customer attitude towards the adoption of online banking (Suh and Han, 2002). 2.1.Trust It is the attitude, which along with trust and PU influences online banking usage intentions. It is worth noting that attitude towards adoption is also understood as acceptance which is an antecedent to adoption intentions which in-turn might influence continuance usage intentions. Contrary to this,Kesharwani and Bisht (2012) found trust as an impediment towards online banking transactions in the Indian context when used as an extra dimension to TAM. 2.4.Attitude Situation of risk and uncertainty can be reduced by the involvement of trust (Mayer et al., 1995). Such challenging situations are common in the context of online banking. An increase in trust would result into positive attitude towards the online banking and hence would influence the intention to adopt the online mode for carrying out banking transactions. This point has also been reiterated by Jarvenpaa et al (1999) in the context of an internet store and also in the context of online banking (Liu et al, 2005). In the present study, the authors also contemplate that satisfaction and commitment also influence the attitude in a similar manner. 2.3.Satisfaction Satisfaction holds a great deal of importance in the online context (Nusair and Kandampully, 2008).An evaluation of seller’sperformance by the buyer results into the level of satisfaction. In addition, the confirmation or disconfirmation from the expectation of the consumers is another conceptualization of satisfaction(Hennig-Thurau and Hansen, 2000) which arises from expectancy-disconfirmation (Oliver, 1981). However, this theory has limitations due to the lack of understanding by the customers about their expectations especially owing to the complexities of the online environment (Allagui and Temessek, 2004). This calls for measuring satisfaction through customer opinions which is accumulated together to provide a measure of overall satisfaction based on a holistic experience (Garbarino and Johnson, 1999). For online banking environments it is the online contact with the customers over the period of time, which defines satisfaction (Liang et al,2008) which comes from the interactive online interfaces provided for the customers (Allagui and Temessek, 2004). A high level of RQ is a function of a high level of satisfaction. Satisfaction being a construct of RQ is highly likely in increasing consumer loyalty (Zhang et al, 2011). According to Mols (1998), online banking environment provides more satisfaction and satisfaction with the existing offering would provide more scope for adapting to the online banking channels. Therefore, past performance acts as an assurance of future performance and hence adoption becomes imminent as it is an additional benefit and more convenient due to the deployment of the technology (Rexha, 2003). In addition, in the context of banking, loyalty is key issue and hence customer satisfaction requires monitoring (Aurier and N’Goala, 2010) in order to achieve a high level of RQ. 2.2.Commitment In the context of consumer behaviour,commitment is a long-term wish for consumer to maintain the relationships (Moorman et al., 1993). Commitment can take three forms viz, attitudinal, instrumental and temporal (Gundlach et al., 1995) and also commitment is revealed in adoption of online banking (Rexha et al., 2003). It is also affects behaviour and attitude in a positive manner (Johnson et al., 2006). In fact, behaviour and attitude are the pointers to loyalty (Chiu et al., 2009). Behaviour can be elaborated as intention to purchase and intention to adopt in the online context. Attitude has been however understood differently in the literature where Chiu et al(2009) discuss it as willingness to recommend while in TAM it is an antecedent to intention to adopt. Looking into the RQ literature it is seen that commitment has found place as a dimension of RQ in different service industry settings. These include electronic products retailing (Heenig-Thurau, 2000), apparel retailing (De Wulf et al, 2003; De Cannie`re et al. 2009), travel agency (Moliner et al, 2007; Beatson et al, 2008). Studies pertaining to the banking industry involving the conceptualization of commitment as a dimension of RQ include the works of Lang and Colgate (2003) and Ndubisi (2007) and both these studies focus more on the intangible aspects with the former focussing on the online activities. 4 4 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 2.5.Continuance Usage Intention Continuance usage intention is related with the loyal behaviour of the customers towards the online banking. This is connected with continuing the relationship in future and not dropping it. Sometimes this phenomenon is also referred as relational continuity. Satisfaction is understood to be a key predictor of relational continuity (Leuthesser and Kohli, 1995). It is highly expected that relationships between the bank and the customer should develop with time which is contingent to high satisfaction levels due to which the customers are loyal and don’t switch their service providers (Mols, 1999). A variety of researches exist on online adoption related to banking and other e-commerce environments, which also include influence of RQ factors on intention to adopt online medium for aforementioned situations.A holistic perspective on the key RQ factors exhibiting their 5 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 influence on online adoption needs profound exploration and analysis. Trust, commitment and satisfaction have been discussed under various research works sometimes together but in most cases the three dimensions did not appear together for the purpose of evaluation of their consequences towards online adoption. For the present study, Trust Commitment and Satisfaction are the three factors considered which constitute RQ. In other words, the concept of RQ in the online context has not been widely discussed in a holistic manner. Brun et al (2014) work is perhaps the first significant study, which discusses online RQ comprising of trust, commitment and satisfaction. In addition, Rajaobelina et al (2013) have also included the very same three dimensions of RQ for classifying the online banking customers. 3.1.Sample and Data Collection Data was collected from the youth using mall-intercept method from various malls of Delhi-NCR. In addition, students from selected colleges in the region were also targeted for data collection. The students selected were either registered for undergraduate or postgraduate courses and both were mainly from the management streams. Other respondents who were contacted through the mall intercept method were mostly the working professionals. Mall intercept method has been found useful in various studies (O'Cass andGrace, 2008) and also in Indian context (Prasad and Aryasri, 2011, Tripathi and Dave, 2013). A qualifier question was included in the survey instrument to ascertain that the respondents were either casual or frequent users of online banking while at the same time non-users were excluded from the analysis. About 1000 respondents were contact, which resulted into 466 usable questionnaires for further analysis thereby giving a response rate of about 46%. The low response rate is quite usual in survey responses in Indian context especially among the youth who are at times casual users of the service. 3.2.Instruments and Measures The scale items for the three RQ factors along with attitude, adoption intention and continuance usage intentions were extracted from the literature. Attitude towards adoption and behavioural intention were culled from the study by Taylor and Todd (1995) with some modifications. The items for trust, commitment, satisfaction and continuance usage intention were extracted and adapted from the work of Liang et al (2012) 3.3.Hypothesis Based on the review of literature the following hypothesis are proposed – Based on the review of literature the following hypothesis are proposed H1: Trust directly and positively influences Attitude towards Adoption of online banking H2: Commitment directly and positively influences Attitude towards Adoption of online banking H3: Satisfaction directly and positively influences Attitude towards Adoption of online banking H4: Attitude towards Adoption directly and positively influences Intention to Adopt online banking 6 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 H5: Intention to Adopt online banking directly and positively influences Continuance Usage Intention Figure 1: Structural model Trust Behavioural Intentions Continuance Usage Intentions Commitment Satisfaction Attitude H3 H2 H1 H4 H5 Figure 1: Structural model 4.ANALYSIS AND DISCUSSION First, an Exploratory Factor Analysis was conducted using SPSS for all the variables. The KMO measure was 0.901 and was found significant. The communalities score were all above 0.50 and the total variance extracted was 74% resulting into six factors. The reliability scores were measured through Cronbach’s alpha. Then the measurement model was evaluated using CFA through AMOS software first for the RQ factors only and then for the whole model together. For the RQ factors only the model fit was found within the acceptable levels with CMIN/DF=2.314, RMR= 0.061, GFI= 0.951, CFI=0.981, RMSEA= 0.075. All items loaded significantly on their respective constructs. Next, CFA was subjected on RQ factors along with attitude towards adoption, intention to adopt and continuance intention. In this case, all the items loaded significantly on their respective dimensions. The model fit was acceptable with CMIN/DF=2.529, RMR= 0.053, GFI= 0.908, CFI=0.948, RMSEA 0 051 Trust Behavioural Intentions Continuance Usage Intentions Commitment Satisfaction Attitude H3 H2 H1 H4 H5 Trust Satisfaction Commitment H3 H1 Attitude Behavioural Intentions Figure 1: Structural model 4.2.Construct Validity It measures the level to which an observed variable measures the unobserved factors for which they are created. It includes face validity, convergent validity and discriminant validity. Face validity is established by extracting the scale items from the extant literature and the same have been adjusted to fit with the present study. Convergent validity was established by investigating the factor loadings and AVE. All the factor loadings (or standardized estimates) were significant at p < 0.001. Reliability scores were pre- established and were all greater than 0.7. All these estimates were higher than 0.7. AVE figures were all greater than 0.50. In addition, AVE was greater than square of the inter-construct correlation. Hence, discriminant validity was established. 4.ANALYSIS AND DISCUSSION First, an Exploratory Factor Analysis was conducted using SPSS for all the variables. The KMO measure was 0.901 and was found significant. The communalities score were all above 0.50 and the total variance extracted was 74% resulting into six factors. The reliability scores were measured through Cronbach’s alpha. Then the measurement model was evaluated using CFA through AMOS software first for the RQ factors only and then for the whole model together. For the RQ factors only the model fit was found within the acceptable levels with CMIN/DF=2.314, RMR= 0.061, GFI= 0.951, CFI=0.981, RMSEA= 0.075. All items loaded significantly on their respective constructs. Next, CFA was subjected on RQ factors along with attitude towards adoption, intention to adopt and continuance intention. In this case, all the items loaded significantly on their respective dimensions. The model fit was acceptable with CMIN/DF=2.529, RMR= 0.053, GFI= 0.908, CFI=0.948, RMSEA= 0.051. 7 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 4.1.Scale Reliability All the latent constructs show signs of high reliability with Cronbach’s Alpha > 0.70 (Hair et al., 2006) and hence the scale is reliable. In addition, composite reliability and average variance extracted scores are taken for all the RQ factors and for the dependent constructs. These are provided in the Table # 1.Composite reliability obtained for all the latent variables is greater than 0.70 and hence it crosses the minimum acceptable values (Carmines and Zeller, 1988). AVE for each factor is greater than 0.5 and hence it is above the minimum acceptable values (Fornell and Larcker, 1981). 4.3.Testing the Structural model The structural model provides an adequate fit with CMIN/DF= 3.279, RMR= 0.057, GFI= 0.905, CFI= 0.918, RMSEA= 0.059. The hypotheses results are provided in Table #2. Except for H2, which is aimed at testing the influence of commitment onattitude towards adopting online banking, all other hypotheses are significant at p<0.001. Results indicate that trust and satisfaction directly and positively influence attitude towards adoption of online banking, which in turn directly and positively influence Intention towards online adoption, which further directly and positively influences intention to continue the online banking usage in future. Table# 1: Measurement model statistics Constructs Measurement Items Estimate AVE CR Trust Trust1 0.745 0.730 0.889 Trust2 0.918 Trust3 0.890 Commitment Commitment1 0.829 0.697 0.889 Commitment2 0.885 Table# 1: Measurement model statistics International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 Commitment3 0.787 Satisfaction Satisfaction1 0.948 0.839 0.940 Satisfaction2 0.879 Satisfaction3 0.919 Attitude Adoption3 0.918 0.813 0.929 Adoption2 0.910 Adoption1 0.877 Behavioural Intentions B_Int1 0.763 0.759 0.903 B_Int2 0.941 B_Int3 0.899 Continuance Usage CU_Int1 0.873 0.692 0.871 CU_Int2 0.834 CU_Int3 0.786 Table #2: Hypothesis testing Hypotheses Hypothesised Paths Standardised Path Co- efficients p- value Results H1 Adoption <--- Trust 0.417 * Hypothesis Supported H2 Adoption <--- Commitment -0.036 0.671 Hypothesis Not Supported H3 Adoption <--- Satisfaction 0.393 * Hypothesis Supported H4 BI <--- Adoption 0.749 * Hypothesis Supported H5 CUI <--- BI 0.764 * Hypothesis Supported Table #2: Hypothesis testing 5.DISCUSSION AND MANAGERIAL IMPLICATIONS 9 The present study is aimed at testing how relationship quality can influence the adoption of online banking by the customers. Insights are gained into the key relationship quality factors, which are identified and validated using the CFA approach. These dimensions can be used by the managers 9 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 to understand the consumer behaviour related to online adoption by the bank customers in the Indian context where the penetration levels of online banking are limited despite a relatively large number of internet users. These factors also bring forward a very important aspect of decision- making process, which comes from attitude towards adoption, which is influenced, by trust and satisfaction. Attitude acts an influencer to behavioural intention to adopt online mode for banking. The difference between attitude and intention to adopt is critical but the managerial understanding about the online adoption process is important at one-step further. The intention to adopt might be casual but if the experience werepositive, it would help in adopting online mode for banking more regularly. This focus is necessary for the banks, as online mode is more cost effective than other modes for transactions. 6. LIMITATIONS AND SCOPE FOR FURTHER RESEARCH The current study is limited in its geographical scope of Delhi-NCR, which shall not be considered as a reference for understanding the consumer behaviour for the whole country. In addition, more RQ factors can be derived from the literature, which might help in expanding the academic horizon in the context of online banking adoption. There might be a possibility of a second order constructs particularly comprising of trust and satisfaction especially in the case of online environments (Zhang et al., 2011). Further research can also be conducted regarding the transition from internet banking adoption to mobile banking adoption. REFERENCES 1. Allagui, A. and Temessek, A. (2004). Testing an e-loyalty conceptual framework. Journal of E- Business, 4(1), 44-56. 2. Aurier, P., &N’Goala, G. (2010). The differing and mediating roles of trust and relationship commitment in service relationship maintenance and development. Journal of the Academy of Marketing Science, 38(3), 303-325. 3. Baker, T. L., Simpson, P. M., &Siguaw, J. A. (1999). The impact of suppliers' perceptions of reseller market orientation on key relationship constructs.Journal of the Academy of Marketing Science, 27(1), 50-57. 4. Beatson, A., Lings, I., &Gudergan, S. (2008). Employee behaviour and relationship quality: imp on customers. The Service Industries Journal,28(2), 211-223. 5. Brun, I., Rajaobelina, L., & Line, R. (2014). Online relationship quality: scale development and initial testing. International Journal of Bank Marketing, 32(1), 5-27. g f g 6. Carmines, E.G. and Zeller, R.A. (1988). Reliability and Validity Assessment, Sage, Beverly Hills, CA. 7. Carr, C. L. (2006). Reciprocity: the golden rule of IS-user service relationship quality and cooperation. Communications of the ACM, 49(6), 77-83. p f 8. Chiou, J. S., &Shen, C. C. (2012). The antecedents of online financial service adoption: the impact of physical banking services on Internet banking acceptance. Behaviour& Information Technology, 31(9), 859-871. 9. Chiu, C. M., Lin, H. Y., Sun, S. Y., & Hsu, M. H. (2009). Understanding customers' loyalty intentions towards online shopping: an integration of technology acceptance model and fairness theory. Behaviour& Information Technology, 28(4), 347-360. y f gy 10. Chong, A. Y. L., Ooi, K. B., Lin, B., & Tan, B. I. (2010). Online banking adoption: an empirical analysis. International Journal of Bank Marketing, 28(4), 267-287. 11. Crosby, L. A., Evans, K. R., & Cowles, D. (1990). Relationship quality in services selling: An interpersonal influence perspective. The Journal of Marketing, 54(3), 68-81. 12. Cyr, D., Hassanein, K., Head, M., &Ivanov, A. (2007). The role of social presence in establishing loyalty in e-Service environments. Interacting with Computers, 19(1), 43-56. 10 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 13. De Cannière, M. H., De Pelsmacker, P., &Geuens, M. (2009). Relationship quality and the theory of planned behavior models of behavioral intentions and purchase behavior. Journal of Business Research, 62(1), 82-92. 14. De Wulf, K., Odekerken-Schröder, G., & Van Kenhove, P. (2003). Investments in consumer relationships: a critical reassessment and model extension. The International Review of Retail, Distribution and Consumer Research, 13(3), 245-261. 15. Dorsch, M. J., Swanson, S. REFERENCES R., & Kelley, S. W. (1998). The role of relationship quality in the stratification of vendors as perceived by customers. Journal of the Academy of Marketing Science, 26(2), 128-142. ( ) 16. Eriksson, K., Kerem, K., & Nilsson, D. (2005). Customer acceptance of internet banking in Estonia. International Journal of Bank Marketing, 23(2), 200-216. 17. Esmaeili, M., Nazarpoori, A., &Najafi, M. (2013). An investigation on loyalty formation model in e- banking customers: A case study of banking industry. Management Science Letters, 3(3). 18. Floh, A., &Treiblmaier, H. (2006). What keeps the e-banking customer loyal? A multigroup analysis of the moderating role of consumer characteristics on e-loyalty in the financial service industry. Journal of Electronic Commerce Research, 7(2), 97-110. y 19. Fornell, C., &Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of marketing research, 18(1), 39-50. 20. Garbarino, E., & Johnson, M. S. (1999). The different roles of satisfaction, trust, and commitment in customer relationships. the Journal of Marketing, 63(2), 70-87. 21. Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: an integrated model. MIS quarterly, 27(1), 51-90. 22. Ghane, S., Fathian, M., Gholamian, M. R., Zbakh, M., El Kettani, M. D., Jambari, H., & Chen, C. (2011). Full Relationship Among E-Satisfaction, E-Trust, E-Service Quality, and E-Loyalty: the case of Iran E-Banking. Journal of Theoretical and Applied Information Technology, 33(1), 1-6. 23. Gundlach, G. T., Achrol, R. S., &Mentzer, J. T. (1995). The structure of commitment in exchange. The Journal of Marketing, 59(1), 78-92. 24. Hair, J.F. Jr, Black, C.W., Babin, J.B., Anderson, R.E. and Tatham, L.R. (2006), Multivariate Data Analysis, 6th ed., Pearson, Upper Saddle River, NJ. 25. Hennig‐Thurau, T., & Klee, A. (1997). The impact of customer satisfaction and relationship quality on customer retention: A critical reassessment and model development. Psychology & Marketing, 14(8), 737-764. 26. Hennig-Thurau, T. (2000). Relationship quality and customer retention through strategic communication of customer skills. Journal of Marketing Management, 16(1-3), 55-79. 27. Hennig-Thurau, T., & Hansen, U. (2000). Relationship marketing–some reflections on the state-of- the-art of the relational concept. Relationship marketing: Gaining competitive advantage through 26. Hennig-Thurau, T. (2000). Relationship quality and customer retention through strategic communication of customer skills. Journal of Marketing Management, 16(1-3), 55-79. 27. Hennig-Thurau, T., & Hansen, U. (2000). Relationship marketing–some reflections on the state-of- the-art of the relational concept. REFERENCES Relationship marketing: Gaining competitive advantage through customer satisfaction and customer retention, 3-27. f 28. Järvenpää, S. L., Tractinsky, N., Saarinen, L., & Vitale, M. (1999). Consumer trust in an Internet store: A cross-cultural validation. Journal of Computer-Mediated Communication, 5(2), 1-5. 29. Johnson, M. D., Herrmann, A., & Huber, F. (2006). The evolution of loyalty intentions. Journal of 29. Johnson, M. D., Herrmann, A., & Huber, F. (2006). The evolution of loyalty intentions. Journa Marketing, 70(2), 122-132. g ( ) 30. Johnson, J. L., Sohi, R. S., &Grewal, R. (2004). The role of relational knowledge stores in interfirm partnering. Journal of Marketing, 68(3), 21-36. 31. Kim, Y. J., Han, S. M., & Lee, E. (2013). The role of trust in online shopping malls: different types of trust and how they affect consumer intention. International Journal of Electronic Commerce, 4(2). 385-388 32. Kesharwani, A., &Bisht, S. S. (2012). The impact of trust and perceived risk on internet banking adoption in India: An extension of technology acceptance model. International Journal of Bank Marketing, 30(4), 303-322. 33. Lang, B., & Colgate, M. (2003). Relationship quality, on-line banking and the information technology gap. International Journal of Bank Marketing, 21(1), 29-37. 34. Leonidou, L. C., Barnes, B. R., &Talias, M. A. (2006). Exporter–importer relationship quality: The inhibiting role of uncertainty, distance, and conflict.Industrial Marketing Management, 35(5), 576- 588. 11 11 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 35. Leuthesser, L., &Kohli, A. K. (1995). Relational behavior in business markets: Iplications relationship management. Journal of Business Research, 34(3), 221-233. 36. Liang, C. J., Chen, H. J., & Wang, W. H. (2008). Does online relationship marketing enha customer retention and cross-buying?. The service industries journal, 28(6), 769-787. 37. Liang, T. P., Ho, Y. T., Li, Y. W., & Turban, E. (2011). What drives social commerce: The role of social support and relationship quality. International Journal of Electronic Commerce, 16(2), 69-90. 38. Lin, H. H., & Wang, Y. S. (2006). An examination of the determinants of customer loyalty in mobile commerce contexts. Information & Management, 43(3), 271-282. 39. Liu, C., Marchewka, J. T., Lu, J., & Yu, C. S. (2004). Beyond concern: a privacy–trust–behavioral intention model of electronic commerce. Information & Management, 42(1), 127-142. 40. Loureiro, S. M. C., Kaufmann, H. R., &Rabino, S. (2014). Intentions to use and recommend to others: an empirical study of online banking practices in Portugal and Austria. REFERENCES Online Information Review, 38(2), 2-2. ( ) 41. Mayer, R. C., Davis, J. H., &Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of management review, 20(3), 709-734. 42. Meuter, M. L., Ostrom, A. L., Roundtree, R. I., &Bitner, M. J. (2000). Self-service technologies: understanding customer satisfaction with technology-based service encounters. Journal of Marketing, 64(3), 50-64. 43. Moliner, M. A., Sánchez, J., Rodríguez, R. M., &Callarisa, L. (2007). Relationship quality with a travel agency: The influence of the postpurchase perceived value of a tourism package. Tourism and Hospitality Research, 7(3-4), 194-211. 44. Mols, N. P. (1998). The Internet and the banks’ strategic distribution channel decisions. Internet Research, 8(4), 331-337. , ( ), 45. Moorman, C., Zaltman, G., &Deshpande, R. (1992). Relationships Between Providers and Users of Market Research: The Dynamics of Trust. Journal of marketing research, 29(3), 314-28. 46. Mukherjee, A., &Nath, P. (2003). A model of trust in online relationship banking. International Journal of Bank Marketing, 21(1), 5-15. 47. Ndubisi, N. O. (2007). Relationship quality antecedents: the Malaysian retail banking perspective. International Journal of Quality & Reliability Management, 24(8), 829-845. 48. Nusair, K. K., &Kandampully, J. (2008). The antecedents of customer satisfaction with online tra services: a conceptual model. European Business Review, 20(1), 4-19. 49. O'Cass, A., & Grace, D. (2008). Understanding the role of retail store service in light of self‐image– store image congruence. Psychology & Marketing, 25(6), 521-537. 50. Oliver, R.L. (1981). Measurement and evaluation of satisfaction processes in retail settings. Journal of Retailing, 57(3), 25-48. 51. Pikkarainen, T., Pikkarainen, K., Karjaluoto, H., &Pahnila, S. (2004). Consumer acceptance of online banking: an extension of the technology acceptance model. Internet research, 14(3), 224-235. 52. Prasad, C. J., &Aryasri, A. R. (2011). Effect of shopper attributes on retail format choice behaviour for food and grocery retailing in India. International Journal of Retail & Distribution Management, 39(1), 68-86. g ( ) 53. Rajaobelina, L., Brun, I., &Toufaily, É. (2013). A relational classification of online banking customers. International Journal of Bank Marketing, 31(3), 187-205. 54. Ratnasingham, P. (1998). The importance of trust in electronic commerce. Internet research, 8(4), 313-321. 55. Rexha, N., Kingshott, R. P. J., & Aw, A. S. S. (2003). The impact of the relational plan on adoption of electronic banking. Journal of Services Marketing, 17(1), 53-67. 56. Ritter, D. S. (1993). Relationship banking: cross-selling the bank's products & services to meet your customer's every financial need. International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 REFERENCES Bankers Publishing Company/Probus Publishing Company. 57. Smith, J. B. (1998). Buyer–seller relationships: similarity, relationship management, and quality. Psychology & Marketing, 15(1), 3-21. q y y gy g 8. Storbacka, K., Strandvik, T., &Grönroos, C. (1994). Managing customer relationships for profit: the dynamics of relationship quality. International journal of service industry management, 5(5), 21-38. 12 International Journal of Business Information Systems Strategies (IJBISS) Vol.5, No.1,February 2016 59. Suh, B., & Han, I. (2003). Effect of trust on customer acceptance of Internet banking. Electro Commerce research and applications, 1(3), 247-263. 60. Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: A test of compe models. Information systems research, 6(2), 144-176. 61. Tripathi, G., & Dave, K. (2013). Store format choice and relationship quality in apparel retail: A study of young and early-middle aged shoppers in New Delhi region. Journal of Retailing and Consumer Services, 20(5), 479-487. 62. Ulaga, W., &Eggert, A. (2006). Relationship value and relationship quality: Broadening the nomological network of business-to-business relationships. European Journal of Marketing, 40(3/4), 311-327. 63. Van Bruggen, G. H., Kacker, M., &Nieuwlaat, C. (2005). The impact of channel function performance on buyer–seller relationships in marketing channels.International Journal of Research in Marketing, 22(2), 141-158. g ( ) 64. Walter, A., Müller, T. A., Helfert, G., & Ritter, T. (2003). Functions of industrial supplier relationships and their impact on relationship quality. Industrial Marketing Management, 32(2), 159- 169. 65. Warrington, T. B., & Caldwell, H. M. (2000). Building trust to develop competitive advantage in e- business relationships. Competitiveness Review: An International Business Journal incorporating Journal of Global Competitiveness, 10(2), 160-168. 66. Wong, A., &Sohal, A. (2002). An examination of the relationship between trust, commitment and relationship quality. International Journal of Retail & Distribution Management, 30(1), 34-50. 67. Yee, B. Y., &Faziharudean, T. M. (2010). Factors affecting customer loyalty of using Internet banking in Malaysia. Journal of Electronic Banking Systems, 2010(2010), 21. g y f g y 68. Yousafzai, S., Pallister, J., &Foxall, G. (2009). Multi-dimensional role of trust in Internet banking adoption. The Service Industries Journal, 29(5), 591-605. 69. Zhang, Y., Fang, Y., Wei, K. K., Ramsey, E., McCole, P., & Chen, H. (2011). Repurchase intention in b2c e-commerce—a relationship quality perspective.Information& Management, 48(6), 192-200. 13
https://openalex.org/W2190558261	https://www.frontiersin.org/articles/10.3389/fpsyg.2015.01880/pdf	English	null	The Complexity of Neuroenhancement and the Adoption of a Social Cognitive Perspective	Frontiers in psychology	2,015	cc-by	5,672	PERSPECTIVE published: 01 December 2015 doi: 10.3389/fpsyg.2015.01880 The Complexity of Neuroenhancement and the Adoption of a Social Cognitive Perspective Arnaldo Zelli 1, Fabio Lucidi 2 and Luca Mallia 1 1 Department of Movement, Human and Health Sciences, University of Rome “Foro Italico”, Rome, Italy, 2 Department of Psychology of Development and Socialization Processes, Sapienza University of Rome, Rome, Italy This contribution attempts to provide a broad perspective to the psychological study of neuroenhancement (NE). It departs from the assumption that, as the use of performance enhancing substances in sport, the use of substances with the aim of improving one’s cognitive, motivational and affective functioning in academic domains is a goal-directed behavior. As such, its scientific study may very well benefit from an analysis taking into account the psychological processes regulating people’s behavioral intentions and decisions. Within this broad framework, this contribution addresses several issues that currently seem to characterize the debate in the literature on neuroenhancement substances (NES) use. The first conceptual issue seeks to determine and define the “boundaries” of the phenomenon. The second issue concerns the empirical evidence on the prevalence of using certain substances for the purpose of NE. Finally, there is a debate around the ethical and moral implications of NE. Along these lines, the existing psychological research on NE has adopted mainly sociological and economic decision-making perspectives, greatly contributing to the psychological discourse about the phenomenon of NE. However, we argue that the existing psychological literature does not offer a common, explicit and integrated theoretical framework. Borrowing from the framework of doping research, we recommend the adoption of a social cognitive model for pursuing a systematic analysis of the psychological processes that dynamically regulate students’ use of NES over time. Edited by: Wanja Wolff, University of Potsdam, Germany Specialty section: This article was submitted to Performance Science, a section of the journal Frontiers in Psychology Received: 21 September 2015 Accepted: 20 November 2015 Published: 01 December 2015 Citation: Zelli A, Lucidi F and Mallia L (2015) The Complexity Specialty section: This article was submitted to Performance Science, a section of the journal Frontiers in Psychology Received: 21 September 2015 Accepted: 20 November 2015 Published: 01 December 2015 Specialty section: This article was submitted to Performance Science, a section of the journal Frontiers in Psychology Keywords: neuroenhancement, students, social cognitive models, doping, substance use PREMISES The use of pharmacological substances to enhance performance is an issue psychologists have thoroughly investigated in the sport context. In this context, a broad psychological perspective focusing on the social-cognitive processes regulating one’s intentions and use of performance enhancement substances has been largely adopted by many scholars in recent years (see Ntoumanis et al., 2014, for a review). Similarly, the use of substances with the aim of improving one’s cognitive, motivational and affective functioning in academic and work contexts has also recently emerged and been debated as a critical research issue in the literature on neuroenhancement (NE) and cognitive enhancement (e.g., Zohny, 2015). Keywords: neuroenhancement, students, social cognitive models, doping, substance use Edited by: Wanja Wolff, University of Potsdam, Germany Reviewed by: Sylvain Laborde, German Sport University Cologne, Germany Chris Englert, University of Bern, Switzerland Correspondence: Arnaldo Zelli arnaldo.zelli@uniroma4.it Citation: 156), we feel that this definition, despite being extremely clear, still needs further consideration or clarification. “.. . . the misuse of prescription drugs, other illicit drugs, or alcohol for the purpose of enhancing cognition, mood, or prosocial behavior in academic or work related contexts” (e.g., Maier and Schaub, 2015, p. 156), we feel that this definition, despite being extremely clear, still needs further consideration or clarification. First, some NE studies distinguish among prescribed- substances (e.g., Methylphenidate, Modafinil, Amphetamines, etc), substances of abuse (e.g., Alcohol, Cannabis, Cocaine) and over-the-counter substances or drugs (e.g., caffeinated products and food supplements), the so called “soft enhancers” or “life style” drugs (e.g., Franke et al., 2014; Maier and Schaub, 2015). Second, clarification seems warranted when one considers the extent to which NE must or needs to be conceived with respect to behavioral rather than cognitive performance criteria (e.g., a substance enhances one’s memory which, in turn, affects and positively contributes to one’s exam grades). With this in mind, some scholars (e.g., Zohny, 2015) distinguish substances’ effects on mood or motivational processes from their effects on other processes, such as attention or memory, and go on in suggesting that the latter type of effects specifically constitutes cognitive enhancement. Whether “cognitive” only refers to what is traditionally seen as “cold” cognition or, rather, whether motivational and emotional processes legitimately represent parts of one’s cognitions, is an issue that has been long debated in classical work (e.g., Pessoa, 2008). g In the context of the present contribution, it seems important to us to highlight another issue that perhaps has relevant assonances with the distinction between “cold” and “hot” cognitions. One’s use of cognitive enhancement substances legitimately may call upon two broadly alternative cases. The first envisions the possibility that one may use a given substance to improve his or her “effort” as a means of performance (e.g., Ritalin to stay awake and study for a longer time). The second envisions the possibility that one may use a given substance to improve specific cognitive functions or tasks (e.g., memory recall or problem solving). Both cases highlight a critical issue in any psychological analysis, that of one’s goals for choosing a particular course of action. At any rate, to what extent any or both of these cases must be considered “cognitive enhancement” has not yet been addressed by the existing literature. Citation: At a minimum, however, it seems plausible to hypothesize that users of cognitive enhancement substances might primarily be interested in achieving their best (academic) performance outcomes, rather than in the processes underlying any particular outcome. Finally, it is important to note that the present contribution unfolds with an exclusive focus on academic or educational contexts. Typically, these contexts offer clear-cut and broadly acknowledged behavioral criteria and protocols for referring to and observing individuals’ performances. Furthermore, as also suggested by Kipke (2013), academic examinations and testing might warrant special attention, as performance outcomes clearly rely on one’s cognitive functioning and capacities. Third, NE in these settings also raises issues regarding the integrity and validity of academic examinations and testing results. As a concluding note, academic and educational settings are also the contexts which have often been the target of empirical studies on NE (e.g., Smith and Farah, 2011; Franke et al., 2014). We think the above issues, despite their peculiarities, offer some ground for consensus. The use of neuroenhancement substances (NES), by students or professionals, reflects a person’s conscious or deliberate intentions, at least in the case of unsupervised use of psychoactive substances by healthy individuals. Furthermore, no matter what the NES chemical and medical properties are with respect to the enhancement of specific cognitive capacities (e.g., memory), we think that there is consensus in the literature on the general view that individuals pursue enhancement goals with the intention of influencing actual behavioral performance. Citation: Zelli A, Lucidi F and Mallia L (2015) The Complexity of Neuroenhancement and the Adoption of a Social Cognitive Perspective. Front. Psychol. 6:1880. doi: 10.3389/fpsyg.2015.01880 December 2015 \| Volume 6 \| Article 1880 1 Frontiers in Psychology \| www.frontiersin.org A Social Cognitive Perspective for Neuroenhancement Zelli et al. We believe this debate is currently focusing on three clearly distinct—and yet intimately related—issues. There is a conceptual focus which seeks to determine and define the “boundaries” of the phenomenon. Some boundaries stress the distinction between pharmacological and non-pharmacological substances to enhance performance. Other boundaries instead refer to the distinction between “hot” (e.g., mood, motivation) and “cold” (e.g., attention, memory) cognitions, and to the general notion that cognitive enhancement seemingly only matters for the latter type of cognitions (Zohny, 2015). Finally, there are boundaries stressing the contexts in which it is plausible or relevant to discuss NE (e.g., work places, educational settings) and those to which the term of NE instead does not apply (e.g., recreational settings, sport settings). There also is an empirical debate seeking to clarify the prevalence and social relevance of using certain substances for the purpose of NE (e.g., Maier and Schaub, 2015). There is also a debate around the ethical and moral implications of NE, with the literature primarily addressing issues ranging from personal safety, to the social responsibility of institutions, agencies or firms promoting or contributing to NE, to issues about a person’s character and his or her right to seek a good life (Schermer, 2008). This contribution briefly summarizes the key elements of these debates and, while recognizing the undisputable value these debates have for scientific progress, also argues that they are undermined by a lack of explicit reference to a clear theoretically-grounded psychological perspective. We believe that the adoption of a theoretical psychological perspective, as in the case of existing doping research, would favor a shift from insightful and yet seemingly endless debates to prospective research and intervention programs that could clarify and possibly resolve some of these debates. In the remaining sections of this paper, we attempt to sustain and justify this core belief. “.. . . the misuse of prescription drugs, other illicit drugs, or alcohol for the purpose of enhancing cognition, mood, or prosocial behavior in academic or work related contexts” (e.g., Maier and Schaub, 2015, p. THE PREVALENCE OF NEUROENHANCEMENT SUBSTANCE USE conduct policies that the use of prescription medications aimed at enhancing academic performance falls in the category of “academic dishonesty” (e.g., Duke University: Policy on academic dishonesty; URL: https://studentaffairs.duke.edu/conduct/z- policies/academic-dishonesty), even though policies of this sort are still a matter of debate (e.g., Schermer, 2008; Dubljević, 2013). Interestingly, Dodge et al. (2012) have separately assessed how individuals judge others who use performance enhancing drugs both in athletic and academic domains. Not surprisingly, their findings suggest that people tend to consider the use of NES to enhance academic performance as more acceptable than doping substance use in sport. A large number of recent empirical NE studies have estimated the prevalence of NES use. However, it seems difficult to draw a precise and reliable map of its diffusion, as prevalence estimates often vary widely depending upon sampling criteria, measurements, and demographic or contextual factors. For instance, Smith and Farah (2011), in reviewing 28 epidemiological studies on the prevalence of non-medical prescription drug use in American and Canadian students, reported a lifetime use of stimulants for non-medical purposes ranging from 5.3 to 55%. More recently, Franke et al. (2014) have reviewed studies reporting prevalence rates for NES use that range from 1 to 20%. One could reasonably argue that the lack of clear-cut norms and regulations for the use of NES makes the latter unfit for being treated as a case of cheating. Nonetheless, there are some actions or behaviors that, despite not being clear violations of explicit rules or norms, allow one to gain some advantages over others and, as such, might be considered unfair. In the sport context, these behaviors fall under the rubric of “gamesmanship” (e.g., Lee et al., 2007). According to Vallerand et al. (1996), in order to approach the ethical evaluations of a given behavior, one needs to recognize the social origins of these evaluations, that is, the notion that they emerge over time by consensus within a social context. How individuals perceive the misuse of substances has important implications for prevention efforts. Thus, the use of NES might be evaluated positively when the emphasis and judgment criteria focus on one’s effort to perform well, and negatively when the emphasis and judgment criteria focus on one’s attempt to increase one’s own academic performance through the help of pharmacological aids, thus altering the integrity and validity of (his or her) academic examinations and testing results. DEFINITIONAL AND CONCEPTUAL ISSUES CONCERNING NEUROENHANCEMENT Should NE be considered a complex property of some substances currently is a matter of debate, and the scientific evidence and general understanding of this proposition seems far from having been ascertained or confirmed (Zohny, 2015). Even if one departed from the definition of NE that in recent years has been shared by scholars and referred to as December 2015 \| Volume 6 \| Article 1880 Frontiers in Psychology \| www.frontiersin.org 2 A Social Cognitive Perspective for Neuroenhancement Zelli et al. ETHICAL AND MORAL ISSUES CONCERNING NEUROENHANCEMENT There is an important debate concerning the ethical issue related to the use of NES. Some scholars argue that, especially in the context of examinations, this behavior might be considered cheating, because its use may alter performance (Schermer, 2008), as in the well-known case of doping in sports. There are a number of parallels between the misuse of NES in academic settings and doping in sport. In both contexts, an individual is misusing a substance that has legitimate medical value with the purpose of increasing one’s own performance. As in the field of doping research (see, for istance, Petroczi, 2013), several scholars have debated the ethical and moral implications of using NES in academic or educational settings (e.g., Kipke, 2013; Zohny, 2015). At the same time, there are also some clear differences between the use of NES and the use of doping substances. In sport contexts, there is a clear and well-accepted distinction between which substances and protocols are illicit (illegal performance enhancing substances) and which are not (legal performance enhancing substances). In educational and academic contexts, at least until recently, law or binding regulations concerning the use of cognitive enhancing substances were lacking. Some universities, in fact, have recently clarified in their own academic There is an important debate concerning the ethical issue related to the use of NES. Some scholars argue that, especially in the context of examinations, this behavior might be considered cheating, because its use may alter performance (Schermer, 2008), as in the well-known case of doping in sports. There are a number of parallels between the misuse of NES in academic settings and doping in sport. In both contexts, an individual is misusing a substance that has legitimate medical value with the purpose of increasing one’s own performance. As in the field of doping research (see, for istance, Petroczi, 2013), several scholars have debated the ethical and moral implications of using NES in academic or educational settings (e.g., Kipke, 2013; Zohny, 2015). THE PREVALENCE OF NEUROENHANCEMENT SUBSTANCE USE The issue of reliably assessing prevalence rates has also characterized doping research, and the distinction between legal and illegal substances has definitely contributed to establishing valid estimates of doping use in sport settings (Mallia et al., 2013). In a similar fashion, it is plausible that the distinction among non- medical prescription drugs, drugs of abuse, and soft-enhancers (e.g., caffeine) in the NE literature might contribute to a correct assessment of prevalence estimates. Generally speaking, however, the estimation of prevalence of NES use, as for performance enhancing substances (PES) use in sport, remains a complex process and many methodological issues could influence it and lead to increasing variability and differences in findings across studies. For instance, while social desirability biases might easily come into play in the assessment of doping substance use in the face of explicit sport law regulations against their adoption, the lack of any clear-cut social or legal norms about NES may pose complex challenges for correct or agreed-upon prevalence rates. Faulmuller et al. (2013) emphasize that the indirect psychological costs of the use of NES is related to the ways people attribute performance to agents. Given that people tend to exaggerate the efficacy of cognitive enhancers, they might perceive NES users’ performance as not fully attributable to them. At any rate, individuals contemplating the use of NES may very well dwell upon the moral implications of using these substances and utilize their personal self-sanctions as internal deterrents. These possibilities imply and presuppose a strong link between NES use and moral reasoning, and this link is consistent with a well-grounded psychological literature addressing the relations between moral reasoning and the use of performance enhancing substances in sport-related contexts (e.g., Lucidi et al., 2008, 2013; Zelli et al., 2010). A SOCIAL COGNITIVE PERSPECTIVE ON NEUROENHANCEMENT At the same time, there are also some clear differences between the use of NES and the use of doping substances. In sport contexts, there is a clear and well-accepted distinction between which substances and protocols are illicit (illegal performance enhancing substances) and which are not (legal performance enhancing substances). In educational and academic contexts, at least until recently, law or binding regulations concerning the use of cognitive enhancing substances were lacking. Some universities, in fact, have recently clarified in their own academic The Theoretical Framework: Its General Principles and Hypotheses Frontiers in Psychology \| www.frontiersin.org A Social Cognitive Research Program for Neuroenhancement As an initial note, we believe that a social-cognitive model of NES use might nicely integrate some of the theoretical propositions that seem to have variously characterized recent NES studies. One proposition calls upon an incremental-functional view of NES use, and the hypothesis that students might be motivated and involved in performance enhancing practices that, over time, increasingly acquire high instrumental value (e.g., Sattler and Wiegel, 2013; Wolff and Brand, 2013). Another proposition calls upon belief systems which may build upon a link between one’s performance enhancement goals and the functional or moral implications of NES use as a purposive, goal-driven behavior. The focus on instrumental decisions also seems to characterize other NE research stressing the need for psychological theorizing (e.g., Wolff and Brand, 2013; Wolff et al., 2014). This research hypothesizes that NE is “. . .the medically unsupervised use of presumably psychoactive substances by healthy individuals who expect this substance to be a functional means of enhancing their cognitive capacity. . .” (Wolff et al., 2014, p. 2). This research very recently has moved on and utilized principles and constructs borrowed from occupational theories (e.g., demands, strain, burnout) to address the “means-to-end” NE hypothesis in educational settings (Wolff et al., 2014). This research has shown that the use of lifestyle drugs and prescribed NE drugs is more likely among university students who experience burnout, and that the use of NES worsens students’ psychological experience of academic demands and interferes with their motivational resources. These existing contributions have greatly contributed to the psychological discourse about NE. We also believe that, at least in educational settings, research attention to constructs such as (a) students’ attitudes about NES, (b) prospective intentions toward NES use, (c) efficacy and self- regulatory beliefs about one’s own capacity to counteract social and internal pressures to use NES, (d) personal standards and justifications in favor or against NES use, and (e) students’ appraisals of the self-relevance of interpersonal situations eliciting NES use would have high scientific value. It would acknowledge and be consistent with the above theoretical propositions, as these social cognitive constructs recognize and encompass the dynamic and functional properties of one’s life and behavioral experiences with NES that existing literature has highlighted. The Theoretical Framework: Its General Principles and Hypotheses From the previous sections of this contribution, it appears clear to us that the use of NES falls under the rubric of a goal-directed behavior and, as such, its scientific study may very well benefit from a psychological analysis presuming that NES use depends on self-regulation and on the mental processes intervening in behavioral intentions and decisions bounded to specific social December 2015 \| Volume 6 \| Article 1880 3 A Social Cognitive Perspective for Neuroenhancement Zelli et al. systems. Illustratively, this broad view has found clear and distinct expressions in research that variously adopted either a “theory of planned behavior” approach (e.g., Lucidi et al., 2004; Lazuras et al., 2010; Mallia et al., 2013), a motivational orientation approach (e.g., Barkoukis et al., 2013) or an explicit social- cognitive integrative approach (e.g., Lucidi et al., 2008, 2013; Zelli et al., 2010; Lazuras et al., 2015). All these cases typically refer to belief structures, and these beliefs may specifically refer to either outcome beliefs guiding one’s behavioral attitudes about doping use, behavioral control beliefs concerning the means for reaching one’s own goals, personal and self-regulatory efficacy beliefs, or moral disengagement beliefs that one may adopt to counteract personal self-sanctions against doping use (e.g., Lucidi et al., 2008, 2013; Lazuras, 2015). contexts or situations. So stated, our view endorses key tenets of a social cognitive perspective on NES use, insofar the latter “. . .. entails not only behavioral skill in self-managing environmental contingencies, but also the knowledge and the sense of personal agency to enact this skill in relevant contexts. Self-regulation refers to self-generated thoughts, feelings, and actions that are planned and cyclically adapted to the attainment of personal goals. . ..” (Zimmerman, 2000, pp. 13–14). contexts or situations. So stated, our view endorses key tenets of a social cognitive perspective on NES use, insofar the latter “. . .. entails not only behavioral skill in self-managing environmental contingencies, but also the knowledge and the sense of personal agency to enact this skill in relevant contexts. Self-regulation refers to self-generated thoughts, feelings, and actions that are planned and cyclically adapted to the attainment of personal goals. . ..” (Zimmerman, 2000, pp. 13–14). These general notions seem to be shared at least in part by psychological research that has adopted sociological and economic decision-making perspectives (see Sattler et al., 2014, for a thorough review). The Theoretical Framework: Its General Principles and Hypotheses Much of this research (e.g., Müller and Schumann, 2011; Sattler and Wiegel, 2013; Wolff and Brand, 2013) broadly argues that the use of (or willingness to use) cognitive enhancement substances reflects an instrumental decision individuals make on the basis of the degree to which substance use “fits” their personal preferences, perceived opportunities and constraints. Consistent with this general hypothesis, empirical studies have focused on several classes of variables, ranging from considerations about the risks and benefits of particular cognitive enhancement drugs’ characteristics (e.g., Castaldi et al., 2012), to forms of social environmental effects (e.g., forms of social control, social pressure from significant others) influencing decisions about the use of enhancement substances (e.g., Glannon, 2008; Bavarian et al., 2013) to, finally, personal characteristics (e.g., cognitive test anxiety, lack of academic competencies) that may make individuals more vulnerable at the time of deciding whether to use cognitive enhancement substances (e.g., Tice and Baumeister, 1997; Klassen et al., 2008; Weyandt et al., 2009). This belief-based social cognitive doping research has more recently been integrated by an additional social cognitive component, namely, one’s self-relevance appraisals of interpersonal and social situations eliciting doping use (Zelli et al., 2010, 2015). Theoretically, over time, this component would interact with belief systems in increasing the probability that people would show doping intentions and actual doping use. h h h l d l d f d This belief-based social cognitive doping research has more recently been integrated by an additional social cognitive component, namely, one’s self-relevance appraisals of interpersonal and social situations eliciting doping use (Zelli et al., 2010, 2015). Theoretically, over time, this component would interact with belief systems in increasing the probability that people would show doping intentions and actual doping use. We argue that the theoretical and empirical advances of doping research stand as a mature and plausible model for moving forward on NE research. In the following section, we describe, albeit in broad terms, some key elements of a possible social cognitive research program for the study of NE use. We argue that the theoretical and empirical advances of doping research stand as a mature and plausible model for moving forward on NE research. In the following section, we describe, albeit in broad terms, some key elements of a possible social cognitive research program for the study of NE use. Frontiers in Psychology \| www.frontiersin.org A Social Cognitive Research Program for Neuroenhancement More importantly, it would provide a single, unified, framework for theory development and assessment, allowing scholars to pursue a systematic analysis of the psychological processes that dynamically regulate students’ use of NES over time. This notwithstanding, it seems difficult to identify in this literature a common and explicit theoretical framework. On the contrary, and interestingly, doping-related psychological research has in recent years been able to adopt a broad social cognitive view that clearly and systematically put the study of performance enhancement substances on a qualitatively different level of theoretical analysis. According to this social cognitive view, doping substance use is a goal-directed behavior that is the expression of one’s intentional processes, and these intentions reflect the influence of socially construed belief December 2015 \| Volume 6 \| Article 1880 4 A Social Cognitive Perspective for Neuroenhancement Zelli et al. In our view, such a novel research focus should rely on and pursue some key research objectives. The first is concerned with the possibility of clearly establishing the empirical relations between people’s behavioral intentions and actual NES use. This first objective necessarily calls upon a second objective, namely, the adoption of longitudinal research designs allowing scholars to establish how behavioral intentions contribute to changes in NES use over time (i.e., controlling for behavioral stability). The third objective is concerned with the possibility of identifying the set of key social-cognitive variables regulating people’s NES behavioral intentions. As these variables operate in a system of dynamic relations, the empirical focus cannot merely address their unique contribution to behavioral NES intentions. Rather, it also needs to address how changes in the model of effects on behavioral intentions correspond to changes in the interrelations among key social cognitive variables and in their unique contributions. Consistent with a social-cognitive view of NES, the hypothesis of a system of interrelated variables influencing one’s behavioral intentions also calls upon the empirical possibility that this system is dynamically linked to the meaning people assign to relevant social and interpersonal situations possibly soliciting NES use. We believe this is a fourth critical objective for NES research, insofar as one’s intention to use NES might be strengthened or, alternatively, weakened by the degree to which social and interpersonal situations acquire personal relevance. REFERENCES Lazuras, L., Barkoukis, V., Rodafinos, A., and Tsorbatzoudis, H. (2010). Predictors of doping intentions in elite level athletes: a social cognition approach. J. Sport Exerc. Psychol. 32, 694–710. Barkoukis, V., Lazuras, L., Tsorbatzoudis, H., and Rodafinos, A. (2013). Motivational and social cognitive predictors of doping intentions in elite sports: an integrated approach. Scand. J. Med. Sci. Sports 23, e330–e340. doi: 10.1111/sms.12068 Lazuras, L., Barkoukis, V., and Tsorbatzoudis, H. (2015). Toward an integrative model of doping use: an empirical study with adolescent athletes. J. Sport Exerc. Psychol. 37, 37–50. doi: 10.1123/jsep.2013-0232 Bavarian, N., Flay, B. R., Ketcham, P. L., and Smit, E. (2013). Illicit use of prescription stimulants in a college student sample: a theory-guided analysis. Drug Alcohol Depend 132, 665–673. doi: 10.1016/j.drugalcdep.2013.04.024 Lucidi, F., Grano, C., Leone, L., Lombardo, C., and Pesce, C. (2004). Determinants of the intention to use doping substances: an empirical contribution in a sample of Italian adolescents. Int. J. Sport Psychol. 35, 133–148. Castaldi, S., Gelatti, U., Orizio, G., Hartung, U., Moreno-Londono, A. M., Nobile, M., et al. (2012). Use of cognitive enhancement medication among northern Italian university students. J. Addict. Med. 6, 112–117. doi: 10.1097/ADM.0b013e3182479584 Lucidi, F., Zelli, A., and Mallia, L. (2013). The contribution of moral disengagement to adolescents’ use of doping substances. Int. J. Sport Psychol. 44, 331–350. doi: 10.1037/a0034488 Lucidi, F., Zelli, A., Mallia, L., Grano, C., Russo. P. M., and Violani, C. (2008). The social-cognitive mechanisms regulating adolescents’ use of doping substances. J. Sports Sci. 26, 447–456. doi: 10.1080/02640410701579370 Dodge, T., Williams, K. J., Marzell, M., and Turrisi, R. (2012). Judging cheaters: is substance misuse viewed similarly in the athletic and academic domains? Psychol. Addict. Behav. 26, 678–682. doi: 10.1037/a0027872 Maier, L. J., and Schaub, M. P. (2015). The use of prescription drugs and drugs of abuse for neuroenhancement in Europe: not widespread but a reality. Eur. Psychol. 20, 155–166. doi: 10.1027/1016-9040/a000228 Dubljević, V. (2013). Prohibition or coffee shops: regulation of amphetamine and methylphenidate for enhancement use by healthy adults. Am. J. Bioeth. 13, 23–33. doi: 10.1080/15265161.2013.794875 Mallia, L., Lucidi, F., Zelli, A., and Violani, C. (2013). Doping attitudes and the use of legal and illegal performance-enhancing substances among Italian adolescents. J. Child Adolesc. Subst. Abuse 22, 179–190. doi: 10.1080/1067828X.2012.733579 Faulmuller, N., Maslen, H., and de Sio, F. S. (2013). The indirect psychological costs of cognitive enhancement. Am. J. Bioeth. 13, 45–47. doi: 10.1080/15265161.2013.794880 Müller, C. P., and Schumann, G. (2011). A Social Cognitive Research Program for Neuroenhancement As a concluding note, we firmly believe that the social cognitive research perspective that has been briefly outlined above can provide, whatever its findings might be, the specific contours for any educational program that is interested in effectively addressing NES use and its implications in people’s daily lives and experiences. AUTHOR CONTRIBUTIONS AZ, FL, and LM substantially have equally contributed to the development and preparation of the manuscript. Furthermore, all authors have approved the final version of the manuscript. Finally, the authors have agreed to be accountable for all aspects of the manuscript in ensuring that questions related to the accuracy or integrity of any part of it are appropriately investigated and resolved. REFERENCES The contribution of interpersonal appraisals to a social-cognitive analysis of adolescents’ doping use. Psychol. Sport Exerc. 11, 304–311. doi: 10.1016/j.psychsport.2010.02.008 Smith, M. E., and Farah, M. J. (2011). Are prescription stimulants smart pills? The epidemiology and cognitive neuroscience of prescription stimulant use by normal healthy individuals. Psychol. Bull. 137, 717–741. doi: 10.1037/ a0023825 Zelli, A., Mallia, L., and Lucidi, F. (2015). “‘I am not sure what you mean. . .’: the possible contribution of interpersonal appraisals to social-cognitive accounts of doping use,” in The Psychology of Doping in Sport, eds V. Brkoukis, L. Lazuras, and H. Tsorbatzoudis (Oxon: Routledge), 18–32. Tice, D. M., and Baumeister, R. F. (1997). Longitudinal study of procrastination, performance, stress, and health: the costs and benefits of dawdling. Psychol. Sci. 8, 454–458. doi: 10.1111/j.1467-9280.1997.tb00460.x Vallerand, R. J., Deshaies, P., Cuerrier, J.-P., Brière, N. M., and Pelletier, L. G. (1996). Toward a multidimensional definition of sportsmanship. J. Appl. Sport Psychol. 8, 89–101. Zimmerman, B. J. (2000). “Attaining self-regulation: a social cognitive perspective,” in Handbook of Self-Regulation, eds M. Boekaerts, P. R. Pintrich, and M. Zeidner (San Diego: Academic Press), 13–39. Weyandt, L. L., Janusis, G., Wilson, K. G., Verdi, G., Paquin, G., Lopes, J., et al. (2009). Nonmedical prescription stimulant use among a sample of college students: relationship with psychological variables. J. Atten. Disord. 13, 284–296. doi: 10.1177/1087054709342212 Zohny, H. (2015). The myth of cognitive enhancement drugs. Neuroethics 8, 257–269. doi: 10.1007/s12152-015-9232-9 Conflict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Wolff, W., and Brand, R. (2013). Subjective stressors in school and their relation to neuroenhancement: a behavioral perspective on students’ everyday life “doping”. Subst. Abuse Treat. Prev. Policy 8, 23. doi: 10.1186/1747- 597X-8-23 Copyright © 2015 Zelli, Lucidi and Mallia. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Wolff, W., Brand, R., Baumgarten, F., Lösel, J., and Ziegler, M. (2014). REFERENCES To use or not to use: expanding the view on non-addictive psychoactive drug consumption and its implications. Behav. Brain Sci. 34, 328–347. doi: 10.1017/S0140525X1100135X Franke, A. G., Bagusat, C., Rust, S., Engel, A., and Lieb, K. (2014). Substances used and prevalence rates of pharmacological cognitive enhancement among healthy subjects. Eur. Arch. Psychiatry Clin. Neurosci. 264, 83–90. doi: 10.1007/s00406- 014-0537-1 Ntoumanis, N., Ng, J. Y. Y., Barkoukis, V., and Backhouse, S. (2014). Personal and psychosocial predictors of doping use in physical activity settings: a meta-analysis. Sports Med. 44, 1603–162. doi: 10.1007/s40279-014- 0240-4 Glannon, W. (2008). Psychopharmacological enhancement. Neuroethics 1, 45–54. doi: 10.1007/s12152-008-9005-9 Pessoa, L. (2008). On the relationship between emotion and cognition. Nat. Rev. Neurosci. 9, 148–158. doi: 10.1038/nrn2317 Kipke, R. (2013). “What is cognitive enhancement and is it justified to point out this kind of enhancement within the ethical discussion?,” in Cognitive Enhancement: An Interdisciplinary Approach, eds E. Hildt and A. G. Franke (Dordrecht: Springer), 145–157. Petroczi, A. (2013). The doping mindset—part I: implications of the functional use theory on mental representations of doping. Perform. Enhanc. Health 2, 53–63. doi: 10.1016/j.peh.2014.06.001 Klassen, R. M., Krawchuk, L. L., and Rajani, S. (2008). Academic procrastination of undergraduates: low self-efficacy to self-regulate predicts higher levels of procrastination. Contemp. Educ. Psychol. 33, 915–931. doi: 10.1016/j.cedpsych.2007.07.001 Sattler, S., Mehlkop, G., Graeff, P., and Sauer, C. (2014). Evaluating the drivers of and obstacles to the willingness to use cognitive enhancement drugs: the influence of drug characteristics, social environment, and personal characteristics. Subst. Abuse Treat. Prev. Policy 9, 8. doi: 10.3109/10826084.2012.751426 Lee, M. J., Whitehead, J., and Ntoumanis, N. (2007). Development of the attitudes to moral decision-making in youth sport questionnaire (AMDYSQ). Psychol. Sport Exerc. 8, 369–392. doi: 10.1016/j.psychsport.2006.12.002 buse Treat. Prev. Policy 9, 8. doi: 10.3109/10826084.2012.751426 Sattler, S., and Wiegel, C. (2013). Test anxiety and cognitive enhancement: the influence of students’ worries on their use of performance-enhancing drugs. Subst. Use Misuse 48, 220–232. doi: 10.3109/10826084.2012.751426 Lazuras, L. (2015). “Social-cognitive predictors of doping use: an integrative approach,” in The Psychology of Doping in Sport, eds V. Brkoukis, L. Lazuras, and H. Tsorbatzoudis (Oxon: Routledge), 44–62. Schermer, M. (2008). On the argument that enhancement is “cheating”. J. Med. Ethics 34, 85–88. doi: 10.1136/jme.2006.019646 Frontiers in Psychology \| www.frontiersin.org December 2015 \| Volume 6 \| Article 1880 5 A Social Cognitive Perspective for Neuroenhancement Zelli et al. Zelli, A., Mallia, L., and Lucidi, F. (2010). Frontiers in Psychology \| www.frontiersin.org REFERENCES Modeling students’ instrumental (mis-) use of substances to enhance cognitive performance: neuroenhancement in the light of job demands- resources theory. Biopsychosoc. Med. 8, 12. doi: 10.1186/1751- 0759-8-12 Frontiers in Psychology \| www.frontiersin.org December 2015 \| Volume 6 \| Article 1880 6
https://openalex.org/W2060207753	https://lirias.kuleuven.be/bitstream/123456789/458831/2/KV-Molecules.pdf	English	null	Antifungal Plant Defensins: Mechanisms of Action and Production	Molecules/Molecules online/Molecules annual	2,014	cc-by	11,655	Kim Vriens 1, Bruno P. A. Cammue 1,2,* and Karin Thevissen 1 1 Centre of Microbial and Plant Genetics, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, Heverlee 3001, Belgium 2 Department of Plant Systems Biology, VIB, Technologiepark 927, Ghent 9052, Belgium 2 Department of Plant Systems Biology, VIB, Technologiepark 927, Ghent 9052, Belgium * Author to whom correspondence should be addressed; E-Mail: bruno.cammue@biw.vib-kuleuven.be; Tel.: +32-1632-9682; Fax: +32-1632-1966. Received: 18 July 2014; in revised form: 29 July 2014 / Accepted: 4 August 2014 / Published: 14 August 2014 Received: 18 July 2014; in revised form: 29 July 2014 / Accepted: 4 August 2014 / Published: 14 August 2014 Abstract: Plant defensins are small, cysteine-rich peptides that possess biological activity towards a broad range of organisms. Their activity is primarily directed against fungi, but bactericidal and insecticidal actions have also been reported. The mode of action of various antifungal plant defensins has been studied extensively during the last decades and several of their fungal targets have been identified to date. This review summarizes the mechanism of action of well-characterized antifungal plant defensins, including RsAFP2, MsDef1, MtDef4, NaD1 and Psd1, and points out the variety by which antifungal plant defensins affect microbial cell viability. Furthermore, this review summarizes production routes for plant defensins, either via heterologous expression or chemical synthesis. As plant defensins are generally considered non-toxic for plant and mammalian cells, they are regarded as attractive candidates for further development into novel antimicrobial agents. Keywords: mechanism of action; antimicrobial peptide; plant defensin; RsAFP2; NaD1; MsDef1; MtDef4; Psd1; heterologous protein expression; chemical protein synthesis Molecules 2014, 19, 12280-12303; doi:10.3390/molecules190812280 molecules ISSN 1420-3049 www.mdpi.com/journal/molecules Review Antifungal Plant Defensins: Mechanisms of Action and Production Kim Vriens 1, Bruno P. A. Cammue 1,2,* and Karin Thevissen 1 1 Centre of Microbial and Plant Genetics, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, Heverlee 3001, Belgium 2 Department of Plant Systems Biology, VIB, Technologiepark 927, Ghent 9052, Belgium * Author to whom correspondence should be addressed; E-Mail: bruno.cammue@biw.vib-kuleuven.be; Tel.: +32-1632-9682; Fax: +32-1632-1966. Received: 18 July 2014; in revised form: 29 July 2014 / Accepted: 4 August 2014 / Published: 14 August 2014 OPEN ACCESS Molecules 2014, 19, 12280-12303; doi:10.3390/molecules190812280 Antifungal Plant Defensins: Mechanisms of Action and Production Kim Vriens 1, Bruno P. A. Cammue 1,2,* and Karin Thevissen 1 Molecules 2014, 19 Molecules 2014, 19 AMPs can be constitutive in e.g., storage organs and reproductive tissues or can be induced systemically as well as locally, in e.g., leaves, during microbial invasion or injury [1–3]. Plant AMPs are small cationic peptides that exert biological activity against a broad range of organisms. Their activity is primarily directed against fungi, but bactericidal and insecticidal actions are also reported. These defence-related peptides have a compact structure that is stabilized by intramolecular disulphide bridges, enhancing structural and thermodynamic stability [2]. Based on their tertiary structure, they are subdivided into distinct classes, being thionins, defensins, knottins, lipid transfer proteins, heveins, snakins and cyclotides (reviewed in [2,4–7]). As the scope of this review is focused on plant defensins, the other classes of AMPs will not be discussed. For more detailed information on antifungal AMPs, the reader is referred to [8–10]. 2. Plant Defensins Plant defensins are present in all plant families, including the Brassicaceae, Fabaceae and Solanaceae. These peptides were primarily found in the seeds, but leaves and flowers are also common sources [11–16]. They are either constitutively expressed in storage and reproductive organs or produced upon pathogenic attack or injury as part of a systemic defence response [2]. In addition, production of plant defensins is also induced in response to environmental stress, such as drought [17,18], and signalling molecules, including methyl jasmonate, ethylene and salicylic acid [19,20]. 1. Introduction Like all living organisms, plants are repeatedly confronted with attacks by for instance insects, fungi and bacteria. In order to cope with these pests and pathogens, plants have developed a number of defence mechanisms, including the production of antimicrobial peptides (AMPs). Expression of these 12281 2.2. Biological Activity Plant defensins possess a variety of biological activities (reviewed in [27–29]). They have been reported to inhibit protein synthesis, enzyme activity and ion channel function. Some plant defensins even exhibit antiproliferative activity towards cancer cells or proved effective against HIV reverse transcriptase. To date, only a few plant defensins are shown to inhibit the growth of bacteria, whereas their antifungal activity has been studied extensively [27,28]. It has become clear that, although a similar activity might be observed for several defensins, their mode of action can be extremely diverse with regard to target molecules and (sub)cellular localization [30]. In order to illustrate this variety, this review will focus on the mode of antifungal action of plant defensins isolated from Raphanus sativus (RsAFP2), Pisum sativum (Psd1), Medicago spp. (MsDef1 and MtDef4) and Nicotiana alata (NaD1). Molecules 2014, 19 Molecules 2014, 19 12282 2.1. Structure Plant defensins are small, cationic peptides with a length of approximately 45–54 amino acids. Their structure typically comprises a cysteine-stabilized αβ-motif (CSαβ) with a prominent α-helix and a triple-stranded antiparallel β-sheet that is stabilized by four disulphide bridges [21–23]. A subclass of the plant defensin family comprises defensins with 10 cysteine residues, resulting in a total of five disulphide bonds. The fifth disulphide bond seems to reinforce a conserved hydrogen bond and is likely to confer additional thermodynamic stability of the defensin, as compared to other defensins, by replacing non-covalent hydrophobic interactions or hydrogen bonds with a covalent bond [24]. To our knowledge, this extra pair of cysteines has only been reported for PhD1 and PhD2, both floral defensins isolated from Petunia hybrida [16,24]. According to the structure of their precursor protein, plant defensins can be subdivided into two groups. A first group comprises defensins in which the precursor is composed of a signal sequence and a mature defensin domain. The signal sequence targets the protein to the endoplasmic reticulum, where it is folded and subsequently enters the secretory pathway. In a second and less common group, the precursor protein contains an additional C-terminal prodomain that is proteolytically removed during or after transit through the secretory pathway [16,25]. This type of defensins have been identified in solanaceous plants, such as Nicotiana alata and Petunia hybrida [16]. Recently, Lay and co-workers assigned a cytoprotective and subcellular targeting function to this prodomain [26]. 3. Mode of Action of Plant Defensins In addition, interaction of Sd5 and Psd1 with respectively fungal GlcCer- [42] and phosphatidylcholine- (PC) [43] containing vesicles further highlights the importance of fungal sphingolipids as interaction sites for plant defensins. Upon interaction with their target, plant defensins are either internalized by the fungal cell and interact with intracellular targets, or they stay at the cell surface and induce cell death through induction of a signalling cascade. While the latter has been reported for RsAFP2 [44], cellular uptake was observed for NaD1 [45,46], MtDef4 [47] and Psd1 [48]. Recently, Sagaram and colleagues identified a RGFRRR motif in the MtDef4 sequence that is thought to be a translocation signal required for fungal cell entry, since replacement of this motif with AAAARR or RGFRAA abolished the ability of the peptide to enter the cell [47]. However, this sequence is, to our knowledge, not present in other plant defensins that enter the fungal cell. This suggests multiple mechanisms by which defensins are internalized by the cell. Proposed mechanisms include receptor-mediated internalization, membrane translocation (i.e., transient membrane permeabilization and lipid-assisted pore formation) and membrane permeabilization (reviewed in [49]). Membrane permeabilization has been described for plant defensins, however, it is suggested to be a secondary effect rather than the key to microbial killing, since it only occurs at concentrations well above the concentration required for growth inhibition [50]. Alternatively, ROS production and hence oxidative stress, most often play a role in defensin-mediated cell death, as has been reported for many plant defensins including RsAFP2 [51], HsAFP1 [52], DmAMP1 [53] and NaD1 [45]. Another common feature of plant defensins is the loss of their antifungal activity by an increase in ionic strength of the growth medium. Especially divalent cations, such as Ca2+ and Mg2+, seem to play an important role in this phenomenon [11,13,15,37,54]. Since the antagonistic effect strongly depends on the test fungus and type of the defensin, it is suggested that electrostatic interactions alter the target site on the fungal membrane, and hence reduce the affinity of the defensin to bind the membrane, rather than altering the conformation of the defensin itself [55]. However, Oard and Karki proposed another mechanism for inhibition of antimicrobial activity by cations which is in contrast with this interpretation. 3. Mode of Action of Plant Defensins Extensive research during the past decades has allowed us to identify a number of key features by which plant defensins exert their antimicrobial activity. It has been demonstrated that defensins can specifically interact with host membrane compounds, such as bacterial lipid II receptors (reviewed in [31]), fungal sphingolipids (reviewed in [31]) and fungal phospholipids [32]. Fungal sphingolipids are classified into two groups, i.e., phosphosphingolipids and glycosphingolipids (GSLs). The most common GSL is glucosylceramide (GlcCer), which is synthesized by glucosylceramide synthase, encoded by the GCS gene, and glycoinositolphosphorylceramide (GIPC) [33,34]. The latter can be further subdivided into inositolphosphorylceramide (IPC), mannosyl-IPC (MIPC) and mannosyldi-IPC (M(IP)2C), which are formed by sequential addition of inositolphosphate and mannose. The final step in this reaction, i.e., converting MIPC to M(IP)2C, requires inositolphosphate transferase, encoded by the IPT1 gene (reviewed in [35]). Different plant defensins have been shown to interact with different classes of sphingolipids: the plant defensin RsAFP2 from radish [11] interacts with GlcCer [36], whereas the plant defensin DmAMP1 from dahlia [37] interacts with M(IP)2C [38,39]. In contrast, the plant defensins NaD1 from tobacco was recently shown to interact with a variety of phospholipids, including phosphatidylinositol mono-/bis-/tri-phosphates, phosphatidylserine and phospatidic acid, but not with sphingolipids [32]. In case of plant defensins that interact with sphingolipids, it was found that the presence of specific sphingolipids was essential mediating cell death of fungi and yeast, since yeast mutants deficient in genes involved in sphingolipid biosynthesis were resistant towards these peptides. For instance, S. cerevisiae strains with a non-functional IPT1 allele, and thus lacking M(IP)2C in their membranes, were found highly resistant towards DmAMP1 as compared to the wild type (WT) [38]. In line, deletion of the GSC gene resulted in an increased resistance towards RsAFP2 for Pichia pastoris and Candida albicans knockout strains as compared to the corresponding WTs [36]. Similar observations were made for MsDef1 and a Δgcs Fusarium graminearum strain [40]. Furthermore, Neurospora crassa mutants displaying structurally different GlcCer, novel glycosphingolipids and an altered level of steryl glucosides in their membranes were found more resistant towards RsAFP2, DmAMP1 and HsAFP1 when compared to the WT, suggesting that the specific structure of sphingolipids in the fungal membrane is crucial for sensitivity towards plant Molecules 2014, 19 12283 defensins [41]. 3. Mode of Action of Plant Defensins A schematic overview of the proposed mechanism of action of the plant defensins discussed in this review is given in Figure 2. A schematic overview of the proposed mechanism of action of the plant defensins discussed in this review is given in Figure 2. 3. Mode of Action of Plant Defensins They found that the structure of β-purothionin, a thionin purified from wheat, is altered by the presence of K+ and Mg2+, making the peptide more rigid and impairing interaction with its target. Structural changes in this case include elongation of α1-helix, unfolding of α2-helix and an overall change in loop conformation [56]. Inhibition of antimicrobial activity by the presence of cations seems to be a common theme for AMPs in general, since it is also observed for thionins, insect defensins and mammalian defensins and is not solely associated with antifungal activity [56–60]. In order to illustrate the variety of mechanisms of action by which plant defensins exhibit their antifungal activity, four case studies will be discussed in the next sections. The plant defensins discussed include RsAFP1 and RsAFP2 from radish [11], Psd1 from pea pods [13], MsDef1 from alfalfa [14] and MtDef4 from barrel clover [61], and NaD1 from tobacco [16] and are listed in Table 1. A multiple alignment of the amino acid sequences of these defensins is presented in Figure 1, in which the regions important for their antifungal activity are highlighted. Molecules 2014, 19 Molecules 2014, 19 Molecules 2014, 19 12284 Table 1. Overview of the plant defensins discussed in this review. NA: not available. Defensin Name Source UNIPROT Accession Number Protein Data Bank Accession Number Reference RsAFP1 Radish seeds P69241 1AYJ [11] RsAFP2 Radish seeds P30230 NA [11] MsDef1 Alfalfa seeds Q9FPM3 1H3R (theoretical model) [14] MtDef4 Barrel clover seeds G7L736 2LR3 [61] Psd1 Pea pods P81929 1JKZ [13] NaD1 Tobacco flowers Q8GTM0 1MR4 [62] Figure 1. Amino acid sequence alignment of RsAFP1, RsAFP2, MsDef1, MtDef4, Psd1 and NaD1. Multiple alignment was performed using the alignment tool from UniProt. Cysteine-pairing is shown at the top of the figure. Highly conserved residues are shown in grey; (-) denote gaps in the alignment. Arrows represent the position of the β-strands; the helix represents the position of the α-helix. Purple boxes indicate regions important for antifungal activity: boxes with dashed and full lines in the RsAFP1 sequence represent the first and second site in the tertiary structure, resp., important for antifungal activity [15,47,63]; blue boxes represent peptide fragments that exhibit antifungal activity similar to the parental peptide and hence, are important for antifungal activity [43,47,63,64]. A schematic overview of the proposed mechanism of action of the plant defensins discussed in this review is given in Figure 2. Molecules 2014, 19 Molecules 2014, 19 12286 These two regions might constitute two sites contacting a single receptor/interaction site or might indicate two binding sites that interact with two receptor/interaction sites [66]. The membrane target of RsAFP2 has been identified as fungal GlcCer [36] and since sphingolipids, such as GlcCers, are clustered with other membrane compounds to form lipid rafts [67], the hypothesis of RsAFP2-interaction with multiple interaction sites is plausible. Interaction of RsAFP2 with its membrane target is essential, however not sufficient for antifungal activity, as [Y38G]RsAFP2, an RsAFP2 variant devoid of antifungal activity, is able to interact with GlcCer [36,66]. Other RsAFP2-associated aspects leading to fungal cell death have been described. Ion fluxes take part in RsAFP2-induced cell death, since a rapid K+ efflux and Ca2+ influx was observed in RsAFP2-treated N. crassa hyphae [50]. Moreover, Ca2+ influx has been correlated with the antifungal potency of RsAFP2 [66], however, no blockage of L-type Ca2+ channels is observed [15]. In addition, production of ROS was shown in RsAFP2-treated C. albicans cells, suggesting a downstream signalling cascade of RsAFP2-binding to GlcCer [51]. In line with these findings, Aerts and colleagues demonstrated that RsAFP2 induces programmed cell death or apoptosis in C. albicans. Moreover, RsAFP2-induced apoptosis involves caspases, but not metacaspase Mca1 [68]. Cell wall stress has been associated with RsAFP2 activity as well. As GlcCers are also abundant in the cell wall (i.e., approx. 40% of GlcCer is located in the cell wall) [44], this is not surprising. RsAFP2 activates the cell wall integrity pathway by increasing the level of dually phosphorylated Mkc1p [44], a MAP kinase in C. albicans associated with oxidative stress, changes in osmotic pressure, cell wall damage and a decrease in growth temperature [69]. In line, RsAFP2 was shown to activate MAP kinase signalling cascades in F. graminearum, without involvement of the Hog1 MAP kinase pathway [61]. Furthermore, RsAFP2 was found to induce accumulation of membrane phytoC24-ceramides, affect septin formation and localization and impair the yeast-to-hyphae transition in C. albicans [44]. The exact binding site of RsAFP2 to GlcCer is still unknown, however, the region encompassing the β2-β3 loop is suggested to play a role in binding to the fungal membrane, as synthetic derivatives of this region, displayed in Figure 1, exhibited antifungal activity and binding to the membrane is essential to induce fungal cell death [64,66,70]. 3.1. Plant Defensins from Radish: RsAFP1 and RsAFP2 RsAFP1 and RsAFP2 are antifungal defensins found in the seeds of radish [11,36,65]. Regarding their antifungal activity, it was shown that RsAFP2 is more potent than RsAFP1 (2-30-fold dependent on the test fungus), although differences between the two peptides solely consist of two amino acid substitutions [11]. Analysis of the RsAFP2 primary structure showed two adjacent sites involved in antifungal action. 285 sAFP1 and , the reader sAFP1 and , the reader sAFP1 and s, the reader roposed mechanisms of action of the plant defensins discussed in this review. (A) Rs Def4; (E) NaD1. For detailed information on the Roman numerals displayed in the figures, (C), 3.4 (D) and 3.5 (E) of this review. ensins discussed in this review. (A) R Roman numerals displayed in the figures A) R osed mechanisms of action of the p ; (E) NaD1. For detailed information , 3.4 (D) and 3.5 (E) of this review. 3.2. Plant Defensin from Pea: Psd1 Psd1 is an antifungal defensin isolated from pea seeds [13]. Almeida and colleagues estimated other biological activities of Psd1 based on in silico analysis of the surface topology using various well-characterized defensins and toxins. Although no correlation could be demonstrated between antifungal or antibacterial activity and the surface electrostatic potential of the considered peptides, the study showed a clear correlation between K+ channel inhibitory activity and peptide surface topology. Since Psd1 shows a surface topology similar to that of peptides and toxins belonging to the latter group, it is therefore implied to act as a K+ channel inhibitor [71]. The fungal membrane target of Psd1 has not yet been identified, but is suggested to be GlcCer and/or ergosterol [43,72]. Analysis of the Psd1 lipid selectivity by fluorescence spectroscopy revealed that the peptide is likely to be adsorbed on or slightly inserted into the fungal membrane during initial interaction. Furthermore, this study showed that Psd1 does not interact with membranes containing GlcCer derived from soybean, nor with cholesterol-enriched lipid bilayers, such as mammalian cell membranes, which highlights its therapeutic potential [72]. The exact binding site for membrane interaction is thought to be Psd1 Loop1, displayed in Figure 1, since the peptide corresponding to this region still interacts with GlcCer-containing vesicles. Moreover, this region was found to exhibit significant conformational changes upon binding with GlcCer, as compared to the conformational accommodation during nonspecific binding to phosphocholine. Hence, interaction of Psd1 with the fungal membrane is probably due to conformational selection [43]. Upon membrane binding, Psd1 is internalized by the cell and interacts with nuclear proteins, as was shown via a GAL4-based yeast two-hybrid assay by Lobo and colleagues [48]. Cyclin F was identified as the main target of Psd1, which plays a key role in nuclear translocation of Cyclin B and cell cycle progression [48,73–75]. In vivo studies with Psd1 on retinal neuroblasts showed that Psd1 affects interkinetic nuclear migration and hence, impairs cell cycle progression [48]. Taken together, the mechanism of action of Psd1 is not yet completely understood, however, a proposed mode of action includes (i) adsorption on the fungal membrane surface; (ii) interaction with GlcCer and/or ergosterol in the fungal membrane; (iii) insertion of Psd1 into the fungal membrane; (iv) nuclear translocation and interaction with Cyclin F, resulting in cell cycle impairment and (v) fungal cell death. 3.2. Plant Defensin from Pea: Psd1 In which order step (iii) to (v) take place has yet to be identified. A schematic overview of the proposed mechanism of action of Psd1 is represented in Figure 2B. Molecules 2014, 19 Furthermore, position 38 and 39 were shown critical for antifungal action, as amino acid substitutions at these positions significantly altered the potency of the peptide in vitro [66]. Interestingly, [Y38G]RsAFP2, which is devoid of antifungal activity, is still able to interact with its membrane target, and hence, target binding and antifungal activity seem not controlled by the same peptide region [36]. Noteworthy is the fact that loop regions have been demonstrated to be important for antifungal activity in other plant defensins as well, including Psd1, MsDef1 and MtDef4 [15,43,63]. In conclusion, the mode of action of RsAFP2 is suggested to involve (i) recognition of and binding with GlcCer in the fungal membrane and cell wall; (ii) activation of the CWI pathway and MAP kinase signalling pathways, excluding the Hog1 MAP kinase pathway; (iii) production of ROS; (iv) induction of ion fluxes; (v) activation of caspases, but not metacaspase; (vi) induction of cell wall stress; (vii) accumulation of membrane phytoC24-ceramides; (viii) abnormal septin formation and localization; (ix) impairment of the yeast-to-hyphae transition and (x) fungal cell death. In which order step (ii) to (ix) take place, has yet to be elucidated. A schematic overview of the proposed mechanism of action of RsAFP2 can be found in Figure 2A. Molecules 2014, 19 12287 3.3. Plant Defensins from Medicago spp.: MsDef1 and MtDef4 MsDef1 and MtDef4 are antifungal defensins found in Medicago spp. [14,61]. MsDef1 is suggested to display a similar mode of antifungal action as the virally encoded toxin KP4 from the fungus Ustilago maydis, since both peptides strongly block the mammalian L-type Ca2+ channel in a specific manner [15]. Interaction of toxins with K+ channels has been reported by Zhu and colleagues, who hypothesized that defensins with a Lys-Cys4-Xaa-Asn motif interact with K+ channels when the flexible loop of the peptide is deleted and hence, steric hindrance is reduced [76]. The same might hold true for other ion channels, such as Ca2+ channels. Similarities in modes of action between KP4 and MsDef1 are further demonstrated by a more pronounced hyperbranching of fungal hyphae upon Molecules 2014, 19 12288 treatment with these peptides as compared to treatment with RsAFP2 [15]. Moreover, the antifungal activity of MsDef1 and KP4 is strongly abrogated by addition of exogenous Ca2+ and since addition of other metals (K+, Mg2+ and Na+) did not affect peptide activity, Ca2+ is suggested to be involved in MsDef1 mode of action [15]. Recently, Muñoz and colleagues confirmed that MsDef1, as well as MtDef4, disrupt Ca2+ signalling and/or homeostasis and that this phenomenon is not caused by direct membrane permeabilization [77]. GCS was found essential in MsDef1-mediated growth inhibition, which implies the involvement of GlcCer as a fungal membrane target [40]. This is in line with recent findings by Muñoz and colleagues who concluded that MsDef1-sensitivity is mediated by GlcCer in filamentous fungi [77]. Molecular targets involved in MsDef1-tolerance include MAP kinase signalling cascades, with exception of the Hog1 MAP kinase pathway, and were similar to the observations made for RsAFP2. Immunoblot analysis revealed rapid activation of Gpmk1 and Mgv1 MAP kinases upon MsDef1 treatment [61], which regulate processes related to cell wall integrity, sexual reproduction and pathogenicity [78,79]. Interestingly, RsAFP2 activates the cell wall integrity pathway and exerts its antifungal activity from the extracellular space [44]. Since MsDef1 and RsAFP2 seem to activate the same MAP kinase signalling cascades [61], they are suggested to have a similar mode of action. Hence, MsDef1 is suggested to induce fungal cell death through activation of signalling cascades involving MAP kinases without entering the fungal cell. In contrast to MsDef1 activity, GCS is not important for MtDef4-mediated fungal growth inhibition and MtDef4 activity seems independent of MAP kinase signalling cascades. 3.3. Plant Defensins from Medicago spp.: MsDef1 and MtDef4 Furthermore, MsDef1 induces extensive hyperbranching of fungal hyphae, whereas MtDef4 does not, suggesting different mechanisms of action for MsDef1 and MtDef4 [61]. Both peptides were shown to induce membrane permeabilization, however, membrane permeabilization was significantly higher in hyphae treated with MtDef4 as compared to treatment with MsDef1, which is consistent with the in vitro antifungal potency of the peptides [63]. Sagaram and colleagues reported internalization of MtDef4 in the fungal cell and identified the RGFRRR motif in the MtDef4 sequence as a translocation signal that is required for fungal cell entry. In addition, MtDef4 was shown to interact with cytosolic phosphatidic acid (PA), dependent on the presence of the RGFRRR motif [47]. These results are consistent with previous findings that correlate the antifungal activity of MtDef4 with the presence of that motif [63]. Recently, Muñoz and colleagues reported that MsDef1 and MtDef4 affect conidial germination, conidial anastomosis tube fusion and conidial cell death in N. crassa in significantly different ways. High fungicidal concentrations of MtDef4 caused rapid and complete inhibition of germination, with cell death occurring rather fast, whereas treatment of conidia with high MsDef1 concentrations resulted in a delayed cell death and complete inhibition of germination was not reached. Moreover, it was shown that the RGFRRR motif in MtDef4 is not only important for cell entry and binding to PA, but it furthermore plays a role in inhibition of cell fusion [77]. As is the case for various plant defensins, structure-activity studies have been performed for MsDef1 and MtDef4. Both the N-terminal and the C-terminal region of MsDef1, displayed in Figure 1, were found important for antifungal activity and, in line with the observations made for RsAFP2, position 38 is critical for antifungal action [15,66]. The MsDef1 and MtDef4 C-terminal domains are also suggested to play a role in Ca2+ homeostasis [77]. Furthermore, Spelbrink and colleagues demonstrated the importance of the loops for antifungal activity in the MsDef1 tertiary structure, 12289 Molecules 2014, 19 which has been reported for Psd1, RsAFP2 and MtDef4 as well [15,43,63,66,70]. Structure-activity determinants of MtDef4 demonstrated the importance of the γ-core motif, composed of β2 and β3 strands and the interposed loop, for antifungal activity. Here, cationic and hydrophobic residues are considered important for antifungal action, as the MtDef4 γ-core alone is able to inhibit fungal growth, whereas the γ-core of MsDef1 is not. 3.3. Plant Defensins from Medicago spp.: MsDef1 and MtDef4 Both F37 and R38 are critical for antifungal activity in MtDef4, since the hexapeptide RGFRRR present in the MtDef4 γ-core is capable of inducing growth inhibition and membrane permeabilization, while RGARRR and RGFARR are not [63]. In addition, replacement of RGFRRR with RGFRAA or AAAARR abolished the ability of the peptide to enter the fungal cell as well as to interact with intracellular PA, suggesting that both fungal cell entry and PA binding are mediated by the RGFRRR loop [47], again highlighting the importance of peptide loops for antifungal activity. In conclusion, MsDef1 and MtDef4 seem to have different mechanisms of antifungal action, however, their complete mode of action remains to be elucidated. The proposed mechanism of antifungal action for MsDef1 includes: (i) interaction with the fungal membrane, presumably GlcCer; (ii) activation of Gpmk1 and Mgv1 MAP kinase signalling cascades; (iii) (partial) inhibition of conidial germination; (iv) disruption of Ca2+ signalling and/or homeostasis and (v) delayed fungal cell death. In which order step (ii) to (v) take place has yet to be identified. The hypothesized mechanism of action of MtDef4 comprises (i) recognition of the fungal membrane; (ii) translocation of the peptide to the cytosol via the RGFRRR motif; (iii) interaction with cytosolic PA and supposedly subsequent interference with PA signalling and/or biosynthesis; (iv) disruption of Ca2+ signalling and/or homeostasis; (v) inhibition of conidial germination; (vi) inhibition of cell fusion and (vii) rapid fungal cell death. In which order (ii) to (vii) take place remains to be elucidated. A schematic overview of the proposed mechanism of action of MsDef1 and MtDef4 is given in Figure 2C,D, respectively. Molecules 2014, 19 Molecules 2014, 19 12290 whereas other MAP kinase signalling cascades play a key role in tolerance to these defensins [61]. These findings clearly suggest distinct modes of action of MsDef1 and RsAFP2 on the one hand, and NaD1 on the other. Another novel finding in the NaD1 mechanism of action is the identification of Agp2p as a regulator of the potency of the peptide [81]. Agp2p is a plasma membrane protein that regulates the transport of positively charged molecules. Upon NaD1 treatment, cells lacking AGP2 show a delayed membrane permeabilization, reduced uptake of NaD1 and are overall more resistant to NaD1 treatment compared to the WT. Deletion of AGP2 probably results in an accumulation of positive charges on the surface of the cell, thereby repelling cationic peptides from the surface [81]. Taken together, the NaD1 mechanism of action includes (i) interaction with the fungal cell membrane; (ii) translocation to the cytoplasm; (iii) PIP2-mediated oligomerization of NaD1 (14-mer) (iv) membrane permeabilization; (v) possible interaction with intracellular targets; (vi) ROS and NO production, i.e., oxidative stress; (vii) activation of the HOG pathway; (viii) membrane disruption and (ix) fungal cell death. In which order (ii) to (ix) take place has yet to be identified. A schematic overview of the proposed mechanism of action of NaD1 is shown in Figure 2E. 4. Production of Plant Defensins Due to their selective toxicity towards microbial cells and their unique mode of action, plant defensins are attractive candidates for further development as novel antimicrobial agents. However, development of plant defensins for medicinal or biotech purposes requires large amounts of purified peptides. Extraction of plant defensins from natural sources is rather complicated due to their low abundance and the presence of a variety of other compounds in these plant parts. Chemical synthesis and heterologous production are therefore convenient alternatives to obtain large amounts of functional peptides. In addition, these approaches allow for the production of mutant peptides, which are interesting to include in structure-activity studies. In the following part, these techniques as well as their advantages and drawbacks are discussed. 3.4. Plant Defensin from Tobacco: NaD1 NaD1 is a floral defensin from the tobacco plant and exhibits antifungal properties [16]. The membrane target of NaD1 was recently identified by Poon and colleagues as the phospholipid PIP2, which is present in eukaryotic cell membranes. It was shown that NaD1 forms 14-mer oligomers, mediated by PIP2, and that this oligomerization is important for membrane permeabilization and lysis of the fungal cell [32]. Upon interaction and permeabilization of the cell membrane, NaD1 enters fungal hyphae and is localized to the cytoplasm, where it causes granulation of the cytoplasm and induces ROS production [45]. NaD1 is implied to induce cell death via oxidative stress, as ROS and nitric oxide (NO) production was observed upon treatment of C. albicans cells [80]. These findings are consistent with results reported by Bleackley and colleagues, in which mitochondrial genes are implicated in NaD1 mode of action [81]. Recently, Hayes and colleagues reported the importance of the high-osmolarity glycerol (HOG) pathway in tolerance to NaD1, being the sole stress-responsive pathway involved in NaD1 action that was screened in this study [80]. Although the HOG pathway is mainly involved in protection against osmotic stress and osmotic stress does not contribute significantly to NaD1 mode of action [80], these findings are not surprising, since Hog1p is also known to participate in tolerance to oxidative stress [82]. Interestingly, the HOG pathway is excluded in the mode of action of MsDef1 and RsAFP2, Molecules 2014, 19 12291 Molecules 2014, 19 NCL. Subsequent cycles of NCL results in the formation of a linear polypeptide chain [83,90,91]. A major drawback of this technique is the mandatory use of an N-terminal cysteine residue, as cysteines are seldom conveniently distributed throughout a peptide sequence. Furthermore, the polypeptide is generated from the C-terminus towards the N-terminus and not vice versa [85]. Another approach to couple two peptides includes enzyme-assisted ligation, in which enzymes are used as catalysts to promote peptide bond formation. This approach complements chemical ligation strategies and has great significance since these enzymes are naturally involved in protein modifications in vivo and are rated nontoxic, whereas many chemicals employed in NCL are unfavourable for food applications of the peptides. Nonetheless, the use of enzymes in protein crosslinking is still in its infancy and further research is essential to improve enzyme-assisted peptide ligation (extensively reviewed in [84]). Proteins often require post-translational and conformational modifications in order to render biological activity. As post-translational modifications cannot be provided via chemical synthesis, a strategy termed expressed protein ligation (EPL) or intein-mediated protein ligation (IPL) is employed. Following this method, semisynthetic peptides are created by fusion of recombinant peptide fragments to synthetic peptide fragments. Since post-translational modifications mainly occur at the termini of peptides, EPL is a plausible approach to obtain functional peptides [91,92]. When a disulphide bond pattern is essential, (re-)folding of the protein is advised and oxidative folding is performed [93–96]. In a first attempt, oxidative folding is often employed in a direct manner, i.e., a one-step oxidative folding procedure, as it allows for a spontaneous fold in which the protein is energy-stable and assumed to acquire its native conformation. In addition, it is less expensive and time-consuming as compared to regioselective oxidative folding in which individual cysteine pairs are deprotected and oxidized sequentially to allow subsequent formation of disulphide bonds [97]. Functional peptides can be synthesized following SPPS and NCL methods, as was reported for conotoxins, snakins and defensins [94,95,97–99]. For instance, the insect defensin lucifensin, synthesized by Fmoc-SPPS and folded using a one-step oxidative folding technique, showed biological activity against Bacillus subtilis, Micrococcus luteus and Staphylococcus aureus with MIC-values of 1.2 µM, 0.6 µM and 41 µM, resp., whereas linear unfolded lucifensin and lucifensin analogues folded by 1 out of 3 disulphide bridges were inactive (MIC > 100 µM) [99]. 4.1. Chemical Synthesis of Proteins Synthesis of proteins by chemical means has recently gained interest, as it allows the generation of proteins that cannot be produced biologically, e.g., labelled peptides. Various strategies have been developed in which proteins can be synthesized, often consisting of combinations of solid-phase peptide synthesis (SPPS), native chemical ligation (NCL) and enzyme-catalyzed ligation (reviewed in [83–86]). In SPPS, a peptide is synthesized in a stepwise manner on a polymeric resin through sequential steps of coupling and deprotection of protected amino acids. Both Fmoc (9-fluorenylmethyl carbamate) and Boc (di-tert-butyl dicarbonate) strategies can be used in SPPS to protect the N-terminus of the amino acid being coupled and hence prevent polymerization or non-specific reactions (reviewed in [87–89]. SPPS plays a key role in peptide synthesis, however, studies have shown that following this method, only peptides containing less than 50 amino acids can be reliably prepared with acceptable yields and purity [83]. Hence, other strategies have to be implemented when synthesizing larger peptides and proteins. To this end, a strategy of peptide segment condensation is used in which unprotected peptide fragments, often produced by SPPS, are subjected to ligation methods such as 4.2.1. Heterologous Expression of Proteins in E. coli 4.2.1. Heterologous Expression of Proteins in E. coli Protein expression in E. coli is relatively simple and inexpensive and the variety of available plasmids, fusion partners and strains makes it often the preferred method for production of AMPs [103,104]. However, major drawbacks have been reported using bacteria for effective AMP production as discussed below. First of all, the recombinant protein often needs to be fused to a carrier protein to neutralize its toxicity towards the host and to increase its solubility to avoid formation of inclusion bodies [105]. This fusion partner needs to be released during or after purification of the protein of interest via enzymatic or chemical cleavage to render functional proteins, which results in a decreased yield [102]. In addition, fusion proteins are not necessarily properly folded and production of these proteins can result in so-called “soluble inclusion bodies” [105]. Recently, it was shown that co-expression of the human Quiescin Sulfhydryl Oxidase (QSOX), a chaperone with thiol/disulphide oxidase activity, in the cytoplasm of E. coli can counter protein misfolding and increase the yield of soluble cysteine-rich proteins [106,107]. Although such approach improves protein production in E. coli, other obstacles remain. Direct secretion mechanisms are not present in E. coli strains used for recombinant protein production, which complicates protein purification. Protein secretion can be obtained, however, by destabilization of the E. coli structural components or by using leaky strains that lack certain structural components or mutant strains in which secretion modules derived from pathogenic E. coli or other species are incorporated [103,108–111]. Finally, and most importantly, the protein of interest often requires complex folding, including the formation of multiple disulphide bonds and/or glycosylation. In both cases, an eukaryotic system is preferred [112]. In addition, Puertas and colleagues reported that the protein yield, when using E. coli as a host for recombinant production, is inversely proportional to the cysteine content of the protein [113]. This indicates the need for other expression systems when producing proteins with a high cysteine content, such as plant defensins. Yet, production of functional defensins in E. coli has been reported. For instance, using E. coli both the spruce defensin PgD5 and the Scots pine defensin PsDef1 were produced while displaying a high antifungal activity [114,115]. In line, functional potato snakin-1 (SN1) and defensin-1 (PTH1) were generated using E. coli as a host [116]. Molecules 2014, 19 Furthermore, human β-defensin 4 (HBD4) and HBD4 analogues, synthesized using Fmoc-SPPS and folded employing a three-step oxidative folding procedure, showed antimicrobial activity against Escherichia coli, Pseudomonas aerigunosa, S. aureus and C. albicans. However, only the completely folded peptide showed a similar or a 2-fold decreased activity, depending on the test organism, as commercially available HBD4 [98]. Although it seems that plant defensins can be produced by chemical means, multidisulphide-containing peptides are not always successfully produced, as multiple isoforms are generated during folding [96]. In addition, chemical synthesis is rather expensive due to a high cost of reagents, and peptide aggregation and formation of by-products renders this method often unfavourable [100,101]. These observations highlight the need for other systems to generate functional proteins, as will be discussed in the next section. Molecules 2014, 19 12292 4.2. Heterologous Expression of Proteins Heterologous expression of proteins is a widely used technique and different expression systems have been reported to date. The main host systems used for recombinant production of AMPs include Escherichia coli and Pichia pastoris (reviewed in [102]). 4.2.2. Heterologous Expression of Proteins in P. pastoris Yeasts are largely used for production of recombinant proteins due to their eukaryotic nature. Unlike bacteria, yeasts have the ability to implement many post-translation modifications, including disulphide bond formation, glycosylation and processing of signal sequences. These features make them attractive hosts for AMP production. Recently, the yeast P. pastoris has gained interest as a host for protein production for several reasons, as discussed below [120–126]. Firstly, P. pastoris displays a high growth rate and allows for high cell densities to be reached, resulting in a higher protein yield as compared to yields obtained with other eukaryotic systems (reviewed in [102]). P. pastoris is of particular interest for large-scale productions of recombinant proteins, since the growth media are cheap, universal and well defined. Furthermore, when handling fermenter setups in which pH, aeration, feed rate, etc. are controlled, P. pastoris can easily grow to ultra-high cell densities, which in turn leads to an increased protein yield [127]. Further enhancement of the protein yield can be obtained by using multicopy transformants during protein production (reviewed in [128]). Multiple copies of the plasmid are often incorporated in the Pichia genome through crossover events and are integrated in a head-to-tail manner at the same locus. When using proper plasmids such as pPICZ plasmids containing a ZeocinTM resistance gene, multicopy transformants are selected in a straightforward manner by modulating the antibiotic concentration and screening for an increased antibiotic resistance [129]. Thirdly, protein production in P. pastoris can be initiated by exogenous addition of inducing agents (reviewed in [126]). Initiation of protein production at any given time point is an asset, since biomass generation, and hence protein yield, is not affected by the potential toxicity of the protein towards the host [127]. Finally, proteins produced by P. pastoris are easily exported to the culture medium using signal sequences, such as the S. cerevisiae α-factor sequence, which facilitates downstream processing [122,126]. Other signal sequences that direct the protein of interest to the secretory pathway are reviewed by Ahmad and colleagues [126]. A minor drawback associated with the use of the α-factor secretion signal is the presence of protein isoforms in which additional amino acids are incorporated at the N-terminus of the protein due to incomplete processing of the STE13 protease [126,130]. 4.2.1. Heterologous Expression of Proteins in E. coli Nevertheless, peptides produced in eukaryotic systems are often more active, i.e., characterized by a lower MIC-value, as compared to peptides produced in prokaryotic systems. The latter is possibly due to structural defects or misfolding. For instance, Kant and co-workers found that the corn defensin PDC1 exhibited a 2-fold higher antifungal activity when produced in P. pastoris as compared to its production in E. coli. In addition, Fourier transform infrared spectroscopy (FTIR) revealed more β-sheets and less random structures when PDC1 was produced in P. pastoris [117]. Similar observations were made for other proteins. Both human adiponectin and alkaline phosphatase from 12293 Molecules 2014, 19 archaea were found more active when produced in P. pastoris as compared to their counterparts in E. coli [118,119]. Hence, these observations highlight the advantages of using P. pastoris for generation of functional and properly folded proteins. 5. Conclusions Plant defensins are interesting candidates for use in medicinal and biotech purposes and can be produced via heterologous expression in eukaryotic hosts. PDFs are generally considered non-toxic to plant and mammalian cells and have distinct modes of action, involving specific interactions with the cell surface [39]. They are therefore suggested to have great therapeutic potential, however, literature falls short on studies reporting the in vivo performance of PDFs in an animal model. To our knowledge, only RsAFP2 has been reported to show in vivo efficacy in a murine candidiasis model upon intravenous administration [65]. In addition, the mechanisms of action of various PDFs have not yet been identified or are not yet fully understood. Hence, further research is needed to demonstrate the therapeutic potential of PDFs and to elucidate their mechanisms of action. 4.2.2. Heterologous Expression of Proteins in P. pastoris A non-native N-terminus can influence the biological activity of the protein, as was reported for the pea defensin Psd1 [130,131] and the shrimp AMP Ch-penaeidin [132], and is therefore inadmissible. Addition of an alanine or protease cleavage site at the N-terminus of the protein is recommended as it allows for successful cleavage of the signal sequence [126,130,133,134]. P. pastoris has been successfully used for the production of AMPs, including hPAB-β, a variant of human β-defensin, and shrimp Ch-penaeidin [132,135]. Likewise, defensins from pea, tomato, mungbean, Mexican turnip, corn, tobacco, radish, alfalfa and barrel clover were successfully produced in P. pastoris [15,40,117,130,133,136–138]. These observations highlight the ease of using P. pastoris for production of AMPs, and more specifically, plant defensins. Molecules 2014, 19 Molecules 2014, 19 Molecules 2014, 19 12294 Acknowledgments This work was supported by funds from FWO-Vlaanderen (to B.P.A.C. and K.T.) and from Industrial Research Fund, KU Leuven (IOF/KP/12/002). K.T. and K.V. acknowledge the receipt of a postdoctoral grant from Industrial Research Fund, KU Leuven, and a predoctoral grant from IWT-Vlaanderen, respectively. Conflicts of Interest The authors declare no conflict of interest. The authors declare no conflict of interest. References wles, D.J. Defense-related proteins in higher plants. Annu. Rev. Biochem. 1990, 59, 873– 2. Broekaert, W.F.; Cammue, B.P.A.; de Bolle, M.F.C.; Thevissen, K.; de Samblanx, G.W.; Osborn, R.W.; Nielson, K. Antimicrobial Peptides from Plants. Crit. Rev. Plant Sci. 1997, 16, 297–323. 2. Broekaert, W.F.; Cammue, B.P.A.; de Bolle, M.F.C.; Thevissen, K.; de Samblanx, G.W.; Osborn, R.W.; Nielson, K. Antimicrobial Peptides from Plants. Crit. Rev. Plant Sci. 1997, 16, 297–323. 3. Heil, M.; Bostock, R.M. Induced systemic resistance (ISR) against pathogens in the context of induced plant defences. Ann. Bot. 2002, 89, 503–512. 3. Heil, M.; Bostock, R.M. Induced systemic resistance (ISR) against pathogens in the context of induced plant defences. Ann. Bot. 2002, 89, 503–512. 4. Ts Egorov, A.; Odintsova, T.I. Defense peptides of plant immune system. Bioorg. Khim. 2012, 38, 7–17. 4. Ts Egorov, A.; Odintsova, T.I. Defense peptides of plant immune system. Bioorg. Khim. 2012, 38, 7–17. 5. Pelegrini, P.B.; del Sarto, R.P.; Silva, O.N.; Franco, O.L.; Grossi-de-Sa, M.F. Antibacterial peptides from plants: What they are and how they probably work. Biochem. Res. Int. 2011, 2011, doi:10.1155/2011/250349. 6. Stotz, H.U.; Waller, F.; Wang, K. Innate Immunity in Plants: The Role of Antimicrobial Peptides. In Antimicrobial Peptides and Innate Immunity; Hiemstra, P.S., Zaat, S.A.J., Eds.; Springer: Basel, Switzerland, 2013; pp. 29–51. 7. Wong, J.H.; Ng, T.B., Cheung, R.C.; Ye, X.J.; Wang, H.X.; Lam, S.K.; Lin, P.; Chan, Y.S.; Fang, E.F.; Ngai, P.H.; et al. Proteins with antifungal properties and other medicinal applications from plants and mushrooms. Appl. Microbiol. Biotechnol. 2010, 87, 1221–1235. 8. Van der Weerden, N.L.; Bleackley, M.R.; Anderson, M.A. Properties and mechanisms of action of naturally occurring antifungal peptides. Cell Mol. Life Sci. 2013, 70, 3545–3570. Molecules 2014, 19 12295 9. Hegedus, N.; Marx, F. Antifungal proteins: More than antimicrobials? Fungal Biol. Rev. 2013, 26, 132–145. 9. Hegedus, N.; Marx, F. Antifungal proteins: More than antimicrobials? Fungal Biol. Rev. 2013, 26, 132–145. 10. De Cândido, E.S.; Cardoso, M.H.S.; Sousa, D.A.; Viana, J.C.; de Oliveira-Júnior, N.G.; Miranda, V.; Franco, O.L. The use of versatile plant antimicrobial peptides in agribusiness and human health. Peptides 2014, 55, 65–78. 11. Terras, F.R.; Schoofs, H.M.; de Bolle, M.F.; van Leuven, F.; Rees, S.B.; Vanderleyden, J.; Cammue, B.P.; Broekaert, W.F. Analysis of two novel classes of plant antifungal proteins from radish (Raphanus sativus L.) seeds. J. Biol. Chem. 1992, 267, 15301–15309. 12. References Structure of Petunia hybrida defensin 1, a novel plant defensin with five disulfide bonds. Biochemistry 2003, 42, 8214–8222. 24. Janssen, B.J.; Schirra, H.J.; Lay, F.T.; Anderson, M.A.; Craik, D.J. Structure of Petunia hybrida defensin 1, a novel plant defensin with five disulfide bonds. Biochemistry 2003, 42, 8214–8222. 25. Lay, F.T.; Anderson, M.A. Defensins--components of the innate immune system in plants. Curr. Protein Pept. Sci. 2005, 6, 85–101. 26. Lay, F.T.; Poon, S.; McKenna, J.A.; Connelly, A.A.; Barbeta, B.L.; McGinness, B.S.; Fox, J.L.; Daly, N.L.; Craik, D.J.; Heath, R.L.; et al. The C-terminal propeptide of a plant defensin confers cytoprotective and subcellular targeting functions. BMC Plant Biol. 2014, 14, 1471–2229. 27. Carvalho Ade, O.; Gomes, V.M. Plant defensins—Prospects for the biological functions and biotechnological properties. Peptides 2009, 30, 1007–1020. 28. Carvalho Ade, O.; Gomes, V.M. Plant defensins and defensin-like peptides—Biological activities and biotechnological applications. Curr. Pharm. Des. 2011, 17, 4270–4293. 29. De Coninck, B.; Cammue, B.P.A.; Thevissen, K. Modes of antifungal action and in planta functions of plant defensins and defensin-like peptides. Fungal Biol. Rev. 2013, 26, 109–120. 30. Van der Weerden, N.L.; Anderson, M.A. Plant defensins: Common fold, multiple functions. Fungal Biol. Rev. 2013, 26, 121–131. 31. Wilmes, M.; Cammue, B.P.; Sahl, H.G.; Thevissen, K. Antibiotic activities of host defense peptides: More to it than lipid bilayer perturbation. Nat. Prod. Rep. 2011, 28, 1350–1358. 32. Poon, I.; Baxter, A.A.; Lay, F.T.; Mills, G.D.; Adda, C.G.; Payne, J.A.; Phan, T.K.; Ryan, G.F.; White, J.A.; Veneer, P.K.; et al. Phosphoinositide-mediated oligomerization of a defensin induces cell lysis. eLife 2014, 3, e01808. 33. Del Poeta, M.; Nimrichter, L.; Rodrigues, M.L.; Luberto, C. Synthesis and biological properties of fungal glucosylceramide. PLoS Pathog. 2014, 10, 9. 34. Merrill, A.H., Jr.; Schmelz, E.M.; Dillehay, D.L.; Spiegel, S.; Shayman, J.A.; Schroeder, J.J.; Riley, R.T.; Voss, K.A.; Wang, E. Sphingolipids--the enigmatic lipid class: Biochemistry, physiology, and pathophysiology. Toxicol. Appl. Pharmacol. 1997, 142, 208–225. 35. Montefusco, D.J.; Matmati, N.; Hannun, Y.A. The yeast sphingolipid signaling landscape. Chem. Phys. Lipids 2014, 177, 26–40. 36. Thevissen, K.; Warnecke, D.C.; Francois, I.E.; Leipelt, M.; Heinz, E.; Ott, C.; Zahringer, U.; Thomma, B.P.; Ferket, K.K.; Cammue, B.P. Defensins from insects and plants interact with fungal glucosylceramides. J. Biol. Chem. 2004, 279, 3900–3905. 37. Osborn, R.W.; de Samblanx, G.W.; Thevissen, K.; Goderis, I.; Torrekens, S.; van Leuven, F.; Attenborough, S.; Rees, S.B.; Broekaert, W.F. References Terras, F.R.; Eggermont, K.; Kovaleva, V.; Raikhel, N.V.; Osborn, R.W.; Kester, A.; Rees, S.B.; Torrekens, S.; van Leuven, F.; Vanderleyden, J.; et al. Small cysteine-rich antifungal proteins from radish: Their role in host defense. Plant Cell 1995, 7, 573–588. 13. Almeida, M.S.; Cabral, K.M.; Zingali, R.B.; Kurtenbach, E. Characterization of two novel defense peptides from pea (Pisum sativum) seeds. Arch. Biochem. Biophys. 2000, 378, 278–286. 14. Gao, A.G.; Hakimi, S.M.; Mittanck, C.A.; Wu, Y.; Woerner, B.M.; Stark, D.M.; Shah, D.M.; Liang, J.; Rommens, C.M. Fungal pathogen protection in potato by expression of a plant defensin peptide. Nat. Biotechnol. 2000, 18, 1307–1310. 15. Spelbrink, R.G.; Dilmac, N.; Allen, A.; Smith, T.J.; Shah, D.M.; Hockerman, G.H. Differential antifungal and calcium channel-blocking activity among structurally related plant defensins. Plant Physiol. 2004, 135, 2055–2067. 16. Lay, F.T.; Brugliera, F.; Anderson, M.A. Isolation and properties of floral defensins from ornamental tobacco and petunia. Plant Physiol. 2003, 131, 1283–1293. 17. Maitra, N.; Cushman, J.C. Characterization of a drought-induced soybean cDNA encoding a plant defensin. Plant Physiol. 1998, 118, 1536. 18. Maitra, N.; Cushman, J.C. Isolation and characterization of a drought-induced soybean cDNA encoding a D95 family late-embryogenesis-abundant protein. Plant Physiol. 1994, 106, 805–806. 19. Hanks, J.N.; Snyder, A.K.; Graham, M.A.; Shah, R.K.; Blaylock, L.A.; Harrison, M.J.; Shah, D.M. Defensin gene family in Medicago truncatula: Structure, expression and induction by signal molecules. Plant Mol. Biol. 2005, 58, 385–399. 20. Manners, J.M.; Penninckx, I.A.; Vermaere, K.; Kazan, K.; Brown, R.L.; Morgan, A.; Maclean, D.J.; Curtis, M.D.; Cammue, B.P.; Broekaert, W.F. The promoter of the plant defensin gene PDF1.2 from Arabidopsis is systemically activated by fungal pathogens and responds to methyl jasmonate but not to salicylic acid. Plant Mol. Biol. 1998, 38, 1071–1080. 21. Bruix, M.; Gonzalez, C.; Santoro, J.; Soriano, F.; Rocher, A.; Mendez, E.; Rico, M. 1H-nmr studies on the structure of a new thionin from barley endosperm. Biopolymers 1995, 36, 751–763. 22. Fant, F.; Vranken, W.; Broekaert, W.; Borremans, F. Determination of the three-dimensional solution structure of Raphanus sativus antifungal protein 1 by 1H NMR. J. Mol. Biol. 1998, 279, 257–270. 23. Fant, F.; Vranken, W.F.; Borremans, F.A. The three-dimensional solution structure of Aesculus hippocastanum antimicrobial protein 1 determined by 1H nuclear magnetic resonance. Proteins 1999, 37, 388–403. Molecules 2014, 19 Molecules 2014, 19 12296 24. Janssen, B.J.; Schirra, H.J.; Lay, F.T.; Anderson, M.A.; Craik, D.J. References Isolation and characterisation of plant defensins from seeds of Asteraceae, Fabaceae, Hippocastanaceae and Saxifragaceae. FEBS Lett. 1995, 368, 257–262. 38. Thevissen, K.; Cammue, B.P.; Lemaire, K.; Winderickx, J.; Dickson, R.C.; Lester, R.L.; Ferket, K.K.; van Even, F.; Parret, A.H.; Broekaert, W.F. A gene encoding a sphingolipid biosynthesis enzyme determines the sensitivity of Saccharomyces cerevisiae to an antifungal plant defensin from dahlia (Dahlia merckii). Proc. Natl. Acad. Sci. USA 2000, 97, 9531–9536. 39. Thevissen, K.; Ferket, K.K.A.; François, I.E.J.A.; Cammue, B.P.A. Interactions of antifungal plant defensins with fungal membrane components. Peptides 2003, 24, 1705–1712. Molecules 2014, 19 12297 40. Ramamoorthy, V.; Cahoon, E.B.; Li, J.; Thokala, M.; Minto, R.E.; Shah, D.M. Glucosylceramide synthase is essential for alfalfa defensin-mediated growth inhibition but not for pathogenicity of Fusarium graminearum. Mol. Microbiol. 2007, 66, 771–786. 41. Ferket, K.K.; Levery, S.B.; Park, C.; Cammue, B.P.; Thevissen, K. Isolation and characterization of Neurospora crassa mutants resistant to antifungal plant defensins. Fungal Genet. Biol. 2003, 40, 176–185. 42. De Paula, V.S.; Razzera, G.; Barreto-Bergter, E.; Almeida, F.C.; Valente, A.P. Portrayal of complex dynamic properties of sugarcane defensin 5 by NMR: Multiple motions associated with membrane interaction. Structure 2011, 19, 26–36. 43. De Medeiros, L.N.; Angeli, R.; Sarzedas, C.G.; Barreto-Bergter, E.; Valente, A.P.; Kurtenbach, E.; Almeida, F.C. Backbone dynamics of the antifungal Psd1 pea defensin and its correlation with membrane interaction by NMR spectroscopy. Biochim. Biophys. Acta 2010, 2, 105–113. 44. Thevissen, K.; de Mello Tavares, P.; Xu, D.; Blankenship, J.; Vandenbosch, D.; Idkowiak-Baldys, J.; Govaert, G.; Bink, A.; Rozental, S.; de Groot, P.W.; et al. The plant defensin RsAFP2 induces cell wall stress, septin mislocalization and accumulation of ceramides in Candida albicans. Mol. Microbiol. 2012, 84, 166–180. 45. Van der Weerden, N.L.; Lay, F.T.; Anderson, M.A. The plant defensin, NaD1, enters the cytoplasm of Fusarium oxysporum hyphae. J. Biol. Chem. 2008, 283, 14445–14452. 46. Van der Weerden, N.L.; Hancock, R.E.; Anderson, M.A. Permeabilization of fungal hyphae by the plant defensin NaD1 occurs through a cell wall-dependent process. J. Biol. Chem. 2010, 285, 37513–37520. 47. Sagaram, U.S.; El-Mounadi, K.; Buchko, G.W.; Berg, H.R.; Kaur, J.; Pandurangi, R.S.; Smith, T.J.; Shah, D.M. Structural and functional studies of a phosphatidic acid-binding antifungal plant defensin MtDef4: Identification of an RGFRRR motif governing fungal cell entry. PLoS One 2013, 8, e82485. 48. Lobo, D.S.; Pereira, I.B.; Fragel-Madeira, L.; Medeiros, L.N.; Cabral, L.M.; Faria, J.; Bellio, M.; Campos, R.C.; Linden, R.; Kurtenbach, E. References Antifungal Pisum sativum defensin 1 interacts with Neurospora crassa cyclin F related to the cell cycle. Biochemistry 2007, 46, 987–996. 49. Nicolas, P. Multifunctional host defense peptides: Intracellular-targeting antimicrobial peptides. FEBS J. 2009, 276, 6483–6496. 50. Thevissen, K.; Ghazi, A.; de Samblanx, G.W.; Brownlee, C.; Osborn, R.W.; Broekaert, W.F. Fungal membrane responses induced by plant defensins and thionins. J. Biol. Chem. 1996, 271, 15018–15025. 51. Aerts, A.M.; Francois, I.E.; Meert, E.M.; Li, Q.T.; Cammue, B.P.; Thevissen, K. The antifungal activity of RsAFP2, a plant defensin from raphanus sativus, involves the induction of reactive oxygen species in Candida albicans. J. Mol. Microbiol. Biotechnol. 2007, 13, 243–247. 52. Aerts, A.M.; Bammens, L.; Govaert, G.; Carmona-Gutierrez, D.; Madeo, F.; Cammue, B.P.; Thevissen, K. The Antifungal plant defensin HsAFP1 from Heuchera sanguinea induces apoptosis in Candida albicans. Front. Microbiol. 2011, 2, 47. 53. Aerts, A.M.; Francois, I.E.; Bammens, L.; Cammue, B.P.; Smets, B.; Winderickx, J.; Accardo, S.; de Vos, D.E.; Thevissen, K. Level of M(IP)2C sphingolipid affects plant defensin sensitivity, oxidative stress resistance and chronological life-span in yeast. FEBS Lett. 2006, 580, 1903–1907. Molecules 2014, 19 12298 54. Terras, F.R.; Torrekens, S.; van Leuven, F.; Osborn, R.W.; Vanderleyden, J.; Cammue, B.P.; Broekaert, W.F. A new family of basic cysteine-rich plant antifungal proteins from Brassicaceae species. FEBS Lett. 1993, 316, 233–240. 55. Broekaert, W.F.; Terras, F.R.; Cammue, B.P.; Osborn, R.W. Plant defensins: Novel antimicrobial peptides as components of the host defense system. Plant Physiol. 1995, 108, 1353–1358. 56. Oard, S.; Karki, B. Mechanism of beta-purothionin antimicrobial peptide inhibition by metal ions: Molecular dynamics simulation study. Biophys. Chem. 2006, 121, 30–43. 57. Cociancich, S.; Ghazi, A.; Hetru, C.; Hoffmann, J.A.; Letellier, L. Insect defensin, an inducible antibacterial peptide, forms voltage-dependent channels in Micrococcus luteus. J. Biol. Chem. 1993, 268, 19239–19245. 58. Lehrer, R.I.; Ganz, T.; Szklarek, D.; Selsted, M.E. Modulation of the in vitro candidacidal activity of human neutrophil defensins by target cell metabolism and divalent cations. J. Clin. Investig. 1988, 81, 1829–1835. 59. Sugiarto, H.; Yu, P.L. Effects of cations on antimicrobial activity of ostricacins-1 and 2 on E. coli O157:H7 and S. aureus 1056MRSA. Curr. Microbiol. 2007, 55, 36–41. 60. Tomita, T.; Hitomi, S.; Nagase, T.; Matsui, H.; Matsuse, T.; Kimura, S.; Ouchi, Y. Effect of ions on antibacterial activity of human beta defensin 2. Microbiol. Immunol. 2000, 44, 749–754. 61. Ramamoorthy, V.; Zhao, X.; Snyder, A.K.; Xu, J.R.; Shah, D.M. References Two mitogen-activated protein kinase signalling cascades mediate basal resistance to antifungal plant defensins in Fusarium graminearum. Cell. Microbiol. 2007, 9, 1491–1506. 62. Lay, F.T.; Schirra, H.J.; Scanlon, M.J.; Anderson, M.A.; Craik, D.J. The three-dimensional solution structure of NaD1, a new floral defensin from Nicotiana alata and its application to a homology model of the crop defense protein alfAFP. J. Mol. Biol. 2003, 325, 175–188. 63. Sagaram, U.S.; Pandurangi, R.; Kaur, J.; Smith, T.J.; Shah, D.M. Structure-activity determinants in antifungal plant defensins MsDef1 and MtDef4 with different modes of action against Fusarium graminearum. PLoS One 2011, 6, 0018550. 64. De Samblanx, G.W.; del Carmen, F.A.; Sijtsma, L.; Plasman, H.H.; Schaaper, W.M.; Posthuma, G.A.; Fant, F.; Meloen, R.H.; Broekaert, W.F.; van Amerongen, A. Antifungal activity of synthetic 15-mer peptides based on the Rs-AFP2 (Raphanus sativus antifungal protein 2) sequence. Pept. Res. 1996, 9, 262–268. 65. Tavares, P.M.; Thevissen, K.; Cammue, B.P.; Francois, I.E.; Barreto-Bergter, E.; Taborda, C.P.; Marques, A.F.; Rodrigues, M.L.; Nimrichter, L. In vitro activity of the antifungal plant defensin RsAFP2 against Candida isolates and its in vivo efficacy in prophylactic murine models of candidiasis. Antimicrob. Agents Chemother. 2008, 52, 4522–4525. 66. De Samblanx, G.W.; Goderis, I.J.; Thevissen, K.; Raemaekers, R.; Fant, F.; Borremans, F.; Acland, D.P.; Osborn, R.W.; Patel, S.; Broekaert, W.F. Mutational analysis of a plant defensin from radish (Raphanus sativus L.) reveals two adjacent sites important for antifungal activity. J. Biol. Chem. 1997, 272, 1171–1179. 67. Klose, C.; Ejsing, C.S.; Garcia-Saez, A.J.; Kaiser, H.J.; Sampaio, J.L.; Surma, M.A.; Shevchenko, A.; Schwille, P.; Simons, K. Yeast lipids can phase-separate into micrometer-scale membrane domains. J. Biol. Chem. 2010, 285, 30224–30232. Molecules 2014, 19 12299 68. Aerts, A.M.; Carmona-Gutierrez, D.; Lefevre, S.; Govaert, G.; Francois, I.E.; Madeo, F.; Santos, R.; Cammue, B.P.; Thevissen, K. The antifungal plant defensin RsAFP2 from radish induces apoptosis in a metacaspase independent way in Candida albicans. FEBS Lett. 2009, 583, 2513–2516. 68. Aerts, A.M.; Carmona-Gutierrez, D.; Lefevre, S.; Govaert, G.; Francois, I.E.; Madeo, F.; Santos, R.; Cammue, B.P.; Thevissen, K. The antifungal plant defensin RsAFP2 from radish induces apoptosis in a metacaspase independent way in Candida albicans. FEBS Lett. 2009, 583, 2513–2516. 69. Navarro-Garcia, F. The MAP kinase Mkc1p is activated under different stress conditions in Candida albicans. Microbiology 2005, 151, 2737–2749. 70. Schaaper, W.M.; Posthuma, G.A.; Plasman, H.H.; Sijtsma, L.; Fant, F.; Borremans, F.A.; Thevissen, K.; Broekaert, W.F.; Meloen, R.H.; van Amerongen, A. References The Hog1 mitogen-activated protein kinase is essential in the oxidative stress response and chlamydospore formation in Candida albicans. Eukaryot. Cell 2003, 2, 351–361. 82. Alonso-Monge, R.; Navarro-Garcia, F.; Roman, E.; Negredo, A.I.; Eisman, B.; Nombela, C.; Pla, J. The Hog1 mitogen-activated protein kinase is essential in the oxidative stress response and chlamydospore formation in Candida albicans. Eukaryot. Cell 2003, 2, 351–361. 83. Zheng, J.S.; Tang, S.; Huang, Y.C.; Liu, L. Development of new thioester equivalents for protein chemical synthesis. Acc. Chem. Res. 2013, 46, 2475–2484. 84. Heck, T.; Faccio, G.; Richter, M.; Thony-Meyer, L Enzyme-catalyzed protein crosslinking. Appl. Microbiol. Biotechnol. 2013, 97, 461–475. 85. Macmillan, D. Evolving strategies for protein synthesis converge on native chemical ligation. Angew. Chem. Int. Ed. Engl. 2006, 45, 7668–7672. 86. Nilsson, B.L.; Soellner, M.B.; Raines, R.T. Chemical synthesis of proteins. Annu. Rev. Biophys. Biomol. Struct. 2005, 34, 91–118. 87. Sureshbabu, V.V.; Narendra, N. Protection Reactions, in Amino Acids, Peptides and Proteins in Organic Chemistry; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2011; pp. 1–97. 88. Cherkupally, P.; Acosta, G.A.; Ramesh, S.; de la Torre, B.G.; Govender, T.; Kruger, H.G.; Albericio, F. Solid-phase peptide synthesis (SPPS), C-terminal vs. side-chain anchoring: A reality or a myth. Amino Acids 2014, 46, 1827–1838. 89. Amblard, M.; Fehrentz, J.A.; Martinez, J.; Subra, G. Methods and protocols of modern solid phase peptide synthesis. Mol. Biotechnol. 2006, 33, 239–254. 90. El-Faham, A.; Albericio, F. Peptide coupling reagents, more than a letter soup. Chem. Rev. 2011, 111, 6557–6602. 91. Muir, T.W. Semisynthesis of proteins by expressed protein ligation. Annu. Rev. Biochem. 2003, 72, 249–289. 92. Taylor, V.E.; Imperiali, B. Native Chemical ligation: Semi-synthesis of post-translationally modified proteins and biological probes. In Protein Engineering; Köhrer, C., RajBhandary, U., Eds.; Springer: Berlin/Heidelberg, Germany, 2009; pp. 65–96. 93. Váradi, G.; Tóth, G.K.; Kele, Z.; Galgóczy, L.; Fizil, Á.; Batta, G. Synthesis of PAF, an antifungal protein from P. chrysogenum, by native chemical ligation: Native disulfide pattern and fold obtained upon oxidative refolding. Chemistry 2013, 19, 12684–12692. 94. Wu, X.; Wu, Y.; Zhu, F.; Yang, Q.; Wu, Q.; Zhangsun, D.; Luo, S. Optimal cleavage and oxidative folding of alpha-conotoxin TxIB as a therapeutic candidate peptide. Mar. Drugs 2013, 11, 3537–3553. 95. Harris, P.W.; Yang, S.H.; Molina, A.; Lopez, G.; Middleditch, M.; Brimble, M.A. Plant antimicrobial peptides snakin-1 and snakin-2: Chemical synthesis and insights into the disulfide connectivity. Chemistry 2014, 20, 5102–5110. 96. References Synthetic peptides derived from the beta2-beta3 loop of Raphanus sativus antifungal protein 2 that mimic the active site. J. Pept. Res. 2001, 57, 409–418. 71. Almeida, M.S.; Cabral, K.M.; Kurtenbach, E.; Almeida, F.C.; Valente, A.P. Solution structure of Pisum sativum defensin 1 by high resolution NMR: Plant defensins, identical backbone with different mechanisms of action. J. Mol. Biol. 2002, 315, 749–757. 72. Goncalves, S.; Teixeira, A.; Abade, J.; de Medeiros, L.N.; Kurtenbach, E.; Santos, N.C. Evaluation of the membrane lipid selectivity of the pea defensin Psd1. Biochim. Biophys. Acta 2012, 5, 1420–1426. 73. Kong, M.; Barnes, E.A.; Ollendorff, V.; Donoghue, D.J. Cyclin F regulates the nuclear localization of cyclin B1 through a cyclin-cyclin interaction. EMBO J. 2000, 19, 1378–1388. 74. Lew, D.J.; Reed, S.I. Morphogenesis in the yeast cell cycle: Regulation by Cdc28 and cyclins. J. Cell Biol. 1993, 120, 1305–1320. 75. Mendenhall, M.D.; Hodge, A.E. Regulation of Cdc28 cyclin-dependent protein kinase activity during the cell cycle of the yeast Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 1998, 62, 1191–1243. 76. Zhu, S.; Peigneur, S.; Gao, B.; Umetsu, Y.; Ohki, S.; Tytgat, J. Experimental conversion of a defensin into a neurotoxin: Implications for origin of toxic function. Mol. Biol. Evol. 2014, 31, 546–559. 77. Munoz, A.; Chu, M.; Marris, P.I.; Sagaram, U.S.; Kaur, J.; Shah, D.M.; Read, N.D. Specific domains of plant defensins differentially disrupt colony initiation, cell fusion and calcium homeostasis in Neurospora crassa. Mol. Microbiol. 2014, 92, 1357–1374. 78. Lengeler, K.B.; Davidson, R.C.; D’Souza, C.; Harashima, T.; Shen, W.C.; Wang, P.; Pan, X.; Waugh, M.; Heitman, J. Signal transduction cascades regulating fungal development and virulence. Microbiol. Mol. Biol. Rev. 2000, 64, 746–785. 79. Xu, J.R. Map kinases in fungal pathogens. Fungal Genet. Biol. 2000, 31, 137–152. 80. Hayes, B.M.; Bleackley, M.R.; Wiltshire, J.L.; Anderson, M.A.; Traven, A.; van der Weerden, N.L. Identification and mechanism of action of the plant defensin NaD1 as a new member of the antifungal drug arsenal against Candida albicans. Antimicrob. Agents Chemother. 2013, 57, 3667–3675. 81. Bleackley, M.R.; Wiltshire, J.L.; Perrine-Walker, F.; Vasa, S.; Burns, R.L.; van der Weerden, N.L.; Anderson, M.A. Agp2p, the plasma membrane transregulator of polyamine uptake, regulates the antifungal activities of the plant defensin NaD1 and other cationic peptides. Antimicrob. Agents Chemother. 2014, 58, 2688–2698. Molecules 2014, 19 12300 82. Alonso-Monge, R.; Navarro-Garcia, F.; Roman, E.; Negredo, A.I.; Eisman, B.; Nombela, C.; Pla, J. References Okumura, M.; Shimamoto, S.; Hidaka, Y. Chemical methods for producing disulfide bonds in peptides and proteins to study folding regulation. Curr. Protoc. Protein Sci. 2014, 76, 1–28. 97. Bulaj, G.; Olivera, B.M. Folding of conotoxins: Formation of the native disulfide bridges during chemical synthesis and biosynthesis of Conus peptides. Antioxid. Redox Signal. 2008, 10, 141–155. 98. Sharma, H.; Nagaraj, R. Antimicrobial activity of human β-defensin 4 analogs: Insights into the role of disulfide linkages in modulating activity. Peptides 2012, 38, 255–265. Molecules 2014, 19 12301 99. Cerovsky, V.; Slaninova, J.; Fucik, V.; Monincova, L.; Bednarova, L.; Malon, P.; Stokrova, J. Lucifensin, a novel insect defensin of medicinal maggots: Synthesis and structural study. ChemBioChem 2011, 12, 1352–1361. 100. Chan, W.C.; White, P.D. Fmoc Solid Phase Peptide Synthesis: A Practical Approach; Oxford University Press: Oxford, UK, 2000. 101. Collins, J.M.; Leadbeater, N.E. Microwave energy: A versatile tool for the biosciences. Org. Biomol. Chem. 2007, 5, 1141–1150. 102. Parachin, N.S.; Mulder, K.C.; Viana, A.A.; Dias, S.C.; Franco, O.L. Expression systems for heterologous production of antimicrobial peptides. Peptides 2012, 38, 446–456. 103. Sorensen, H.P.; Mortensen, K.K. Advanced genetic strategies for recombinant protein expression in Escherichia coli. J. Biotechnol. 2005, 115, 113–128. 104. Gopal, G.J.; Kumar, A. Strategies for the production of recombinant protein in Escherichia coli. Protein J. 2013, 32, 419–425. 105. Sorensen, H.P.; Mortensen, K.K. Soluble expression of recombinant proteins in the cytoplasm of Escherichia coli. Microb. Cell Fact. 2005, 4, doi:10.1186/1475-2859-4-1. 106. Abskharon, R.N.; Ramboarina, S.; el Hassan, H.; Gad, W.; Apostol, M.I.; Giachin, G.; Legname, G.; Steyaert, J.; Messens, J.; Soror, S.H. A novel expression system for production of soluble prion proteins in E. coli. Microb. Cell Fact. 2012, 11, 1475–2859. 107. Nguyen, V.D.; Hatahet, F.; Salo, K.E.; Enlund, E.; Zhang, C.; Ruddock, L.W. Pre-expression of a sulfhydryl oxidase significantly increases the yields of eukaryotic disulfide bond containing proteins expressed in the cytoplasm of E.coli. Microb. Cell Fact. 2011, 10, 1475–2859. 108. Manting, E.H.; Driessen, A.J. Escherichia coli translocase: The unravelling of a molecular machine. Mol. Microbiol. 2000, 37, 226–238. 109. Shokri, A.; Sanden, A.M.; Larsson, G. Cell and process design for targeting of recombinant protein into the culture medium of Escherichia coli. Appl. Microbiol. Biotechnol. 2003, 60, 654–664. 110. Hochkoeppler, A. Expanding the landscape of recombinant protein production in Escherichia coli. Biotechnol. Lett. 2013, 35, 1971–1981. 111. Berlec, A.; Strukelj, B. References Current state and recent advances in biopharmaceutical production in Escherichia coli, yeasts and mammalian cells. J. Ind. Microbiol. Biotechnol. 2013, 40, 257–274. 112. Viana, A.A.; Pelegrini, P.B.; Grossi-de-Sa, M.F. Plant biofarming: Novel insights for peptide expression in heterologous systems. Biopolymers 2012, 98, 416–427. 113. Puertas, J.M.; Caminal, G.; Gonzalez, G. Expression of metallocarboxypeptidase inhibitors in Escherichia coli: Effect of cysteine content and protein size in the secretory production of disulfide-bridged proteins. J. Ind. Microbiol. Biotechnol. 2011, 38, 1553–1560. 114. Picart, P.; Pirttila, A.M.; Raventos, D.; Kristensen, H.H.; Sahl, H.G. Identification of defensin-encoding genes of Picea glauca: Characterization of PgD5, a conserved spruce defensin with strong antifungal activity. BMC Plant Biol. 2012, 12, doi:10.1186/1471-2229-12-180. 115. Kovaleva, V.; Krynytskyy, H.; Gout, I.; Gout, R. Recombinant expression, affinity purification and functional characterization of Scots pine defensin 1. Appl. Microbiol. Biotechnol. 2011, 89, 1093–1101. Molecules 2014, 19 12302 116. Kovalskaya, N.; Hammond, R.W. Expression and functional characterization of the plant antimicrobial snakin-1 and defensin recombinant proteins. Protein Expr. Purif. 2009, 63, 12–17. 116. Kovalskaya, N.; Hammond, R.W. Expression and functional characterization of the plant antimicrobial snakin-1 and defensin recombinant proteins. Protein Expr. Purif. 2009, 63, 12–17. 117. Kant, P.; Liu, W.Z.; Pauls, K.P. PDC1, a corn defensin peptide expressed in Escherichia coli and Pichia pastoris inhibits growth of Fusarium graminearum. Peptides 2009, 30, 1593–1599. 118. Rothan, H.A.; Teh, S.H.; Haron, K.; Mohamed, Z. A Comparative Study on the expression, purification and functional characterization of human adiponectin in Pichia pastoris and Escherichia coli. Int. J. Mol. Sci. 2012, 13, 3549–3562. 119. Zappa, S.; Hasche, A.; Boudrant, J. Expression of Pyrococcus abyssi recombinant alkaline phosphatase: Influences of Escherichia coli rare codons and secretion by the methylotrophic yeast Pichia pastoris. Enzym. Microb. Technol. 2003, 32, 751–756. 120. Cregg, J.M.; Cereghino, J.L.; Shi, J.; Higgins, D.R. Recombinant protein expression in Pichia pastoris. Mol. Biotechnol. 2000, 16, 23–52. 121. De Bruin, E.C.; Duitman, E.H.; de Boer, A.L.; Veenhuis, M.; Bos, I.G.; Hack, C.E. Pharmaceutical proteins from methylotrophic yeasts. Methods Mol. Biol. 2005, 308, 65–76. 122. Macauley-Patrick, S.; Fazenda, M.L.; McNeil, B.; Harvey, L.M. Heterologous protein production using the Pichia pastoris expression system. Yeast 2005, 22, 249–270. 123. Darby, R.A.; Cartwright, S.P.; Dilworth, M.V.; Bill, R.M. Which yeast species shall I choose? Saccharomyces cerevisiae versus Pichia pastoris (review). Methods Mol. Biol. 2012, 866, 11–23. 124. References Gasser, B.; Prielhofer, R.; Marx, H.; Maurer, M.; Nocon, J.; Steiger, M.; Puxbaum, V.; Sauer, M.; Mattanovich, D. Pichia pastoris: Protein production host and model organism for biomedical research. Future Microbiol. 2013, 8, 191–208. 125. Mattanovich, D.; Branduardi, P.; Dato, L.; Gasser, B.; Sauer, M.; Porro, D. Recombinant protein production in yeasts. Methods Mol. Biol. 2012, 824, 329–358. 126. Ahmad, M.; Hirz, M.; Pichler, H.; Schwab, H. Protein expression in Pichia pastoris: Recent achievements and perspectives for heterologous protein production. Appl. Microbiol. Biotechnol. 2014, 98, 5301–5317. 127. Cereghino, G.P.; Cereghino, J.L.; Ilgen, C.; Cregg, J.M. Production of recombinant proteins in fermenter cultures of the yeast Pichia pastoris. Curr. Opin. Biotechnol. 2002, 13, 329–332. 128. Aw, R.; Polizzi, K.M. Can too many copies spoil the broth? Microb. Cell Fact. 2013, 12, 1475–2859. 129. Sunga, A.J.; Tolstorukov, I.; Cregg, J.M. Posttransformational vector amplification in the yeast Pichia pastoris. FEMS Yeast Res. 2008, 8, 870–876. 130. Cabral, K.M.S.; Almeida, M.S.; Valente, A.P.; Almeida, F.C.L.; Kurtenbach, E. Production of the active antifungal Pisum sativum defensin 1 (Psd1) in Pichia pastoris: Overcoming the inefficiency of the STE13 protease. Protein Expr. Purif. 2003, 31, 115–122. 131. Almeida, M.S.; Cabral, K.S.; de Medeiros, N.L.; Valente, A.P.; Almeida, F.C.L.; Kurtenbach, E. cDNA Cloning and heterologous expression of functional cysteine-rich antifungal protein Psd1 in the yeast pichia pastoris. Arch. Biochem. Biophys. 2001, 395, 199–207. 132. Li, L.; Wang, J.X.; Zhao, X.F.; Kang, C.J.; Liu, N.; Xiang, J.H.; Li, F.H.; Sueda, S.; Kondo, H. High level expression, purification, and characterization of the shrimp antimicrobial peptide, Ch-penaeidin, in Pichia pastoris. Protein Expr. Purif. 2005, 39, 144–151. Molecules 2014, 19 12303 133. Lay, F.T.; Veneer, P.K.; Hulett, M.D.; Kvansakul, M. Recombinant expression and purification of the tomato defensin TPP3 and its preliminary X-ray crystallographic analysis. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 2012, 68, 314–316. 134. Tang, X.S.; Tang, Z.R.; Wang, S.P.; Feng, Z.M.; Zhou, D.; Li, T.J.; Yin, Y.L. Expression, purification, and antibacterial activity of bovine lactoferrampin-lactoferricin in Pichia pastoris. Appl. Biochem. Biotechnol. 2012, 166, 640–651. 135. Chen, Z.; Wang, D.; Cong, Y.; Wang, J.; Zhu, J.; Yang, J.; Hu, Z.; Hu, X.; Tan, Y.; Hu, F.; et al. Recombinant antimicrobial peptide hPAB-beta expressed in Pichia pastoris, a potential agent active against methicillin-resistant Staphylococcus aureus. Appl. Microbiol. Biotechnol. 2011, 89, 281–291. 136. Chen, J.J.; Chen, G.H.; Hsu, H.C.; Li, S.S.; Chen, C.S. © 2014 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/). References Cloning and functional expression of a mungbean defensin VrD1 in Pichia pastoris. J. Agric. Food Chem. 2004, 52, 2256–2261. 137. Song, X.; Wang, J.; Wu, F.; Li, X.; Teng, M.; Gong, W. cDNA cloning, functional expression and antifungal activities of a dimeric plant defensin SPE10 from Pachyrrhizus erosus seeds. Plant Mol. Biol. 2005, 57, 13–20. 138. Lay, F.T.; Mills, G.D.; Poon, I.K.; Cowieson, N.P.; Kirby, N.; Baxter, A.A.; van der Weerden, N.L.; Dogovski, C.; Perugini, M.A.; Anderson, M.A.; et al. Dimerization of plant defensin NaD1 enhances its antifungal activity. J. Biol. Chem. 2012, 287, 19961–19972. © 2014 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/). © 2014 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).
https://openalex.org/W2420882273	https://hal.science/hal-01473783/document	English	null	The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe	Biodiversity and conservation	2,016	cc-by	12,393	The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe C. Nesshöver, M. Vandewalle, H. Wittmer, Estelle V. Balian, Esther Carmen, Ilse R. Geijzendorffer, Christoph Görg, Rob Jongman, Barbara Livoreil, Luis Santamaria, et al. To cite this version: C. Nesshöver, M. Vandewalle, H. Wittmer, Estelle V. Balian, Esther Carmen, et al.. The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe. Biodiversity and Conservation, 2016, 25 (7 spec), pp.1215-1233. 10.1007/s10531- 016-1127-5. hal-01473783 To cite this version: C. Nesshöver, M. Vandewalle, H. Wittmer, Estelle V. Balian, Esther Carmen, et al.. The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe. Biodiversity and Conservation, 2016, 25 (7 spec), pp.1215-1233. 10.1007/s10531- 016-1127-5. hal-01473783 To cite this version: C. Nesshöver, M. Vandewalle, H. Wittmer, Estelle V. Balian, Esther Carmen, et al.. The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe. Biodiversity and Conservation, 2016, 25 (7 spec), pp.1215-1233. 10.1007/s10531- 016-1127-5. hal-01473783 Distributed under a Creative Commons Attribution 4.0 International License HAL Id: hal-01473783 https://hal.science/hal-01473783v1 Submitted on 9 Jan 2024 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License Biodivers Conserv (2016) 25:1215–1233 DOI 10.1007/s10531-016-1127-5 ORIGINAL PAPER The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe Carsten Nessho¨ver1 • Marie Vandewalle1 • Heidi Wittmer2 • Estelle V. Balian4 • Esther Carmen17 • Ilse R. Geijzendorffer5,6 • Christoph Go¨rg3,7 • Rob Jongman5 • Barbara Livoreil8,9 • Luis Santamaria10 • Stefan Schindler11,12 • Josef Settele3,13 • Isabel Sousa Pinto14 • Katalin To¨ro¨k15 • Jiska van Dijk16 • Allan D. Watt17 • Juliette C. Young17 • Klaus Peter Zulka12,18 • the KNEU Project Team Received: 27 November 2015 / Revised: 8 April 2016 / Accepted: 7 May 2016 / Published online: 14 June 2016 Received: 27 November 2015 / Revised: 8 April 2016 / Accepted: 7 May 2016 / Published online: 14 June 2016 Received: 27 November 2015 / Revised: 8 April 2016 / Accepted: 7 May 2016 / Published online: 14 June 2016 The Author(s) 2016 This article is published with open access at Springerlink com The Author(s) 2016. This article is published with open access at Springerlink.com Abstract The absence of a good interface between scientiﬁc and other knowledge holders and decision-makers in the area of biodiversity and ecosystem services has been recog- nised for a long time. Despite recent advancements, e.g. with the Intergovernmental Communicated by David Hawksworth. This is part of the special issue on Networking Biodiversity Knowledge. Team of the KNEU Project is provided in acknowledgement section. & Carsten Nessho¨ver carsten.nesshoever@ufz.de 1 Department of Conservation Biology, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Permoserstr.15, 04318 Leipzig, Germany 1 Department of Conservation Biology, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Permoserstr.15, 04318 Leipzig, Germany 2 Department of Environmental Politics, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Permoserstr. 15, 04318 Leipzig, Germany 2 Department of Environmental Politics, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Permoserstr. 15, 04318 Leipzig, Germany 3 Department of Community Ecology, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Theodor-Lieser-Str. 4, 06120 Halle, Germany 3 Department of Community Ecology, UFZ–Helmholtz Centre for Environmental Research, UFZ Science-Policy Expert Group, Theodor-Lieser-Str. 4, 06120 Halle, Germany 4 Royal Belgian Institute of Natural Sciences (RBINS), Brussels, Belgium 5 Alterra, Wageningen University and Research Centre, P.O. The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe Among other functions, the NoK provides consolidated scientiﬁc views on contested topics, identiﬁca- tion of research gaps to support relevant policies, and horizon scanning activities to anticipate emerging issues. The NoK includes a capacity building component on inter- facing activities and contains mechanisms to ensure its credibility, relevance and legiti- macy. Such a network would need to ensure credibility, relevance and legitimacy of its work by maximizing transparency and ﬂexibility of processes, quality of outputs, the link to data and knowledge provision, the motivation of experts for getting involved and sound communication and capacity building. Keywords Science-policy-interface European policy Research networking IPBES Participatory methods Information and data mobilisation The Network of Knowledge approach: improving the science and society dialogue on biodiversity and ecosystem services in Europe Box 47, 6700 AA Wageningen, The Netherlands 6 Institut Me´diterrane´en de Biodiversite´ et d’Ecologie marine et continentale (IMBE) Aix Marseille Universite´, CNRS, IRD, Avignon Universite´, Technopoˆle Arbois-Me´diterrane´e, 13545 Aix-en-Provence, France 7 Present Address: Institute of Social Ecology, University of Klagenfurt, Schottenfeldgasse 29, 1070 Vienna, Austria 8 Centre for Evidence Based Conservation, Bangor University, Bangor, Gwynedd LL57 2DG, UK 12 3 3 1216 Biodivers Conserv (2016) 25:1215–1233 Platform on Biodiversity and Ecosystem Services (IPBES), challenges remain, particularly concerning the timely provision of consolidated views from different knowledge domains. To address this challenge, a strong and ﬂexible networking approach is needed across knowledge domains and institutions. Here, we report on a broad consultation process across Europe to develop a Network of Knowledge on biodiversity and ecosystem services (NoK), an approach aiming at (1) organising institutions and knowledge holders in an adaptable and responsive framework and (2) informing decision-makers with timely and accurate biodiversity knowledge. The consultation provided a critical analysis of the needs that should be addressed by a NoK and how it could complement existing European initiatives and institutions at the interface between policy and science. Among other functions, the NoK provides consolidated scientiﬁc views on contested topics, identiﬁca- tion of research gaps to support relevant policies, and horizon scanning activities to anticipate emerging issues. The NoK includes a capacity building component on inter- facing activities and contains mechanisms to ensure its credibility, relevance and legiti- macy. Such a network would need to ensure credibility, relevance and legitimacy of its work by maximizing transparency and ﬂexibility of processes, quality of outputs, the link to data and knowledge provision, the motivation of experts for getting involved and sound communication and capacity building. Platform on Biodiversity and Ecosystem Services (IPBES), challenges remain, particularly concerning the timely provision of consolidated views from different knowledge domains. To address this challenge, a strong and ﬂexible networking approach is needed across knowledge domains and institutions. Here, we report on a broad consultation process across Europe to develop a Network of Knowledge on biodiversity and ecosystem services (NoK), an approach aiming at (1) organising institutions and knowledge holders in an adaptable and responsive framework and (2) informing decision-makers with timely and accurate biodiversity knowledge. The consultation provided a critical analysis of the needs that should be addressed by a NoK and how it could complement existing European initiatives and institutions at the interface between policy and science. Introduction At the same time, biodiversity and ecosystem services issues are complex and often inﬂuenced by a multitude of drivers and pressures, which require a broad array of knowledge to understand and address them (Spierenburg 2012; Young et al. 2013; Young et al. 2014). p g g g Although a number of established approaches to synthesize scientiﬁc knowledge on speciﬁc issues exist (e.g., Pullin and Stewart 2006; Pullin et al. 2009; Sutherland et al. 2014; Dicks et al. 2014), and a number of institutions and processes (e.g. Service contracts) provide knowledge for policy processes, these rarely include the variety of existing knowledge and its holders. Consequently, networking and communication components are not adequately reﬂected in many existing science-policy-interfaces (SPI). Here we deﬁne SPIs quite broadly as organizations, initiatives or projects that work at the boundary of science, policy and society to enrich decision making, shape their participants’ and audiences’ understandings of problems, and eventually produce outcomes regarding decisions and behaviours (Sarkki et al. 2015). More ﬂexible approaches that complement and enrich the available scientiﬁc evidence by taking into account the changing needs and constraints of knowledge users, as well as their own knowledge are needed (Sarkki et al. 2013; Nessho¨ver et al. 2014; Young et al. 2014). In this context, the work carried out within the scientiﬁc community and jointly with other actors can increasingly be under- stood as an interfacing activity between knowledge domains, rather than a mere translation of knowledge from providers to requesters (Funtowicz and Ravetz 1994; Pielke 2007). g p q To address the challenge of improving the SPI on biodiversity and ecosystem services, we developed the Network of Knowledge (NoK) approach to provide a better linkage and organisation of the knowledge-holder community to render it more capable to respond to knowledge needs from decision-making. The Network of Knowledge concept originated in the guidelines proposed by the European Platform for Biodiversity Research Strategy (EPBRS) of ‘‘bringing together existing organisations and processes in a ﬂexible, responsive and broad-based way […] helping to focus the support of science and scientists on the needs of those setting policy and taking decisions’’ through ‘‘temporary, ad hoc associations of diverse organisations to assemble and communicate knowledge adapted to the needs of clients’’ (EPBRS 2009). The concept was further developed in a broad consultation with knowledge holders and decision-makers across Europe. Introduction Policy development related to biodiversity and ecosystem services conservation, man- agement and use, requires availability of credible, timely and relevant scientiﬁc knowl- edge. Such claims are based on the perception that policies and decision-making are sometimes not adequately informed by existing knowledge, or that the processes to make 9 Present Address: Foundation for Research on Biodiversity (FRB), 195 rue Saint Jacques, 75005 Paris, France 10 Don˜ana Biological Station (EBD-CSIC), C/Americo Vespucio s/n, Isla de la Cartuja, 41092 Seville, Spain 10 Don˜ana Biological Station (EBD-CSIC), C/Americo Vespucio s/n, Isla de la Cartuja, 41092 Seville, Spain 11 Department of Conservation Biology, Vegetation & Landscape Ecology, University of Vienna, Rennweg 14, 1030 Vienna, Austria 12 Environment Agency Austria, Spittelauer La¨nde 5, 1090 Vienna, Austria 13 German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Deutscher Platz 5e, 04103 Leipzig, Germany 14 CIIMAR–Interdisciplinary Centre for Marine and Environmental Research and Faculty of Sciences University of Porto, R. dos Bragas, 289, 4050-123 Porto, Portugal 14 CIIMAR–Interdisciplinary Centre for Marine and Environmental Research and Faculty of Sciences University of Porto, R. dos Bragas, 289, 4050-123 Porto, Portugal 15 MTA Centre for Ecological Research, Institute of Ecology and Botany Va´cra´to´t, Alkotma´ny u. 2-4, Va´cra´to´t 2163, Hungary 16 Norsk Institutt for Naturforskning (NINA), 7485 Trondheim, Norway 17 NERC-CEH Edinburgh, Bush Estate, Penicuik EH26 0QB, UK 18 Department of Integrative Zoology, University of Vienna, Althanstr. 14, 1090 Vienna, Austria 123 123 1217 Biodivers Conserv (2016) 25:1215–1233 such knowledge available to policy- and decision-makers are insufﬁciently structured. A good example for this is the recent discussion on the effect of neonicotinoid pesticides on bees and other pollinators, where interests from society strongly interfered with the syn- thesis of potentially relevant knowledge (Walters 2013; van der Sluijs et al. 2015). At the same time, biodiversity and ecosystem services issues are complex and often inﬂuenced by a multitude of drivers and pressures, which require a broad array of knowledge to understand and address them (Spierenburg 2012; Young et al. 2013; Young et al. 2014). such knowledge available to policy- and decision-makers are insufﬁciently structured. A good example for this is the recent discussion on the effect of neonicotinoid pesticides on bees and other pollinators, where interests from society strongly interfered with the syn- thesis of potentially relevant knowledge (Walters 2013; van der Sluijs et al. 2015). Methods and approach used We carried on an iterative process to develop and test a Network of Knowledge on Biodiversity and Ecosystem Services in Europe and address the questions outlined above. Core elements were broad consultation activities with individuals or groups of experts We carried on an iterative process to develop and test a Network of Knowledge on Biodiversity and Ecosystem Services in Europe and address the questions outlined above. Biodiversity and Ecosystem Services in Europe and address the questions outlined above. Core elements were broad consultation activities with individuals or groups of experts from research, research networks, policy and practice. Throughout the project in total over 400 experts participated actively. Their composition included about 75 % of scientists from different ﬁelds and backgrounds (universities and research institutes across Europe), 15 % practitioners (e.g., from NGOs and management agencies)l, and 10 % policy makers. Both groups included actors from the European (e.g., Directorates General on Environment and on Research and Innovation, European NGOs) as well as the national level (e.g., national agencies, biodiversity managers). The process started with an assessment of existing European science-policy interfaces on biodiversity and ecosystem services to obtain an overview on one hand of European knowledge holders. (Section 3.1). The assessment also served to sketch a preliminary set of functions that a Network of Knowledge should fulﬁl to complement existing SPIs (Sect. 4). This ﬁrst assessment was based on an internal survey involving all partners in the KNEU project, who represent and are involved in a very broad set of networks and projects in Europe and on the national scale (see acknowledgements). Further insights about connections between knowledge holders were received from a number of targeted inter- views using the Net-map approach (for results, see Hauck et al. 2015). The results were presented at regional workshops and at the KNEU ﬁrst conference (see below) to gain additional information on the knowledge holder landscape, but we had to acknowledge that the number and diversity of knowledge holders in the 28 EU countries was too broad and diverse to be properly mapped, so we focused on actors on the European level. To identify needs from the policy side, a number of interviews with policy makers and practitioners on the European level were conducted. Out of 45 experts contacted, 24 agreed to be inter- viewed (for details, see Balian et al. 2012). Introduction Here we present this iterative consultation process and, as a result of this process, identify the criteria and challenges, which should be addressed by a NoK approach. We do this by following the key questions of the consultation process: (1) How is the landscape of knowledge on biodiversity and ecosystem services in Europe currently shaped, and what are the current policy needs—in terms of SPI activities? (Section 3); (1) How is the landscape of knowledge on biodiversity and ecosystem services in Europe currently shaped, and what are the current policy needs—in terms of SPI activities? (Section 3); (2) Which functions should a Network of Knowledge provide to complement the existing institutions? (Section 4) ) What are the main challenges in establishing such a networking approach in a ﬂexible and transparent way? (Section 5) (4) How could it ensure that knowledge holders and decision-makers share their knowledge mutually in a credible, relevant and legitimate way? (Section 5) With addressing these questions, we critically analysed the potential added value required of a networking approach in the current science-policy landscape in Europe. 3 123 1218 Biodivers Conserv (2016) 25:1215–1233 123 Methods and approach used The next step in the consultation were three regional workshops, which addressed potential challenges and needs for a networking approach, also taking into account regional perspectives. To initiate discussions, the project team had developed a set of functions and principles of the NoK and sorted potential challenges it might face. This included a preliminary list of functions, a number of ethical principles (see also Tremblay et al. 2016), and a dedicated process design on how policy requests could be addressed (Livoreil et al. 2016). These functions and principles were further discussed with actors across the knowledge landscape through an online-consultation, and a larger European conference. About 250 experts participated in these activities, which resulted in a preliminary ‘‘green paper’’ outlining how a NoK could work (Livoreil et al. 2012). As an additional element of the consultation, the NoK approach as developed in the ﬁrst step for the green paper was tested with three trial assessments on biodiversity topics with policy relevance: (1) Current trends in kelp forests in Europe and evidence that these trends will affect the ecosystems biodiversity and the provision of ecosystem services (Arau´jo et al. 2016); (2) Effectiveness of interventions aiming at manipulating non-crop habitat or landscape features to maintain or support natural (indigenous) population of pest control agents (Dicks et al. 2016); and (3) Impact of multifunctional ﬂoodplain management on biodiversity (Schindler et al. 2016b). These trial knowledge assessments delivered pre- liminary results on the speciﬁc research questions they dealt with (e.g. Schindler et al. 2014) as well as direct feedback on the applied functionality of the NoK. The test cases 123 123 Biodivers Conserv (2016) 25:1215–1233 1219 were independently evaluated by a separate team that regularly visited the meetings and workshops, and conducted interviews with participants (Carmen et al. 2015). The aim of the evaluation was to identify strengths, weaknesses and complementarities of the different methodologies used, and to gain further insights about potential challenges in conducting the knowledge reviews. About 100 experts were involved in this test phase. were independently evaluated by a separate team that regularly visited the meetings and workshops, and conducted interviews with participants (Carmen et al. 2015). The aim of the evaluation was to identify strengths, weaknesses and complementarities of the different methodologies used, and to gain further insights about potential challenges in conducting the knowledge reviews. About 100 experts were involved in this test phase. Methods and approach used The lessons learned from the trial assessments were used to advance the development of the Network of Knowledge green paper into a white paper (KNEU Team 2014) which was presented to an online consultation and discussed during a second KNEU conference with 80 participants (September 2013), and during the ﬁnal project event at the European parliament (April 2014), mobilising in total about 250 experts who provided feedback on the proposed approach (results included in Sects. 4 and 5). During the project, the notions of credibility, relevance and legitimacy (CRELE) were used as an analytic framework (van der Sluijs et al. 2010; Sarkki et al. 2013, 2015). These notions are widely accepted and used in analysing SPIs (Cash et al. 2003; Sarkki et al. 2013). Although separating the conceptual from the potentially practical perspective of these attributes is difﬁcult (Heink et al. 2015), they were considered as a useful guiding framework for reﬂecting on the challenges faced by an SPI (e.g. in the Intergovernmental Platform on Biodiversity and Ecosystem Services, IPBES (UNEP 2010)): • Credibility of the NoK is the perceived quality, validity and expertise of the people, processes and knowledge exchanged at the interface. It should be ensured by the rigour of the process, the skills of the participants and the transparency of all processes and decisions. • Relevance or saliency, represent the responsiveness of the NoK to needs of policy and society, i.e. to the users of an interface activity. • Relevance or saliency, represent the responsiveness of the NoK to needs of policy and society, i.e. to the users of an interface activity. • Legitimacy is the perceived fairness and balance of knowledge holders’ perspectives within the SPI processes, including inclusiveness of all relevant stakeholders and fairness in treatment of diverging values, beliefs, and interests. In addition to the above three attributes, independence (i.e. avoiding inﬂuence of speciﬁc groups e.g. from donors, political parties and vested interest groups) was con- sidered important by many contributors to the various consultations. It was therefore added as a fourth attribute of the NoK framework, although it could as well be understood as part of legitimacy or credibility. The following sections summarize the main insights obtained from the process of developing and testing the Network of Knowledge on Biodiversity and Ecosystem Services in Europe as well as the challenges arising from them with regard to its credibility, relevance and legitimacy. The current science-policy interface landscape for biodiversity and ecosystem services When discussing the need for new interfacing approaches, it is crucial to analyse and to understand the existing knowledge landscape and its potential gaps. Section 3.1 brieﬂy summarizes the developments in the biodiversity and ecosystem services knowledge landscape in Europe based predominantly on scientiﬁc knowledge-holders. Section 3.2 addresses the needs of policy- and decision-makers, how the current landscape addresses them and which gaps remain. 123 1220 Biodivers Conserv (2016) 25:1215–1233 6 See http://www.epbrs.org. p g g p j EUBON—Building the European Biodiversity Observation Network, EU project in the 7th framework programme (http://www.eubon.eu); GEO BON—The Group on Earth Observations Biodiversity Observa- tion Network (http://www.geobon.org). 3 ALARM—Assessing Large scale Risks for biodiversity with tested methods, EU project in the 6th framework programme (http://www.alarmproject.net); SCALES—Securing the conservation of biodiversity across administrative levels and spatial, temporal, and ecological scales, EU project in the 7th framework programme (http://www.scales-project.net); OpenNESS–Operationalisation of Natural Capital and Ecosystem Services (http://www.openness-project.eu/), OPERAs—Ecosystem science for policy and practice http://operas-project.eu/), OpenNESS and OPERAs are EU projects in the 7th framework pro- gramme and jointly host the webplatform OPPLA (http://www.oppla.eu). 1 EBONE—European Biodiversity Observation Network, EU project in the 6th framework programme (http://www.wageningenur.nl/en/Expertise-Services/Research-Institutes/alterra/Projects/EBONE-2.htm); 5 ALTER-Net–A long-term biodiversity, ecosystem and awareness research network (http://www.alter-net. info/); EuroMarine–Integration of European Marine Research Networks of Excellence (http://www. euromarineconsortium.eu/); both are network of research institutions working on biodiversity and related topics, developed from EU-funded Networks of Excellence. 1 EBONE—European Biodiversity Observation Network, EU project in the 6th framework programme (http://www.wageningenur.nl/en/Expertise-Services/Research-Institutes/alterra/Projects/EBONE-2.htm); EUBON—Building the European Biodiversity Observation Network, EU project in the 7th framework programme (http://www.eubon.eu); GEO BON—The Group on Earth Observations Biodiversity Observa- tion Network (http://www.geobon.org). 2 The European network of LTER sites (http://www.lter-europe.net/) and its national members is also p of an international network, ILTER (http://www.ilternet.edu/). The evolving science-policy interface landscape The analysis of European knowledge holders and science-policy interfaces in the ﬁeld of biodiversity and ecosystem services showed a diversity of initiatives, some of which have an explicit mandate, others have an intrinsic, self-given motivation to provide policy- relevant knowledge, and there are those that have the potential to contribute but currently refrain from acting. Environmental and conservation agencies (e.g., the European Envi- ronment Agency), or the Joint Research Centre (Institute for Environment and Sustain- ability) have ofﬁcial mandates by the European Commission. In many European countries, advisory committees of governments and other authorities might play a role besides national agencies or academies. Often, especially in the case of the European Commission, environmental consultancies play a role in compiling existing knowledge on speciﬁc questions and policies. Additional policy-relevant knowledge emerges from research projects funded by the European Commission Framework Programs, particularly large- scale data and knowledge infrastructures that represent the European contributions to broader international programmes such as the European EBONE and EUBON projects, which contribute to the global GEOBON program of biodiversity observation data1; (Hoffmann et al. 2014; Jongman 2013) and the long-term Ecological Research network (LTER).2 Large-scale projects addressing biodiversity and ecosystem services (e.g., ALARM (Settele et al. 2005), SCALES (Henle et al. 2010), OPENness and OPERAs)3; as well as projects funded by thematic European funding networks of national funders (ERA- Nets),4 such as BiodivERsA (Durham et al. 2014), also actively engage in science-policy processes. Networks of Excellence (research networks formerly funded by Framework Programmes) such as ALTER-Net and EUROmarine5 continue to play an active role in networking in the landscape of knowledge holders in biodiversity and reaching out to policy. Since 1999, the European Platform for Biodiversity Research Strategy (EPBRS), a forum at which natural and social scientists, policy-makers and other stakeholders identify structured and strategically important biodiversity research, serves as an interface on research policy issues (Nessho¨ver et al. 2008).6 Learned societies on different levels are 3 ALARM—Assessing Large scale Risks for biodiversity with tested methods, EU project in the 6th framework programme (http://www.alarmproject.net); SCALES—Securing the conservation of biodiversity across administrative levels and spatial, temporal, and ecological scales, EU project in the 7th framework programme (http://www.scales-project.net); OpenNESS–Operationalisation of Natural Capital and Ecosystem Services (http://www.openness-project.eu/), OPERAs—Ecosystem science for policy and practice http://operas-project.eu/), OpenNESS and OPERAs are EU projects in the 7th framework pro- gramme and jointly host the webplatform OPPLA (http://www.oppla.eu). 4 For details on the ERA-Net scheme, see http://ec.europa.eu/research/fp7/index_en.cfm?pg= projects-home, for the ERA-Net BiodiERsA, see: http://www.biodiversa.org/. The evolving science-policy interface landscape 4 For details on the ERA-Net scheme, see http://ec.europa.eu/research/fp7/index_en.cfm?pg=eran projects-home, for the ERA-Net BiodiERsA, see: http://www.biodiversa.org/. 5 ALTER-Net–A long-term biodiversity, ecosystem and awareness research network (http://www.alter-net. info/); EuroMarine–Integration of European Marine Research Networks of Excellence (http://www. euromarineconsortium.eu/); both are network of research institutions working on biodiversity and related topics, developed from EU-funded Networks of Excellence. 3 12 Biodivers Conserv (2016) 25:1215–1233 1221 also active at the science-policy interface, though their role is often limited by lack of resources (e.g., European Ecological Federation and its national members, Society of Conservation Biology, European chapter of the Society for Ecological Restoration, the International Association of Landscape Ecology). also active at the science-policy interface, though their role is often limited by lack of resources (e.g., European Ecological Federation and its national members, Society of Conservation Biology, European chapter of the Society for Ecological Restoration, the International Association of Landscape Ecology). The analysis also pointed to major developments in the organization of networks and research programs at the global level. Some of these programs, such as GBIF and GEO BON, were linked via growing networks with European and national activities, but their role at the interface with society is still under development (Hobern et al. 2013; Gei- jzendorffer et al. 2015). Other global initiatives, such as the Future Earth programme, represent an explicit step towards this direction with a much broader thematic focus (Mauser et al. 2013; Future Earth 2014). In addition to these mainly science-driven activities, European and international NGOs play an increasing role in providing knowledge-based advice and input into policy pro- cesses. IUCN has an international mandate for this, and NGOs like Birdlife and WWF (and many more) are regularly using in their policy work own data and knowledge as well as other sources. Expertise in all science-policy processes comes mainly from individual experts of research institutes, universities and other knowledge holding organisations at national level. Links across the SPIs at national, European and international level are rarely established, but globally initiatives are ongoing. While European research funding has clearly linked the science communities across countries, these links are rarely established at the science-policy interface—apart from speciﬁc activities of some Networks of Excellence, international learned societies or the umbrella organisations of national aca- demies (such as the European Academies Science Advisory Council EASAC and All European Academies ALLEA). The evolving science-policy interface landscape All these major SPIs are complemented by numerous ones with more restricted thematic or geographic focus. They include the interfacing activities of policy institutions such as the European Commission, which regularly invites experts to its meetings (e.g. DG Environment’s working group on the implementation of the 2020 EU Biodiversity Strategy). A new approach to improve the existing landscape needs to be carefully tailored to complement this wide array of existing activities and to minimize conﬂict of interests within and between existing institutions and initiatives, while contributing as much as possible to mainstreaming and coordinating these contributions (Nessho¨ver et al. 2014). Needs for improvement at the science-policy-interface Processes and knowledge outputs of the NoK on biodiversity and ecosystem services should take into account the challenges of feeding relevant and timely knowledge to the different phases of the policy cycle (namely policy design, monitoring and revision). Knowledge outputs should also include explicitly the knowledge of practitioners and managers to guarantee relevance. Production of concerted views from the knowledge community (cross- and intra- disciplinary) Different interests or knowledge holder groups can present opposing or seemingly contradictory scientiﬁc evidence, particularly on contentious and/or emerging issues. Often this is caused by a combination of factors including a focus on only a part of a complex interaction, biased assumptions, or time and budget constraints. Failure to acknowledge and adequately address such potential biases can undermine the credibility of the scientiﬁc knowledge used and its usefulness for the policy questions at stake. The NoK on biodiversity and ecosystem services should include structures and processes that prevent inappropriate questions, help gain a clearer picture on available knowledge, make explicit the complex causal links as well as the uncertainties behind it, and outline the implications for different policy contexts. Production of concerted views from the knowledge community (cross- and intra- disciplinary) Different interests or knowledge holder groups can present opposing or seemingly contradictory scientiﬁc evidence, particularly on contentious and/or emerging issues. Often this is caused by a combination of factors including a focus on only a part of a complex interaction, biased assumptions, or time and budget constraints. Failure to acknowledge and adequately address such potential biases can undermine the credibility of the scientiﬁc knowledge used and its usefulness for the policy questions at stake. The NoK on biodiversity and ecosystem services should include structures and processes that prevent inappropriate questions, help gain a clearer picture on available knowledge, make explicit the complex causal links as well as the uncertainties behind it, and outline the implications for different policy contexts. • Timely advice and contact with relevant experts In many situations policy-makers need to respond very quickly. If they happen to know someone knowledgeable on the issue, they will often rely on his/her opinion—even though they are aware that the opinions of single experts might be biased (Dicks et al. 2014). When no expert is readily available, searching the internet might be the chosen option. Needs for improvement at the science-policy-interface Different actors in the policy context may need different information and knowledge (Balian et al. 2012), which requires adapted approaches and formats as stated in interviews and a focus group with policy-makers and practitioners (Balian et al. 2012; Young et al. 2013; Nessho¨ver et al. 2014). In addressing different user groups, there is a need to improve how the different SPIs and individual knowledge holders interact and in what ways they are enabled to make their knowledge available. In the consultation process, several requirements for successfully addressing the existing diversity of knowledge users were identiﬁed (KNEU Team 2014; Nessho¨ver et al. 2014): • Joint formulation of questions and challenges between knowledge requesters and knowledge holders The process of jointly formulating questions between decision- makers and knowledge holders may often be the most important part of science-policy 12 3 3 1222 Biodivers Conserv (2016) 25:1215–1233 interactions, as they facilitate mutual understanding and may yield early insights about what knowledge on the topic is actually available (Pullin et al. 2009). Without such a participatory approach early on in the process, expectations by policy-makers might diverge from the understanding of knowledge providers. There are a number of recent approaches for collaboratively identifying research needs (Sutherland et al. 2009; Dicks et al. 2013; Ingram et al. 2013; Jones et al. 2015) that could also further inform and support this joint formulation of questions requiring synthesis of existing knowledge (see below). • Focus on knowledge for implementation and evaluation (policy design, monitoring and revision) Scientiﬁc outputs aimed at informing policies often remain at a overarching and strategic level, not taking into account the explicit needs and views of policy design and revision. Processes and knowledge outputs of the NoK on biodiversity and ecosystem services should take into account the challenges of feeding relevant and timely knowledge to the different phases of the policy cycle (namely policy design, monitoring and revision). Knowledge outputs should also include explicitly the knowledge of practitioners and managers to guarantee relevance. • Focus on knowledge for implementation and evaluation (policy design, monitoring and revision) Scientiﬁc outputs aimed at informing policies often remain at a overarching and strategic level, not taking into account the explicit needs and views of policy design and revision. 2009; Dicks et al. 2013; Ingram et al. 2013; Jones et al. 2015) Needs for improvement at the science-policy-interface Several of the interviewees stressed the usefulness of a ‘‘one-stop shop’’ where they could be sure to ﬁnd relevant information and useful links (Balian et al. 2012). • Horizon scanning and foresight Many activities have been undertaken recently to better scan for emerging issues, but these attempts were either focussed on speciﬁc topics (e.g., Sutherland et al. 2015) or, engaged in much broader perspectives (e.g. the foresight activities of the Joint Research Centre for Europe). In the biodiversity and ecosystem services context, an approach combining both hasn’t been applied so far, but would be a helpful tool to create relevant outcomes (see for example Cook et al. 2014). • Facilitate the identiﬁcation of research needs Many questions arising from policy cannot be answered directly/sufﬁciently based on existing knowledge, but require additional research. To be able to timely provide new research-based knowledge, regular scoping activities identifying research gaps arising from policy needs should support the set-up of research programmes. Often, the compilation of knowledge on a particular subject via the functions mentioned above will identify major research gaps. As mentioned above this could also make use of recent collaborative identiﬁcation of key research questions related to biodiversity and ecosystem services (Sutherland et al. 2009; Dicks et al 2013; Ingram et al 2013; Jones et al 2015) 12 123 3 1223 Biodivers Conserv (2016) 25:1215–1233 Although various institutions currently address some of these needs, the consultation process pointed out that a networking approach could help to improve their links and provide a more coherent framing for interface activities. For example, the trial assessment that conducted a knowledge synthesis on management issues of ﬂoodplains brought together knowledge holders from science and management practice (Schindler et al. 2014). The consultations also showed, that the knowledge holder community is increasingly interested in getting engaged in SPI processes, and that they could beneﬁt from a frame- work that allows them to identify the best way of getting active and build their capacity on SPI activities (see also Sect. 5). Knowledge synthesis function This function aims to contribute to decision-making by providing relevant knowledge on a request-driven basis. Whenever a topic requires an in-depth analysis and a consolidated view from the community of knowledge holders, speciﬁc activities to analyse and syn- thesize existing knowledge will be needed. A main characteristic of the approach devel- oped is that it systematically envisages joint scoping of the questions with the requesters and other relevant actors. Based on this, knowledge holders from different scientiﬁc dis- ciplines and other relevant experts, including practitioners and other knowledge holders, are identiﬁed and invited to jointly synthesize available knowledge on all the aspects speciﬁed (for details, see KNEU Team 2014; Livoreil et al. 2016; Schindler et al. 2016a). This may also include different levels of synthesis products, from broader reports to short synthesis paper and briefs, as for example described by Dicks et al. (2014). It may also include direct, short term contacts, as for example recently established by the OPPLA platform.7 p The added value of the developed NoK is to create a one-entry point for requests that need direct input from science but cannot currently easily be tackled via existing pathways. The NoK approach would also enable broad engagement of knowledge holders in synthesis activities, a factor largely missing from most current assessment processes (Beck et al. 2014; Vohland and Nadim 2015), and would be more ﬂexible in accessing knowledge at appropriate scales (e.g. on the Member State level via local networks and institutions). Given the broad expertise in Europe for different methodologies to synthesize knowledge, the NoK would allow the use of a broad range of methodologies that go beyond the approach of writing of peer-reviewed reports and assessments (Pullin et al. 2016). Applied methodologies may include for instance systematic reviews (Pullin and Stewart 2006), short synthesis or briefs (Dicks et al. 2014), adaptive management frameworks (Armitage et al. 2007; Westgate et al. 2013), expert based approaches (see Bergmann et al. 2012 for an overview) or a combination of them. This diversity of potential methodologies was perceived as a major value of a NoK approach, as it allows for a better integration of different perspectives and types of knowledge (Tengo¨ et al. 2014). However, this per- ception proved very difﬁcult to achieve in practice in the trial assessments (Carmen et al. 2015). 7 The ‘‘Ask OPPLA’’ function at http://oppla.eu/ask-oppla is a ‘‘crowd-sources enquiry service’’ that allows requesters to pose questions related to ecosystem services management and receive an answer within 48 h from suitable members of the community. The Network of Knowledge approach, its functions and added value To complement the landscape of existing institutions and address the needs identiﬁed in the consultation, the Network of Knowledge would fulﬁl four interrelated functions: It would (A) address the knowledge needs of environmental policy-makers through the synthesis of relevant knowledge on a request-driven basis, (B) proactively contribute to policy on research and innovation by identifying upcoming research needs, (C) enhance the net- working capacity of existing institutions and facilitate engagement in SPI activities for Fig. 1 Flowchart of the working methodology of a Network of Knowledge by indicating entry points where the Network of Knowledge is supposed to support decision making in identifying and collating relevant knowledge. Numbered arrows represent steps of logical pathways of knowledge search starting at simple questions that can be responded via web search to the most complex knowledge generation by conducting research Fig. 1 Flowchart of the working methodology of a Network of Knowledge by indicating entry points where the Network of Knowledge is supposed to support decision making in identifying and collating relevant knowledge. Numbered arrows represent steps of logical pathways of knowledge search starting at simple questions that can be responded via web search to the most complex knowledge generation by conducting research 3 1224 Biodivers Conserv (2016) 25:1215–1233 individuals and their institutions, (D) link European networks to the international context, to ensure a European ‘‘added-value’’ for international developments. The developed four functions in the NoK are described in more detail below (see also Fig. 1). Knowledge synthesis function Particularly when dealing with contentious issues, but also to establish credibility in general, it will be crucial that in addressing any given request, all knowledge synthesis processes are conducted and documented in an entirely transparent manner and open for a broad engagement from the start (Sarkki et al. 2013). Clear and transparent procedures should allow for a broad participation and open up to different perspectives in science and beyond (Young et al. 2014), while maintaining the advantages of scientiﬁc documentation and methodological rigor. For example, the protocols of systematic reviews will be made publicly available for comment before the review is conducted. 12 3 1225 Biodivers Conserv (2016) 25:1215–1233 Research strategy function This function aims to identify upcoming policy-relevant research gaps and emerging issues and how the knowledge holders could be supported to address them. Although this function is currently provided by individual institutions for instance in the form of publishing sector speciﬁc research agendas, it could be greatly improved by facilitating a broader partici- pation of experts and broadening the scope of topics beyond its narrow focus on biodi- versity and ecosystem services. Besides identifying knowledge needs directly upon requests from policy-makers, a NoK would identify emerging issues from science and stakeholders via horizon scanning and other approaches (e.g., Sutherland et al. 2011). When answering requests from decision-makers, research gaps and/or the need for further integration of data, infrastructure, and institutions will often be identiﬁed. Efﬁ- ciency gains can therefore be achieved by linking the knowledge synthesis and the research strategy functions, especially for complex requests where different depths of existing knowledge are readily apparent. Linking both functions is also important for engaging researchers to join science-policy interactions. Having the possibility to point to further research needs derived from knowledge assessment processes has often been mentioned in the consultation as incentive increasing the motivation of researchers to participate in the Network of Knowledge. International collaboration function A NoK should support knowledge integration at the European level, enhancing collabo- ration and encouraging openness within Europe and beyond. In doing so, it would optimize the input of European knowledge in international science-policy processes and multilateral agreements (e.g. the Convention on Biological Diversity or IPBES), foster European links to global research efforts (e.g. Future Earth, GEOBON, GBIF), and at the same time proﬁt from international inputs. To give an example of the potential of a NoK, it could facilitate regional activities planned/started within the IPBES work programme, e.g. the regional assessment on biodiversity and ecosystem services in Europe and Central Asia, expected to be ﬁnished in 2017. Regional inputs via regional networks that go beyond the IPBES internal process, such as a NoK, will be crucial to ensure credibility and relevance and may also help to translate back IPBES results into national/regional contexts (Beck et al. 2014). Networking and capacity building function This function aims at improving the existing networks of knowledge holders and plays a key role for the other functions. The added value of a NoK lies in its ability to identify and involve a much broader set of experts, over a wider thematic and geographic scope than most current SPIs. Networking should be understood in its broad sense and include a strong element of capacity building for enabling individuals to participate in activities on the other functions. Networking is necessary to strengthen the community of knowledge holders and to increase their ability to engage in structured approaches to integrate and synthesize knowledge from different sources, disciplines and sectors. Today, the limited opportunities for engagement of knowledge holders and decision makers in many SPI projects and processes is experienced as a major challenge for existing SPIs (Nessho¨ver et al. 2013; Bednarek et al. 2015) as veriﬁed with the trial assessments conducted to test the NoK (Schindler et al. 2016a). This function contains an institutional dimension that addresses institutes, existing networks, learned societies and other knowledge holders to get them engaged in the wider network, as well as an individual dimension that addresses the capabilities of individual knowledge holders to get actively involved in interface activities (Go¨rg et al. 2016). Participants of the consultation process indicated that the added-value from this function include the strengthening and better linking of existing networks, as well as the possibility to better identify the relevant target audiences. Network approaches on the national scale (e.g. in the national biodiversity platforms in Belgium, France and Germany) have shown these ingredients to be essential for success at the SPI. Networking also further enhances collaboration and encourages openness in bringing together different disciplines and expertise across countries on a speciﬁc topic. Also, it makes the link between knowledge forms more explicit and enhances responsiveness, an issue seen already at the level of individual research projects (Nessho¨ver et al. 2013). 123 1226 Biodivers Conserv (2016) 25:1215–1233 Integrated approach to functions To address this set of functions that summarized different elements discussed during the consultation, a Network of Knowledge would need to develop an integrated and active community of practice that supports knowledge-informed policy-making (Fig. 1, ‘‘NoK operation space’’). Different kinds of requests could be addressed by these functions. If a knowledge requester (e.g., a policy-maker from the EU Commission) can gain a satis- factory answer via existing means, such as a search on the Biodiversity Information System for Europe (BISE) portal of the European Environment Agency (EEA) (white arrow in Fig. 1), the NoK does not need to be invoked. If more knowledge is needed, the NoK might be mobilised with its knowledge synthesis function, to call for further expertise, e.g. speciﬁc experts to help, or point to synthesis or reviews already available to meet the request . If an even more detailed assessment of knowledge is needed, new synthesis processes could be set up . If the available knowledge proves insufﬁcient and further research is needed (or the request directly addresses research needs), the research strategy function of the NoK would then be addressed . All of these steps are valid for European- scale questions, or for questions of international collaboration, for example in the context of the Convention of Biological Diversity or IPBES. Discussion Form the consultation it became clear that addressing the diverse needs at the science- policy-interface in an integrated way, as outlined with the NoK approach, poses a number of challenges with respect to maximizing credibility, relevance and legitimacy, as in many other interface processes (Koetz et al. 2012; Sarkki et al. 2013). But from the consultation we also see potentials strengths of the integrated NoK approach to address them. Challenges for credibility, relevance and legitimacy in a Network of Knowledge The consultations in the project led to the identiﬁcation of ﬁve essential issues for developing a credible, relevant and legitimate NoK that fulﬁls the four functions described above: quality assurance; data standards and sharing; connecting, motivating and acknowledging the knowledge holders and requesters involved; communication; and Biodivers Conserv (2016) 25:1215–1233 1227 capacity building. All these issues are strongly interlinked and must be addressed in an integrated manner. Many of them are, of course, a challenge to SPIs in general. capacity building. All these issues are strongly interlinked and must be addressed in an integrated manner. Many of them are, of course, a challenge to SPIs in general. Quality assurance in SPIs covers a broad range of issues, some of them directly tied to scientiﬁc work (see also next section on data standards) and others related to the SPI process, where quality stands for effective and transparent procedures and ensures credi- bility. The NoK approach enables independent internal and external feedback loops and other means for evaluating and increasing quality. Particularly for conﬂicting issues, linking the different perspectives into a common process can help to bring more knowledge into the decision-making process and make the underlying conﬂicts explicit. The NoK approach includes an explicit choice of the best available methodologies to compile and assess the available evidence for addressing requests from policy (see KNEU- Team 2014, Pullin et al. 2016). This ranges from evidence-focused methodologies such as systematic reviews to different forms of moderated expert consultations to transdisci- plinary approaches such as collaborative adaptive management, with the possibility to combine these approaches depending on the needs identiﬁed (Pullin et al. 2016; Schindler et al. 2016a). The choice process will be made transparent through pre-established pro- tocols that lay out the circumstances under which each the methods is recommended, as well as their strengths and limitations, and required type of information, expert involve- ment and resources. Despite recent improvements (e.g. by GBIF, LifeWatch, LTER and EUBON/GEO- BON), answering questions and producing knowledge that require interpretation of bio- diversity data is still hampered by lack of harmonized, reliable and publicly-accessible databases. The lack of agreement in relatively simple matters, such as the use of stan- dardized protocols, can result in multiple experts disagreeing with each other already at the data integration level (Bendix et al. 2012; Enke et al. 2012). Challenges for credibility, relevance and legitimacy in a Network of Knowledge Involving well-known and respected contributors will improve visibility and credibility, but the processes should incorporate mechanisms to remain open to new, less experienced contributors to ensure capacity building (see below). In addition, continuity in the commitment from the com- munity of interest will be required to ensure long-term functioning of the mechanism. A NoK approach—through explicitly reaching out to the whole community across knowledge types, disciplines, regions and backgrounds—has a clear added value with regard to credibility, relevance and legitimacy as it can reach out in a timely and efﬁcient manner to relevant expertise and enables broader participation in terms of knowledge types and sources (Carmen et al. 2015). This is rarely the case in many science-policy approa- ches that restrict the input to certain groups, institutions or individuals, such as most consultancy contracts and research projects. Of course, there is a trade-off between enlarging the potential relevant expertise and ensuring credibility through a restricted number of high-level experts. In speciﬁc situations, the credibility of the knowledge produced relies more predominantly on direct evidence than on perceived inclusion of a wider science or policy context. The added value of the NoK lies in the open and trans- parent way in which such trade-offs are addressed and that the choice is made in relation to the needs and requirements of the requesters. Meet the challenges outlined above will require a high level of effective two-way communication on issues such as policy needs, processes of the NoK, data and method- ologies. This is especially true at the initial phases of the NoK, when the approach and its procedures (Fig. 1) will be new to most actors, particularly since it requires a high level of understanding of the different processes and does not always follow ‘‘classical’’ approa- ches of science-policy interactions. Communications in the NoK will need to balance the needs of communicating results, engaging people and fostering capacity building. This holds especially true as many biodiversity and ecosystem service related issues are cross-sectoral issues (Tittensor et al. 2014), so a continuous broad outreach is needed to engage and make aware the relevant knowledge holders and requesters from all areas, including other policy sectors (e.g., agriculture, forestry, ﬁsheries, climate and transport, ﬁnance) as well as different scientiﬁc disciplines (Jolibert and Wesselink 2012; Young et al. 2014). Challenges for credibility, relevance and legitimacy in a Network of Knowledge This is a serious constraint to transparent and easy-to-understand communication with requesters at a later stage of knowledge compilation, and may weaken the credibility of the information provided by the scientiﬁc community. The reluctance of many researchers to openly share data often arises from complex issues like conﬁdentiality, ownership (data owners often do not agree to publish their data due to legal issues), or data sensitivity (red list data for instance), and may severely hinder the timely and constant integration of new data into shared databases (Moritz et al. 2011; Enke et al. 2012). This underlying challenge cannot be tackled directly by the NoK, but rather by specialized processes in science like GBIF, LifeWatch, LTER and GEOBON, or through speciﬁc agreements between science, environmental agencies (or other continuously working institutions), and sometimes society (e.g. in the context of citizen science). It will, however, remain an obstacle for better informed policy-making in both science as well as in policy, and preclude the use of data-demanding methods to analyze existing knowledge (Wetzel et al. 2015). In today’s science as well as in the policy world, lack of time is the most crucial constraint for getting engaged in interface activities (Nessho¨ver et al. 2013; Sarkki et al. 2013). Many experts raised this concern during the consultations. At the same time, par- ticipation reached a critical mass in many processes, as soon as potential beneﬁts (for a ‘‘higher’’ target such as better biodiversity policy, as well as personal targets such as learning via involvement) became clear and experts were addressed directly (Carmen et al. 2015; Schindler et al. 2016a). Nonetheless, acknowledgment of science-policy activities of experts in their institutions and by funders and policy are still seen as a challenge and must be strengthened as one aspect of expert performance in order to raise the proﬁle and acceptance of such work (Carmen et al. 2015). 123 1228 Biodivers Conserv (2016) 25:1215–1233 The main challenge is then to connect enough knowledge holders for a comprehensive representation of the existing disciplinary and interdisciplinary knowledge on a topic that is going to be tackled. To enhance credibility and legitimacy, the NoK will have to work in a complementary process of networking people with excellent skills and the latest knowl- edge as well as integrating different types of knowledge. 123 Strengths of a Network of Knowledge Based on the consultations carried out and resulting identiﬁcation of needs and require- ments, we suggest that the NoK approach is an appropriate option to address several obstacles facing current science-policy interactions on complex issues like biodiversity and ecosystem services. These include a broader stakeholder involvement, the potential to include different forms of knowledge (if needed) and the ﬂexibility and transparency in using synthesis methods. (UNEP 2009; Koetz et al. 2012; Beck et al. 2014). The potential strengths of the NoK approach are highlighted below and further elaborated in other contributions to this Special Issue: • Integration of ‘‘networking’’ and ‘‘working’’ Following the approach of trans- and interdisciplinarity and jointly developing relevant knowledge across knowledge domains, the NoK approach explicitly links the issue of networking with the actual work on synthesising existing knowledge and identifying research needs. Inherently this involves power dynamics on one hand and the need for participation and openness on the other (see also Tremblay et al. 2016). • Transparency and openness throughout processes The NoK approach will only work if transparency and openness are major elements. Decision as well as work processes need to be as transparent as possible, including the choice of methods used in knowledge assessments, the persons and institutions involved, and the potential reviews of activities. A single communication strategy addressing these different aspects might be of great help here. It could also help raising awareness on the need for improved communication of complex environmental topics. In terms of credibility, this is a major issue that requires dedicated resources including dedicated intermediaries as experts for such processes (Bednarek et al. 2015). • Flexibility of methodological approaches European experts share a broad set of knowledge synthesis methods, and thus can adapt synthesis work to the types of questions posed and the quality and amount to knowledge available. To our knowledge, such a toolbox of methods has not been used in a coherent manner for science-policy- interfacing in the environmental sector (see also Pullin et al. 2016). By making such an overview available, the NoK can also stress the value of methods already in wide use (e.g., systematic reviews) and how they can be complemented by newer ones. • Reﬂexivity and iterativity. In complex situations and settings, the attributes of credibility, relevance and legitimacy might not be enough to properly describe and analyse whether an SPI functions in a desired way (Heink et al. Challenges for credibility, relevance and legitimacy in a Network of Knowledge Here, a major challenge lies in the translation of problems to be tackled and the results achieved into the language and mindset of those sectors and disciplines. This will require dialogue with the policy ‘requesters’ to understand their needs in terms of process and outputs from the NoK, which appears to be one of the biggest challenge as the trial assessments have shown (Schindler et al. 2016a). Building capacity at the science-policy interface, especially through the networking function, involves developing understanding, fostering trust, creating new links, applying new skills and developing shared knowledge. Hence, it is a process that may inﬂuence attitudes, behaviors and actions of individuals, institutions and the system as a whole (Cash et al. 2003; van den Hove 2007). Skills such as facilitation and conﬂict resolution may be crucial to implementing the NoK processes and should be a key component of its capacity building program (Nessho¨ver et al. 2013). Training in understanding the policy- and decision-making processes are also essential for experts getting involved in NoK pro- cesses—particularly those originating from the scientiﬁc community, where major misunderstandings still prevail regarding policy processes (Pielke 2007). 123 123 1229 Biodivers Conserv (2016) 25:1215–1233 Expert groups working in the NoK will include a wide range of perspectives, skills, expertise and knowledge sources from the start, requiring the building of capacities of experts facing different languages, theories and methodologies being from social sciences or natural sciences (see for example Tengo¨ et al. 2014). A key challenge of the NoK will be to ensure the building of some common grounds and trust among all experts and actors engaged in the processes. As such, it will require support not only from funders, but also strengthening links with all kinds of knowledge hubs—organisations, networks and ini- tiatives—at both the European and the national levels. A process of reﬂection and learning must be central to the NoK to help build bridges and reduce gaps between groups and move ever closer to collaborative working and information sharing (Carmen et al. 2015). Strengths of a Network of Knowledge 2015, Sarkki et al. 2015). Rather, an additional attribute might be needed, that includes the learning process within an SPI and its intention to systematically improve its internal processes, 3 3 1230 Biodivers Conserv (2016) 25:1215–1233 so that it remains ﬂexible and reﬂexive about its work. Sarkki et al. (2015) describe this attribute as iterativity. The NoK approach integrates this idea in its work and captures the importance (perceived by experts in the consultation) of ensuring capacity building and learning on processes. From a governance perspective, this could be included in a process of continuous formative evaluation, which should be embedded in the network processes and continuously informing and updating their functioning (see also Go¨rg et al. 2016). so that it remains ﬂexible and reﬂexive about its work. Sarkki et al. (2015) describe this attribute as iterativity. The NoK approach integrates this idea in its work and captures the importance (perceived by experts in the consultation) of ensuring capacity building and learning on processes. From a governance perspective, this could be included in a process of continuous formative evaluation, which should be embedded in the network processes and continuously informing and updating their functioning (see also Go¨rg et al. 2016). • Independence versus strong mandate In the consultation, the issue of independence of the Network of Knowledge and its work from policy and other stakeholders was perceived as extremely important. Many experts consulted stressed that the work should be independent of vested interests. At the same time, it was recognized that a strong political mandate, and thus also a strong role of speciﬁc policy-makers (or institutions) would be important, not the least in terms of motivation of experts to get involved. However, also in this case, political independence must be ensured. • Governance model To address the challenges outlined, a proper governance model of the NoK will be crucial (KNEU Team 2014). Governance needs to involve institutions from the network, as well as individual experts, and balance the contrasting interests attached to different knowledge-holder groups, to allow for a maximum of potential engagement and transparency (see Go¨rg et al. 2016). The experience of the KNEU project, including the remaining contributions to this Special Issue, indicates that a NoK model could be suitable to address the most pressing needs of the science-policy interface on biodiversity and ecosystem services, comple- menting and even making use of existing SPI approaches. Acknowledgments The KNEU project was funded by the 7th Framework Programme of the European Commission (Contract No. 265299). We thank the numerous experts and stakeholders that participated in the project’s activities, its workshop, conferences, survey, questionnaires and test cases. The authors also thank all project partners in the KNEU partners from the following institutions: UFZ (Germany), NERC- CEH (UK), RBINS (Belgium), CIIMAR (Portugal), NIOZ (Netherlands), FRB (France), ALTERRA (Netherlands), University of Vienna (Austria), NINA (Norway), CSIC (Spain), MTA (Hungary), ECNC (Netherlands), Bangor University (UK), INBO (Belgium), EAA (Austria), SYKE (Finland), BEC (Ireland) and VLIZ (Belgium). I.R.G. was partly funded by the EU BON project (EU FP 7 Grant No. 308454) and contributes to the Labex OT-Med (No. ANR-11-LABX-0061) funded by the French Government through the AMIDEX Project (No. ANR-11-IDEX-0001-02). Strengths of a Network of Knowledge Such model is to some extent, the only appropriate one considering the speciﬁc European situation, with a high number of experts already engaged in pre-existing networks, including several SPIs, and a broad array of approaches and methodologies already available to assess knowledge. Yet, such a model requires that all actors work in a novel, more open, ﬂexible and transparent manner that is often called for in discussion on the future knowledge society (Felt et al. 2007; Cornell et al. 2013). For decision-makers, it would yield a new, ﬂexible access to existing knowledge which would be complementary to existing channels. For the biodiversity and ecosystem services research community it means further work on developing networks as well as getting engaged even more in the science-policy interface and thus raising the proﬁle and the policy uptake of their work. Acknowledgments The KNEU project was funded by the 7th Framework Programme of the European Commission (Contract No. 265299). We thank the numerous experts and stakeholders that participated in the project’s activities, its workshop, conferences, survey, questionnaires and test cases. The authors also thank all project partners in the KNEU partners from the following institutions: UFZ (Germany), NERC- CEH (UK), RBINS (Belgium), CIIMAR (Portugal), NIOZ (Netherlands), FRB (France), ALTERRA (Netherlands), University of Vienna (Austria), NINA (Norway), CSIC (Spain), MTA (Hungary), ECNC (Netherlands), Bangor University (UK), INBO (Belgium), EAA (Austria), SYKE (Finland), BEC (Ireland) and VLIZ (Belgium). I.R.G. was partly funded by the EU BON project (EU FP 7 Grant No. 308454) and contributes to the Labex OT-Med (No. ANR-11-LABX-0061) funded by the French Government through the AMIDEX Project (No. ANR-11-IDEX-0001-02). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 Inter- national License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 123 12 Biodivers Conserv (2016) 25:1215–1233 1231 References u´jo RM et al (2016) Status, trends, drivers and effects of change of kelp forests in Europe: a compre- hensive expert consultation study. Biodivers Conserv. doi:10.1007/s10531-016-1141-7 Armitage D, Berkes F, Doubleday N (2007) Adaptive co-management: collaboration, learning and multi- level governance. UBC Press, Vancouver Balian E, Berhault A, Jones-Walters L, Torre-Martin A, Nessho¨ver C, Vandewalle M (2012) Overview of experts and requesters of a potential NoK: mapping knowledge holders, identifying requesters and barriers on how to link them (Deliverable 1.1). http://biodiversityknowledge.eu/images/Documents/ D li bl /KNEU D1 1 li t dh ld i db i Fi l df A d 8 A il 2016 Deliverables/KNEU-D1-1_clientsandholdersoverviewandbarriers_Final.pdf. Accessed 8 April 201 Beck S et al (2014) Towards a reﬂexive turn in the governance of global environmental expertise the cases of the IPCC and the IPBES. Gaia 23:80–87 Bednarek AT, Shouse B, Hudson CG, Goldburg R (2015) Science–policy intermediaries from a practi- tioner’s perspective: the Lenfest Ocean Program experience. Sci Public Policy. doi:10.1093/scipol/ scv008 Bendix J, Nieschulze J, Michener WK (2012) Data platforms in integrative biodiversity research. Ecol Inform 11:1–4. doi:10.1016/j.ecoinf.2012.04.001 Bergmann M, Jahn T, Knobloch T, Krohn W, Pohl C, Faust RC (2012) Methods for transdisciplinary research: a primer for practice. Campus Verlag, Frankfurt Carmen E, Nessho¨ver C, Saarikoski H, Vandewalle M, Watt A, Wittmer H, Young J (2015) Creating a biodiversity science community: experiences from a European Network of Knowledge. Env Sci Pol 54:497–504 Cash DW et al (2003) Knowledge systems for sustainable development. Proc Natl Acad Sci 100:8086 Cash DW et al (2003) Knowledge systems for sustainable development. Proc Natl Acad Sci 100:8086–8091 Cook CN, Inayatullah S, Burgman MA, Sutherland WJ, Wintle BA (2014) Strategic foresight: how planning for the unpredictable can improve environmental decision-making. Trends Ecol Evol 29:531–541. doi:10.1016/j.tree.2014.07.005 j nell S et al (2013) Opening up knowledge systems for better responses to global environmental change. Env Sci Pol 28:60–70. doi:10.1016/j.envsci.2012.11.008 Dicks LV et al (2013) What do we need to know to enhance the environmental sustainability of agricultural production? A prioritisation of knowledge needs for the UK food system. Sustainability 5:3095–3115 production? A prioritisation of knowledge needs for the UK food system. Sustainability 5:3095–3115 Dicks LV et al (2014) A transparent process for ‘‘Evidence-Informed’’ policy making. Cons Lett 7:119–125. doi:10.1111/conl.12046 Dicks LV et al (2014) A transparent process for ‘‘Evidence-Informed’’ policy making. Cons Lett 7:119–125. doi:10.1111/conl.12046 Dicks LV et al (2016) What works in conservation? References Gaia 19:187–193 Hobern D et al (2013) Global biodiversity informatics outlook: delivering biodiversity knowledge in the information age. Global Biodiversity Information Facility (Secretariat), Copenhagen Hoffmann A et al (2014) Improved access to integrated biodiversity data for science, pract the European Biodiversity Observation Network (EU BON). Nat Conserv 6:49–65 odiversity Observation Network (EU BON). Nat Co Ingram JS et al (2013) Priority research questions for the UK food system. Food Secur 5:617–636 Jolibert C, Wesselink A (2012) Research impacts and impact on research in biodiversity conservation: the inﬂuence of stakeholder engagement. Env Sci Pol 22:100–111. doi:10.1016/j.envsci.2012.06.012 g g j Jones AC et al (2015) Prioritization of knowledge needs for sustainable aquaculture: a national and glo perspective. Fish Fish 16:668–683 p p Jongman R (2013) Biodiversity observation from local to global. Ecol Ind 33:1–4 Koetz T, Farrell KN, Bridgewater P (2012) Building better science-policy interfaces for internatio environmental governance: assessing potential within the Intergovernmental Platform for Biodiver d i i Koetz T, Farrell KN, Bridgewater P (2012) Building better science-policy interfaces for international environmental governance: assessing potential within the Intergovernmental Platform for Biodiversity and Ecosystem Services. Int Environ Agreem 12:1–21 and Ecosystem Services. Int Environ Agreem 12 KNEU Team (2014) A recommended design for ‘‘BiodiversityKnowledge’’, a Network of Knowledge to support decision making on biodiversity and ecosystem services in Europe. Leipzig. http:// biodiversityknowledge.eu/images/PDF/WhitePaper_web.pdf Accessed 11 November 2015 Livoreil B et al (2012) Prototype NOK mechanism (Deliverable 2.1 of the KNEU project). KNEU Project, Leipzig. http://www.biodiversityknowledge.eu/images/Documents/Deliverables/KNEU-D2-1NoKpro totype.pdf Accessed 11 Nov 2015 Livoreil B, Geijzendorffer IR, Pullin AS, Schindler S, Vandewalle M, Nessho¨ver C (2016) Biodiversity knowledge synthesis at the European scale: actors and steps. Biodivers Conserv. doi:10.1007/s10531- 016-1143-5 Mauser W, Klepper G, Rice M, Schmalzbauer BS, Hackmann H, Leemans R, Moore H (2013) Transdis- ciplinary global change research: the co-creation of knowledge for sustainability. Curr Opin Environ Sustain 5:420–431 Moritz T et al (2011) Towards mainstreaming of biodiversity data publishing: recommendations of the GBIF Data Publishing Framework. Task Group BMC Bioinform 12:S1 Nessho¨ver C, Mu¨ssner R, Henle K, Pinto IS (2008) Linking biodiversity research and policy in Europe. Ambio 37:138–141 Nessho¨ver C et al (2014) Summary report and recommendations on Improving the Science-Policy Interface for Biodiversity and Ecosystem Services in Europe (contract: Ref No 07-0307/2013/661961/SER/B2). References Using expert assessment of summarised evidence to identify practices that enhance natural pest control in agriculture. Biodivers Conserv. doi:10.1007/ s10531-016-1133-7 Durham E, Baker H, Smith M, Moore E, Morgan V (2014) The BiodivERsA stakeholder engagement handbook. BiodivERsA, Paris Enke N, Thessen A, Bach K, Bendix J, Seeger B, Gemeinholzer B (2012) The user’s view on biodiversity data sharing—investigating facts of acceptance and requirements to realize a sustainable use of research data. Ecol Informatics 11:25–33 EPBRS (2009) Concept note: Network of Knowledge for biodiversity governance. European Platform for Biodiversity Research Strategy, Brussels. http://www.epbrs.org/PDF/2009%2009%2010%20Concept %20note%20on%20the%20network%20of%20knowledge_version%202-1.pdf Accessed 8 April 2016 Biodiversity Research Strategy, Brussels. http://www.epbrs.org/PDF/2009%2009%2010%20Concept %20note%20on%20the%20network%20of%20knowledge_version%202-1.pdf Accessed 8 April 2016 %20note%20on%20the%20network%20of%20knowledge_version%202-1.pdf Accessed 8 April 2016 Felt U, Wynne B, Stirling A, Callon M, Goncalves ME (2007) Science and governance: taking European knowledge society seriously. European Commission, Brussels. http://ec.europa.eu/research/science- society/document_library/pdf_06/european-knowledge-society_en.pdf Accessed 11 November 2015 Felt U, Wynne B, Stirling A, Callon M, Goncalves ME (2007) Science and governance: taking European knowledge society seriously. European Commission, Brussels. http://ec.europa.eu/research/science- society/document_library/pdf_06/european-knowledge-society_en.pdf Accessed 11 November 2015 F t i SO R t JR (1994) U t i t l it d t l i E i T i l Ch y _ y p _ p g y_ p Funtowicz SO, Ravetz JR (1994) Uncertainty, complexity and post-normal science. Environ Toxicol Chem 13:1881–1885 Future Earth (2014) Strategic research Agenda—priorities for a global sustainability research strategy. ICSU, Paris Geijzendorffer IR et al (2015) Bridging the gap between biodiversity data and policy reporting needs: Essential Biodiversity Variables perspective. J Appl Ecol. doi:10.1111/1365-2664.12417 Go¨rg C et al (2016) The governance of science-policy interf divers Conserv. doi:10.1007/s10531-016-1132-8 Go¨rg C et al (2016) The governance of science-policy interfaces—Network vs. Platform approach. Bio- divers Conserv. doi:10.1007/s10531-016-1132-8 Hauck J, Stein C, Schiffer E, Vandewalle M (2015) Seeing the forest and the trees: facilitating participatory network planning in environmental governance. Glob Environ Change 35:400–410 Heink U et al (2015) Conceptualizing credibility, relevance and legitimacy for evaluating the effectiveness of science–policy interfaces: challenges and opportunities. Sci Pub Pol 43:676–689. doi:10.1093/ scipol/scu082 123 3 3 1232 Biodivers Conserv (2016) 25:1215–1233 Henle K et al (2010) Securing the conservation of biodiversity across administrative levels and spatial, temporal, and ecological scales–research needs and approaches of the SCALES project. Gaia 19:187–193 Henle K et al (2010) Securing the conservation of biodiversity across administrative levels and spatial, temporal, and ecological scales–research needs and approaches of the SCALES project. References Conserv Biol 23:557–567 y Sutherland WJ, Fleishman E, Mascia MB, Pretty J, Rudd MA (2011) Methods for collaboratively identifying research priorities and emerging issues in science and policy. Meth Ecol Evol 2:238–247 y Sutherland WJ, Fleishman E, Mascia MB, Pretty J, Rudd MA (2011) Methods for collaboratively identifying research priorities and emerging issues in science and policy. Meth Ecol Evol 2:238–247 Sutherland WJ et al (2014) Solution scanning as a key policy tool: identifying management interventions to help maintain and enhance regulating ecosystem services. Ecol Soc. doi:10.5751/ES-06082-190203 Sutherland WJ et al (2015) A horizon scan of global conservation issues for. TREE 30:17–24. doi:10.1016/j. Sutherland WJ et al (2014) Solution scanning as a key policy tool: identifying management interventions to help maintain and enhance regulating ecosystem services. Ecol Soc. doi:10.5751/ES-06082-190203 help maintain and enhance regulating ecosystem services. Ecol Soc. doi:10.5751/ES-06082-190203 Sutherland WJ et al (2015) A horizon scan of global conservation issues for. TREE 30:17–24. doi:10.1016/j. tree.2014.11.002 Sutherland WJ et al (2015) A horizon scan of global conservation issues for. TREE 30:17–24. doi:10.1016/j. tree.2014.11.002 Tengo¨ M, Brondizio ES, Elmqvist T, Malmer P, Spierenburg M (2014) Connecting diverse knowledge systems for enhanced ecosystem governance: the multiple evidence base approach-. Ambio 43:579–591. doi:10.1007/s13280-014-0501-3 Tremblay M, Vandewalle M, Wittmer H (2016) Ethical challenges in an open system: the Network of Knowledge’s ethical risks assessment and its ethical infrastructure. Biodivers Conserv. doi:10.1007/ s10531-016-1123-9 References Leipzig http://ec europa eu/environment/nature/knowledge/pdf/EU%20Mechanism%20Summary%20 Nessho¨ver C et al (2014) Summary report and recommendations on Improving the Science-Policy Interface for Biodiversity and Ecosystem Services in Europe (contract: Ref No 07-0307/2013/661961/SER/B2). Leipzig http://ec.europa.eu/environment/nature/knowledge/pdf/EU%20Mechanism%20Summary%20 Report%202015.pdf Accessed 11 Nov 2015 Leipzig http://ec.europa.eu/environment/nature/knowledge/pdf/EU%20Mechanism%20Summary%20 Report%202015 pdf Accessed 11 Nov 2015 Leipzig http://ec.europa.eu/environment/nature/knowledge/pdf/EU%20Mechanism%20Summary%20 Report%202015.pdf Accessed 11 Nov 2015 p p Nessho¨ver C et al (2013) Improving the science-policy interface of biodiversity research projects. Gaia 22:99–103 Pielke RA (2007) The honest broker: making sense of science in policy and politics. Cambridge Univ Press, Cambridge Pullin AS, Stewart GB (2006) Guidelines for systematic review in conservation and environmental man- agement. Conserv Biol 20:1647–1656 Pullin AS, Knight TM, Watkinson AR (2009) Linking reductionist science and holistic policy using sys- tematic reviews: unpacking environmental policy questions to construct an evidence-based framework. J Appl Ecol 46:970–975 pp Pullin AS et al (2016) Selecting appropriate methods of knowledge synthesis to inform biodiversity policy. Biodivers Conserv. doi:10.1007/s10531-016-1131-9 Sarkki S, Niemela¨ J, Tinch R, van den Hove S, Watt A, Young J (2013) Balancing credibility, relevance and legitimacy: a critical assessment of trade-offs in science–policy interfaces. Sci Public Policy 41:194–206. doi:10.1093/scipol/sct046 p Sarkki S et al (2015) Adding ‘iterativity’to the credibility, relevance, legitimacy: a novel scheme to highlight dynamic aspects of science–policy interfaces. Env Sci Pol 54:505–512 y p p y Schindler S et al (2016a) The network Biodiversity Knowledge in practice: insights from three trial assessments. Biodivers Conserv. doi:10.1007/s10531-016-1128-4 Schindler S et al (2016a) The network Biodiversity Knowledge in pr assessments. Biodivers Conserv. doi:10.1007/s10531-016-1128-4 Schindler S et al (2016b) Multifunctional ﬂoodplain management and b European countries. Biodivers Conserv. doi:10.1007/s10531-016-1 Schindler S et al (2016b) Multifunctional ﬂoodplain management and biodiversity effects: lessons from six European countries. Biodivers Conserv. doi:10.1007/s10531-016-1129-3 Schindler S et al (2014) Multifunctionality of ﬂoodplain landscapes: relating management options to ecosystem services. Landsc Ecol 29:229–244. doi:10.1007/s10980-014-9989-y Settele J et al (2005) ALARM: assessing LArge-scale environmental risks for biodiversity with tested methods. Gaia 14:69–72 3 12 1233 Biodivers Conserv (2016) 25:1215–1233 Spierenburg M (2012) Getting the message across biodiversity science and policy interfaces–a review. Gaia 21:125–134 Sutherland W et al (2009) One hundred questions of importance to the conservation of global biological diversity. Conserv Biol 23:557–567 Sutherland W et al (2009) One hundred questions of importance to the conservation of global biological diversity. g s10531-016-1123-9 UNEP (2009) Gap analysis for the purpose of facilitating the discussions on how to improve and strengthen the science-policy interface on biodiversity and ecosystem services (UNEP/IPBES/2/INF/1). UNEP, Nairobi UNEP (2010) Report of the third ad hoc intergovernmental and multi-stakeholder meeting on an inter- governmental science-policy platform on biodiversity and ecosystem services (‘‘Busan Outcome’’)— UNEP/IPBES/3/3. UNEP, Nairobi. http://www.ipbes.net/images/stories/documents/K1061514_IPB ES-3-3-REPORT.pdf Accessed 11 Nov 2015 p van den Hove S (2007) A rationale for science–policy interfaces. Futures 39:807–826. doi:10.1016/j.futures. 2006.12.004 van der Sluijs JP, van Est R, Riphagen M (2010) Beyond consensus: reﬂections from a democratic per- spective on the interaction between climate politics and science. Curr Opin Env Sust 2:409–415. doi:10.1016/j.cosust.2010.10.003 van der Sluijs JP et al (2015) Conclusions of the Worldwide Integrated Assessment on the risks of neon- icotinoids and ﬁpronil to biodiversity and ecosystem functioning. Environ Sci Pollut Res 22:148–154. doi:10.1007/s11356-014-3229-5 Vohland K, Nadim T (2015) Ensuring the success of IPBES: between interface, market place and parlia- ment. Philos T R Soc B 370:1662. doi:10.1098/rstb.2014.0012 Walters K (2013) Data, data everywhere but we don’t know what to think? Neonicotinoid insecticides and pollinators. Outlooks on Pest Management 24:151–155. doi:10.1564/v24_aug_02 Walters K (2013) Data, data everywhere but we don’t know what to think? Neonicotinoid insecticid lli t O tl k P t M t 24 151 155 d i 10 1564/ 24 02 p g g Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive management of biological systems: a review. Biol Conserv 158:128–139. doi:10.1016/j.biocon.2012.08.016 Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive management of biological systems: a revi Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive manage Biol Conserv 158:128–139. doi:10.1016/j.biocon.2012.08.016 Biol Conserv 158:128–139. doi:10.1016/j.biocon.2012.08.016 Wetzel FT et al (2015) The roles and contributions of Biodiversity Observation Networks (BONs) in better tracking progress to 2020 biodiversity targets: a European case study. Biodiversity 16:137–149. doi:10. 1080/14888386.2015.1075902 Young JC, Watt AD, van den Hove S, SPIRAL Project Team (2013) Effective interfaces between science, policy and society: the SPIRAL project handbook. http://www.spiral-project.eu/content/documents/ Accessed 8 April 2016 Young JC et al (2014) Improving the science-policy dialogue to meet the challenges of biodiversity con- servation: having conversations rather than talking at one-another. Biodivers Conserv 23:387–404 123
https://openalex.org/W3212671474	https://www.researchsquare.com/article/rs-136961/latest.pdf	English	null	Comparing Nephrolithiasis with Nephrocalcinosis in Children; A Study From Two Tertiary Centers in Saudi Arabia	Research Square (Research Square)	2,021	cc-by	5,189	Comparing Nephrolithiasis with Nephrocalcinosis in Children; A Study From Two Tertiary Centers in Saudi Arabia Khalid Alhasan King Saud University Mohamed Shalaby King Abdulaziz University Amr Albanna King Saud bin Abdulaziz University fo Mohamad-Hani Temsah King Saud University Zainab Alhaik King Saud University Zaher Zaher King Abdulaziz University Najlaa Alotaibi King Abdulaziz University Nada Kalakattawi King Abdulaziz University Mohammed Abdallah King Saud University Jameela Kari (  jkari@kau.edu.sa ) King Abdulaziz University Research Article Page 1/17 Keywords: Children, nephrolithiasis, renal stones, nephrocalcinosis, outcome. Posted Date: January 18th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-136961/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/17 Keywords: Children, nephrolithiasis, renal stones, nephrocalcinosis, outcome. Posted Date: January 18th, 2021 DOI: https://doi.org/10.21203/rs.3.rs-136961/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License eywords: Children, nephrolithiasis, renal stones, nephrocalcinosis, outcome. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/17 Page 1/17 Abstract Background: Nephrolithiasis and nephrocalcinosis is uncommon in children; however, its incidence is increasing. Patients and Methods: A multicenter retrospective study of the clinical presentation, etiology, and outcome of childhood nephrolithiasis and compare it with nephrocalcinosis. Results: The study included 144 children; 93 with nephrolithiasis (formation of stones within renal pelvis or tubular lumen) and 51 with nephrocalcinosis. (deposition of calcium in the renal parenchyma) Mean age at presentation were 72 months and 54 months for nephrolithiasis and nephrocalcinosis, respectively. In 64.8% of the nephrolithiasis and 76% of nephrocalcinosis cases, a history of consanguinity was found. Congenital anomalies of the kidneys and urinary tract were present in 28% and 9.8% of those with nephrolithiasis and nephrocalcinosis, respectively. The most common symptoms of nephrolithiasis were flank pain (29%), hematuria (15%), and dysuria (11%). Urinary tract infection was the primary presentation in the nephrocalcinosis group (18%) followed by failure to thrive (16%), polyuria (12%), and dehydration (12%). The majority of nephrolithiasis cases were caused by metabolic disorders. In contrast, the most common underlying disorders for nephrocalcinosis were familial hypomagnesemia hypercalciuria nephrocalcinosis (35%), distal renal tubular acidosis (23%), and Bartter syndrome (6%). The majority of nephrolithiasis cases were caused by metabolic disorders. In contrast, the most common underlying disorders for nephrocalcinosis were familial hypomagnesemia hypercalciuria nephrocalcinosis (35%), distal renal tubular acidosis (23%), and Bartter syndrome (6%). Clinical outcomes were significantly better in children with nephrolithiasis than those with nephrocalcinosis who had radiological evidence of worsening/persistent calcinosis and progressed more frequently to chronic kidney disease (stage II-IV) and end stage kidney disease. Clinical outcomes were significantly better in children with nephrolithiasis than those with nephrocalcinosis who had radiological evidence of worsening/persistent calcinosis and progressed more frequently to chronic kidney disease (stage II-IV) and end stage kidney disease. Conclusion: The etiology of nephrolithiasis can be identified in many children. Nephrocalcinosis is associated with worse clinical outcomes related to kidney function and disease resolution as compared to nephrolithiasis. Study Design This multicenter retrospective study was conducted at two tertiary centers in KSA (King Abdulaziz University Hospital and King Saud University Medical City). Introduction Nephrolithiasis is relatively uncommon in the pediatric population; however, cases in children are increasing [1]. It is associated with significant long-term sequelae, including the morbidity caused by recurrent stones as well as the development of chronic kidney disease (CKD) and renal impairment. The annual incidence of pediatric nephrolithiasis has increased from 6–10% over the past 20 years in the United States [1], with an observation of a greatest increase among adolescent black girls [2]. The etiology is metabolic in most children, with hematuria and urinary tract infections as the most common presentations [3]. However, the clinical presentation varies with age, as flank pain may be seen in older children or adolescents whereas vague symptoms such as nausea, vomiting, and irritability are Page 2/17 Page 2/17 typical in younger children. Incidental findings on imaging studies have also been reported in a considerable proportion of affected children [4]. typical in younger children. Incidental findings on imaging studies have also been reported in a considerable proportion of affected children [4]. Kidney stone formation requires urine that has a higher solute concentration than its solubility [1]. Crystallization occurs due to an imbalance of promoters and inhibitors, and the attachment and growth of crystals into nephroliths due to epithelial abnormalities [5]. The management of pediatric nephrolithiasis includes urinary decompression, medical treatment of specific risk factors, and surgical intervention [6]. In this study, we report the epidemiology, etiology and outcome of childhood nephrolithiasis from two tertiary centers in the Kingdom of Saudi Arabia (KSA) and compare it with nephrocalcinosis. Patients And Methods The metabolic workup of nephrolithiasis was done using a spot urine sample and was interpreted as solute/creatinine ratio [8]. The test was repeated twice as a confirmatory method for hypercalciuria, hyperoxaluria, cystinuria, hypocitraturia, and hyperuricosuria. Hypercalciuria, hyperoxaluria, and hyperuricosuria were defined as a urinary solute: creatinine ratio greater than the 95th percentile as a function of age (Table 2). We also examined the associated acid-base and electrolytes disturbances such as metabolic acidosis and alkalosis, hypo- or hypernatremia, hypo- or hyperkalemia, hypo- or hyperchloremia, hypo- or hypercalcemia, and hypo- or hyperphosphatemia. Hypernatremia and hyponatremia were defined as a serum sodium level greater than 145 mmol or less than 135 mmol, respectively. Hyperkalemia was defined as a serum potassium greater than 5.5 meq/L in children and greater than 6 meq/L in neonates, and hypokalemia was defined as potassium level less than 3.5 meq/L. Hypercalcemia was defined as a serum calcium level greater than 2.6 mmoL, and hyperphosphatemia as a serum phosphate level greater than 1.58 mmol/L. Metabolic acidosis was defined as a pH level less than 7.35 and a serum bicarbonate less than 18 meq/L, whereas metabolic alkalosis was defined as a pH level greater than 7.45 and a serum bicarbonate level greater than 25 meq/L. Renal imaging (renal ultrasound, X-ray, and CT) was reviewed and used for classification, as previously mentioned. Genetic tests and stone analysis were performed when possible and were used as confirmatory tests to diagnose the underlying cause of kidney disease. All confirmed cases of renal stones or nephrocalcinosis were followed-up, and serial imaging studies (renal ultrasound and/or X-ray and/or helical CT) were performed to evaluate the outcome. eGFR was calculated using the Schwartz formula [7][9]. All clinical data about the stones such as the number, location, laterality, and grades of nephrocalcinosis were collected. All associated congenital anomalies of the kidney and urinary tract were reported. Stone analysis was performed for those who spontaneously passed their stones. We documented the used treatment modalities, which included conservative, extracorporeal shock wave lithotripsy (ECSL), or surgical therapies. The outcomes of nephrolithiasis or nephrocalcinosis were monitored and categorized as follows: spontaneous resolution, post intervention resolution, persistence, worsening or with new stone formation, or nephrocalcinosis. Cases that were missed during the follow-up were identified and documented. Spontaneous remission was defined as the disappearance of stones and/or nephrocalcinosis in two or more serial imaging studies. Patients And Methods Children diagnosed with renal stones or nephrocalcinosis were included and followed in the recruiting pediatric nephrology units during a period of eight years (between January 2010 and December 2018). Children diagnosed with renal stones or nephrocalcinosis were included and followed in the recruiting pediatric nephrology units during a period of eight years (between January 2010 and December 2018). The inclusion criteria included children (defined as those aged 14 years or younger) with a radiologic di i f l h l i i The inclusion criteria included children (defined as those aged 14 years or younger) with a radiologic diagnosis of renal stones or nephrocalcinosis. Renal stone (nephrolithiasis) was defined as the presence of a stone in two images excluding artifacts. Children who had spontaneous passage of a stone or had a stone removed by surgical intervention were included if the stone analysis was available. Nephrocalcinosis was defined as medullary calcification in a portion of the renal medulla without shadowing. We used either ultrasound or computerized tomography (CT) in detecting faint calcifications. Nephrocalcinosis was classified based on renal ultrasound findings as follows: mild (early hyper echogenicity in the periphery of the pyramids), moderate (diffuse hyperechoic pyramids), and severe (clumps of renal pyramids)[7]. All demographic and clinical data were collected from the patients’ electronic hospital records and included age at presentation, gender, creatinine level, estimated glomerular filtration rate (eGFR) at presentation, medical and surgical history, presence of a family history of nephrolithiasis or CKD, consanguinity, symptoms and signs at presentation, number and localization of stones as well as grades of nephrocalcinosis. All patients presenting with symptoms and/or signs of urinary tract infection (UTI) were screened using urine analysis and urine culture. Urine samples were collected either via transurethral catheterization for children younger than 3 years old or clean catch samples if more than 3 years old. We defined UTI as the presence of more than 5 white blood cells per high power field (hpf) and a positive Page 3/17 urine culture with more than 1,000,000 bacterial colony counts per ml. Urine collection was obtained from midstream urine or through a transient clean catheter for younger children. To reach the diagnosis of metabolic renal stones or nephrocalcinosis, we combined the data of urine metabolic work up, genetic tests with the relevant clinical data, and stone analysis. Patients And Methods Renal outcome was used as an indicator of morbidity and was assessed by measuring the reduction in eGFR compared with the initial GFR and determining the presence and severity of proteinuria. The incidence of mortality was reported. Results We have identified 144 patients who met the inclusion criteria (93 presented mainly with nephrolithiasis and 51 with nephrocalcinosis). The average age at presentation for children with nephrolithiasis was greater than that of those with nephrocalcinosis (mean of 72 months versus 54 months). Most patients with nephrolithiasis were male (66.7%), representing a male-to-female ratio of 2:1. On the contrary, nephrocalcinosis occurred mainly in females (60.4%), representing a male-to-female ratio of 0.66. Furthermore, a greater proportion of children with nephrolithiasis had normal kidney function at the time of presentation. The mean eGFR at time of presentation was higher in children with nephrolithiasis (148.4 versus 122.72 ml/min/1.73 m2) (Table 1). A history of consanguinity was found in 64.8% of the patients with nephrolithiasis (56% of the patients had parents who were first degree cousins). A quarter of this group had a family history of nephrolithiasis and a third had a family history of CKD (Table 1). In the nephrocalcinosis group, consanguinity was found in 76% of the cases. Of these, 35% had a family history of nephrolithiasis and 23.7% had a family history of CKD (Table 1). Congenital anomalies of the kidney and urinary tract were reported in 28% of children with renal stones and in 9.8% of those with nephrocalcinosis (Table 1). The most frequent presenting symptom of nephrolithiasis was flank pain (29%), followed by hematuria (15%) and dysuria (11%). Approximately 19% of the children presented with signs and symptoms of UTI, which was confirmed with a urine culture in 15% of the cases. Nephrolithiasis were incidentally discovered during routine investigations in 14% of the patients. Other symptoms such as failure to thrive, dehydration, and polyuria were reported less frequently (Fig. 1). UTI symptoms and signs were the main presentation in the nephrocalcinosis group (31%); however, UTI was confirmed with a urine culture in only 18% of the cases. Failure to thrive, polyuria, and dehydration were the presentation in 16%, 12%, and 12%, respectively, of the children with tubulopathy with nephrocalcinosis (Bartter syndrome and distal renal tubular acidosis). Nephrocalcinosis was incidentally diagnosed during routine investigations in 18% of the children. Other less frequent presentations included rickets, dysuria, and hematuria (Fig. 1). The etiology of nephrolithiasis was determined in 78% of the cases, whereas it was idiopathic in 22% of the children. Statistical analysis: Page 4/17 Page 4/17 All analyses were performed using STATA (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP) software. The proportion and mean for dichotomous and continuous variables, respectively, were measured to describe patients’ characteristics. Comparative analyses were performed using chi-square test for categorical variables and student t test for continuous variables. Statistical significance was determined using the 95% confidence interval and p-value of 0.05. All analyses were performed using STATA (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP) software. The proportion and mean for dichotomous and continuous variables, respectively, were measured to describe patients’ characteristics. Comparative analyses were performed using chi-square test for categorical variables and student t test for continuous variables. Statistical significance was determined using the 95% confidence interval and p-value of 0.05. Results Metabolic disorders were observed in 69% of the cases and included the following conditions: hyperoxaluria (18%), cystinuria (18%), hypercalciuria (12%), and hyperuricosuria (2%). A UTI Page 5/17 was documented in 20% of the cases. Distal renal tubular acidosis and the use of diuretics were reported in less than 5% of the children in this group. On the other hand, the underlying cause of nephrocalcinosis was determined in 86% of the cases (familial hypomagnesemia hypercalciuria nephrocalcinosis [FHHNC] 35%), distal renal tubular acidosis (23%), and Bartter’s syndrome (7%) (Fig. 2). was documented in 20% of the cases. Distal renal tubular acidosis and the use of diuretics were reported in less than 5% of the children in this group. On the other hand, the underlying cause of nephrocalcinosis was determined in 86% of the cases (familial hypomagnesemia hypercalciuria nephrocalcinosis [FHHNC] 35%), distal renal tubular acidosis (23%), and Bartter’s syndrome (7%) (Fig. 2). The results of urinary metabolites (solute/creatine ratio) and serum electrolytes for both the renal stone and nephrocalcinosis groups were represented as mean of solute/creatine ratio and 95% confidence interval (CI), which are summarized in Tables 3 and 4. The high frequency of FHHNC explains the higher incidence of hypermagnesuria represented as fractional excretion of magnesium (FEMg%) and hypercalciuria in the nephrocalcinosis group as compared to those with the nephrolithiasis group (p-value ≤ 0.001; Fig. 3A). Similarly, the incidence of hypokalemia, hypomagnesemia, and acid-base disturbance were significantly higher in the nephrocalcinosis group (p-value < 0.001, Fig. 3B). The clinical outcomes were significantly better in children with nephrolithiasis; among those a greater proportion showed spontaneous or post intervention improvement. In contrast, a greater proportion of patients with nephrocalcinosis had a radiological evidence of worsening or persistent calcinosis (Fig. 4A). In addition, patients with nephrocalcinosis progressed more frequently to CKD (stage II-IV) and end stage kidney disease compared to those with renal stone (Fig. 4B). Discussion A larger proportion of children with nephrocalcinosis presented with symptoms and signs of a UTI, which was confirmed in 18% of the cases. In literature, UTIs were reported in 10 to 17% of children with urolithiasis [11, 12]. The etiology of kidney stones in our study differs from that in a previous report that included 85 children with urolithiasis [11]. A-Rasheed et al., in their study, reported that 60% of the children had formed idiopathic stones, contrary to the 22% in our patients with renal stones[11]. However, they found that approximately 12% of the children in their study had hypercalciuria which is similar to the cases in our study. Additionally, a metabolic etiology was implicated in 10% of the children in their study, mainly in the form of cystinuria and primary hyperoxaluria[11]. While we found that metabolic causes were implicated in 69% of our patients, we also found that these were mainly in the form of hyperoxaluria (18%) and cystinuria (18%). These findings suggest the need to perform a metabolic screening test in all children with nephrolithiasis because a UTI, which is a common finding in these children, can mask an underlying metabolic etiology, which may be the primary cause [4]. A strong family history of urolithiasis has been reported in children with kidney stones [14, 15]. Although consanguinity has been reported in stone formers, these were reported in studies of adult patients [16, 17]. Studies conducted worldwide found that pediatric stone formers had a strong family history of urolithiasis. In a study conducted at a tertiary center in Brazil, the investigators reported 85% of the children in their study has a family history of urolithiasis [18]. Furthermore, the investigators found that 83.3% of the patients with metabolic changes had a family history of kidney stone disease. In other studies, it was reported that about 40% of pediatric stone formers had a family history of urolithiasis [19– 21]. In another study conducted in Iran [22], a family history of urolithiasis was reported in 62.7% of the 142 children with kidney stones. We found that 36.3% of children had a family history of nephrolithiasis and 23.1% had a family history of CKD, confirming the importance of family history in the occurrence of pediatric urolithiasis. The clinical outcome of kidney disease in children is poorly understood. Discussion Although the regional incidence of kidney stones is high in Saudi Arabia [8], pediatric nephrolithiasis is uncommon. However, in a country where children constitute approximately 40% of the population [9], it is essential to study the epidemiological and clinical features of this disease. Moreover, pediatric nephrolithiasis is associated with significant morbidity, mainly due to the potential of stone recurrence and, consequently, it should not be overlooked. Unfortunately, a paucity of reports has described the epidemiological and clinical features of nephrolithiasis in Saudi pediatric patients. This report attempts to describe the epidemiology and underlying causes of pediatric nephrolithiasis and nephrocalcinosis. According to a study that evaluated urolithiasis in the Middle East, there has been a change in the pattern and etiology of pediatric nephrolithiasis in Saudi Arabia [10]. In one study, it was reported that cases of pediatric nephrolithiasis constitute < 1% of all kidney stones [11, 12]. The mean age at diagnosis was 12 years, with a male-to-female ratio of 2:1. In a single-center study conducted in Jordan, it was reported that pediatric urolithiasis constituted 1.85% of all cases of stones [13]. In our report, the mean age at presentation for children with renal stones was 72 months while that of nephrocalcinosis cases was 54 months. Of note is that the male-to-female ratio among our patients with nephrolithiasis was 2:1, which is consistent with those of other investigators [10, 11]. On the contrary, nephrocalcinosis occurred mainly in females, with a male-to-female ratio of 0.66. In a report from Jordan, the mean age of occurrence of pediatric urolithiasis was 14 years, with a male-to-female ratio at 2:1 [13]. Page 6/17 Page 6/17 In another hospital-based study conducted in Iraq, the investigators found that kidney stones occurred at an early age, with most cases diagnosed in children less than 5 years old [14]. Similar to our study, the authors reported a higher prevalence among males, with a male-to-female ratio of 2.8:1. In another hospital-based study conducted in Iraq, the investigators found that kidney stones occurred at an early age, with most cases diagnosed in children less than 5 years old [14]. Similar to our study, the authors reported a higher prevalence among males, with a male-to-female ratio of 2.8:1. In our study, approximately 19% of the children with nephrolithiasis presented with signs and symptoms of a UTI, which was confirmed with a urine culture in 15% of the cases. Discussion According to a recent report, male and female patients have similar hospitalization rates and frequency of stone episodes [23]. In an older study that attempted to investigate clinical outcomes in children with urolithiasis, the investigators were unable to comment on the outcome of urolithiasis in their patients [21]. Although we found that clinical outcomes were better in children with renal stones, children with nephrocalcinosis showed radiological evidence of worsening of the disease. However, we believe these results only provide preliminary evidence of the disease course in pediatric kidney stone formers. There are several limitations to this study that merit consideration. One of the main limitations is retrospective nature of the study and relatively small sample size. Page 7/17 Page 7/17 Page 7/17 Availability of data and materials: available Competing interests: The authors report no conflicts of interest. Competing interests: The authors report no conflicts of interest. Funding: This work was supported by the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah (G:257-140-1439) Funding: This work was supported by the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah (G:257-140-1439) Conclusion The etiology of nephrolithiasis was identified in most of our study population. This was achieved through metabolic screening of all suspected cases of pediatric nephrolithiasis, as metabolic causes are implicated in most of these patients. We also showed that most pediatric stone formers have better clinical outcomes compared to patients with nephrocalcinosis which was associated with worse outcomes related to kidney function and disease resolution. Consent for publication: Not required. Consent for publication: Not required. Declarations Ethics approval and consent to participate: The study was approved by the Biomedical Ethics Research Committees at college of medicine at King Abdulaziz University and also approved by institutional review board at college of medicine at King Saud University.Consent from participants was not required as this was a retrospective study using data collected for routine clinical practice. All methods were carried out in accordance with relevant guidelines and regulation Consent for publication: Not required. Authors' contributions: Page 8/17 KAA: Idea, coordination of the study , writing and editing the manuscript MAS: Writing and editing the manuscript ASA; Analysis of the data, writing up the results and revising the manuscript MHT: Writing and editing the manuscript ZA; Collecting data and review the manuscript MSA; Collecting data and review the manuscript NGA; Collecting data , review the manuscript NMK; Collecting data , review the manuscript ZFZ; Writing and editing the manuscript Page 8/17 JAK: Idea, application for grant, coordination of the study, writing and editing the manuscript JAK: Idea, application for grant, coordination of the study, writing and editing the manuscript Acknowledgements The authors acknowledge the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah, KSA for technical and financial support (G:257-140-1439) This work was also supported by the College of Medicine, Research Center, Deanship of Scientific Research, King Saud University, Riyadh, KSA. The authors acknowledge the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah, KSA for technical and financial support (G:257-140-1439) This work was also supported by the Jeddah, KSA for technical and financial support (G:257-140-1439) This work was also supported by the College of Medicine, Research Center, Deanship of Scientific Research, King Saud University, Riyadh, KSA. College of Medicine, Research Center, Deanship of Scientific Research, King Saud University, Riyadh, KSA. College of Medicine, Research Center, Deanship of Scientific Research, Kin References Sepahi MA, Heidari A, Shajari A: Clinical manifestations and etiology of renal stones in children less than 14 years age. Saudi J Kidney Dis Transpl 2010, 21(1):181–184. 15. Sepahi MA, Heidari A, Shajari A: Clinical manifestations and etiology of renal stones in children less than 14 years age. Saudi J Kidney Dis Transpl 2010, 21(1):181–184. 16. Abbassene F, Maizia A, Messaoudi N, Bendahmane L, Boukharouba H, Daudon M, Addou A: Adult urolithiasis in Western Algeria: a study of 1104 cases. Tunis Med 2020, 98(5):396–403. 17. Dello Strologo L, Pras E, Pontesilli C, Beccia E, Ricci-Barbini V, de Sanctis L, Ponzone A, Gallucci M, Bisceglia L, Zelante L et al: Comparison between SLC3A1 and SLC7A9 cystinuria patients and carriers: a need for a new classification. J Am Soc Nephrol 2002, 13(10):2547–2553. 18. Amancio L, Fedrizzi M, Bresolin NL, Penido MG: Pediatric urolithiasis: experience at a tertiary care pediatric hospital. J Bras Nefrol 2016, 38(1):90–98. 18. Amancio L, Fedrizzi M, Bresolin NL, Penido MG: Pediatric urolithiasis: experience at a tertiary care pediatric hospital. J Bras Nefrol 2016, 38(1):90–98. 19. Alemzadeh-Ansari MH, Valavi E, Ahmadzadeh A: Predisposing factors for infantile urinary calculus in south-west of Iran. Iran J Kidney Dis 2014, 8(1):53–57. 19. Alemzadeh-Ansari MH, Valavi E, Ahmadzadeh A: Predisposing factors for infantile urinary calculus in south-west of Iran. Iran J Kidney Dis 2014, 8(1):53–57. 20. Hoppe B, Kemper MJ: Diagnostic examination of the child with urolithiasis or nephrocalcinosis. Pediatr Nephrol 2010, 25(3):403–413. 20. Hoppe B, Kemper MJ: Diagnostic examination of the child with urolithiasis or nephrocalcinosis. Pediatr Nephrol 2010, 25(3):403–413. 21. VanDervoort K, Wiesen J, Frank R, Vento S, Crosby V, Chandra M, Trachtman H: Urolithiasis in pediatric patients: a single center study of incidence, clinical presentation and outcome. J Urol 2007, 177(6):2300–2305. 21. VanDervoort K, Wiesen J, Frank R, Vento S, Crosby V, Chandra M, Trachtman H: Urolithiasis in pediatric patients: a single center study of incidence, clinical presentation and outcome. J Urol 2007, 177(6):2300–2305. 22. Naseri M, Varasteh AR, Alamdaran SA: Metabolic factors associated with urinary calculi in children. Iran J Kidney Dis 2010, 4(1):32–38. 22. Naseri M, Varasteh AR, Alamdaran SA: Metabolic factors associated with urinary calculi in children. Iran J Kidney Dis 2010, 4(1):32–38. 23. Schwaderer AL, Raina R, Khare A, Safadi F, Moe SM, Kusumi K: Comparison of Risk Factors for Pediatric Kidney Stone Formation: The Effects of Sex. Front Pediatr 2019, 7:32. Tables References 1. Van Batavia JP, Tasian GE: Clinical effectiveness in the diagnosis and acute management of pediatric nephrolithiasis. Int J Surg 2016, 36(Pt D):698–704. 2. Tasian GE, Ross ME, Song L, Sas DJ, Keren R, Denburg MR, Chu DI, Copelovitch L, Saigal CS, Furth SL: Annual Incidence of Nephrolithiasis among Children and Adults in South Carolina from 1997 to 2012. Clin J Am Soc Nephrol 2016, 11(3):488–496. 3. Alpay H, Ozen A, Gokce I, Biyikli N: Clinical and metabolic features of urolithiasis and microlithiasis in children. Pediatr Nephrol 2009, 24(11):2203–2209. 3. Alpay H, Ozen A, Gokce I, Biyikli N: Clinical and metabolic features of urolithiasis and microlithiasis in children. Pediatr Nephrol 2009, 24(11):2203–2209. 4. Chu DI, Tasian GE, Copelovitch L: Pediatric Kidney Stones - Avoidance and Treatment. Curr Treat Options Pediatr 2016, 2(2):104–111. 5. Ratkalkar VN, Kleinman JG: Mechanisms of Stone Formation. Clin Rev Bone Miner Metab 2011, 9(3– 4):187–197. 6. Hernandez JD, Ellison JS, Lendvay TS: Current Trends, Evaluation, and Management of Pediatric Nephrolithiasis. JAMA Pediatr 2015, 169(10):964–970. 7. Salvador CL, Tøndel C, Rowe AD, Bjerre A, Brun A, Brackman D, Mørkrid L: Estimating glomerular filtration rate in children: evaluation of creatinine- and cystatin C-based equations. Pediatr Nephrol 2019, 34(2):301–311. 8. El-Faqih SR: Epidemiology of Stone Disease in Saudi Arabia with an Overview of the Regional Differences. In: Urolithiasis. edn.: Springer; 2012: 77–83. 9. General Authority for Statistics, Kingdom of Saudi Arabia. https://www.stats.gov.sa/en Accessed 26 Dec 2020 10. El-Faqih SR, Hussain I: Urolithiasis in the Middle East: Epidemiology and pathogenesis. In: The Management of Lithiasis. edn.: Springer; 1997: 35–41. 11. Al-Rasheed SA, el-Faqih SR, Husain I, Abdurrahman M, al-Mugeirin MM: The aetiological and clinical pattern of childhood urolithiasis in Saudi Arabia. Int Urol Nephrol 1995, 27(4):349–355. 12. Al-Rasheed S, Al Jurayyan NA, Al Nasser MN, Al-Mugeiren MM, Al-Salloum AA, Petterson BA: Nephrolithiasis in children and adolescents in the South Western region of saudi arabia. Saudi J Kidney Dis Transpl 1995, 6(4):396–399. 13. Alsheyab F, Bani Hani I, Yousef M: Chemical composition of urinary calculi in North Jordan. J Biol Sci 2007, 7(7):1290–1292. 13. Alsheyab F, Bani Hani I, Yousef M: Chemical composition of urinary calculi in North Jordan. J Biol Sci 2007, 7(7):1290–1292. Page 9/17 Page 9/17 14. Aboud MJ, Kadhim MM: Review for urolithiasis in patients attending the paediatric surgery unit at Al- Qadisiya Governorate, Iraq. N Iraqi J Med 2010, 6(3):17–23. 15. Tables Page 10/17 Table 1 Patients’ baseline demographic and disease characteristics. Characteristics Renal Stone Nephrocalcinosis Estimate 95% CI Estimate 95% CI Age at presentation (mean, month) 72.22 62.7–84.7 54.25 39.1–63.6 Male sex (%) 66.7 56.6–75.7 39.2 26.6–53.0 Creatinine (mean, at presentation) 60.24 39.6–80.7 83.90 41.2–126.5 GFR (mean, at presentation) 148.31 136.2–160.3 122.72 104.4–140.9 Consanguinity (%) 64.8 54.6–74.1 78.7 65.3–88.7 Family history of nephrolithiasis 23.1 15.3–32.6 59.6 45.2–72.8 Family history of renal disease 36.3 26.9–46.5 51.1 33.7–60.7 Associated kidney and urinary tract (%) 28 23.0–44.6 9.80 0.3–20.3 Initial renal status Normal (%) 86 77.0–92.0 72 59.2–82.6 Stage (II-IV) (%) 12 6.40–19.6 24 13.4–36.6 ESRD (%) 2 0.4–6.90 4 0.7–12.4 Abbreviations: GFR, ml/min/1.73 m2, Creatinine µmol Table 1 Table 1 Patients’ baseline demographic and disease characteristics. Page 11/17 Page 11/17 Table 2 Table 2 Cut-off level of solute:creatine ratio in relation to age group Cut-off level of solute:creatine ratio in relation to age group Item Range (Month) Range (year) Normal Value Ca/Creatine (mmol/mmol) 0–12 0–1 2.2 12–36 1–3 1.5 36–60 3–5 1.1 60–84 5–7 0.8 > 84 > 7 0.6 Oxalate/Creatine (mmol/mmol) 0–12 0–1 0.17 12–24 1–2 0.13 24–36 2–3 0.1 36–60 3–5 0.08 60–84 5–7 0.07 > 84 > 7 0.06 Cystine/Creatine (mmol/mol) 0–1 0->1 85 1–6 53 > 6 18 Citrate/Creatine (mmol/mmol) 0–60 0–5 0.12 > 60 > 5 0.08 Uric acid/Creatine (mmol/mmol) 0–12 0–1 1.5 12–36 1–3 1.3 36–60 3–5 1 60–120 5–10 0.6 > 120 > 10 0.4 Range (Month) Range (year) Normal Value Page 12/17 Page 12/17 Page 12/17 Table 3 Urine chemistry results among patients with nephrolithiasis and nephrocalcinosis Characteristics Renal Stone Nephrocalcinosis P- value Estimate 95% CI Estimate 95% CI Recurrent UTI (%) 36.71% 25.8– 47.6 17.24% 0.3–20.3 0.09 Ca/Creatinine (mean, mmol/ mmol) 1.35 0.7–2.1 2.07 1.3–2.7 0.168 Oxalate/Creatinine (mean, mmol/ mmol) 2.18 0–4.9 7.12 0–19.3 0.433 Cystine/Creatinine (mean, mmol/ mol) 29.47 17.2– 41.6 5.28 0–11.7 0.408 Citrate/Creatinine (mean, mmol/ mmol) 6.86 4.7–8.9 8.63 4.2–13.1 0.449 Uric acid/Creatinine (mean, mmol/ mmol) 0.63 0.4–0.8 0.74 0.3–1.1 0.718 FEMg% (mean) 2.07 1.7–2.3 8.78 5.7–11.7 < 0.001 TRP% (mean) 91.82 90.8– 92.0 87.76 84.4– 91.0 0.02 Abbreviations: CI, confidence interval; FEMg%: fraction of excretion of magnesium; TRP: tubular reabsorption of phosphate %; UTI, urinary tract infection. Table 3 TRP: tubular reabsorption of phosphate %; UTI, urinary tract infection. Tables Page 13/17 Page 13/17 Table 4 Table 4 Serum electrolyte results among patients with nephrolithiasis and nephrocalcinosis Characteristics Renal Stone Nephrocalcinosis P-value Estimate 95% CI Estimate 95% CI Serum sodium (mean) 139.04 138.2–139.8 139.43 138.3–140.4 0.567 Serum K (mean) 4.15 4.04–4.25 3.78 3.58–3.95 0.001 Serum chloride(mean) 103.66 102.83–104.46 103.18 101.37–104.97 0.58 pH (mean) 7.37 7.36–7.37 7.38 7.33–7.42 0.505 Serum HCO3 (mean) 23.59 23.15–24.16 25.06 23.53–26.58 0.086 Serum Ca (mean) 2.38 2.35–2.40 2.28 2.21–2.32 0.001 Serum phosphate (mean) 1.57 1.49–1.62 1.47 1.32–1.59 0.184 Serum Mg (mean) 0.83 0.63–1.02 0.73 0.67–0.76 < 0.001 Page 14/17 Serum Mg (mean) 0.83 0.63–1.02 0.73 0.67–0.76 < 0.001 Figures Figure 1 Page 14/17 Figures Figure 1 Figures Figure 1 Page 14/17 Page 14/17 Clinical presentations of patients with nephrolithiasis and nephrocalcinosis Abbreviations: FTT: failure to thrive, UTI: urinary tract infection. Clinical presentations of patients with nephrolithiasis and nephrocalcinosis Abbreviations: FTT: failure to thrive, UTI: urinary tract infection. Figure 2 Figure 2 Comorbidities associated with renal stones and nephrocalcinosis Abbreviations: FHHNC: familial hypomagnesemia hypercalciuria nephrocalcinosis. Page 15/17 Page 15/17 Figure 3 Figure 3 Figure 3 Laboratory outcomes (urinary [A] and serum [B]) among patients with nephrolithiasis and nephrocalcinosis Page 16/17 Page 16/17 Page 16/17 gure 4 nical outcomes among patients with nephrolithiasis and nephrocalcinosis Abbreviations Figure 4 Clinical outcomes among patients with nephrolithiasis and nephrocalcinosis Abbreviations: CKD: chronic kidney disease, ESRD : end stage of renal disease. Page 17/17 Page 17/17
https://openalex.org/W2938629602	https://www.nature.com/articles/s41598-019-42308-5.pdf	English	null	Gene Expression Profiles Associated with Brain Aging are Altered in Schizophrenia	Scientific reports	2,019	cc-by	7,971	www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Sarven Sabunciyan Sarven Sabunciyan Existence of aging associated transcriptional differences in the schizophrenia brain was investigated in RNA sequencing data from 610 postmortem Dorso-Lateral Pre-Frontal Cortex (DLPFC) samples in the CommondMind Consortium (CMC) and the psychENCODE cohorts. This analysis discovered that the trajectory of gene expression changes that occur during brain aging differed between schizophrenia cases and unaffected controls. Mainly, the identified gene expression differences between the diagnosis groups shrank in magnitude following 60 years of age. A differential expression analysis restricted to the 40 to 60 year age group identified 556 statistically significant loci that replicated and had highly consistent gene expression fold changes in the two cohorts. An interaction between age and diagnosis in the wider psychENCODE cohort was also detected. Gene set enrichment analysis discovered disruptions in mitochondria, RNA splicing and phosphoprotein gene pathways. The identified differentially expressed genes in the two cohorts were also significantly enriched in genomic regions associated with schizophrenia although no enrichment was observed for differentially expressed genes identified in the 40 to 60 year age group. This work implicates disruptions to the normal brain aging processes in the pathology of schizophrenia and demonstrates the need for age stratification in schizophrenia postmortem brain gene expression studies. Gene expression in the postmortem schizophrenia brain has been extensively characterized by microarray and next generation sequencing studies1–6. The consensus from these studies is that there are many subtle expression differences in individual genes that likely alter the functioning of various gene pathways in schizophrenia2,4,7,8. The standard analysis approach employed by these studies assumes that gene expression and age have a linear relationship. Thus, appropriate statistical methods are used to adjust for the effect of age on gene expression. However, epidemiological studies consistently find excess early mortality8,9 in disease. In addition, studies that estimate brain age based on neuroanatomical structures10 and integrity of the white matter11,12 have found accel- erated aging in schizophrenia. As transcriptional changes are likely to accompany neuroanatomical changes in the brain, the relationship between age and disease is likely to be more complex than assumed in current analysis pipelines. Therefore, in this work publicly available RNA sequencing data was reanalyzed to determine whether the transcriptome of the aging brain differs between unaffected controls and schizophrenia cases. Received: 15 August 2018 Accepted: 27 March 2019 Published: xx xx xxxx epartment of Pediatrics, Johns Hopkins University, Baltimore, MD, 21287, USA. Correspondence and requests fo materials should be addressed to S.S. (email: ssabunc1@jhmi.edu) Gene Expression Profiles Associated with Brain Aging are Altered in Schizophrenia Received: 15 August 2018 Accepted: 27 March 2019 Published: xx xx xxxx Received: 15 August 2018 Accepted: 27 March 2019 Published: xx xx xxxx Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 Results G Gene expression profiles in the aging schizophrenia brain is altered. In order to study aging in schizophrenia, RNA sequencing data generated from the BA9 of the CMC and BA46 of the psychENCODE post- mortem brain collections were reanalyzed. As CMC has the largest number of brain samples with RNA sequenc- ing data and gene expression differences in schizophrenia have already been identified in this cohort1,13, we reasoned that this cohort likely has the necessary statistical power to detect differences in our analysis. Following the exclusion criteria (see Methods) 237 unaffected controls and 239 schizophrenia subjects between the ages of 25 to 90 remained (Fig. 1A). In the psychENCODE cohort only BA46 samples from schizophrenia and control cases that were between the ages of 25 and 65 were included (See Methods). This left a total of 67 cases and 67 control samples (Fig. 1B). Gene count tables were generated for the CMC cohort using the Ensembl annotation for the hg19 human genome build and the DESeq214 package from the bioconductor project was used to identify differentially Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 www.nature.com/scientificreports/ Figure 1. Demographic Data. Histogram of diagnoses by age for (A) the CMC and (B) the psychENCODE Cohorts. The psychENCODE cohort has fewer samples and a narrower age range. Figure 1. Demographic Data. Histogram of diagnoses by age for (A) the CMC and (B) the psychENCODE Cohorts. The psychENCODE cohort has fewer samples and a narrower age range. expressed loci. The following statistical design that assumes an interaction between diagnoses and age was u (see Methods for details): + + + + + + + ~Diagnosis Age Age: Diagnosis Gender Ethnicity PMI RIN RuvFactors PMI refers to Post Mortem Interval whereas RIN is the RNA Integrity Number, which is an estimate of RNA quality15,16. RuvFactors were calculated using the RuvSeq package which is a statistical method to remove unwanted variation from RNA sequencing data17. Based on these criteria, Wald tests identified 7434 differentially expressed genes in which the adjusted p-value (corrected for multiple testing) was <0.1 and each gene on average had at least 20 reads per sample. Similar to previous studies, we found moderate differences in fold change in the statistically significant differences we identified (Fig. 2 and Supplementary Fig. 1 plots the top 100 loci with the lowest p-values). Results G These graphs revealed that the difference in gene expression levels between schizophrenia cases and unaffected controls vary with age (Fig. 2). The maximum gene expression difference in the CMC was consistently between 40 and 60 years of age (Supplementary Figure 1) with the caveat that there are relatively few samples under the age of 40 in the CMC cohort. For many loci the expression differences at ages older than 60 gets smaller and the direction of the change is reversed. In order to summarize the results of the differential expression analysis, the normalized difference between schizophrenia cases and unaffected controls was plotted for each locus (Fig. 3 and Supplementary Fig. 2. See Normalized Difference Plots in Methods). Briefly, schizo- phrenia expression levels (the blue lines in Fig. 2) were subtracted from control expression levels (the red lines in Fig. 2) for each age. The resulting differences in expression at each age were plotted for every locus (Fig. 3 - See methods for details). In order to avoid over plotting, the differentially expressed loci were subdivided into those with a maximum difference at ages 60 or less and were 1) more expressed (3114 loci) or 2) less expressed (3369 loci) in controls (Fig. 3) and those with a maximum difference at ages over 60 (Supplementary Fig. 2) that were 3) more expressed (405 loci) or 4) less expressed (546 loci) in controls. These plots verified that the difference in gene expression between cases and controls is not constant across age. Although, the CMC cohort contains 307 Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 2 www.nature.com/scientificreports/ Figure 2. Gene Expression Differences Between Schizophrenia and Control Brains During Aging Representative plots of the most common gene expression trajectories found in the CMC analysis. Observed normalized read counts were plotted vs age and the loess function was used to fit the best line through the data points. Figure 2. Gene Expression Differences Between Schizophrenia and Control Brains During Aging Representative plots of the most common gene expression trajectories found in the CMC analysis. Observed normalized read counts were plotted vs age and the loess function was used to fit the best line through the data points. Figure 3. Results G Gene Expression Differences Between Controls and Schizophrenia At Each Age The normalized difference in gene expression between control and schizophrenia samples is plotted for ages 25 to 90 for genes identified to be differentially expressed in the CMC by Wald tests using DESeq2. As the values at leach locus are normalized to the observed maximum difference, only values between 1 and −1 are possible. In order to avoid over plotting, the loci were grouped into those with (A) positive maximum difference (expression higher in controls) occurring at age 60 or less (3114 loci), (B) negative maximum difference (expression higher in schizophrenia) occurring at age 60 or less (3369 loci). Figure 3. Gene Expression Differences Between Controls and Schizophrenia At Each Age The normalized difference in gene expression between control and schizophrenia samples is plotted for ages 25 to 90 for genes identified to be differentially expressed in the CMC by Wald tests using DESeq2. As the values at leach locus are normalized to the observed maximum difference, only values between 1 and −1 are possible. In order to avoid over plotting, the loci were grouped into those with (A) positive maximum difference (expression higher in controls) occurring at age 60 or less (3114 loci), (B) negative maximum difference (expression higher in schizophrenia) occurring at age 60 or less (3369 loci). samples over the age of 60 there are only 33 samples under the age of 40. Therefore, the smaller gene expression differences observed between cases and controls under the age of 40 needs to be interpreted with caution. Age stratification improves consistency between studies. As the largest expression differences between schizophrenia and control samples occurred consistently between 40 and 60 years of age in the CMC analysis, differential expression analysis for this age group in the CMC and psychENCODE cohorts was per- formed. Eighty-two cases and 54 controls were present in the CMC and 51 cases and 48 controls were present in the psychENCODE cohorts. Gene count tables were generated based on the Ensembl hg19 annotation for both cohorts. For this analysis the interaction term was removed and the following regression formula was used: + + + + + + ~Diagnosis Age RIN PMI Race Sex RuvFactors The resulting analysis identified 2167 differentially expressed genes in the CMC and 4086 in psychENCODE that had an adjusted p-value of less than 0.1 and at least 20 reads per gene. Results G Category Term Gene Count Fold Enrichment Bonferroni SP_PIR_KEYWORDS mitochondrion 43 2.16 1.74E-03 SP_PIR_KEYWORDS transit peptide 30 2.64 1.78E-03 SP_PIR_KEYWORDS respiratory chain 11 6.48 2.84E-03 UP_SEQ_FEATURE transit peptide:Mitochondrion 30 2.67 4.61E-03 GOTERM_CC_FAT GO:0005739~mitochondrion 51 1.88 5.30E-03 GOTERM_CC_FAT GO:0070469~respiratory chain 11 5.88 5.54E-03 GOTERM_CC_FAT GO:0044429~mitochondrial part 33 2.22 1.20E-02 GOTERM_BP_FAT GO:0030198~extracellular matrix organization 13 5 1.95E-02 SP_PIR_KEYWORDS mitochondrion inner membrane 16 3.47 2.64E-02 KEGG_PATHWAY hsa05010:Alzheimer’s disease 14 3.31 3.17E-02 KEGG_PATHWAY hsa05012:Parkinson’s disease 12 3.61 5.05E-02 KEGG_PATHWAY hsa00190:Oxidative phosphorylation 12 3.56 5.75E-02 GOTERM_CC_FAT GO:0005746~mitochondrial respiratory chain 9 5.63 6.11E-02 GOTERM_CC_FAT GO:0019866~organelle inner membrane 21 2.56 7.50E-02 GOTERM_CC_FAT GO:0044455~mitochondrial membrane part 12 3.85 9.58E-02 Table 1. Gene Set Enrichment Analysis For Differentially Expressed Loci In The 40 To 60 Year Age Group. Table 1. Gene Set Enrichment Analysis For Differentially Expressed Loci In The 40 To 60 Year Age Group. genes are mostly associated with mitochondrial functions (fold enrichment 1.88–6.48, p-value 0.0017–0.1), a biological pathway implicated to be involved in the pathology of schizophrenia (Table 1). Gene set enrichment analysis implicates similar gene pathways in the CMC and The psychEN- CODE cohorts. Although the psychENCODE cohort was much smaller and only spanned an age range of 25 to 65, the existence of an interaction between age and diagnosis was investigated in this cohort. Thus the analysis performed on the CMC cohort using the formula ~Diagnosis + Age + Age: Diagnosis + Gender + Et hnicity + PMI + RIN + RuvFactors was repeated. The threshold for statistical significance was again set at an adjusted p-value of <0.1 and a minimum average read count of 20/sample. Based on these criteria we identified 29 loci to be differentially expressed (Supplementary Table 2). Fifteen out of these loci were also differentially expressed in the CMC cohort (11 in the same direction). Gene set enrichment analysis on the 29 identified loci did not yield statistically significant results. However, a plot of the p-values for all genes tested in the psychEN- CODE cohort revealed that there are many nominally significant genes (unadjusted p-value <0.05) that failed to reach the threshold of statistical significance following multiple testing (Supplementary Fig. 3) suggesting that this small cohort might be underpowered. Results G After excluding results driven by out- liers (using cooks distance) 556 loci replicated between the cohorts (Fig. 4, Supplementary Table 1). The p-values ranged from 1.25e-05 to 0.1 (mean 0.045) for the CMC cohort and 2.71e-09 to 0.1 (mean 0.027) for the psychEN- CODE cohort whereas the absolute fold change ranged from 1.05–1.86 (mean 1.18) for the CMC and 1.04–2.16 (mean 1.17) for psychENCODE. In order to determine the consistency of the changes between the two cohorts, the fold change difference in the two studies was plotted (Fig. 4A). The fold change difference between the two studies were highly correlated (r2 = 0.82) Boxplots were used to gage the variability between diagnosis groups (Fig. 4B). Gene set enrichment analysis performed using the DAVID annotation tool18 revealed that these 556 Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 3 www.nature.com/scientificreports/ Figure 4. Gene Expression Differences For the 40 to 60 Age Group. (A) The fold change differences detected in the psychENCODE plotted against the fold change differences detected in the CMC for the 556 loci identified to be significantly differentially expressed in both cohorts at the 40 to 60 year old age group. (B) Boxplots of differentially expressed loci selected based on significance level and/or relevance to disease. The x-axis label CM-C is CommonMind Control, CM-S is CommonMind Schizophrenia, PE-C is psychENCODE Control, PE-S is psychENCODE Schizophrenia. Figure 4. Gene Expression Differences For the 40 to 60 Age Group. (A) The fold change differences detected in the psychENCODE plotted against the fold change differences detected in the CMC for the 556 loci identified to be significantly differentially expressed in both cohorts at the 40 to 60 year old age group. (B) Boxplots of differentially expressed loci selected based on significance level and/or relevance to disease. The x-axis label CM-C is CommonMind Control, CM-S is CommonMind Schizophrenia, PE-C is psychENCODE Control, PE-S is psychENCODE Schizophrenia. Discussion Th f The process of brain aging differs significantly between schizophrenia cases and unaffected controls. The level of differential expression between diagnoses groups in the brain varies with age highlighting the complex affect aging has on disease gene expression profiles. As the largest gene expression level differences in disease for the CMC were in the 40 to 60 year age range, we compared this age group between the CMC and psychENCODE. Five hundred fifty six loci enriched in mitochondrial functions were identified that surpassed the threshold for multiple testing in both the CMC and psychENCODE cohorts. The consistency of fold change differences (r2 = 0.82) between the two cohorts provides further confidence in this finding. In comparison, the analysis of the entire CMC cohort yields much smaller gene expression difference compared to other cohorts - a mean of 1.09 and a range of 1.03–1.33 fold1. The similar differences observed between the CMC and psychENCODE when the analysis is restricted to the 40 to 60 year age group strongly suggests that age stratification in schizophrenia gene expression studies is beneficial. Although the largest gene expression differences observed in the CMC data set was between the ages of 40 and 60, the scarcity of samples under 40 years of age prevented us from characterizing gene expression changes in younger subjects. Clearly, ages from late adolescence to early adulthood are extremely important to schizophrenia pathophysiology as disease onset occurs at this stage. However, since most psychiatric brain banks lack samples at these young ages, we are unable to perform a thorough analysis of a key stage in schiz- ophrenia and are likely missing important gene expression changes associated with disease pathology. Evidence for a statistical interaction between age and schizophrenia was discovered in both the CMC and the psychEN- CODE cohorts. Gene set enrichment analysis revealed disruptions in splicing and phosphoprotein pathways to be common to both cohorts. In addition, the differentially expressed genes identified were enriched in the genomic regions associated with schizophrenia. This finding was especially strong in the CMC cohort and encompassed more than half of all genic regions associated with disease (54 out of 95).i g g ( ) Modeling an interaction between age and diagnosis did not eliminate discrepant findings between the CMC and psychENCODE cohorts. One reason for this discrepancy may be related to the differences between the cohorts. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Category Term Gene Count Fold Enrichment Bonferroni CommonMind SP_PIR_KEYWORDS acetylation 1070 1.50 4.75E-57 SP_PIR_KEYWORDS phosphoprotein 2375 1.21 3.55E-40 GOTERM_CC_FAT GO:0005739~mitochondrion 458 1.58 6.12E-28 GOTERM_CC_FAT GO:0030529~ribonucleoprotein complex 250 1.82 3.61E-24 SP_PIR_KEYWORDS mitochondrion 359 1.60 9.83E-22 SP_PIR_KEYWORDS alternative splicing 2325 1.15 4.17E-21 GOTERM_CC_FAT GO:0070013~intracellular organelle lumen 647 1.36 1.38E-19 GOTERM_CC_FAT GO:0044429~mitochondrial part 266 1.67 3.02E-19 GOTERM_CC_FAT GO:0005840~ribosome 125 2.18 3.11E-19 SP_PIR_KEYWORDS ribonucleoprotein 152 2.01 3.53E-19 PsychENCODE SP_PIR_KEYWORDS alternative splicing 1006 1.19 1.25E-10 UP_SEQ_FEATURE splice variant 1002 1.18 5.28E-09 SP_PIR_KEYWORDS phosphoprotein 957 1.16 1.66E-07 SP_PIR_KEYWORDS coiled coil 280 1.22 9.15E-02 Table 2. Gene Set Enrichment Analysis Results. Only the 10 most significant results listed for CommonMind. Table 2. Gene Set Enrichment Analysis Results. Only the 10 most significant results listed for CommonMind. the psychENCODE cohort were alternative splicing (p = 1.25E-10), splice variant (p = 5.28-09) phosphoprotein (p = 1.66E-07) and coiled coil pathways (p = 0.09) (Table 2). Differentially Expressed Genes Identified In The CMC And The psychENCODE Cohorts Are Enriched In The 108 Regions Associated With Schizophrenia. In order to determine whether the differentially expressed loci identified in the CMC and the psychENCODE cohorts are enriched near the genomic regions associated with schizophrenia19, Inrich analysis20 was performed. The 556 loci identified in the 40 to 60 year age group were not enriched in genomic regions associated with schizophrenia. However, out of the 95 schiz- ophrenia regions that contain genes (based on the Ensembl gene annotations), the differentially expressed genes identified in the CMC at a threshold of adjusted p-value <0.1 overlapped 54 genomic regions associated with schizophrenia (Inrich analysis adjusted p = 0.01, Supplementary Table 4). Repeating the Inrich analysis for dif- ferentially expressed genes that had an adjusted p-value of <0.05 (overlapped 50 GWAS regions, Inrich adjusted p = 0.0024) and < 0.01 (overlapped 35 GWAS regions, Inrich adjusted p = 0.0016) improved the significance of the enrichment. The nominally significant psychENCODE loci overlapped 33 genomic regions associated with schizophrenia (Inrich p = 0.02) (Supplementary Table 5). Twenty-six out of the thirty-three regions (76% overlap) identified in the psychENCODE cohort were also statistically significant in the Inrich CMC analysis (Supplementary Tables 4 and 5). Results G Gene set enrichment analysis was performed using the DAVID annotation tool18 on the 2976 nominally expressed genes in psychENCODE (unadjusted p < 0.05, minimum mean read count 20 and outlier removal by cooks distance) and the significantly expressed loci in the CMC to determine whether there was any overlap between the two cohorts. Gene pathways associated with acetylation (p = 4.75E-57) and phosphoproteins (p = 3.55E-40) had the most significant enrichment in the CMC analysis. In addition multiple gene pathways associated with mitochondrial functions, splicing and ribosome activity were significantly enriched (Table 2, Supplementary Table 3). The 4 gene pathways that were significantly enriched in Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 4 Discussion Th f It is important to note that our current understanding of brain aging at the molecular level is limited and therefore, we may be failing to recognize patterns associated with accelerated aging in our data. These results are also consistent with previous pathway analysis performed on microarray data that found transcriptional differences in the aging schizophrenia brain33,34. Given that inflammation is associated with old age, the findings of increased inflamma- tion in schizophrenia35–37 might also be related to the aging disruptions we are finding. In summary, our results along with previous findings strongly support the existence of a link between aging and schizophrenia pathology. Additional postmortem brain collections that examine schizophrenia throughout the entire lifespan and are cog- nizant of the importance of age are needed to fully explore the effects of age on gene expression in the disease.f 38 p g y pf g g p Individuals suffering from schizophrenia have a tendency to smoke38, use recreational drugs, and consume excess levels of alcohol compared to the general population39. In addition, schizophrenia patients use various antipsychotic and other psychiatric medications that cause weight gain and other health problems40,41. These factors have the potential to disrupt the aging process in the brain and may account for our findings. Conceivably, the expression differences observed in the 40 to 60 age group are caused by such factors. However, the statistically significant enrichment between age dependent RNA expression changes in the brain and the 108 genomic regions associated with schizophrenia is intriguing. Further research is needed to determine whether some of the age related gene expression differences observed in the schizophrenia brain might be inherited.f g pf p g Duration of illness, along with lifetime antipsychotic use are likely to affect gene expression since they are associated with brain volume changes in schizophrenia42,43. Duration of illness in schizophrenia is of particular interest as it appears to influence treatment response44,45. The effects of these factors on gene expression could not be assessed in the current study because this information was not available for the CommonMind collection. In the smaller psychENCODE cohort, which spans a narrower age range and appears to be statistically under- powered, both duration of illness and lifetime antipsychotic use were highly correlated with age. Generally, older subjects with schizophrenia had lived with the disease longer and had used more antipsychotics. Discussion Th f The CMC project generated sequencing data on BA9 samples whereas the PsychENCODE consortium Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 5 www.nature.com/scientificreports/ sequenced BA46 samples. The psychENCODE cohort was also less than 3.5 times that of the CMC and may have been underpowered. The overlap in gene set enrichment and Inrich analysis between the CMC and the nominally significant genes in the psychENCODE cohorts support this notion. The fact that the psychENCODE cohort spanned ages between 25 and 65 as opposed to the CMC that spanned ages between 25 and 90 may have also contributed to the differences in these findings. Another reason for the discrepant findings is likely related to the heterogeneous nature of schizophrenia. As schizophrenia genetic studies required over a hundred thousand sam- ples in order to overcome the heterogeneity problem19, expecting every loci to be differentially expressed in every single schizophrenia postmortem brain sample is unrealistic. It is likely that the disruption of different genes that belong to the same biological pathway results in similar pathology in different people. This is congruent with the presented gene set enrichment analysis that found disruptions in RNA splicing and phosphoprotein pathways in both the CMC and the psychENCODE cohorts.h p y This work implicates gene pathways associated with mitochondrial function, splicing and phosphoproteins in schizophrenia pathology. Previous schizophrenia postmortem brain studies have reported differential gene expression for mitochondrial genes in disease21,22. Widespread splicing deficits have also been reported in schiz- ophrenia2,23,24 and numerous phosphoproteins such as synapsin II25 and DARPP-3226 have been implicated in disease pathology. It is important to note that mitochondrial changes27,28, widespread alterations in splicing pat- terns29 and altered phosphorylation30,31 also occur in the aging brain. Potentially, the deficits observed in schizo- phrenia might be similar to the changes that occur in the aging brain.h p g g g g This notion is in line with previous studies that have found a link between aging and schizophrenia. Many studies have found schizophrenia patients to have lower life expectancy (for a review see8) and the suggestion has been made that schizophrenia might be a systematic disease of accelerated aging32. Although this work discovered age related changes in schizophrenia, our results do not necessarily support accelerated aging in the disease. Discussion Th f Given this asso- ciation with age, an adequate assessment of the effects of disease duration or the effects of lifetime antipsychotic use will require comparison of early and late disease onset cases that live into old age and comparison of subjects at similar ages that differ in their lifetime antipsychotic use. Each of these comparisons will likely require large cohorts, potentially similar in size to CommonMind, because of the subtle gene expression differences that occur in schizophrenia. It is probable that at least a subset of the differentially expressed genes identified in the current study are associated with the pathology related to duration of illness and/or lifetime antipsychotic use but defin- itive conclusions can not be made. Larger postmortem brain collections will also enable characterization of the associations between gene expression and disease subtypes or assessment scores, neither of which were available for this study. Potentially, an analysis cognizant of disease subtypes or assessment scores will be extremely ben- eficial as it may refine the heterogenous schizophrenia classification and yield larger and more consistent gene expression differences. pf We conclude that age is an important aspect of schizophrenia pathophysiology and special consideration needs to be given to age in gene expression studies of disease. Appropriate treatment of age as a variable is likely to aid in efforts to unravel the molecular etiology of schizophrenia and the identification of biomarkers and ther- apeutic targets for the disease. Materials and Methodsf Samples. Differential expression analysis was performed on the DLPFC RNA sequencing data from CMC1 (N = 476) and the PsychENCODE46 (N = 134) projects. Details regarding the ethical approval process and rele- vant adherence to relevant guidelines and regulations are detailed in the original publications1,46. Access to the RNA sequencing data was approved by the NIMH Respository and Genomic Resource Data Access Committee and deidentified data was downloaded. In the CMC cohort only BA9 samples from schizophrenia cases and unaffected controls were included. Samples from individuals with Klinefelter syndrome were excluded and only a single individual chosen at random from a sibling pair was included in the analysis. Samples below the age of Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 6 www.nature.com/scientificreports/ 25 were excluded since there were only 3 schizophrenia samples in this age group. The age group labeled 90+ was also excluded since an exact age could not be assigned for the regression analysis. Samples less then 25 and more then 65 years of age in the psychENCODE cohort were excluded in order to have a balanced number of cases and controls across all age groups (the 3 samples above 65 were controls and 6 out of the 7 samples below 25 were schizophrenia cases). psychENCODE samples with a PMI of greater than 50 hours were also excluded in order to be consistent with the CMC cohort. Differential expression analysis of schizophrenia and aging. All analysis was performed using R version 3.4.3. Bam files for each corresponding study was downloaded and count tables were generated for genes (considering only exons) using the Ensembl hg19 annotation. Specifically, the readGAlignmentPairs command from the GenomicAlignments package47 was used to read the alignments from the bam files and the resulting data structure along with the Ensembl hg19 annotation was passed as parameters to the summarizeOverlaps command. The summarizeOvelaps command calculates the number of reads that span the exon intervals in the annotation and provide the total number of reads that are present for each gene. RUVSeq package was used to account for hidden batch effects and remove unwanted variation from the samples17. In order to identify loci that are not differentially expressed between the diagnosis groups, which is required by RUVSeq, we performed an initial analysis using DESeq214 in which we only considered diagnosis and age. + + + + + + + ~ ( Diagnosis Age Diagnosis: Age RIN PMI Race Sex RuvFactors) The resulting p-values were corrected (or adjusted) for multiple testing using the Benjamini & Hochberg method implemented in the DESeq2 package. Only loci with a corrected p-value <0.1 on average 20 read counts per gene were considered significantly differentially expressed. An adjusted p-value of 0.1 is recommend over 0.05 since the multiple correction methods used to adjust the p-values are overly stringent14. We repeated the same DESeq2 analysis without RuvFactors and found that the age related trajectory differences are observed even in the absence of batch effect correction (data not shown). Cooks distance was calculated for each differentially expressed locus that surpassed the threshold of statistical significance in order to remove genes that appear to be different because of outliers. The Cooks distance cutoff threshold was identified empirically for each cohort. A similar analysis was performed for the 40 to 60 year old age group using the same formula as above but lacking the interaction term. + + + + + + ~ ( Diagnosis Age RIN PMI Race Sex RuvFactors) New normalization factors were calculated using RuvSeq for the 40 to 60 year age group (an initial DESeq2 analysis was performed with diagnosis only and non-differentially expressed genes were identified as before). Differential expression plots. The raw normalized counts from the DESeq data object were extracted using the ‘counts’ function with the normalized parameter set to true. The obtained normalized read counts were plotted over age using the bioconductor ggplot2 package. We opted to not use log-transformed values in order to provide the most accurate representation of the data. Normalized difference plots. For each differentially expressed locus identified in the CMC analysis, the loess function in R was used to fit gene expression in control or schizophrenia samples and age. The R ‘predict’ function was run to calculate gene expression values for the age range 25–90 based on the results of the loess func- tion. For each locus, the resulting values for the schizophrenia samples were subtracted from control samples and this value was normalized to the absolute maximum difference in all ages using the following formula: Materials and Methodsf Genes that had an unadjusted p-value of >0.7 were considered to be not differentially expressed and passed to the Ruvg command to calculate normalization factors that account for unwanted variation17. The factors calculated from RUVseq were included in the design formula of the DESeq analysis. Wald tests were performed in DESeq2 using the DESeq command, which also performs normalization on the samples, in order to identify differentially expressed genes. The statisti- cal model used for the Wald test included an interaction term between diagnosis and age. This and all subsequent formulas were used as the ‘design’ parameter in DESeq2. − − (control schizophrenia)/max(abs(control schizophrenia)) − − (control schizophrenia)/max(abs(control schizophrenia)) Normalization meant that only values between −1 and 1 were possible. Each locus had a value of 1 or −1 at some age indicating the maximal gene expression difference in disease for the locus. The normalized values for each locus was plotted using a very thin line in order to avoid over plotting. GWAS enrichment and gene set analysis. Differentially expressed genes in the CMC with an adjusted p-value of 0.1 or less and a minimum mean read count of 20 were used for gene set enrichment and Inrich20 anal- ysis. The DAVID 6.7 Bioinformatics Resource (https://david-d.ncifcrf.gov/) was selected for the gene set enrich- ment analysis as it is a robust and widely used tool with over 33000 citations48. DAVID performs a competitive analysis using a modified Fisher’s exact test. The standard DAVID analysis49 was performed using the default conditions of minimum 2 genes per term and an EASE score, which is a modified and more stringent Fisher Exact P-value, of 0.1. The genes in the human genome were used as the background gene list. The databases included in the DAVID analysis were (Selected from Annotation Summary Results page that is generated by DAVID after gene sets are uploaded and mapped): g p pp OMIM (https://www.ncbi.nlm.nih.gov/omim) from the Disease category; COG_ONTOLOGY (https://www. ncbi.nlm.nih.gov/COG/), SP_PIR_KEYWORD (https://proteininformationresource.org/pirwww/dbinfo/ipro- class.shtml) and UP_SEQ_FEATURES (https://www.uniprot.org/) from the Functional Categories category; Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 7 www.nature.com/scientificreports/ GOTERM_BP_FAT, GOTERM_CC_FAT and GOTERM_MF_FAT (https://www.ebi.ac.uk/GOA) from the Gene_Ontology category; BBID50, BIOCARTA (https://cgap.nci.nih.gov/Pathways/BioCarta_Pathways) and KEGG_PATHWAY (https://www.genome.jp/kegg/) from the Pathways category; INTERPRO (https://www.ebi. ac.uk/interpro/), PIR_SUPERFAMILY (https://proteininformationresource.org/pirwww/dbinfo/iproclass.shtml) and SMART (http://smart.embl-heidelberg.de/) from the Protein_Domains category. p g g y Inrich, analysis was performed as described51 to determine whether intervals near schizophrenia associated loci were more likely to be associated with the identified differentially expressed genes. Only genes that fell within the GWAS regions were considered. Adjusted p-value <0.05 from the Inrich program was set as the threshold for significance. For the PsychENCODE cohort, genes with a nominal p-value of 0.05 or less and a minimum read count of 20 were used for the same analyses. The same genes were used for gene set enrichment analysis performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) 6.7 web tool18. In the 40 to 60 age group the 556 genes that replicated between the two cohorts were subjected to DAVID and Inrich analyses. Data Availability Th y The CMC and the psychEncode datasets are available from the NIMH Repository and Genomics Resources NRGR). − − (control schizophrenia)/max(abs(control schizophrenia)) Gene pathways reaching a Bonferroni corrected p-values < 0.1 from the DAVID analysis are reported. References 374(9690), 620–7, https://doi.org/10.1016/S0140-6736(09)60742-X (2009). p g p y ( ) 9. Tiihonen, J. et al. 11-year follow-up of mortality in patients with schizophrenia: a population-based cohort study (FIN11 study). Lancet. 374(9690), 620–7, https://doi.org/10.1016/S0140-6736(09)60742-X (2009). p g 0. Koutsouleris, N. et al. Accelerated brain aging in schizophrenia and beyond: a neuroanatomical marker of psychiatric disorders Schizophr Bull. 40(5), 1140–53 (2014). p 11. Kochunov, P. et al. Testing the hypothesis of accelerated cerebral white matter aging in schizophrenia and major depression. Biol Psychiatry. 73(5), 482–91 (2013). p 11. Kochunov, P. et al. Testing the hypothesis of accelerated cerebral white matter aging in schizophrenia and major depression. Biol Psychiatry. 73(5), 482–91 (2013). y y ( ) ( ) 12. Wright, S. et al. Accelerated white matter aging in schizophrenia: role of white matter blood perfusion. Neurobiology of aging. 35(10), 2411–8 (2014). 3. Gandal, M. J. et al. Shared molecular neuropathology across major psychiatric disorders parallels polygenic overlap. Science 359(6376), 693–7 (2018). 14. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15(12), 550 (2014). 15. Catts, V. S. et al. A microarray study of post-mortem mRNA degradation in mouse brain tissue. Brain research Molecular brain research. 138(2), 164–77 (2005).i ( ) ( ) 6. Gallego Romero, I., Pai, A. A., Tung, J. & Gilad, Y. RNA-seq: impact of RNA degradation on transcript quantification. BMC biology 12, 42 (2014). 7. Risso, D., Ngai, J., Speed, T. P. & Dudoit, S. Normalization of RNA-seq data using factor analysis of control genes or samples. Na Biotechnol. 32(9), 896–902 (2014). 18. Huang, D. W. et al. DAVID Bioinformatics Resources: expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic Acids Res. 35(Web Server issue), W169–75 (2007). g g 9. Biological insights from 108 schizophrenia-associated genetic loci. Nature, 511(7510), 421–7, https://doi.org/10.1038/nature13595 (2014).h 20. Lee P. H., O’Dushlaine C., Thomas B. & Purcell S. M. INRICH: interval-based enrichment analysis for genome-wide association studies. Bioinformatics. 28(13), 1797–9, 10.093/bioinformatics/bts191 (2012). f 21. Iwamoto, K., Bundo, M. & Kato, T. Altered expression of mitochondria-related genes in postmortem brains of patients with bipolar disorder or schizophrenia, as revealed by large-scale DNA microarray analysis. Hum Mol Genet. 14(2), 241–53 (2005). 2. Altar, C. A. et al. Deficient hippocampal neuron expression of proteasome, ubiquitin, and mitochondrial genes in multiple schizophrenia cohorts. Biol Psychiatry. 58(2), 85–96 (2005). Code Availability Analysis code used for this manuscript is available upon request and without restriction. y Analysis code used for this manuscript is available upon request and without restrictio Code Availability Code Availability Analysis code used for this manuscript is available upon request and without restriction. References References 1. Fromer, M. et al. Gene expression elucidates functional impact of polygenic risk for schizophrenia. Nat Neurosci. 19(11), 1442–53 10.038/nn.4399 (2016). 1. Fromer, M. et al. Gene expression elucidates functional impact of polygenic risk for schizophrenia. Nat Neurosci. 19(11), 1442–53, 10.038/nn.4399 (2016). 10.038/nn.4399 (2016). 2. Darby, M. M., Yolken, R. H. & Sabunciyan, S. Consistently altered expression of gene sets in postmortem brains of individuals with major psychiatric disorders. Transl Psychiatry. 6(9), e890, https://doi.org/10.1038/tp.2016.173 (2016). j y y y g 3. Vawter, M. P. et al. Genome scans and gene expression microarrays converge to identify gene regulatory loci relevant in schizophrenia. Hum Genet. 119(5), 558–70 (2006). p ( ), ( ) 4. Fillman, S. G. et al. Increased inflammatory markers identified in the dorsolateral prefrontal cortex of individuals with schizophrenia. Mol Psychiatry (2012). p ( ), ( ) 4. Fillman, S. G. et al. Increased inflammatory markers identified in the dorsolateral prefrontal cortex of individuals with schizophrenia. Mol Psychiatry (2012). p y y ( ) 5. Roussos, P., Katsel, P., Davis, K. L., Siever, L. J. & Haroutunian, V. A System-Level Transcriptomic Analysis of Schizophrenia Using Postmortem Brain Tissue Samples. Arch Gen Psychiatry. 1–11 (2012). Postmortem Brain Tissue Samples. Arch Gen Psychiatry. 1 11 (2012). 6. Vawter, M. P. et al. Gene expression of metabolic enzymes and a protease inhibitor in the prefrontal cortex are decreased in schizophrenia. Neurochem Res. 29(6), 1245–55 (2004). p y y ( ) 6. Vawter, M. P. et al. Gene expression of metabolic enzymes and a protease inhibitor in the prefrontal cortex are decreased in schizophrenia. Neurochem Res. 29(6), 1245–55 (2004). schizophrenia. Neurochem Res. 29(6), 1245–55 (2004). p ( ), ( ) 7. Horvath, S. & Mirnics, K. Schizophrenia as a Disorder of Molecular Pathways. Biol Psychiatry. 10(14), 00007–9 (2014).t 7. Horvath, S. & Mirnics, K. Schizophrenia as a Disorder , , p y y y ( ), ( ) 8. Laursen, T. M., Nordentoft, M. & Mortensenm, P. B. Excess early mortality in schizophrenia. Annu Rev Clin Psychol. 10, 425–48, https://doi.org/10.1146/annurev-clinpsy-032813-153657 (2014). p y y y ( ) ( ) 8. Laursen, T. M., Nordentoft, M. & Mortensenm, P. B. Excess early mortality in schizophrenia. Annu Rev Clin Psychol. 10, 425–48, https://doi.org/10.1146/annurev-clinpsy-032813-153657 (2014). p // g/ / p y ( ) 9. Tiihonen, J. et al. 11-year follow-up of mortality in patients with schizophrenia: a population-based cohort study (FIN11 study). Lancet. www.nature.com/scientificreports/ Mol Psychiatry 23(3), 708–12 (2018). www.nature.com/scientificreports/ www.nature.com/scientificreports/ 0. Foster, T. C. Interaction of rapid signal transduction cascades and gene expression in mediating estrogen effects on memory over the life span. Frontiers in neuroendocrinology. 26(2), 51–64 (2005).h 30. Foster, T. C. Interaction of rapid signal transduction cascades and gene expression in mediating estrogen effects on memory over the life span. Frontiers in neuroendocrinology. 26(2), 51–64 (2005). 31 M i M S G i S B tt i i F & T b hi M Th i b i t i h h l ti t t f h i l p gy 1. Magnoni, M. S., Govoni, S., Battaini, F. & Trabucchi, M. The aging brain: protein phosphorylation as a target of changes in neurona function. Life sciences. 48(5), 373–85 (1991). f ( ), ( ) 32. Kirkpatrick, B., Messias, E., Harvey, P. D., Fernandez-Egea, E. & Bowie, C. R. Is schizophrenia a syndrome of accelerated aging? Schizophr Bull. 34(6), 1024–32 (2008).i p ( ) ( ) 3. Tang, B. et al. Normal human aging and early-stage schizophrenia share common molecular profiles. Aging Cell. 8(3), 339–42 https://doi.org/10.1111/j.474-9726.2009.00468.x (2009).h p g j ( ) 34. Torkamani, A., Dean, B., Schork, N. J. & Thomas, E. A. Coexpression network analysis of neural tissue reveals perturbations in developmental processes in schizophrenia. Genome Res. 20(4), 403–12, https://doi.org/10.1101/gr.101956.109 (2010). p p p p g g 5. Fillman, S. G. et al. Increased inflammatory markers identified in the dorsolateral prefrontal cortex of individuals with schizophrenia. Mol Psychiatry. 18(2), 206–14 (2013).il y y 6. Hwang, Y. et al. Gene expression profiling by mRNA sequencing reveals increased expression of immune/inflammation-related genes in the hippocampus of individuals with schizophrenia. Transl Psychiatry. 3, e321, https://doi.org/10.1038/tp.2013.94 (2013).l g pp p p y y , , p // g/ / p ( ) 37. Potvin, S. et al. Inflammatory cytokine alterations in schizophrenia: a systematic quantitative review. Biol Psychiatry. 63(8), 801–8. Epub 2007 Nov 19 (2008). p 8. de Leon, J. & Diaz, F. J. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schizophr Res. 76(2-3), 135–57, https://doi.org/10.1016/j.schres.2005.02.010 (2005). g p g j 9. Fowler, I. L., Carr, V. J., Carter, N. T. & Lewin, T. J. Patterns of current and lifetime substance use in schizophrenia. Schizophr Bull 24(3), 443–55 (1998). 40. Chen, J., Huang, X. F., Shao, R., Chen, C., Deng, C. Molecular Mechanisms of Antipsychotic Drug-Induced Diabetes. Front Neurosci. References p y y ( ) ( ) 23. Morikawa, T. & Manabe, T. Aberrant regulation of alternative pre-mRNA splicing in schizophrenia. Neurochemistry international. 57(7), 691–704 (2010).i 4. Wu, J. Q. et al. Transcriptome sequencing revealed significant alteration of cortical promoter usage and splicing in schizophrenia PLoS One. 7(4), e36351 (2012). 5. Molinaro, L., Hui, P., Tan, M. & Mishra, R. K. Role of presynaptic phosphoprotein synapsin II in schizophrenia. World journal o psychiatry. 5(3), 260–72 (2015). p y y ( ), ( ) 26. Colantuoni, C. et al. Age-related changes in the expression of schizophrenia susceptibility genes in the human prefrontal cortex. Brain structure & function. 213(1-2), 255–71 (2008). h k l h f d d l ( ) ( ) f Bishop, N. A., Lu, T. & Yankner, B. A. Neural mechanisms of ageing k d d f 27. Bishop, N. A., Lu, T. & Yankner, B. A. Neural mechanisms of ageing and cognitive decline. Nature. 464(7288), 529–35 (2010). 28. Grimm, A. & Eckert, A. Brain aging and neurodegeneration: from a mitochondrial point of view. Journal of neurochemistry. 143(4), 418–31 (2017). ( ) 29. Mazin, P. et al. Widespread splicing changes in human brain development and aging. Molecular systems biology. 9, 633 (2013). 8 Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 Acknowledgements g I thank the Stanley Medical Research Institute for providing funding for this research. I am especially thankful to Dr. Jeff Leek for his guidance and advice on statistical approaches. I would also like to thank Drs Robert Yolken, Tim Moran, Jonathan Pevsner, Russell Margolis, Brion Maher, Fuller Torrey and Maree Webster for their helpful suggestions and Drs Yunjung Kim and Patrick Sullivan for providing code for the Inrich analysis. CMC data were generated as part of the CommonMind Consortium supported by funding from Takeda Pharmaceuticals Company Limited, F. Hoffman-La Roche Ltd and NIH grants R01MH085542, R01MH093725, P50MH066392, P50MH080405, R01MH097276, RO1-MH-075916, P50M096891, P50MH084053S1, R37MH057881 and R37MH057881S1, HHSN271201300031C, AG02219, AG05138 and MH06692. Brain tissue for the study was obtained from the following brain bank collections: the Mount Sinai NIH Brain and Tissue Repository, the University of Pennsylvania Alzheimer’s Disease Core Center, the University of Pittsburgh NeuroBioBank and Brain and Tissue Repositories and the NIMH Human Brain Collection Core. CMC Leadership: Pamela Sklar, Joseph Buxbaum (Icahn School of Medicine at Mount Sinai), Bernie Devlin, David Lewis (University of Pittsburgh), Raquel Gur, Chang-Gyu Hahn (University of Pennsylvania), Keisuke Hirai, Hiroyoshi Toyoshiba (Takeda Pharmaceuticals Company Limited), Enrico Domenici, Laurent Essioux (F. Hoffman-La Roche Ltd), Lara Mangravite, Mette Peters (Sage Bionetworks), Thomas Lehner, Barbara Lipska (NIMH). PsychENCODE data were generated as part of the PsychENCODE Consortium, supported by: U01MH103339, U01MH103365, U01MH103392, U01MH103340, U01MH103346, R01MH105472, R01MH094714, R01MH105898, R21MH102791, R21MH105881, R21MH103877, and P50MH106934 awarded to: Schahram Akbarian (Icahn School of Medicine at Mount Sinai), Gregory Crawford (Duke), Stella Dracheva (Icahn School of Medicine at Mount Sinai), Peggy Farnham (USC), Mark Gerstein (Yale), Daniel Geschwind (UCLA), Thomas M. Hyde (LIBD), Andrew Jaffe (LIBD), James A. Knowles (USC), Chunyu Liu (UIC), Dalila Pinto (Icahn School of Medicine at Mount Sinai), Nenad Sestan (Yale), Pamela Sklar (Icahn School of Medicine at Mount Sinai), Matthew State (UCSF), Patrick Sullivan (UNC), Flora Vaccarino (Yale), Sherman Weissman (Yale), Kevin White (UChicago) and Peter Zandi (JHU). www.nature.com/scientificreports/ 11, 643, https://doi.org/10.3389/fnins.2017.00643 eCollection 2017 (2017). g 41. Silva, A., Ribeiro, M., Sousa-Rodrigues, C. F. & Barbosa, F. T. Association between antipsychotics and cardiovascular adverse events: A systematic review. 1992), 63(3), 261–7, https://doi.org/10.1590/806-9282.63.03.261 (2017). y g 2. Haijma, S. V. et al. Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects. Schizophr Bull. 39(5), 1129–38. (2013). 42. Haijma, S. V. et al. Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects. Schizophr Bull. 39(5), 1129–38. (2013). 43. Hashimoto, N. et al. The effect of duration of illness and antipsychotics on subcortical volumes in schizophrenia: Analysis of 778 subjects. NeuroImage Clinical. 17, 563–9 (2018).i j , p y j p ( ), ( ) 43. Hashimoto, N. et al. The effect of duration of illness and antipsychotics on subcortical volumes in schizophrenia: Analysis of 778 subjects. NeuroImage Clinical. 17, 563–9 (2018).i j g ( ) 4. Szymanski, S. R., Cannon, T. D., Gallacher, F., Erwin, R. J. & Gur, R. E. Course of treatment response in first-episode and chronic schizophrenia. Am J Psychiatry. 153(4), 519–25 (1996).l 45. Altamura, A. C., Serati, M. & Buoli, M. Is duration of illness really influencing outcome in major psychoses? Nordic journal of psychiatry. 69(6), 403–17 (2015).h p y y ( ) ( ) 46. Akbarian, S. et al. The PsychENCODE project. Nat Neurosci. 18(12), 1707–12, 10.038/nn.4156 (2015). l f f d l ( 47. Lawrence, M. et al. Software for computing and annotating genomic ranges. PLoS Comput Biol. 9(8), e1003118, https:/ org/10.1371/journal.pcbi (2013). 48. Research NLB. DAVID Bioinformatics Resource, https://scholar.google.com/citations?user=dMn7gzYAAAAJ Google Sch Google (2019). g 9. Huang da, W., Sherman, B. T. & Lempicki, R. A. Bioinformatics enrichment tools: paths toward the comprehensive functiona analysis of large gene lists. Nucleic Acids Res. 37(1), 1–13 (2009). analysis of large gene lists. Nucleic Acids Res. 37(1), 1–13 (2009 y g g 50. Becker, K. G., White, S. L., Muller, J. & Engel, J. BBID: the biological biochemical image database. Bioinformatics (Oxford, England). 16(8), 745–6 (2000).f y g g 0. Becker, K. G., White, S. L., Muller, J. & Engel, J. BBID: the biological biochemical image database. Bioinformatics (Oxford, England) 16(8), 745–6 (2000).f 51. Kim, Y. et al. Comparative genomic evidence for the involvement of schizophrenia risk genes in antipsychotic effects. Mol Psychiatry. 23(3), 708–12 (2018). 1. Kim, Y. et al. Comparative genomic evidence for the involvement of schizophrenia risk genes in antipsychotic effects. Author Contributions Author Contributions S.S. designed the experiments, performed the analysis and authored the manuscript. Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-019-42308-5. Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 9 www.nature.com/scientificreports/ Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2019 Scientific Reports \| (2019) 9:5896 \| https://doi.org/10.1038/s41598-019-42308-5 10
https://openalex.org/W2768219072	https://bmcmedicine.biomedcentral.com/counter/pdf/10.1186/s12916-017-0975-5	English	null	Depression comorbid with tuberculosis and its impact on health status: cross-sectional analysis of community-based data from 48 low- and middle-income countries	BMC medicine	2,017	cc-by	7,807	* Correspondence: a.koyanagi@pssjd.org 1Research and Development Unit, Parc Sanitari Sant Joan de Déu, Universitat de Barcelona, Fundació Sant Joan de Déu, Dr. Antoni Pujadas, 42, Sant Boi de Llobregat, Barcelona, Spain 2Instituto de Salud Carlos III, Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, Madrid, Spain Full list of author information is available at the end of the article Koyanagi et al. BMC Medicine (2017) 15:209 DOI 10.1186/s12916-017-0975-5 Koyanagi et al. BMC Medicine (2017) 15:209 DOI 10.1186/s12916-017-0975-5 Depression comorbid with tuberculosis and its impact on health status: cross-sectional analysis of community-based data from 48 low- and middle-income countries Depression comorbid with tuberculosis and its impact on health status: cross-sectional analysis of community-based data from 48 low- and middle-income countries Ai Koyanagi1,2* , Davy Vancampfort3,4, André F. Carvalho5, Jordan E. DeVylder6, Josep Maria Haro1,2, Damiano Pizzol7, Nicola Veronese8,9 and Brendon Stubbs10,11,12 Ai Koyanagi1,2* , Davy Vancampfort3,4, André F. Carvalho5, Jordan E. DeVylder6, Josep Maria Haro1,2, Damiano Pizzol7, Nicola Veronese8,9 and Brendon Stubbs10,11,12 © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Background survey was provided by ethical boards at each study site. All participants gave their informed consent. g Tuberculosis (TB) is one of the top 10 causes of deaths glo- bally [1]. In 2015, there were 10.4 million new TB cases and 1.8 million deaths due to TB. Over 95% of TB cases and deaths occur in developing countries [1]. Depression often coexists with TB [2], and this comorbidity is associ- ated with poor adherence to TB treatment and higher mortality [3]. Lack of adherence to anti-TB regimens may lead to higher risk for drug resistance, morbidity, and mortality, as well as community exposure to TB [4, 5]. TB Because the WHS did not include mycobacterial culture or sputum smear examinations, TB was based on past 12-month symptoms of active TB. Specifically, as in pre- vious WHS publications [10–12], those who had both (1) a cough that lasted for 3 weeks or longer and (2) blood in phlegm (or coughed up blood) were considered to have active TB. Previous studies have shown that the presence of these typical symptoms are likely to have a sensitivity and specificity of 65–70% and 55–75%, respectively, in the detection of TB [10]. In low- and middle-income countries (LMICs), the prevalence of depression is high and may be increasing [6]. A recent large prospective study from Korea found that depression at baseline is associated with a higher risk for incident TB [7]. Depression may lead to an increased susceptibility to TB by compromising immunity or through neglected self-care [8]. Thus, depression may be an unrecognized driver of the global TB and multidrug resistant TB (MDR-TB) epidemics [2]. However, the few previous studies on the association between depression and TB from LMICs have only been conducted in clinical settings with small sample sizes, and information from the general population is lacking. Furthermore, there is limited information on the joint effect of TB and depression on health status. We therefore assessed the association be- tween TB and depression, and whether the co-occurrence of TB and depression confers a more pronounced decre- ment in health status and function compared to TB alone using community-based, predominantly nationally repre- sentative data from 48 LMICs that participated in the World Health Survey (WHS). Epidemiological data on the TB/depression comorbidity and its effect on health out- comes are crucial to provide a more accurate assessment of the public health significance of this comorbidity. The survey Subsyndromal depression At least one criterion B symptom with the total number of symptoms being three or less. The criteria of duration of at least 2 weeks and presence of symptoms during most of the day had to be met. d d d The WHS was a cross-sectional survey carried out in 70 countries from 2002 to 2004. Survey details are available elsewhere (http://www.who.int/healthinfo/survey/en/). Briefly, single-stage random sampling and stratified multi-stage random cluster sampling was conducted in 10 and 60 countries, respectively. Eligible participants were those with a valid home address and aged ≥ 18 years. One individual was randomly chosen from the household with the use of Kish tables. The questionnaire was subject to standard translation procedures to ensure comparability between countries. Face-to-face interviews were conducted by trained interviewers. The overall individual response rate was 98.5% [9]. To adjust for non-response, sampling weights were generated using the population distribution as reported by the United Nations Statistical Division. Ethical approval for the None of the above. None of the above. In some analyses, we also dichotomized this variable as the absence or presence of depressive episode. Abstract Background: Depression in tuberculosis increases the risk for adverse health outcomes. However, little is known about comorbid depression and tuberculosis in the general population. Thus, we assessed the association between depression and tuberculosis, and the decrements in health status associated with this comorbidity in 48 low- and middle-income countries. Methods: Cross-sectional, community-based data from the World Health Survey on 242,952 individuals aged ≥18 years were analyzed. Based on the World Mental Health Survey version of the Composite International Diagnostic Interview, past 12-month depression was categorized into depressive episode, brief depressive episode, subsyndromal depression, and no depression. Health status across six domains (cognition, interpersonal activities, sleep/energy, self-care, mobility, pain/discomfort) was assessed. Multivariable logistic and linear regression analyses were performed to assess the associations. Results: The prevalence of depressive episode among those with and without tuberculosis was 23.7% and 6.8%, respectively (P < 0.001). Tuberculosis was associated with a 1.98 (95% CI 1.47–2.67), 1.75 (95% CI 1.26–2.42), and 3.68 (95% CI 3.01–4.50) times higher odds for subsyndromal depression, brief depressive episode, and depressive episode, respectively. Depressive episode co-occurring with tuberculosis was associated with significantly worse health status across all six domains compared to tuberculosis alone. Interaction analysis showed that depression significantly amplifies the association between TB and difficulties in self-care but not in other health domains. Conclusions: Depression is highly prevalent in adults with tuberculosis, and is associated with worse health status compared to tuberculosis without depression. Public health efforts directed to the recognition and management of depression in people with tuberculosis may lead to better outcomes. Keywords: Tuberculosis, Depression, Low- and middle-income countries, Epidemiology © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Koyanagi et al. BMC Medicine (2017) 15:209 Page 2 of 10 Page 2 of 10 Page 2 of 10 Depression Th p The severity of depressive symptoms was established based on the individual questions of the World Mental Health Survey version of the Composite International Diagnostic Interview, which assessed the duration and persistence of depressive symptoms in the past 12 months [13]. Following the algorithms used in a pre- vious WHS publication [14], four mutually exclusive groups were established based on the ICD-10 Diagnostic Criteria for Research (ICD-10-DCR) [15], where criter- ion B referred to symptoms of depressed mood, loss of interest, and fatigability. The algorithms used to define the four mutually exclusive groups were the following: Depressive episode group At least two criterion B symptoms with a total of at least four depressive symptoms lasting 2 weeks most of the day or all of the day. Depressive episode group At least two criterion B symptoms with a total of at least four depressive symptoms lasting 2 weeks most of the day or all of the day. Brief depressive episode group Same criteria as depressive episode but did not meet the 2-week duration criterion. Control variables The selection of the control variables were based on past literature [10]. Sociodemographic variables included age, sex, education (no formal education, primary education, secondary or high school completed, or tertiary education completed), wealth, household size, and setting (rural or urban). Principal component analysis based on 15–20 assets was conducted to establish country-wise wealth quintiles. Current smoking was dichotomized as ‘Yes’ and ‘No’. Respondents were asked how many standard drinks of any alcoholic beverage they had on each day of the past 7 days. Females who reported consuming at least four drinks, and males who reported consuming at least five drinks, on 1 or 2 days in the past 7 days were considered infrequent heavy drinkers, and respondents who drank these amounts at least 3 days in the past 7 days were considered frequent heavy drinkers. All other respondents, apart from lifetime abstainers, were considered non-heavy drinkers [21, 22]. Body mass index (BMI; kg/m2) was based on self-reported weight and height, and was catego- rized as < 18.5 (underweight), 18.5–24.9 (normal weight), 25.0–29.9 (overweight), and ≥30 (obese). Diabetes was based on self-reported diagnosis. Next, we created a four-category variable based on the presence or absence of depressive episode and TB, namely (1) no depression and no TB (n = 183,455); (2) depression without TB (n = 11,440); (3) TB without depression (n = 2617); and (4) TB with depression (n = 687), to assess whether TB with depression is associated with a larger decrement in health status as compared with TB alone. We conducted multivariable linear regression with this this four-category variable as the exposure and the six health status variables as the outcomes (mobility, pain/dis- comfort, self-care, cognition, interpersonal activities, sleep/ energy). We also conducted age-stratified analyses to assess whether the TB and depressive episode comorbidity have different effects on health status by age groups. Age was categorized as 18–44 (young adults; 67.8%), 45–64 (middle- aged adults; 23.6%), and ≥65 (older adults; 8.6%) years, broadly representing distinct life stages [23]. In order to assess whether there is effect modification by depressive episodes in the association between TB and health status, we also conducted interaction analysis by including an interaction term in the model using the overall sample (TB × depressive episode). We did not conduct interaction analysis by age groups due to the small sample size and possibility for lack of statistical power. Health status Health status was assessed with the use of 12 health- related questions pertaining to six different domains, namely (1) mobility, (2) pain and discomfort, (3) self-care, (4) cognition, (5) interpersonal activities, and (6) sleep and energy. These domains correspond to frequently used Koyanagi et al. BMC Medicine (2017) 15:209 Page 3 of 10 (n = 16,163). Information on the individual countries is provided in Additional file 1: Table S2. health outcome measures including the Short Form 12 [16], the Health Utilities Index Mark 3 [17], and the EuroQol 5D [18], and have been used as indicators of health status in prior WHS studies [19, 20]. Each domain consisted of two questions that assessed health function in the past 30 days. The actual questions can be found in Additional file 1: Table S1. Each item was scored on a five-point scale ranging from ‘none’ to ‘extreme/cannot do’. For each separate domain, we used factor analysis with polychoric correlations to obtain a factor score which was later converted to scores ranging from 0 to 100 with higher values representing worse health function [20]. Statistical analyses were performed with Stata 14.1 (Stata Corp LP, College station, Texas). Descriptive analyses included unweighted Ns, and weighted propor- tions and means. Statistical analyses were performed with Stata 14.1 (Stata Corp LP, College station, Texas). Descriptive analyses included unweighted Ns, and weighted propor- tions and means. First, in order to assess the association between TB (exposure) and depression (subtypes; outcome), we con- ducted multivariable multinomial logistic regression analyses using the overall sample. We also assessed the association between TB (exposure) and depressive episode (outcome) using multivariable binary logistic regression while stratify- ing by region (Africa, Americas, Asia, Europe) or country income level (low-income, middle-income). For these strati- fied analyses, we could not assess all depression subtypes as the outcome as the number of individuals with TB was small in some subsamples. Statistical analysis Publically available data of the WHS included 69 countries. The data were nationally representative for all countries with the exception of China, Comoros, the Republic of Congo, Ivory Coast, India, and Russia. We excluded 10 countries as they lacked sampling information. A further 10 high-income countries were deleted as the focus of the study was on LMICs. Finally, Turkey was deleted due to lack of informa- tion on education and diabetes. Thus, a total of 48 countries, of which 21 (n = 105,286) and 27 (n = 137,666) were low- income and middle-income countries, respectively, at the time of the survey (2003) according to the World Bank, were included in the final sample. According to the United Nations’ classification system (http://unstats.un.org/unsd/ methods/m49/m49regin.htm), these corresponded to 20 countries in Africa (n = 82,424), 6 in the Americas (n = 62,732), 13 in Asia (n = 81,633), and 9 in Europe y All regression analyses were adjusted for age, sex, educa- tion, wealth, household size, location, smoking, alcohol consumption, BMI, diabetes, and country. Adjustment for country was performed by including dummy variables in the models, as in previous WHS publications [11, 19]. All variables were included in the models as categorical variables with the exception of age, household size, and the six variables on health status (continuous variables). The sample weighting and the complex study design were taken into account in all analyses. Results from the logistic and linear regression are presented as odds ratios (ORs) and b- coefficients, respectively, with 95% confidence intervals (CIs). The level of statistical significance was set at P < 0.05. Koyanagi et al. BMC Medicine (2017) 15:209 Koyanagi et al. Statistical analysis BMC Medicine (2017) 15:209 Page 4 of 10 Table 1 Sample characteristics Characteristic Category Unweighted N % or Mean (SD) Tuberculosis No 196,417 98.3 Yes 3347 1.7 Depression No depression 205,752 87.7 Subsyndromal depression 5238 2.6 Brief depressive episode 6674 2.9 Depressive episode 13,965 6.9 Age, years Mean (SD) 233,879 38.4 (16.1) Sex Male 104,355 49.2 Female 129,448 50.8 Education No formal 52,116 26.5 Primary 76,193 30.9 Secondary 86,740 33.5 Tertiary 17,860 9.2 Wealth Poorest 51,599 20.1 Poorer 45,893 20.0 Middle 42,317 19.9 Richer 40,128 20.0 Richest 37,724 20.0 Household size Mean (SD) 242,311 5.7 (3.0) Setting Rural 117,556 56.5 Urban 114,825 43.5 Current smoking No 174,814 73.5 Yes 54,746 26.5 Alcohol consumption Lifetime abstainer 142,282 66.4 Non-heavy 74,016 28.8 Infrequent heavy 8817 3.7 Frequent heavy 2411 1.0 Body mass index, kg/m2 <18.5 16,883 13.8 18.5–24.9 95,208 57.9 25.0–29.9 38,700 19.3 ≥30.0 18,287 9.0 Diabetes No 205,671 97.0 Yes 6537 3.0 Data are unweighted N and weighted proportion or mean (SD) SD standard deviation Under 10% of the data were missing for the variables used in the analysis with the exception of TB (17.7%), BMI (30.3%), and diabetes (12.6%). For the regression analyses, we conducted multiple imputation of missing values using the mi commands in Stata using chained equations (20 imputations) [24]. This method uses infor- mation from all other variables except the one being imputed to impute missing values. The variables included in the imputation model were the outcome and all other covariates [12]. The results based on complete case analysis were similar. Table 1 Sample characteristics Results The analytical sample consisted of 242,952 individuals with a mean (SD) age of 38.4 (16.1) years and 50.8% were women (Table 1). The prevalence (95% CI) of TB was 1.7% (1.5–1.8%). All types of depression were more frequent among those with TB, with the difference being particularly pronounced for depressive episode (Fig. 1). The prevalence of depressive episode among those with and without TB was 23.7% (95% CI 20.5–27.1%) and 6.8% (95% CI 6.5–7.1%), respectively (χ2 test P < 0.001). The results of the multivariable multinomial logistic regression using the overall sample showed that TB is associated with a 1.98 (95% CI 1.47–2.67), 1.75 (95% CI 1.26–2.42), and 3.68 (95% CI 3.01–4.50) times higher odds for subsyndro- mal depression, brief depressive episode, and depressive episode, respectively (Table 2). Older age, female sex, lower levels of wealth, smoking, and diabetes were signifi- cant correlates of depressive episode. The association between TB and depressive episode estimated by multivariable binary logistic regression by regions or country income levels are shown in Table 3. TB was associated with a depressive episode across regions and county income levels although the estimates for Europe did not reach statistical significance, possibly due to lack of statistical power (OR, 2.67; 95% CI 0.75–9.52; P = 0.1293). Compared to those with no TB or depressive episode, depression alone, TB alone, and comorbid TB/depres- sion were all significantly associated with worse health status scores in all domains. Comorbid TB/depression was associated with the largest decline (Table 4). The results of the age-stratified analyses are shown in Additional file 1: Table S3. The decline in health status associated with depression alone and co- occurring TB/depression was similar across age groups, but that of TB alone was less pronounced in the oldest age group (i.e., ≥65 years). In order to assess whether the difference between TB alone and comorbid TB/depression is statistically significant, we also conducted the same analysis but changing the reference category to TB alone (overall sample). The b-coefficients (95% CIs) for comorbid TB/depression (vs. TB alone) were mobility 19.17 (14.52–23.81), self- care 18.46 (13.41–23.54), pain/discomfort 20.14 (16.23–24.04), cognition 15.61 (11.13-20.09), interper- sonal activities 15.04 (10.13–19.96), and sleep/energy 19.67 (14.47–24.88) (all P < 0.0001). Overall, the interaction analysis showed that depression signifi- cantly amplifies the association between TB and difficul- ties in self-care but not with other health domains (Additional file 1: Table S4). Koyanagi et al. Discussion We found that TB is associated with the entire depres- sion spectrum in the overall sample, and that the associ- ation between TB and depressive episode is comparable across regions and country income levels. Furthermore, the co-occurrence of depression and TB was associated with a major decrement in all health domains assessed compared to TB alone, with this additive effect being particularly pronounced for difficulties in self-care. The strengths of the study include the large sample size and use of predominantly nationally representative data from approximately one-fourth of the countries in the world obtained by standardized questionnaires across all countries. To the best of our knowledge, this is the first general population study on TB and depression. Further- more, it is one of the very few studies assessing the asso- ciation between TB and depression severity, and is the first to assess the joint effect of TB and depression on a variety of health conditions (i.e., mobility, self-care, pain/ discomfort, cognition, interpersonal activities, sleep/energy). The finding that there may be a synergistic effect between TB and depression in terms of some health outcomes (i.e., self-care) is novel. Other factors which were identified as significant correlates of a depressive episode in our study included sociodemographic factors (older age, female sex, lower levels of wealth), smoking, and diabetes. Previous studies have also found these factors to be associated with depression [32–36]. In particular, diabetes is known to increase risk for TB [37], and may be an important risk factor for TB in LMICs [10] as there is an upward trend in diabetes prevalence mainly driven by changes in lifestyles and diet in this setting [38]. On the other hand, diabetes and depression are often comorbid and com- mon pathophysiological mechanisms (e.g., stress, inflammation) may underlie this co-occurrence [39]. In our study, compared to TB occurring in isolation, co-existing TB/depression was associated with decre- ments in all health domains assessed, while a significant interaction was observed for difficulties in self-care. These results are in line with a small cross-sectional study from Turkey showing that psychiatric comorbidity is associated with a higher rate of disability among TB patients [40]. Depression may lead to poor adherence to The association between TB and depression may be bidirectional [7, 25]. Results BMC Medicine (2017) 15:209 Page 5 of 10 Fig. 1 Prevalence of each type of depression by the presence or absence of tuberculosis. Bars denote 95% confidence intervals. Estimates are based on weighted sample on by the presence or absence of tuberculosis. Bars denote 95% confidence intervals. Estimates are Fig. 1 Prevalence of each type of depression by the presence or absence of tuberculosis. Bars denote 95% confidence intervals. Estimates are based on weighted sample Table 2 Association of tuberculosis and other covariates with depression estimated by multivariable multinomial logistic regression Depression subtypes (Reference = No depression) Subsyndromal depression Brief depressive episode Depressive episode Characteristic Category OR (95% CI) P value OR (95% CI) P value OR (95% CI) P value Tuberculosis Yes vs. No 1.98 (1.47–2.67) <0.0001 1.75 (1.26–2.42) 0.0008 3.68 (3.01–4.50) <0.0001 Age, years Per unit increase 1.02 (1.02–1.03) <0.0001 1.01 (1.00–1.01) 0.0003 1.02 (1.02–1.02) <0.0001 Sex Female vs. Male 1.84 (1.60–2.12) <0.0001 2.19 (1.94–2.46) <0.0001 2.06 (1.87–2.27) <0.0001 Education No formal 1.00 1.00 1.00 Primary 0.86 (0.72–1.03) 0.1101 0.95 (0.82–1.10) 0.4629 0.92 (0.83–1.03) 0.1414 Secondary 0.76 (0.61–0.94) 0.0133 0.98 (0.81–1.18) 0.8148 0.77 (0.67–0.88) 0.0001 Tertiary 0.83 (0.53–1.30) 0.4073 0.80 (0.62–1.02) 0.0686 0.82 (0.53–1.25) 0.3464 Wealth Poorest 1.00 1.00 1.00 Poorer 1.00 (0.80–1.25) 0.9855 0.98 (0.84–1.16) 0.8384 0.95 (0.85–1.07) 0.4023 Middle 1.07 (0.86–1.34) 0.5281 0.90 (0.76–1.06) 0.2022 0.96 (0.84–1.08) 0.4651 Richer 1.05 (0.81–1.36) 0.7289 0.97 (0.82–1.16) 0.7557 0.87 (0.76–1.00) 0.0486 Richest 0.90 (0.64–1.26) 0.5175 0.80 (0.66–0.97) 0.0229 0.69 (0.57–0.83) 0.0001 Household size Per unit increase 1.03 (1.00–1.06) 0.0296 1.01 (0.98–1.03) 0.4946 1.01 (0.99–1.04) 0.2889 Setting Urban vs. Rural 0.93 (0.79–1.11) 0.4373 1.12 (0.98–1.28) 0.0978 1.05 (0.94–1.17) 0.4040 Current smoking Yes vs. No 1.31 (1.10–1.55) 0.0026 1.26 (1.10–1.44) 0.0007 1.29 (1.14–1.45) <0.0001 Alcohol consumption Lifetime abstainer 1.00 1.00 1.00 Non-heavy 1.27 (1.09–1.47) 0.0026 1.51 (1.32–1.73) <0.0001 1.10 (0.99–1.22) 0.0685 Infrequent heavy 1.45 (1.02–2.06) 0.0407 1.72 (1.29–2.28) 0.0002 1.03 (0.83–1.29) 0.7845 Frequent heavy 2.10 (1.15–3.85) 0.0159 1.85 (1.18–2.92) 0.0080 1.14 (0.79–1.65) 0.4887 Body mass index, kg/m2 <18.5 1.02 (0.76–1.36) 0.8991 1.01 (0.83–1.23) 0.8988 1.08 (0.90–1.31) 0.4065 18.5–24.9 1.00 1.00 1.00 25.0–29.9 1.10 (0.89–1.36) 0.3673 0.96 (0.83–1.11) 0.5485 0.99 (0.88–1.11) 0.8586 ≥30.0 1.09 (0.84–1.41) 0.5230 1.05 (0.86–1.29) 0.6136 1.05 (0.91–1.22) 0.4831 Diabetes Yes vs. No 1.14 (0.84–1.54) 0.4000 1.39 (1.09–1.76) 0.0076 1.91 (1.62–2.24) <0.0001 Model is adjusted for all variables in the Table and country Page 6 of 10 Koyanagi et al. Results BMC Medicine (2017) 15:209 Page 6 of 10 Table 3 Association between tuberculosis (exposure) and depressive episode (outcome) by regions or country income level Region or country income level OR (95% CI) P value Africa 3.50 (2.76–4.43) <0.0001 Americas 2.74 (1.80–4.18) <0.0001 Asia 3.75 (2.74–5.14) <0.0001 Europe 2.67 (0.75–9.52) 0.1293 Low-income countries 3.52 (2.74–4.54) <0.0001 Middle-income countries 3.24 (2.40–4.35) <0.0001 Estimates are based on multivariable logistic regression Models are adjusted for age, sex, education, wealth, household size, location, smoking, alcohol consumption, body mass index, diabetes, and country OR odds ratio, CI confidence interval Table 3 Association between tuberculosis (exposure) and depressive episode (outcome) by regions or country income level immunity, leading to an increased risk for TB [8], while increased inflammation in TB may increase risk for depression [26]. Alternatively, depression may be a psy- chological reaction to the symptoms of TB (e.g., chronic cough, fatigue, weight loss) or associated disability [27], while hypoxia in chronic pulmonary diseases may induce depression [28]. It is also possible that patients with TB are perceived as a source of contagion in the commu- nity, which may lead to discrimination, stigma, social isolation, and rejection, and may predispose individuals to a higher risk for depression [27, 29]. Further, some anti-TB drugs can induce depression [27]. Finally, common risk factors, such as compromised immunity, stress, and malnutrition, may underlie the association [8, 26, 30, 31]. Regardless of whether depression and TB are etiologically related, the mere co-existence can complicate the diagnosis and management of these conditions, while it is also pos- sible that they mutually influence each other and lead to the exacerbation of the other, altering the clinical course [27]. Estimates are based on multivariable logistic regression Models are adjusted for age, sex, education, wealth, household size, location, smoking, alcohol consumption, body mass index, diabetes, and country OR odds ratio, CI confidence interval g g g Models are adjusted for age, sex, education, wealth, household size, location, smoking, alcohol consumption, body mass index, diabetes, and country TB tuberculosis CI confidence interval Discussion However, the precise underlying mechanisms or the rea- son why an interaction was only observed for self-care is unclear and warrants further investigation. Finally, delayed diagnosis of TB in people with depression may also partly explain our findings. It has been reported that delayed detection of physical diseases may be common in individuals with depression [41]. Thus, it may be that, when individuals with prior depression are diagnosed with TB, their TB symptoms are more severe compared to those without prior depression. Lack of motivation or social support and cognitive impairment, which may affect decision-making [42], might limit access to health- care among depressed individuals, leading to delayed diagnosis and treatment initiation for TB. g y g p Our results should be interpreted in the light of several limitations. First, we lacked information on HIV, which is known to be associated with higher risk for TB [1] and depression [51]. Thus, some of the association may be attributable to comorbid HIV. However, this may not have been a major limitation as our region-wise analysis showed that TB is associated with depression even in areas with very low HIV prevalence (e.g., the Americas). Second, our study was based on the symp- toms of TB rather than a laboratory confirmed diagnosis. Although we used the identical definition for TB used in previous publications [10–12], it is possible that some level of symptom overlap may exist between respiratory diseases such as pneumonia, bronchitis, and chronic obstructive pulmonary disease, which may also cause cough of long duration and hemoptysis. Thus, our estimates may partially be representing the association between these conditions and depression. Furthermore, the potential misclassification may have led to an underestima- tion of the association between TB and depression. How- ever, it is reassuring that the prevalence of depression in TB was within the previously reported range of estimates among patients with confirmed TB [2]. Additionally, we are not aware of any other population-based data with such a large number of LMICs that can be used to investigate the TB–depression relationship. Third, high-risk groups, such as the institutionalized and homeless, were not included in our study and thus our findings are not generalizable to this population. Finally, the direction of causality cannot be established due to the cross-sectional design. Previous studies have shown that treating the psycho- logical aspects of TB may lead to better clinical outcomes. Discussion Depression itself may compromise Table 4 Association between TB/depressive episode groups and health status estimated by multivariable linear regression TB (-) Depression (+) TB (+) Depression (-) TB (+) Depression (+) b-coefficient (95% CI) P value b-coefficient (95% CI) P value b-coefficient (95% CI) P value Mobility 15.92 (14.81–17.04) <0.0001 8.63 (6.55–10.71) <0.0001 27.80 (23.60–31.99) <0.0001 Self-care 11.96 (10.78–13.13) <0.0001 5.34 (3.20–7.48) <0.0001 23.80 (19.11–28.49) <0.0001 Pain/discomfort 18.70 (17.39–20.00) <0.0001 10.27 (8.24–12.30) <0.0001 30.41 (26.97–33.84) <0.0001 Cognition 16.55 (15.26–17.83) <0.0001 8.63 (6.35–10.91) <0.0001 24.24 (20.19–28.28) <0.0001 Interpersonal activities 12.69 (11.53–13.86) <0.0001 4.41 (2.34–6.48) <0.0001 19.45 (14.81–24.09) <0.0001 Sleep/energy 19.61 (18.37–20.85) <0.0001 10.32 (7.94–12.70) <0.0001 29.99 (25.30–34.68) <0.0001 Reference category is TB (-) Depression (-) Health status was the outcome and scores ranged from 0 to 100 with higher scores corresponding to worse health status Models are adjusted for age, sex, education, wealth, household size, location, smoking, alcohol consumption, body mass index, diabetes, and country TB tuberculosis, CI confidence interval Page 7 of 10 Koyanagi et al. BMC Medicine (2017) 15:209 Page 7 of 10 depression in TB. However, symptoms specific to depression (e.g., low mood, anhedonia) and symptoms of depression that overlap with TB (e.g., fatigue) should be distinguished. Some studies have assessed the validity of depression screeners such as the Center for Epidemiological Studies Depression scale or the General Health Questionnaire 12 among TB patients [40, 49]. These studies found that these screening tools can be used among TB patients to detect de- pression but that there may be a disease-specific optimal cut-off. Future studies on the validity and reliability of such screening tools are warranted, as only scarce data from limited populations are currently available. Finally, patient education and community awareness regarding facts and myths of TB may also be important [50], as discrimination and stigma can be underlying causes of depression in TB. anti-TB drugs, and thereby exacerbate the symptoms of TB and its associated disability. Indeed, a prospective study from Peru showed that co-occurring TB/depres- sion leads to lower adherence to TB treatment and higher mortality when compared to TB without depres- sion [3]. The fact that a significant interaction was observed for self-care may imply that there is a synergis- tic effect between TB and depression. It may be hypoth- esized, for example, that depression leads to poor TB treatment adherence and exacerbation of symptoms, which in turn may lead to a worsening of depression. Discussion For example, a prospective controlled trial in India showed that psychotherapy during TB treatment leads to higher adherence, treatment, and cure rates [43]. Furthermore, a psychological support group intervention for patients with MDR-TB in Peru showed that such an intervention can improve treatment adherence and completion [44]. Additionally, the formation of ‘TB clubs’ in Ethiopia increased treatment completion rates and reduced the stigma associated with TB [45]. Recently, a randomized controlled trial in Ethiopia showed that psychological counseling and educational intervention can substantially improve treatment adherence rates in TB [46]. A multi-faceted approach is likely to be relevant in addressing comorbid depression and TB in LMICs. First, previous studies from LMICs have shown that the treating doctor is often not aware of co-existing psychiatric morbid- ity in TB patients [47]. Thus, training of medical profes- sionals and students on the psychological aspects of TB may lead to early detection and better management of psychiatric complications, and ultimately to a better clinical outcome of TB. Next, a close collaboration between TB and mental health specialists would be important for the early detection and treatment of depression in TB. Previous studies have shown that training of non-mental health specialists in LMICs may only have a limited impact on depression detection rates [48]. Thus, screening for depression may be a cost-effective strategy to improve detection rates of Conclusions In conclusion, individuals with TB have higher odds for depression, and the co-occurrence of TB and depression is associated with decrements in health. Screening for and addressing depression in individuals with TB may lead to better clinical outcomes. However, mental health services and specialists are limited in low-resourced settings where the highest burden of TB is located. Increased recognition Koyanagi et al. BMC Medicine (2017) 15:209 Page 8 of 10 (Continued) Republic of congo Unité de recherche sur les systèmes de santé Ivory Coast Ministère de la Santé Croatia The Croatian National Institute of Public Health Czech Republic Institute of Health Information and Statistics Dominican Republic Centro de Estudios Sociales y Demográficos (CESDEM) Ecuador Fundación Ecuatoriana para la Salud y el Desarrollo (FESALUD) Estonia Saar Poll Ltd. Ethiopia Department of Community Health, Jimma University Georgia Georgian State Medical Academy (GSMA) Ghana Department of Community Health, Ghana Medical School Hungary Johan Bela National Centre for Epidemiology India International Institute of Population Sciences Kazakhstan Kazakstan School of Public Health (KSPH) Kenya Central Bureau of Statistics Laos National Institute of Public Health, Ministry of Health Latvia The Health Promotion Center Malawi Centre for Social Research (CSR) Malaysia Public Health Institute, Ministry of Health Mali Cellule de Planification et de Statistique (CPS) Mauritania Office Nationale de la Statistique (ONS) Mauritius Mauritius Institute of Health Mexico Instituto Nacional de Salud Pública Morocco Ministère de la Santé Myanmar Department of Medical Research, Ministry of Health Namibia Ministry of Health Nepal ORG-MARG Nepal PVT Ltd. Pakistan Ministry of Health Paraguay Fac. de Ciencias Veterinarias, Universidad Nacional/ DGEEC Philippines College of Medicine, University of the Philippines Russia Semashko Institute for Research on Social Hygiene Senegal Direction Etudes, Recherche et Formation (DERF) Slovakia Centre of Biostatistics and Environment South Africa Community Agency for Social Enquiry (CASE) Sri Lanka Ministry of Health Swaziland Faculty of Health Sciences, University of Swaziland Tunisia Institut National de la Santé Publique Ukraine Odessa State Medical University Uruguay Centro de Estudios de Economia y Salud (CEES) Vietnam Ministry of Health Zambia School of Humanities & Social Sciences, University of Zambia Zimbabwe Community Health, University of Zimbabwe of co-existing depression in TB patients by health professionals and the use of non-specialist health workers trained in mental healthcare, especially in resource- limited settings, may be key. Funding AK’s work was supported by the Miguel Servet contract financed by the CP13/ 00150 and PI15/00862 projects, integrated into the National R + D + I and funded by the ISCIII – General Branch Evaluation and Promotion of Health Research, and the European Regional Development Fund (ERDF-FEDER). BS receives funding from the National Institute for Health Research Collaboration for Leadership in Applied Health Research & Care Funding scheme. The views expressed in this publication are those of the authors and not necessarily those of the NHS, the National Institute for Health Research, or the Department of Health. These funders had no role in the design of the study, collection, analysis, and interpretation of data, or in writing the manuscript. Availability of data and materials The dataset supporting the conclusions of this article is available (upon registration) from http://www.who.int/healthinfo/survey/en/. Conclusions However, given that health- care workers are at increased risk of occupationally acquired TB in LMICs [52], sound infection control mea- sures should be implemented to protect these individuals, yet this is a particular challenge in LMICs due to financial constraints. Finally, simultaneously addressing the mental and physical aspects of TB may lead to reduction in TB transmission [53], and also possibly in TB mortality and MDR-TB. This is an area for future research. Additional file Additional file 1: Table S1. Questions used to assess health status. Table S2. Countries included in the analysis and sample size. Table S3. Association between TB/depressive episode groups and health status by age groups estimated by multivariable linear regression. Table S4. Interaction effect of TB and depressive episode on health status. (DOCX 45 kb) Abbreviations BMI: body mass index; CI: confidence intervals; LMIC: low- and middle-income countries; MDR-TB: multidrug resistant tuberculosis; OR: odds ratio; TB: tuberculosis; WHS: World Health Survey Authors’ contributions AK conceived the study idea, analyzed and interpreted the data, and wrote the main body of the text. DV, AFC, JED, DP, JMH, NV, and BS contributed to the drafting of the manuscript, interpreted the data, and commented for intellectual content. All authors read and approved the final manuscript. References 1. World Health Organization. Tuberculosis. http://www.who.int/mediacentre/ factsheets/fs104/en/. Accessed 27 July 2017. 1. World Health Organization. Tuberculosis. http://www.who.int/mediacentre/ factsheets/fs104/en/. Accessed 27 July 2017. 24. StataCorp. http://www.stata.com/manuals13/mimiimputechained.pdf. Accessed 15 Sept 2017. 2. Sweetland A, Oquendo M, Wickramaratne P, Weissman M, Wainberg M. Depression: a silent driver of the global tuberculosis epidemic. World Psychiatry. 2014;13(3):325–6. 25. Yen YF, Chung MS, Hu HY, Lai YJ, Huang LY, Lin YS, Chou P, Deng CY. Association of pulmonary tuberculosis and ethambutol with incident depressive disorder: a nationwide, population-based cohort study. J Clin Psychiatry. 2015;76(4):e505–11. 3. Ugarte-Gil C, Ruiz P, Zamudio C, Canaza L, Otero L, Kruger H, Seas C. Association of major depressive episode with negative outcomes of tuberculosis treatment. PLoS One. 2013;8(7):e69514. 26. Kiecolt-Glaser JK, Glaser R. Depression and immune function: central pathways to morbidity and mortality. J Psychosom Res. 2002;53(4):873–6. 4. Chaudhry LA, Zamzami M, Aldin S, Pazdirek J. Clinical consequences of non- compliance with directly observed therapy short course (DOTS): Story of a recurrent defaulter. Int J Mycobacteriol. 2012;1(2):99–103. 27. Pachi A, Bratis D, Moussas G, Tselebis A. Psychiatric morbidity and other factors affecting treatment adherence in pulmonary tuberculosis patients. Tuberc Res Treat. 2013;2012:489865. recurrent defaulter. Int J Mycobacteriol. 2012;1(2):99–103. 5. Kulkarni P, Akarte S, Mankeshwar R, Bhawalkar J, Banerjee A, Kulkarni A. Non-adherence of new pulmonary tuberculosis patients to anti-tuberculosis treatment. Ann Med Health Sci Res. 2013;3(1):67–74. 28. Mikkelsen RL, Middelboe T, Pisinger C, Stage KB. Anxiety and depression in patients with chronic obstructive pulmonary disease (COPD). A review. Nord J Psychiatry. 2004;58(1):65–70. 6. Patel V, Kleinman A. Poverty and common mental disorders in developing countries. Bull World Health Organ. 2003;81(8):609–15. 29. Lee LY, Tung HH, Chen SC, Fu CH. Perceived stigma and depression in initially diagnosed pulmonary tuberculosis patients. J Clin Nurs. 2017. doi:10.1111/jocn.13837. 7. Oh KH, Choi H, Kim EJ, Kim HJ, Cho SI. Depression and risk of tuberculosis: a nationwide population-based cohort study. Int J Tuberc Lung Dis. 2017; 21(7):804–9. 30. Cegielski JP, McMurray DN. The relationship between malnutrition and tuberculosis: evidence from studies in humans and experimental animals. Int J Tuberc Lung Dis. 2004;8(3):286–98. 8. Reiche EM, Nunes SO, Morimoto HK. Stress, depression, the immune system, and cancer. Lancet Oncol. 2004;5(10):617–25. 31. Vafaei Z, Mokhtari H, Sadooghi Z, Meamar R, Chitsaz A, Moeini M. Malnutrition is associated with depression in rural elderly population. J Res Med Sci. 2013;18 Suppl 1:S15–9. 9. Competing interests 15. World Health Organization. The ICD–10 Classification of Mental and Behavioural Disorders: Diagnostic Criteria for Research. 1993. www.who.int/ classifications/icd/en/GRNBOOK.pdf. Accessed 15 Sept 2017. Received: 1 September 2017 Accepted: 9 November 2017 23. Timsina LR, Willetts JL, Brennan MJ, Marucci-Wellman H, Lombardi DA, Courtney TK, Verma SK. Circumstances of fall-related injuries by age and gender among community-dwelling adults in the United States. PLoS One. 2017;12(5):e0176561. Publisher’s Note classifications/icd/en/GRNBOOK.pdf. Accessed 15 Sept 2017. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 16. Ware Jr J, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34(3):220–33. Author details 1 17. Feeny D, Furlong W, Boyle M, Torrance GW. Multi-attribute health status classification systems. Health Utilities Index. Pharmacoeconomics. 1995; 7(6):490–502. 1Research and Development Unit, Parc Sanitari Sant Joan de Déu, Universitat de Barcelona, Fundació Sant Joan de Déu, Dr. Antoni Pujadas, 42, Sant Boi de Llobregat, Barcelona, Spain. 2Instituto de Salud Carlos III, Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM, Madrid, Spain. 3KU Leuven Department of Rehabilitation Sciences, Leuven, Belgium. 4KU Leuven, University Psychiatric Center KU Leuven, Leuven-Kortenberg, Belgium. 5Department of Clinical Medicine and Translational Psychiatry Research Group, Faculty of Medicine, Federal University of Ceará, Fortaleza, Brazil. 6Fordham University, Graduate School of Social Service, New York, NY, USA. 7Operation Research Unit, Doctors with Africa, Maputo, Mozambique. 8Institute for Clinical Research and Education in Medicine (IREM), Padova, Italy. 9National Research Council, Neuroscience Institute, Aging Branch, Padova, Italy. 10Physiotherapy Department, South London and Maudsley NHS Foundation Trust, Denmark Hill, London, UK. 11Health Service and Population Research Department, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, UK. 12Faculty of Health, Social Care and Education, Anglia Ruskin University, Chelmsford, UK. 18. Kind P. The EuroQoL instrument: an index of health-related quality of life. In: Spilker B, editor. Quality of Life and Pharmacoeconomics in Clinical Trial. Philadelphia, PA: Lippincott-Raven Publishers; 1996. p. 191–201. 19. Koyanagi A, Oh H, Stickley A, Haro JM, DeVylder J. Risk and functional significance of psychotic experiences among individuals with depression in 44 low- and middle-income countries. Psychol Med. 2016;46(12):2655–65. 20. Stubbs B, Koyanagi A, Schuch F, Firth J, Rosenbaum S, Gaughran F, Mugisha J, Vancampfort D. Physical activity levels and psychosis: a mediation analysis of factors influencing physical activity target achievement among 204 186 people across 46 low- and middle-income countries. Schizophr Bull. 2017;43(3):536–45. 21. Hosseinpoor AR, Bergen N, Kunst A, Harper S, Guthold R, Rekve D, d'Espaignet ET, Naidoo N, Chatterji S. Socioeconomic inequalities in risk factors for non communicable diseases in low-income and middle-income countries: results from the World Health Survey. BMC Public Health. 2012;12:912. 22. World Health Organization. International guide for monitoring alcohol consumption and related harm. Geneva: WHO; 2002. Received: 1 September 2017 Accepted: 9 November 2017 Received: 1 September 2017 Accepted: 9 November 2017 Ethics approval and consent to participate Ethical approval was obtained from each of the following committees in the respective countries for the World Health Survey: Country Ethical committee approving the study Bangladesh Mitra and Associates Bosnia and Herzegovina The Federal Public Health Institute Brazil Fundacao Oswaldo Cruz Burkina Faso Institut de Recherche en Sciences de la Santé Chad Faculté des Sciences de la Santé, Univ N'Djamena China Centre for Health Statistics Information Comoros Bureau Comorien de Conseil Page 9 of 10 Page 9 of 10 Koyanagi et al. BMC Medicine (2017) 15:209 Consent for publication Not applicable. International Diagnostic Interview (CIDI). Int J Methods Psychiatr Res. 2004;13(2):93–121. International Diagnostic Interview (CIDI). Int J Methods Psychiatr Res. 2004;13(2):93–121. 14. Ayuso-Mateos JL, Nuevo R, Verdes E, Naidoo N, Chatterji S. From depressive symptoms to depressive disorders: the relevance of thresholds. Br J Psychiatry 2010;196(5):365–71. Competing interests The authors declare that they have no competing interests. Competing interests The authors declare that they have no competing interests. Koyanagi et al. BMC Medicine (2017) 15:209 36. Mommersteeg PM, Herr R, Pouwer F, Holt RI, Loerbroks A. The association between diabetes and an episode of depressive symptoms in the 2002 World Health Survey: an analysis of 231,797 individuals from 47 countries. Diabet Med. 2013;30(6):e208–14. 37. Jeon CY, Murray MB. Diabetes mellitus increases the risk of active tuberculosis: a systematic review of 13 observational studies. PLoS Med. 2008;5(7):e152. 37. Jeon CY, Murray MB. Diabetes mellitus increases the risk of active tuberculosis: a systematic review of 13 observational studies. PLoS Med. 2008;5(7):e152. 38. Popkin BM. Nutrition transition and the global diabetes epidemic. Curr Diab Rep. 2015;15(9):64. 39. Badescu SV, Tataru C, Kobylinska L, Georgescu EL, Zahiu DM, Zagrean AM, Zagrean L. The association between diabetes mellitus and depression. J Med Life. 2016;9(2):120–5. 40. Aydin IO, Ulusahin A. Depression, anxiety comorbidity, and disability in tuberculosis and chronic obstructive pulmonary disease patients: applicability of GHQ-12. Gen Hosp Psychiatry. 2001;23(2):77–83. 41. Kisely S, Crowe E, Lawrence D. Cancer-related mortality in people with mental illness. JAMA Psychiat. 2013;70(2):209–17. 42. Trivedi MH, Greer TL. Cognitive dysfunction in unipolar depression: implications for treatment. J Affect Disord. 2014;152-154:19–27. 43. Janmeja AK, Das SK, Bhargava R, Chavan BS. Psychotherapy improves compliance with tuberculosis treatment. Respiration. 2005;72(4):375–80. 44. Acha J, Sweetland A, Guerra D, Chalco K, Castillo H, Palacios E. Psychosocial support groups for patients with multidrug-resistant tuberculosis: five years of experience. Glob Public Health. 2007;2(4):404–17. 45. Demissie M, Getahun H, Lindtjørn B. Community tuberculosis care through “TB clubs” in rural North Ethiopia. Soc Sci Med. 2003;56(10):2009–18. 46. Tola HH, Shojaeizadeh D, Tol A, Garmaroudi G, Yekaninejad MS, Kebede A, Ejeta LT, Kassa D, Klinkenberg E. Psychological and educational intervention to improve tuberculosis treatment adherence in ethiopia based on health belief model: a cluster randomized control trial. PLoS One. 2016;11(5): e0155147. 47. Aghanwa HS, Erhabor GE. Demographic/socioeconomic factors in mental disorders associated with tuberculosis in southwest Nigeria. J Psychosom Res. 1998;45(4):353–60. 48. Reynolds CFR, Patel V. Screening for depression: the global mental health context. World Psychiatry. 2017;16(3):316–7. 49. Chishinga N, Kinyanda E, Weiss HA, Patel V, Ayles H, Seedat S. Validation of brief screening tools for depressive and alcohol use disorders among TB and HIV patients in primary care in Zambia. BMC Psychiatry. 2011;11:75. 50. Baral SC, Karki DK, Newell JN. Causes of stigma and discrimination associated with tuberculosis in Nepal: a qualitative study. BMC Public Health. 2007;7:211. 51. Sherr L, Cluver L. Koyanagi et al. BMC Medicine (2017) 15:209 Koyanagi et al. BMC Medicine (2017) 15:209 World Health Day focus on HIV and depression - a comorbidity with specific challenges. J Int AIDS Soc. 2017;20(1):21956. 52. Joshi R, Reingold AL, Menzies D, Pai M. Tuberculosis among health-care workers in low- and middle-income countries: a systematic review. PLoS Med. 2006;3(12):e494. 53. Mason PH, Sweetland AC, Fox GJ, Halovic S, Nguyen TA, Marks GB. Tuberculosis and mental health in the Asia-Pacific. Australas Psychiatry. 2016;24(6):553–5. References Nuevo R, Chatterji S, Verdes E, Naidoo N, Arango C, Ayuso-Mateos JL. The continuum of psychotic symptoms in the general population: a cross- national study. Schizophr Bull. 2012;38(3):475–85. 32. Guerra M, Prina AM, Ferri CP, Acosta D, Gallardo S, Huang Y, Jacob KS, Jimenez- Velazquez IZ, Llibre Rodriguez JJ, Liu Z, et al. A comparative cross-cultural study of the prevalence of late life depression in low and middle income countries. J Affect Disord. 2016;190:362–8. 10. Goldhaber-Fiebert JD, Jeon CY, Cohen T, Murray MB. Diabetes mellitus and tuberculosis in countries with high tuberculosis burdens: individual risks and social determinants. Int J Epidemiol. 2011;40(2):417–28. 11. Stubbs B, Koyanagi A, Veronese N, Vancampfort D, Solmi M, Gaughran F, Carvalho AF, Lally J, Mitchell AJ, Mugisha J, et al. Physical multimorbidity and psychosis: comprehensive cross sectional analysis including 242,952 people across 48 low- and middle-income countries. BMC Med. 2016;14:189. 33. Wang K, Lu H, Cheung EF, Neumann DL, Shum DH, Chan RC. "Female preponderance" of depression in non-clinical populations: a meta-analytic study. Front Psychol. 2016;7:1398. 12. Vancampfort D, Koyanagi A, Ward P, Veronese N, Carvalho AF, Solmi M, Mugisha J, Rosenbaum S, De Hert M, Stubbs B. Perceived stress and its relationship with chronic conditions and multimorbidity among 229,293 community-dwelling adults in 44 low- and middle-income countries. Am J Epidemiol. 2017;186(8):979–89. 34. Freeman A, Tyrovolas S, Koyanagi A, Chatterji S, Leonardi M, Ayuso-Mateos JL, Tobiasz-Adamczyk B, Koskinen S, Rummel-Kluge C, Haro JM. The role of socio-economic status in depression: results from the COURAGE (aging survey in Europe). BMC Public Health. 2016;16:1098. 35. Mathew AR, Hogarth L, Leventhal AM, Cook JW, Hitsman B. Cigarette smoking and depression comorbidity: systematic review and proposed theoretical model. Addiction. 2017;112(3):401–12. 13. Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) Composite 13. Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health Organization (WHO) Composite Page 10 of 10 Koyanagi et al. BMC Medicine (2017) 15:209 Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries
https://openalex.org/W4385890702	https://zenodo.org/records/8256873/files/Abduturapova%20Dildora%20Farkhojon%20kizi.pdf	English	null	ANALYSIS OF THE FORMATION OF NEW RISKS IN THE EVELOPMENT OF INSURANCE RELATIONS	Zenodo (CERN European Organization for Nuclear Research)	2,023	cc-by	4,222	ANALYSIS OF THE FORMATION OF NEW RISKS IN THE EVELOPMENT OF INSURANCE RELATIONS https://doi.org/10.5281/zenodo.8256873 Abduturapova Dildora Farkhojon kizi Tashkent State University of Economics Basic doctoral student (PhD) jasdil1995@gmail.com Abduturapova Dildora Farkhojon kizi Tashkent State University of Economics Basic doctoral student (PhD) jasdil1995@gmail.com International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| Annotation. In the article, the concept of risk is explained by the author after studying the opinions of scientists. As a result of the analysis of new risks emerging in the insurance relations of our republic, in particular, natural disasters, economic changes, low water, etc., which have often occurred recently, proposals and recommendations were developed on the impact on human health, its solution in foreign practice. Keywords. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| eat, in the water we drink, and in the work we do towards a goal. He put forward the idea that the occurrence of risk is an event [5]. Based on the above, in our opinion, threats, losses, risks that may occur during any economic activity can be understood as economic risks. From this point of view, despite the fact that the causes of the risk are different, we believe that its consequences are of an economic nature. Research objective and methods. The purpose of the research is to assess the state of new risks in the insurance relations of Uzbekistan and to develop proposals and recommendations aimed at evaluating them based on analyzes and foreign experiences. Logical and structural analysis, grouping, and economic-statistical analysis methods were used in the implementation of the set goals. Analysis аnd results. One of the main conditions for the development of the insurance industry is to ensure the financial stability of the insurance company. The uniqueness of insurance activity is that it is not possible to predict the time of occurrence of the insurance event, and it is not known whether the insurance operation will be successful or vice versa. This abstraction creates complexity in the operation and solvency of insurance companies. According to scientists, "risks reflected in insurance contracts are insurance risks, and the expression of the risk assessment in money is the insurance rate" [6]. According to the legislation of the Republic of Uzbekistan, Article 3 of the Law "On Insurance Activities" states that "insurance risk is an anticipated event, and in the event of this event, insurance is implemented"[7]. As it can be seen, insurance risk and risk are synonymous, so we used the term risk in our work based on foreign scientific literature. In scientific literature, the classification of risks is given in different ways. In particular, Dj.M. Keynes[8] is one of the first to describe risks from the economic side and divide them into several classes (classifications). In his research, he distinguishes between two main types of individual risks, entrepreneurial risk and creditor risk. American scientist A. Marshall's developments [9] covered the main issues of the theory and practice of economic risk assessment. Uzbek scientist Q. Keywords. Risk, insurance contract, insurance risk, insured event, environmental insurance, abnormal weather. Introduction. There is always a possibility that a business that is established will be successful or not. Therefore, it is emphasized by many scientists that the first condition of any entrepreneurship is risk-taking. Literature review. So far, the concept of "risk" has been defined differently by scientists, in particular, our local scientists M. Eshov, A. Burkhanov, H. Khasanov, D. Baratova, J. If the Nosyrovs associate the term risk with the Arabic "risq", Latin "risicum", Greek "rizha"[1], foreign scientists V.M. Granaturov[2] mentioned that it was derived from Spanish "ridsico" (cliff), Italian "risco" (danger, threat). In Europe[3], the term began to be widely used since the Middle Ages, mainly as a threat to maritime trade. Today, it can be seen that most scientists connect the term risk to such concepts as risk, probability, uncertainty, expected damage, risk, probability of failure. S.Kerkegaard[4], a scientist who made a great contribution to the study of risks, divides fear into several classes and evaluates fear in the decision-making process of human nature as a risk. Local scientist Q. And Koldoshev in his researches, "risk is the possibility of the occurrence of some kind of dangerous situation, but not the occurrence of this situation. This risk is present in every step we take, in the food we 749 Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| eat, in the water we drink, and in the work we do towards a goal. He put forward the idea that the occurrence of risk is an event [5]. Based on the above, in our opinion, threats, losses, risks that may occur during any economic activity can be understood as economic risks. From this point of view, despite the fact that the causes of the risk are different, we believe that its consequences are of an economic nature International Journal of Education, Social Science & Humanities. Publishing centre of Finland Keywords. Koldoshev[10] divided risks into types according to the sources of origin and reasons for their occurrence. M. Mahmudova[11] additionally mentions several types of financial risks and divides them into fiscal risk, banking risk, insurance risk and financial literacy risks. It should be noted that although all research scientists have focused on the classification of risks, there is no consensus on this issue in scientific and other 750 Publishing centre of Finland Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| literature. In particular, there are 220 risks and more than 40 types of risks according to the classification system. International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| literature. In particular, there are 220 risks and more than 40 types of risks according to the classification system. Depending on whether one or another risk group is insured, the legislation of each country provides for its classification. According to it, activity in the insurance market of Uzbekistan is divided into two main sectors, general and life insurance, and appropriate classifications have been developed. Fulfillment of the insurer's obligations in insurance relations also depends on the correct formation of insurance reserves. Account groups are developed based on the term distribution of risks by insurance types. According to the legislation of the Republic of Uzbekistan, the insurance activity is divided into four account groups to calculate IMZ (unearned premium reserve) in the formation of insurance reserves of insurers[12]. Risks are equally distributed to the first account group for the entire insurance period of the insurance contract. The second group includes contracts provided that the occurrence of the insured event occurs after the insurance period. The third group includes contracts in which the occurrence of the insurance event and the period of implementation of the insurance obligation are uncertain. The fourth group includes insurance of agricultural products. So, from the above, we have seen the time-related grouping of insurable risks. In this direction, there are many insurance products in the national insurance market. Publishing centre of Finland Keywords. Nevertheless, the expansion of production and its impact on the environment, the outbreak of various diseases and the emergence of new forms due to environmental damage, the emergence of new cyber risks as a result of the acceleration of information processes, and at the same time, the formation of new risks based on the demand for modern approaches in the insurance industry analysis is important. As production and industry develop, along with protection from the risks associated with economic activity, there is a need to identify and eliminate or protect against environmental damage caused by these economic entities. This means the formation of environmental and man-made risks in insurance relations. Today, a number of environmental changes are taking place in the countries of the world due to natural disasters and the human factor. Climate and weather are changing significantly in all regions of the globe, water and soil are becoming increasingly polluted, and plant and animal species are declining. This, in turn, has a negative impact on the environment and the health of all people on our planet [13]. Environmental law (Umweitrecht), one of the developed countries of the world, has a special place for Germany as the regulatory legal documents defining the legal basis of environmental insurance. In 2016, the country's largest reinsurance company 751 Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| MunichRe (2017) identified 750 major global losses, which resulted in about 8,700 deaths and a total of 175 billion. amounted to a dollar loss. In the picture below, we can see the analysis of the previous years 2019-2022. Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| MunichRe (2017) identified 750 major global losses, which resulted in about 8,700 deaths and a total of 175 billion. amounted to a dollar loss. In the picture below, we can see the analysis of the previous years 2019-2022. y p y Figure 1. Publishing centre of Finland Keywords. Dynamics of Munich Re's natural disaster reinsurance coverage for 2019- 2022[14], million euros 2053 906 3139 2430 3124 4689 4304 4173 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 2019 2020 2021 2022 natural disaster losses other losses Figure 1. Dynamics of Munich Re's natural disaster reinsurance coverage for 2019- 2022[14], million euros According to Munich Re[15], the world's largest reinsurance organization, losses from natural disasters in 2022 amounted to 2.4 billion US dollars, of which 44 percent or 120 billion dollars were covered by reinsurance. This figure is $23 billion higher than the last five years. According to Thomas Blank, Chairman of the Board of Management of the company, “According to the results of the latest research, natural disaster figures for 2022 dominated as frequent events. Another concern that we see time and time again is that natural disasters can particularly affect the population of underdeveloped and developing countries. Therefore, more importance should be According to Munich Re[15], the world's largest reinsurance organization, losses from natural disasters in 2022 amounted to 2.4 billion US dollars, of which 44 percent or 120 billion dollars were covered by reinsurance. This figure is $23 billion higher than the last five years. According to Thomas Blank, Chairman of the Board of Management of the company, “According to the results of the latest research, natural disaster figures for 2022 dominated as frequent events. Another concern that we see time and time again is that natural disasters can particularly affect the population of underdeveloped and developing countries. Therefore, more importance should be 752 International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| given to prevention and financial protection, for example, insurance protection” [16]. vention and financial protection, for example, insurance protection” [16]. Figure 2. Global climate indicators (1880-2022)[17] Figure 2. Global climate indicators (1880 2022)[17] According to NCEI (National Center for Environmental Information), the past 9 years have been anomalously warm on record. As a result, the sea level is rising, and glaciers are melting at an average rate of 3 mm per year. Higher temperatures lead to more evaporation, which increases the energy content of the atmosphere. If extreme weather events become more frequent and/or severe, and effective mitigation measures are not taken, losses will continue to increase. Publishing centre of Finland Keywords. According to the law, environmental damage is applied equally to the defendant regardless of when it was caused (that is, before the adoption of this law or after the adoption of the law), and regardless of the degree of environmental pollution, each polluter must fully compensate for the damage caused. International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| Environmental, Compensation and Liability Actions establishes mandatory insurance of hazardous facilities, including activities related to waste collection and processing. According to the law, environmental damage is applied equally to the defendant regardless of when it was caused (that is, before the adoption of this law or after the adoption of the law), and regardless of the degree of environmental pollution, each polluter must fully compensate for the damage caused. Currently, not only in European countries, but also in the national legislation of the Russian Federation, the Republic of Kazakhstan, and the Republic of Azerbaijan, a mandatory type of environmental insurance has been put into practice. Article 36 of the Law on Nature Protection of December 9, 1992 provides that "in the Republic of Uzbekistan, the property and income of enterprises, institutions and organizations, the lives, health and property of citizens, due to environmental pollution and deterioration of the quality of natural resources" Voluntary and mandatory property insurance" serves as the legal basis of environmental insurance, but in current practice, no mechanism has been developed for this type of insurance. It is not a secret to anyone that in recent years, extreme weather events, which are considered global problems, and many losses as a result of them, have been occurring in our country. Especially during the last three years, as a result of environmental problems such as dust storms, heat and as a result drought in some areas, strong winds, floods in settlements, heavy rains, anomalous heat/cold, people's property, agriculture, and business activities are seriously damaged. In 2018 alone, as a result of floods observed in Samarkand, Navoi, Surkhandarya, Kashkadarya regions and Fergana valley of our country, serious damage was done in agriculture. Publishing centre of Finland Keywords. There is scientific evidence of the effects of climate change, such as severe thunderstorms in North America and Europe, major earthquakes in Turkey and Syria, wildfires in California, and heat and drought in Central Asia. In recent years, tropical storms (called hurricanes, typhoons or cyclones depending on the area where they occur) have been accompanied by increasingly heavy rainfall. There are also indications that the proportion of severe storms is increasing. Individual loss events cannot be attributed solely to climate change. However, long-term trends in meteorological data, when analyzed together with insurance companies and socio-economic data, reveal important information about the changing risks posed by weather hazards. USA environmental legislation and its provisions on environmental liability are also based on the "polluter pays" principle. This principle was most fully implemented in the federal Comprehensive Responsibility for Environmental Restoration and Conservation Act of 1980, known as Superfund. In addition, the country's Law on 753 Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| Environmental, Compensation and Liability Actions establishes mandatory insurance of hazardous facilities, including activities related to waste collection and processing. According to the law, environmental damage is applied equally to the defendant regardless of when it was caused (that is, before the adoption of this law or after the adoption of the law), and regardless of the degree of environmental pollution, each polluter must fully compensate for the damage caused. International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| Environmental, Compensation and Liability Actions establishes mandatory insurance of hazardous facilities, including activities related to waste collection and processing. Keywords. This sentence includes collecting information based on biological, chemical, hydrological, geological and other fields, assessing the environmental effects of human activity and dealing with its consequences. In the world's experience, there are mainly three methods of compensation for the damage caused to the economy as a result of anthropogenic activity on the environment, and they are carried out from the state budget, from the personal funds of the person who caused the damage, and from the funds embodied in the insurance business. Levers of state regulation of damage compensation create opportunities for direct and indirect influence on the insurance market and serve to improve relations in the field of environmental insurance, taking into account the real state of the economy and ecology. Environmental risk insurance is also the practice of transferring risk from one entity to another for a fee. Especially if one lives, owns property, or owns a business in a disaster-prone area, it is important to obtain insurance coverage for the various storms, catastrophic events, and natural disasters that can damage real estate. In this context, no one can predict where a natural disaster will occur and the devastation it may cause when it does, or their economic impact, but insurance can help mitigate the damage. Keywords. Levers of state regulation of damage compensation create opportunities for direct and indirect influence on the insurance market and serve to improve relations in the field of environmental insurance, taking into account the real state of the economy and ecology. Environmental risk insurance is also the practice of transferring risk from one entity to another for a fee. Especially if one lives, owns property, or owns a business in a disaster-prone area, it is important to obtain insurance coverage for the various storms, catastrophic events, and natural disasters that can damage real estate. In this context, no one can predict where a natural disaster will occur and the devastation it may cause when it does, or their economic impact, but insurance can help mitigate the damage. International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| minimization of the impact of human activities on nature. Quality natural resources and ecological systems lead to the creation of productive uses for many occupations, including the food industry, tourism, education and other sectors. It is important to solve global problems such as natural problems and diseases occurring around the world, including growth, abnormal heat, drought and the prevention of catastrophic consequences on the ozone layer. environmental protection. This sentence includes collecting information based on biological, chemical, hydrological, geological and other fields, assessing the environmental effects of human activity and dealing with its consequences. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| minimization of the impact of human activities on nature. Quality natural resources and ecological systems lead to the creation of productive uses for many occupations, including the food industry, tourism, education and other sectors. It is important to solve global problems such as natural problems and diseases occurring around the world, including growth, abnormal heat, drought and the prevention of catastrophic consequences on the ozone layer. environmental protection. Keywords. "KAFOLAT insurance company" JSC received only 25 applications as a result of the flood in the village of Uyshun, Chirakchi district, Kashkadarya region (there are 1214 households in this village, approximately 130 of them were damaged as a result of the flood), a total of 25.64 million. we can see that the property interests of the citizens who insured their property have been paid and the insurance coverage has been paid [18]. Conclusion. Environmental insurance, i.e. environmental security, nature protection and self-sustainability is of great importance due to the changes in the life cycle, which has undergone its own changes around the world. The main goal of environmental insurance is to protect against the negative effects of human activities that change nature, in particular, it serves the purpose of nature protection, effective use of natural resources and their sustainable preservation, as well as the improvement of human potential and ecological culture. Also, environmental insurance is more than the economic benefits that lead to the 754 Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| minimization of the impact of human activities on nature. Quality natural resources and ecological systems lead to the creation of productive uses for many occupations, including the food industry, tourism, education and other sectors. It is important to solve global problems such as natural problems and diseases occurring around the world, including growth, abnormal heat, drought and the prevention of catastrophic consequences on the ozone layer. environmental protection. This sentence includes collecting information based on biological, chemical, hydrological, geological and other fields, assessing the environmental effects of human activity and dealing with its consequences. In the world's experience, there are mainly three methods of compensation for the damage caused to the economy as a result of anthropogenic activity on the environment, and they are carried out from the state budget, from the personal funds of the person who caused the damage, and from the funds embodied in the insurance business. Publishing centre of Finland REFERENCES: 1. M. Eshov, A. Burkhanov, H. Khasanov, D. Baratova, J. Nasirov. Basics of insurance work. T.-Spirituality, 2023. p. 30. 1. M. Eshov, A. Burkhanov, H. Khasanov, D. Baratova, J. Nasirov. Basics of insurance work. T.-Spirituality, 2023. p. 30. 2. Granaturov V.M. Economic Risk: Essence, Measurement Methods, Ways to Reduce: Manual / V.M. Granaturov. - 2nd ed., updated and revised. - M .: Business and Service, 1999. - 154 p. 3. Samuelson P.A. Economics. - V.2. - M .: NPO "ALGON", 1994. - 416 p. 3. Samuelson P.A. Economics. - V.2. - M .: NPO "ALGON", 1994. - 416 p. 4. https://www.philosophy.ru/fk/fk-14/ekzistentsialnoe-ponimanie- istiny-u-kerkegora/ 4. https://www.philosophy.ru/fk/fk-14/ekzistentsialnoe-ponimanie- istiny-u-kerkegora/ 5. Q. M. Koldoshev. Improving the methodological foundations of mutual insurance in Uzbekistan. Monograph. 2020, p. 8. 5. Q. M. Koldoshev. Improving the methodological foundations of mutual insurance in Uzbekistan. Monograph. 2020, p. 8. 755 Publishing centre of Finland International Journal of Education, Social Science & Humanities. Finland Academic Research Science Publishers ISSN: 2945-4492 (online) \| (SJIF) = 7.502 Impact factor Volume-11\| Issue-7\| 2023 Published: \|22-07-2023\| 6. 1. I. Abdurakhmanov, M. Abduraimova, N. Abdullaeva; - T. "Economy- Finance", 2020. p. 4. 7. Law of the Republic of Uzbekistan "On Insurance Activities". 23.11.2022. 8. Dj. M. Keynes. Obshchaya theory zanyatosti, protsenta i deneg. / Pod ed. Mileyskogo A.G., Osadchey I.M. - M.: Progress, 1978 9. A. Marshall. Principles of economic science. V 3tt.. – M. : Progress, 1993. 9. A. Marshall. Principles of economic science. V 3tt.. – M. : Progress, 1993. 10. Q. Koldoshev Insurance theory. Textbook. - Ministry of Higher and Secondary Special Education of the Republic of Uzbekistan - T.: "NEW EDITION", 2022 p. 31. 10. Q. Koldoshev Insurance theory. Textbook. - Ministry of Higher and Secondary Special Education of the Republic of Uzbekistan - T.: "NEW EDITION", 2022 p. 31. 11. M. Mahmudova. Risk management and their improvement in the activity of commercial banks. Dissertation written for the degree of Doctor of Philosophy (PhD) in Economics. Tashkent-2022. p. 22 11. M. Mahmudova. Risk management and their improvement in the activity of commercial banks. Dissertation written for the degree of Doctor of Philosophy (PhD) in Economics. Tashkent-2022. p. 22 12. https://lex.uz/docs/1416862 12. https://lex.uz/docs/1416862 13. https://parliament.gov.uz/upload/iblock/074/fpjeggfutkglyz- jzajpbbbodcelhrf%20dwvnbcthwnscyq.pdf. 14. https://www.statista.com/statistics/275288/expenses-of-munich-re- reinsurance-due-to-natural-disaster-damages-since-2006/ 14. https://www.statista.com/statistics/275288/expenses-of-munich-re- reinsurance-due-to-natural-disaster-damages-since-2006/ reinsurance-due-to-natural-disaster-damages-since-2006/ 15. https://www.reinsurancene.ws/munich-re-pegs-global-insured-nat- cat-losses-at-120bn-in- 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 16. https://www.reinsurancene.ws/munich-re-pegs-global-insured-nat- cat-losses-at-120bn-in- 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 17. https://www.ncei.noaa.gov/access/monitoring/monthly- report/global/202113 18. https://www.uzreport.news/insurance/kafolat-sel-natijasida-zarar- kurganlarga-su-urta-oplamalarini-tulab-berdi 15. https://www.reinsurancene.ws/munich-re-pegs-global-insured-nat- cat-losses-at-120bn-in- 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 16. https://www.reinsurancene.ws/munich-re-pegs-global-insured-nat- cat-losses-at-120bn-in- 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 17 h // / / / hl 16. https://www.reinsurancene.ws/munich-re-pegs-global-insured-nat- cat-losses-at-120bn-in- 2022/? lid Cj KCAj t52 BhB5Ei A05YK BT 98 EZ7 WT cat-losses-at-120bn-in- 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 2022/?gclid=CjwKCAjwt52mBhB5EiwA05YKo_BTz98sEZ7gWT_s- bPJUvhTIBrzp6pS9RN2Lc4D174sVNb6zPjLCxoCFF8QAvD_BwE 17. https://www.ncei.noaa.gov/access/monitoring/monthly- report/global/202113 18. https://www.uzreport.news/insurance/kafolat-sel-natijasida-zarar- kurganlarga-su-urta-oplamalarini-tulab-berdi 756
https://openalex.org/W3123077533	https://europepmc.org/articles/pmc7812923?pdf=render	English	null	Effectiveness of modifications to preadjusted appliance prescriptions based on racial dental characteristics assessed by the ABO Cast-Radiograph Evaluation: A propensity score matching study	PeerJ	2,021	cc-by	7,008	ABSTRACT Background. Because racial discrepancies in dental characteristics are known to exist, designing preadjusted appliances according to racial normal occlusion data would be expected to improve treatment results. However, whether modifications based on racial characteristics can improve treatment outcomes in the clinic remains to be investigated. Methods. To study the influence of prescription type on treatment outcomes, 91 patients treated with Chinese or Roth prescription appliances were selected as an initial sample. Two groups of patients were selected by propensity score matching (1:1) to limit the effects of confounding factors, including age, sex, case complexity, and extraction plan. Discrepancy Index and cervical vertebral maturation values were used to quantify case complexity and patient age, respectively. After matching, the final sample of 60 patients consisted of two groups of 30 patients each: group 1 had been treated with a Chinese prescription appliance and group 2 had been treated with a Roth prescription appliance. ABO casts and radiograph evaluation (CR-Eval) and lateral cephalograms were utilized to compare the treatment outcomes of the two groups. Results. The total ABO scores of groups 1 and 2 were 22.03 and 23.87, respectively. There were no significant differences between the two groups in total ABO score or in seven other sub-scores; however, there was a significant difference between the two groups in mandibular canine alignment score. Submitted 21 May 2020 Accepted 28 November 2020 Published 15 January 2021 Corresponding author Yanqin Lu, 213031@csu.edu.cn Academic editor Anne Marie Kuijpers-Jagtman Additional Information and Declarations can be found on page 11 DOI 10.7717/peerj.10605 Conclusions. There are no significant differences in overall treatment outcomes between the Chinese and Roth prescription appliances. The Chinese prescription yielded better alignment results in the mandibular canine for Chinese patients. Subjects Anthropology, Anatomy and Physiology, Dentistry Keywords Preadjusted appliance prescription, Self-ligating appliance, Racial difference, Normal occlusion Keywords Preadjusted appliance prescription, Self-ligating appliance, Racial difference, Normal occlusion Effectiveness of modifications to preadjusted appliance prescriptions based on racial dental characteristics assessed by the ABO Cast-Radiograph Evaluation: A propensity score matching study Yanhao Chu, Lingling Zhang, Yatao Zhao, Fang Yi and Yanqin Lu Department of Orthodontics, Xiangya School of Stomatology, Hunan Key Laboratory of Oral Health Research Xiangya Stomatological Hospital, Central South University, Changsha, Hunan, China Yanhao Chu, Lingling Zhang, Yatao Zhao, Fang Yi and Yanqin Lu Department of Orthodontics, Xiangya School of Stomatology, Hunan Key Laboratory of Oral Health Research, Xiangya Stomatological Hospital, Central South University, Changsha, Hunan, China Yanhao Chu, Lingling Zhang, Yatao Zhao, Fang Yi and Yanqin Lu Department of Orthodontics, Xiangya School of Stomatology, Hunan Key Laboratory of Oral Health Research, Xiangya Stomatological Hospital, Central South University, Changsha, Hunan, China Subjects Anthropology, Anatomy and Physiology, Dentistry Keywords Preadjusted appliance prescription, Self-ligating appliance, Racial difference, Normal occlusion How to cite this article Chu Y, Zhang L, Zhao Y, Yi F, Lu Y. 2021. Effectiveness of modifications to preadjusted appliance prescriptions based on racial dental characteristics assessed by the ABO Cast-Radiograph Evaluation: A propensity score matching study. PeerJ 9:e10605 http://doi.org/10.7717/peerj.10605 Submitted 21 May 2020 Accepted 28 November 2020 Published 15 January 2021 Corresponding author Yanqin Lu, 213031@csu.edu.cn Academic editor Anne Marie Kuijpers-Jagtman Additional Information and Declarations can be found on page 11 DOI 10.7717/peerj.10605 INTRODUCTION Copyright 2021 Chu et al. Andrews (1979) measured crown facial prominence, torque, and tip values of 120 Americans with ideal occlusion. These measurements were then used to design specialized brackets for each tooth in preadjusted appliances. Distributed under Creative Commons CC-BY 4.0 How to cite this article Chu Y, Zhang L, Zhao Y, Yi F, Lu Y. 2021. Effectiveness of modifications to preadjusted appliance prescriptions based on racial dental characteristics assessed by the ABO Cast-Radiograph Evaluation: A propensity score matching study. PeerJ 9:e10605 http://doi.org/10.7717/peerj.10605 According to previous research, craniomaxillofacial hard and soft tissue morphology differs by race and ethnicity (Hideki et al., 2007; Yan et al., 2011). Race has also been shown to be related to dental and occlusal characteristics. Several researchers (Currim & Wadkar, 2004; Lombardo et al., 2015; Watanabe & Koga, 2001) have reported differences in facial prominence, in-out, angulation, and inclination among Japanese, Indian, African, Caucasian, American, and Chinese populations. In theory, prescriptions, including in-out (different base thickness), tip, and torque, should be designed based on racial characteristics. For example, bracket base thickness is determined by crown prominence to obtain the required first order correction. However, differences in prominence, which equate to differences in base thickness, between racial groups likely affects first order alignment. Tip and torque differences between races are also related to anchorage control and aesthetic effect. Several orthodontists have suggested that modifying prescriptions according to differences in race will lead to improved occlusal and aesthetic treatment outcomes (Currim & Wadkar, 2004; Lombardo et al., 2015). However, few follow-up studies on preadjusted appliances based on racial dental characteristics have been reported. Currently, popular prescriptions, such as Roth, MBT, and Andrews, are based on data from patients in Western countries who have normal occlusion. In an early study from our group, a 3D coordinate measuring machine was used to measure dental casts of Chinese patients with normal occlusion (Yang & Zeng, 1998). Compared with American patients who were measured by Andrews, these Chinese patients showed obvious differences in crown facial prominence, torque, angulation, and especially in the in-out between the lateral incisors and the canines, as well as between the second premolars and the first molars, and in the torque, tip values of the anterior teeth (Yang & Zeng, 1998). Subsequently, a conventional ligating appliance was designed (Zeng & Gao, 2008) based on these data (Yang & Zeng, 1998). INTRODUCTION This appliance was later developed into a self-ligating appliance with the same prescription and design concept. Popular prescription appliances are widely used in China and around the world with good treatment outcomes. However, it remains unknown whether modifications to prescriptions based on racial dental characteristics can improve treatment outcomes for occlusal details. The purpose of this study is to compare the effectiveness of a Chinese prescription with that of a popular Western prescription (Roth) in a sample of Chinese patients. The null hypothesis is that there is no significant difference between the treatment outcomes for Chinese patients treated with either Chinese or Roth prescription appliances. METHOD This retrospective study was approved by the Institutional Review Board at the Xiangya Stomatological Hospital Central South University (No.20190018). The production and clinical use of the appliance based on Chinese characteristics was approved by the China Food and Drug Administration. All participants gave written informed consent for inclusion in this study. All patients who were considered for this study received and Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 2/14 completed orthodontic treatment from the same physician at Xiangya Stomatological Hospital between 2015 and 2018. Sample size calculation The American Board of Orthodontics (ABO) Cast-Radiograph Evaluation (CR-Eval) was used to evaluate treatment outcomes. In accordance with several studies (Detterline et al., 2010; Mislik et al., 2016), a difference of five points in the mean ABO CR-Eval score was considered clinically significant in this study. The standard deviation of the ABO CR-Eval score was estimated to be 8.8 according to a previous study with a large sample size (Cansunar & Uysal, 2014). The alpha value and power were set at 0.05 and 80%, respectively. PASS software 11.0 was used to calculate the required sample size, with the results showing that 27 patients were needed for each group in this study. So we included 30 patients for each group. Inclusion and exclusion criteria The inclusion criteria were as follows: (1) the patient had received comprehensive orthodontic treatment with either a Chinese (Z2, 3B Ortho) or a Roth (In-Ovation, Dentsply) prescription self-ligating appliance, both of which are 0.022 inch systems; (2) the patient underwent bonding between 12 and 30 years of age; (3) complete clinical records were available for the patient. The exclusion criteria were: (1) the patient showed craniofacial anomalies or syndromes or underwent orthognathic surgery; (2) the patient had congenitally missing teeth, a tooth deformity, or a fused tooth; (3) the patient was debonded early and showed poor compliance. Propensity score matching After inclusion and exclusion criteria were considered, two patients were excluded because their low compliance and early debonding may affect the results. The initial sample consisted of 91 patients, including 36 patients treated with a Chinese prescription, and 55 patients treated with a Roth prescription. Detailed data regarding the Chinese and Roth prescriptions are shown in Table 1. Propensity score matching (PSM) was used to select the final sample from the initial sample. To prevent several covariates, including sex, age, malocclusion severity, and tooth extraction, from affecting the results, the ABO Discrepancy Index (DI) was used to quantify the severity of malocclusion for every patient. The DI considers overjet, overbite, anterior open bite, lateral open bite, crowding, occlusal relationship, lingual posterior crossbite, buccal posterior crossbite, cephalometric items, and other items (Deguchi et al., 2005). The initial sample was separated into three severity levels according to pre-treatment DI: low (DI < 7), medium (DI 8–16) and high complexity (DI ≥17) (Cansunar & Uysal, 2014). In this study, the cervical vertebral maturation (CVM) method, which separates patients into 6 stages based on a study Baccetti, Franchi & McNamara Jr (2005), was used to quantify the age of the patients according to pre-treatment cephalometric radiographs. Sex of the patients as well as non-extraction or extraction of premolars in the treatment plan were recorded as binary variables. DI, sex, extraction plan, and CVM stage were analysed and recorded for each patient in the two groups. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 3/14 Table 1 Data of the Chinese prescription and the Roth prescription self-ligating appliance. Chinese prescription Roth prescription Torque Angulation Torque Angulation U1 11 4 12 5 U2 7 6 8 9 U3 −3 7 −2 13 U4 −7 2 −7 0 U5 −7 4 −7 0 L1/L2 0 0 −1 2 L3 −3 0 −11 7 L4 −15 3 −17 −1 L5 −23 4 −22 −1 Notes. The unit is degree (◦). Table 1 Data of the Chinese prescription and the Roth prescription self-ligating appliance. A multivariable logistic regression model was calculated using SPSS software for Windows (version 25.0; IBM, Armonk, NY). For the propensity score analysis, prescription type was used as the dependent variable, while CVM stage (Cervical Stage, CS1, CS2, CS3, CS4, CS5, CS6), sex, DI severity level (low, medium, high), and extraction plan (Yes or No) were modelled as covariates. Treatment procedure Direct bonding was used in this study. All brackets were bonded to the facial axis point of the tooth by the same physician. The wire sequence for all patients in this study was: 0.014-inch super elastic nickel-titanium, 0.016-inch super elastic nickel, 0.016-inch stainless steel, 0.018 × 0.025-inch super elastic nickel-titanium, and 0.019 × 0.025-inch stainless steel. Sliding mechanics was used to close the space for cases requiring premolar extraction. Intermaxillary elastic traction was avoided as far as possible. If a bracket got loose and lost, it would be replaced by a new one. All patients were asked to maintain good oral hygiene. All cases completed treatment with straight wires, and no extra wire bending was made throughout the whole clinical process. Impressions were taken to obtain finishing dental casts on the day of debonding. Propensity score matching A one-to-one nearest neighbour matching algorithm was applied using SPSS software, and a calliper with a width of 0.02 times the standard deviation of the logit of the propensity score was applied as the matching criteria (Xiao et al., 2018). After PSM, a final sample of 60 patients was obtained: 30 patients in the Chinese prescription (group 1) and 30 patients in the Roth prescription group (group 2). The baseline data of the two groups were compared using χ 2-tests to ensure that the covariates were balanced. Treatment evaluation In this study, the ABO CR-Eval was selected as the evaluation index to quantify treatment outcomes. The ABO CR-Eval consists of eight measurement items: alignment, marginal ridges, buccolingual inclination, occlusal relationships, occlusal contacts, overjet, 4/14 Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 interproximal contacts, and root angulation. Post-treatment plaster casts and panoramic X-rays were analysed using an ABO calibration kit. In order to compare the torque control of the two prescriptions, lateral cephalograms were used to measure U1/SN for upper incisors and L1/MP for lower incisors. To minimize the error between investigators, the same types of measurements were performed by one investigator. Before ABO measurements were taken, the investigator was trained using ABO instructions and tutorial videos from the ABO website. All the casts and radiographs were assigned a random coded number by an assistant to blind the investigator to the experimental conditions. Error of the method The ICC values for the CVM stage evaluations, the DI scores, and the ABO scores were 0.92, 0.97, and 0.98, respectively. The ICC values in this study were above 0.8, suggesting good consistency and reliability. The Bland-Altman plot (Fig. 1) show that the mean difference between the score of examiner (S1) and the score of another examiner (S0) was −0.15, with limits of agreement (mean ± 1.96 SD) of −2.55 to 2.25. Reliability The validity of the ABO CR-Eval for assessing treatment outcomes of Chinese patients has been established in a previous study (Song et al., 2013). To train for ABO CR-Eval, there are 3 sets of calibration dental casts with collective, agreed-upon scores. To reduce systematic bias, investigators should guarantee that measured ABO scores are consistent with the established measurements of these three casts. To assess the systematic errors, 20 patients were selected at random and another examiner who had also received training for ABO measured the scores again in comparison with the investigator in this study. Bland-Altman plots were used to assess systematic errors by the Graphpad Prism 7 (GraphPad Software Inc, San Diego, CA, USA). To assess intra-examiner reliability and reproducibility in the current study, measurements by the investigator for 15 randomly selected cases were taken one month after the first measurements, and the intraclass correlation coefficient (ICC) was calculated. Statistical analysis Independent samples t-tests were performed to compare the total score of the ABO CR-Eval as well as the sub-scores between group 1 and group 2. Because of differences in the in-out design between the two prescriptions, alignment scores were compared for each tooth type. For each individual patient, the scores for the same tooth type on the left and right sides of the jaw were combined. Independent samples t-tests were also performed for U1/SN and L1/MP. P-values of less than 0.05 were considered to be statistically significant. Final sample In the final sample, the anterior Bolton index were 79.4% ± 1.4% for Chinese group and 79.1% ± 1.3% for Roth group. The total Bolton index were 91.6% ± 0.9% for Chinese group and 91.5% ±0.9% for Roth group. The treatment duration was 22.4 ± 3.6 months for the Chinese group and 23.5 ± 3.0 months for the Roth group. Final sample PSM and baseline data of the final sample are provided in Table 2. Most cases were classified as either medium or high severity according to their DI score. The χ2 tests showed that 5/14 5/14 Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 Figure 1 Bland-Altman plot of ABO score in this study. Bland-Altman plot of ABO score by the investi- gator (S1) and ABO score by another examiner (S0). Full-size DOI: 10.7717/peerj.10605/fig-1 the distributions of CVM stage, DI score, sex, and extraction plan were not significantly different between the two groups (p >0.05). In the final sample, the anterior Bolton index were 79.4% ± 1.4% for Chinese group and 79.1% ± 1.3% for Roth group. The total Bolton index were 91.6% ± 0.9% for Chinese group and 91.5% ±0.9% for Roth group. The treatment duration was 22.4 ± 3.6 months for the Chinese group and 23.5 ± 3.0 months Figure 1 Bland-Altman plot of ABO score in this study. Bland-Altman plot of ABO score by the investi- gator (S1) and ABO score by another examiner (S0). Full-size DOI: 10.7717/peerj.10605/fig-1 the distributions of CVM stage, DI score, sex, and extraction plan were not significantly different between the two groups (p >0.05). In the final sample, the anterior Bolton index were 79.4% ± 1.4% for Chinese group and 79.1% ± 1.3% for Roth group. The total Bolton index were 91.6% ± 0.9% for Chinese group and 91.5% ±0.9% for Roth group. The treatment duration was 22.4 ± 3.6 months for the Chinese group and 23.5 ± 3.0 months Figure 1 Bland-Altman plot of ABO score in this study. Bland-Altman plot of ABO score by the investi- gator (S1) and ABO score by another examiner (S0). Full-size DOI: 10.7717/peerj.10605/fig-1 the distributions of CVM stage, DI score, sex, and extraction plan were not significantly different between the two groups (p >0.05). In the final sample, the anterior Bolton index were 79.4% ± 1.4% for Chinese group and 79.1% ± 1.3% for Roth group. The total Bolton index were 91.6% ± 0.9% for Chinese group and 91.5% ±0.9% for Roth group. The treatment duration was 22.4 ± 3.6 months for the Chinese group and 23.5 ± 3.0 months for the Roth group. the distributions of CVM stage, DI score, sex, and extraction plan were not significantly different between the two groups (p >0.05). ABO CR-Eval scores and incisors inclination The ABO CR-Eval total scores and the eight sub-scores are provided in Table 3. The results show that there were no significant differences in total ABO score or in the sub- scores, including marginal ridges, buccolingual inclination, occlusal relationships, occlusal contacts, overjet, interproximal contacts, and root angulation between the two groups. However, a statistically significant difference in alignment score was observed between group 1 and group 2. The alignment score for each tooth is shown in Table 4. The only significant difference found between the groups was the alignment score for the mandibular canine. There was a lower mean score for the mandibular canine in group 1 (the Chinese prescription group) compared to that of group 2 (the Roth prescription group). According to the ABO CR-Eval analysis of group 2, most of the alignment score for the mandibular canine was due to a mandibular canine lingual deviation from the ideal arch line. As shown in Fig. 2, the mandibular canines were located at the ideal position in the dental arch for patients with the Chinese prescription (Fig. 2D), while the mandibular canines were not arranged in a smooth curving line and deviated lingually from the ideal position for the Roth prescription (Fig. 2B). The incisor inclination including U1/SN and L1/MP of the two groups is shown in Table 5. The statistical analysis showed no significant differences in U1/SN and L1/MP between the two groups. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 6/14 Table 2 Baseline data of the two groups of patients after PSM. The table shows the number of patients in each category. Baseline data after PSM P value Chinese prescription Roth prescription Sex male 14 15 0.796 female 16 15 0.796 Age CVM I-III 1 1 1 CVM IV 9 7 0.559 CVM V 14 16 0.606 CVM VI 6 6 1 Complexity DI low 2 2 1 DI medium 11 12 0.791 DI high 17 16 0.795 Extraction plan Extraction 5 6 0.739 Non-extraction 25 24 0.739 Notes. χ2-test. p < 0.05, statistical significance was set at p < 0.05. Table 2 Baseline data of the two groups of patients after PSM. The table shows the number of patients in each category. Table 3 Independent samples t-test for the eight ABO CR-EVAL items and the total ABO score. Table 3 Independent samples t-test for the eight ABO CR-EVAL items and the total ABO score. ABO CR-Eval scores and incisors inclination Table 3 Independent samples t-test for the eight ABO CR-EVAL items and the total ABO score. Type Mean Stand deviation Mean difference 95% Confidence interval P value Chinese Roth Chinese Roth Lower Upper Total ABO score 22.03 23.87 4.82 5.26 −1.83 −4.44 0.78 0.165 Alignment 4.70 6.13 1.60 2.10 −1.43 −2.40 −0.47 0.004 Marginal Ridges 3.70 3.93 1.86 1.39 −0.23 −1.08 0.61 0.584 Buccolingual inclination 3.40 3.27 1.69 1.96 0.13 −0.81 1.08 0.779 Overjet 3.87 4.10 2.37 1.95 −0.23 −1.36 0.89 0.679 Occlusal contacts 2.77 3.33 1.96 2.02 −0.57 −1.60 0.46 0.275 Occlusal relationships 2.87 2.57 2.73 2.56 0.30 −1.07 1.67 0.662 Interproximal Contacts 0.00 0.03 0.00 0.18 −0.03 −0.10 0.30 0.321 Root angulation 0.73 0.50 0.69 0.62 0.23 −0.11 0.58 0.177 Notes. Independent samples t-test. p < 0.05, statistical significance was set at p < 0.05. DISCUSSION Propensity score analysis has been widely used in many recent clinical studies to compare the therapeutic effects and clinical prognoses of different treatment plans (Wu et al., 2019; Zeng et al., 2019). In orthodontic clinical studies, many confounding factors can affect clinical results, and these have often been overlooked. In this study, PSM was used to adjust Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 7/14 DISCUSSION Propensity score analysis has been widely used in many recent clinical studies to compare the therapeutic effects and clinical prognoses of different treatment plans (Wu et al., 2019; Zeng et al., 2019). In orthodontic clinical studies, many confounding factors can affect clinical results, and these have often been overlooked. In this study, PSM was used to adjust Propensity score analysis has been widely used in many recent clinical studies to compare the therapeutic effects and clinical prognoses of different treatment plans (Wu et al., 2019; Zeng et al., 2019). In orthodontic clinical studies, many confounding factors can affect clinical results, and these have often been overlooked. In this study, PSM was used to adjust 7/14 7/14 Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 Table 4 Independent samples t-test for the alignment score of tooth types by two appliances. Type Mean Stand Deviation Mean Difference 95% Confidence Interval P value Chinese Roth Chinese Roth Lower Upper U1 0.23 0.10 0.50 0.31 0.13 −0.08 0.35 0.220 U2 0.23 0.30 0.50 0.53 −0.07 −0.34 0.20 0.621 U3 0.33 0.60 0.55 0.77 −0.27 −0.61 0.08 0.127 U4 0.53 0.67 0.73 0.88 −0.13 −0.55 0.29 0.527 U5 0.47 0.20 0.73 0.43 0.23 −0.07 0.54 0.137 U6 0.40 0.33 0.50 0.55 0.07 −0.20 0.34 0.623 U7 0.40 0.60 0.56 0.77 −0.20 −0.55 0.15 0.256 L1 0.03 0.10 0.18 0.31 −0.07 −0.20 0.06 0.309 L2 0.20 0.27 0.41 0.58 −0.07 −0.33 0.19 0.610 L3 0.20 1.03 0.41 1.03 −0.83 −1.24 −0.43 0.000 L4 0.20 0.13 0.48 0.35 0.07 −0.15 0.28 0.542 L5 0.17 0.20 0.46 0.48 −0.03 −0.28 0.21 0.786 L6 0.80 0.63 0.96 0.99 0.17 −0.34 0.67 0.513 L7 0.63 0.80 0.81 0.89 −0.17 −0.61 0.27 0.450 Notes. U1-U7: maxillary central incisor to second molar, L1-L7: mandibular central incisor to second molar. p < 0.05, statistical significance was set at p < 0.05. Notes. U1-U7: maxillary central incisor to second molar, L1-L7: mandibular central incisor to second molar. p < 0.05, statistical significance was set at p < 0.05. the baseline data and balance confounding factors, allowing for a more direct comparison of treatment outcomes according to prescription type. Many occlusal indices are used to evaluate therapeutic effects, such as the peer assessment rating (PAR); the index of complexity, outcome and need (ICON); and the ABO Cast- Radiograph Evaluation (CR-Eval). DISCUSSION According to Cansunar & Uysal (2014), the PAR and the ICON are not sufficient to accurately and precisely measure tooth position. The ABO CR-Eval is generally used in clinics to improve treatment outcomes and as a measurement tool for scientific research. In the current study, the total ABO score was 22.03 ± 4.82 for the Chinese prescription appliance and 23.87 ± 5.26 for the Roth prescription appliance. In a previous study by Brown et al. (2015), ABO CR-Eval scores were 28.5 ± 8.5, 32.3 ± 7.8, and 32.3 ± 9.3 for three different self-ligating appliances. Detterline’s (Detterline et al., 2010) experimental results showed that scores following treatment with a 0.018-inch bracket and a 0.022-inch bracket were 26.3 ± .0.0 and 28.5 ± 21.3, respectively. Thus, compared with the scores reported in other studies, the Chinese patients in the current study treated with either the Chinese or the Roth prescription achieved satisfactory total scores. As the Roth prescription appliance is one of the most effective and popular appliances in clinical use, the similar scores between the Chinese prescription group and the Roth prescription group suggest that both prescriptions are effective for overall treatment outcomes as well as the outcomes of the above-mentioned seven items. In our study, the top four items with the highest scores, listed in descending order, were alignment, marginal ridges, buccolingual inclination, and overjet for the Chinese prescription group and alignment, overjet, marginal ridges, and occlusal contacts for the Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 8/14 8/14 Figure 2 The upper and lower occlusal pictures of patients treated with Roth prescription and Chinese prescription appliance. (A) The upper occlusal picture of Roth prescription. (B) The lower occlusal pic- ture of Roth prescription. (C) The upper occlusal picture of Chinese prescription. (D) The lower occlusal picture of Chinese prescription. Full-size DOI: 10.7717/peerj.10605/fig-2 Table 5 Independent samples t-test for the upper and lower incisors inclination of two groups of patients. Type Mean Stand Deviation Mean Difference 95% Confidence Interval P value Chinese Roth Chinese Roth Lower Upper U1/SN 105.1 106.0 7.4 6.9 −0.8 −4.5 2.9 0.68 L1/MP 96.0 95.0 6.4 6.0 1.0 −2.2 4.3 0.52 Notes. Independent samples t-test. p < 0.05, statistical significance was set at p < 0.05. Roth prescription group. DISCUSSION In previous studies, the items with the highest scores were reported to be alignment (Cook, Harris & Vaden, 2005; Yassir et al., 2018), occlusal contacts (Aszkler et al., 2014), and occlusal relationships (Okunami et al., 2007). The item with the lowest Figure 2 The upper and lower occlusal pictures of patients treated with Roth prescription and Chinese prescription appliance. (A) The upper occlusal picture of Roth prescription. (B) The lower occlusal pic- ture of Roth prescription. (C) The upper occlusal picture of Chinese prescription. (D) The lower occlusal picture of Chinese prescription. Full-size DOI: 10.7717/peerj.10605/fig-2 t for the upper and lower incisors inclination of two groups of patients. Stand Deviation Mean Difference 95% Confidence Interval P value h Chinese Roth Lower Upper 0 7.4 6.9 −0.8 −4.5 2.9 0.68 6.4 6.0 1.0 −2.2 4.3 0.52 Figure 2 The upper and lower occlusal pictures of patients treated with Roth prescription and Chin prescription appliance. (A) The upper occlusal picture of Roth prescription. (B) The lower occlusal pic ture of Roth prescription. (C) The upper occlusal picture of Chinese prescription. (D) The lower occlus picture of Chinese prescription. Full-size DOI: 10.7717/peerj.10605/fi Figure 2 The upper and lower occlusal pictures of patients treated with Roth prescription and Chinese prescription appliance. (A) The upper occlusal picture of Roth prescription. (B) The lower occlusal pic- ture of Roth prescription. (C) The upper occlusal picture of Chinese prescription. (D) The lower occlusal picture of Chinese prescription. Full-size DOI: 10.7717/peerj.10605/fig-2 Full-size DOI: 10.7717/peerj.10605/fig-2 Table 5 Independent samples t-test for the upper and lower incisors inclination of two groups of patients. Type Mean Stand Deviation Mean Difference 95% Confidence Interval P value Chinese Roth Chinese Roth Lower Upper U1/SN 105.1 106.0 7.4 6.9 −0.8 −4.5 2.9 0.68 L1/MP 96.0 95.0 6.4 6.0 1.0 −2.2 4.3 0.52 Notes. Independent samples t-test. p < 0.05, statistical significance was set at p < 0.05. Roth prescription group. In previous studies, the items with the highest scores were reported to be alignment (Cook, Harris & Vaden, 2005; Yassir et al., 2018), occlusal contacts (Aszkler et al., 2014), and occlusal relationships (Okunami et al., 2007). The item with the lowest score in this study was interproximal contacts, and this result was consistent with that of Yassir et al. (2018), and Aszkler et al. (2014). This finding suggests that closing the space is not very challenging for most orthodontic cases. DISCUSSION In contrast, orthodontists should pay close attention to other items, such as alignment, marginal ridges, and occlusal relationships, from the beginning to the end of treatment. The alignment scores of the mandibular canines in the Chinese prescription group were significantly lower than those in the Roth prescription group. In a study by Yang & Zeng (1998), the discrepancies in prominence between the maxillary lateral incisor and the Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 9/14 Table 5 Independent samples t-test for the upper and lower incisors inclination of two groups of patients. Type Mean Stand Deviation Mean Difference 95% Confidence Interval P value Chinese Roth Chinese Roth Lower Upper U1/SN 105.1 106.0 7.4 6.9 −0.8 −4.5 2.9 0.68 L1/MP 96.0 95.0 6.4 6.0 1.0 −2.2 4.3 0.52 Notes. Independent samples t-test. *p < 0.05, statistical significance was set at p < 0.05. ble 5 Independent samples t-test for the upper and lower incisors inclination of two groups of patients. Roth prescription group. In previous studies, the items with the highest scores were reported to be alignment (Cook, Harris & Vaden, 2005; Yassir et al., 2018), occlusal contacts (Aszkler et al., 2014), and occlusal relationships (Okunami et al., 2007). The item with the lowest score in this study was interproximal contacts, and this result was consistent with that of Yassir et al. (2018), and Aszkler et al. (2014). This finding suggests that closing the space is not very challenging for most orthodontic cases. In contrast, orthodontists should pay close attention to other items, such as alignment, marginal ridges, and occlusal relationships, from the beginning to the end of treatment. The alignment scores of the mandibular canines in the Chinese prescription group were significantly lower than those in the Roth prescription group. In a study by Yang & Zeng (1998), the discrepancies in prominence between the maxillary lateral incisor and the Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 9/14 maxillary canine, the mandibular lateral incisor and the mandibular canine, the maxillary second premolar and the maxillary first molar, and the mandibular second premolar and the mandibular first molar of Chinese individuals with normal occlusion were considerably different from those of American patients who were measured by Andrews. Due to these differences in crown prominence, there exist differences in the bracket thicknesses between the two prescriptions. DISCUSSION Because the prominence discrepancies between the mandibular canine and the lateral incisor of the Chinese patient group were greater than those of the American patients measured by Andrews, the Roth prescription base thickness may lead to the mandibular canines deviating from the ideal buccal-lingual position in Chinese patients treated with the Roth appliance. In addition, the ideal torque of the mandibular canine for Chinese patients is −3◦(Yang & Zeng, 1998), but the Roth’s torque prescription value is −11◦, meaning that the Roth prescription appliance produces excessive lingual-crown torque for Chinese patients. These differences can explain why the Chinese prescription group achieved better alignment of the mandibular canine than the Roth group did in our study. In contrast, the in-out and torque value differences were not as large for the other teeth as they were for the mandibular canines. This is the most likely explanation for the absence of significant differences in the alignments of the other teeth. Only misalignments of greater than 0.5 mm were scored by the ABO CR-Eval, and it is difficult to accurately measure discrepancies below 0.5 mm using the ABO gauge and the naked eye. Theoretically, real full-size arch wires can give more expression of the bracket prescriptions, however, it is not realistic clinically because there still exists clearance between finishing arch wire and the bracket slot. In this study, both groups were finished with the same arch wire, and thus the interaction between finishing arch wire and bracket slot was comparable. In this study, satisfactory treatment outcomes were achieved by the Chinese prescription appliance for Chinese patients, and the Chinese prescription had the advantage of avoiding lingual malposition of the mandibular canine and thus showed improved alignment compared to the Western prescription. It might be possible that customized brackets will solve the problem of dental anatomical differences between patients in the future. However, at the moment customized brackets are more expensive and the production of an individualized bracketr for each patient takes time. Preadjusted appliances based on mean values of normal occlusion are absolutely the most popular brackets now. The purpose of this study was not to create a ‘‘fit for all’’ appliance. Sometimes bending wires cannot be avoided, but prescriptions based on Chinese normal occlusion may reduce wire bending for the mandibular canines of a significant number of Chinese patients. Our results have significance for other ethnic groups as well. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 Limitations There are several limitations to this study. This was a retrospective study which means there are limitations with respect to patient selection and non-randomization. A prospective study will be performed in the future to both explore and revise the Chinese prescription appliance. Direct bonding of brackets was performed in this study. The indirect bonding technique may be more rigorous. Yet the direct bonding technique is the most common and popular technique in clinics around the world. Moreover, a recent paper (Li et al., 2019) showed that there were no significant differences in the mean errors by the two bonding techniques. All the participants were treated by one clinician in this study, which increases the internal validity and may affect the external validity. However, different clinicians have different treatment preferences, even different levels of clinical skills. If patients from different clinicians had been included in this retrospective study, it would be hard to compare the treatment outcomes of two prescriptions and clinician could have been a confounder. The angulation and torque values in the completed cases could not be directly measured by the ABO CR-Eval. To compare the degrees of angulation and magnitudes of torque, it may be necessary to use customized measurement software to measure these values with 3D digital models in a follow-up study. CONCLUSIONS There were no significant differences in the total ABO CR-Eval score or the scores of seven other subitems between Chinese patients treated with a Chinese prescription and those treated with a Western prescription appliance. However, there was a significant difference in the alignment score of the mandibular canine between the two groups, with the Chinese self-ligating prescription group showing a better score than the Western prescription appliance group. DISCUSSION The dental characteristics of a sample of patients with normal occlusion should be measured for each ethnic group. If there exist significant differences between individuals of different races, it is better to modify the bracket prescriptions based on racial dental characteristics. The resulting overall treatment outcomes may not demonstrate obvious improvements, but the position of certain teeth can be controlled more precisely, closer to normal occlusion. Furthermore, if patients must be treated with appliances which were developed using data from individuals of other races, orthodontists should pay special attention to the positions of certain teeth with regards to dentition and torque magnitudes in the finishing stage. Sometimes Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 10/14 additional wire bending of the first, second, and third order are necessary to realize proper alignment, occlusion, and aesthetic effects for certain teeth. additional wire bending of the first, second, and third order are necessary to realize proper alignment, occlusion, and aesthetic effects for certain teeth. ACKNOWLEDGEMENTS We thank Professor Xianglong Zeng for his help in this study. Funding The authors received no funding for this work. Ethics The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers): The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers): The Xiangya Stomatological Hospital,Central South University granted Ethical approval to carry out the study (No. 20190018). The Xiangya Stomatological Hospital,Central South University granted Ethical approval to carry out the study (No. 20190018). Data Availability The following information was supplied regarding data availability: The following information was supplied regarding data availability: The raw measurements are available as a Supplemental File. Supplemental Information Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.10605#supplemental-information. Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.10605#supplemental-information. peerj.10605#supplemental-information. Author Contributions Author Contributions • Yanhao Chu conceived and designed the experiments, performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. g p p p authored or reviewed drafts of the paper, and approved the final draft. • Lingling Zhang performed the experiments, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft. • Yatao Zhao analyzed the data, authored or reviewed drafts of the paper, and approved the final draft. • Fang Yi performed the experiments, prepared figures and/or tables, and approved the final draft. • Yanqin Lu conceived and designed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Competing Interests The authors declare there are no competing interests. The authors declare there are no competing interests. 11/14 Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 Author Contributions REFERENCES Andrews LF. 1979. The straight-wire appliance. British Journal of Orthodontics 6:125–143 DOI 10.1179/bjo.6.3.125. Andrews LF. 1979. The straight-wire appliance. British Journal of Orthodontics 6:125–143 DOI 10.1179/bjo.6.3.125. Aszkler RM, Preston CB, Saltaji H, Tabbaa S. 2014. Long-term occlusal changes assessed by the American Board of Orthodontics’ model grading system. American Journal of Orthodontics and Dentofacial Orthopedics 145:173–178 DOI 10 1016/j ajodo 2013 10 010 Baccetti T, Franchi L, McNamara Jr JA. 2005. The cervical vertebral maturation (CVM) method for the assessment of optimal treatment timing in dentofacial orthopedics. Seminars in Orthodontics 11(3):119–129. Brown MW, Koroluk L, Ko CC, Zhang K, Chen M, Nguyen T. 2015. Effectiveness and efficiency of a CAD/CAM orthodontic bracket system. American Journal of Orthodontics and Dentofacial Orthopedics 148:1067–1074 DOI 10 1016/j j d 2015 07 029 Brown MW, Koroluk L, Ko CC, Zhang K, Chen M, Nguyen T. 2015. Effectiveness and efficiency of a CAD/CAM orthodontic bracket system. American Journal of Orthodontics and Dentofacial Orthopedics 148:1067–1074 d Cansunar HA, Uysal T. 2014. Relationship between pretreatment case complexity and orthodontic clinical outcomes determined by the American Board of Orthodontics criteria. Angle Orthodontist 84:974–979 DOI 10.2319/010114-001.1. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 Cook DR, Harris EF, Vaden JL. 2005. Comparison of university and private-practice orthodontic treatment outcomes with the American Board of Orthodontics objective grading system. American Journal of Orthodontics and Dentofacial Orthopedics 127:707–712 DOI 10.1016/j.ajodo.2004.12.014. Currim S, Wadkar PV. 2004. Objective assessment of occlusal and coronal characteristics of untreated normals: a measurement study. American Journal of Orthodontics and Dentofacial Orthopedics 125:582–588 DOI 10.1016/j.ajodo.2003.06.008. Deguchi T, Honjo T, Fukunaga T, Miyawaki S, Roberts WE, Takano-Yamamoto T. 2005. Clinical assessment of orthodontic outcomes with the peer assessment rating, discrepancy index, objective grading system, and comprehensive clinical assess- ment. American Journal of Orthodontics and Dentofacial Orthopedics 127:434–443 DOI 10.1016/j.ajodo.2004.03.030. Detterline DA, Isikbay SC, Brizendine EJ, Kula KS. 2010. Clinical outcomes of 0.018- inch and 0.022-inch bracket slot using the ABO objective grading system. Angle Orthodontist 80:528–532 DOI 10.2319/060309-315.1. Hideki I, Shunsuke N, Akihiko N, Counts AL. 2007. Comparison of cephalometric norms between Japanese and Caucasian adults in antero-posterior and vertical dimension. European Journal of Orthodontics 29:493–499 DOI 10.1093/ejo/cjm059. Li Y, Mei L, Wei J, Yan X, Zhang X, Zheng W, Li Y. 2019. Effectiveness, efficiency and adverse effects of using direct or indirect bonding technique in orthodon- tic patients: a systematic review and meta-analysis. BMC Oral Health 19:137 DOI 10.1186/s12903-019-0831-4. REFERENCES Lombardo L, Perri A, Arreghini A, Latini M, Siciliani G. 2015. Three-dimensional assessment of teeth first-, second- and third-order position in Caucasian and African subjects with ideal occlusion. Progress in Orthodontics 16:11 DOI 10.1186/s40510-015-0086-9. Mislik B, Konstantonis D, Katsadouris A, Eliades T. 2016. University clinic and private practice treatment outcomes in Class I extraction and nonextraction patients: a comparative study with the American Board of Orthodontics Objective Grading System. American Journal of Orthodontics and Dentofacial Orthopedics 149:253–258 DOI 10.1016/j.ajodo.2015.10.012. Okunami TR, Kusnoto B, BeGole E, Evans CA, Sadowsky C, Fadavi S. 2007. Assessing the American Board of Orthodontics objective grading system: digital vs plaster den- tal casts. American Journal of Orthodontics and Dentofacial Orthopedics 131:51–56 DOI 10.1016/j.ajodo.2005.04.042. Song GY, Baumrind S, Zhao ZH, Ding Y, Bai YX, Wang L, He H, Shen G, Li WR, Wu WZ, Ren C, Weng XR, Geng Z, Xu TM. 2013. Validation of the American Board of Orthodontics Objective Grading System for assessing the treatment outcomes of Chinese patients. American Journal of Orthodontics and Dentofacial Orthopedics 144:391–397 DOI 10.1016/j.ajodo.2013.04.018. Watanabe K, Koga M. 2001. A morphometric study with setup models for bracket design. Angle Orthodontist 71:499–511. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 13/14 Wu J, Kim C, Bai HX, Tang L, Wang Y, Li X, Zhang Z, Karakousis G, Huang R, Zhang PJ, Xiao B, Yang L. 2019. Comparison of radiation therapy alone and chemother- apy alone for low-grade gliomas without surgical resection. World Neurosurgery 122:e108–e120 DOI 10.1016/j.wneu.2018.09.109 . Xiao H, Liu W, Quan H, Ouyang Y. 2018. Peri-operative blood transfusion does not influence overall and disease-free survival after radical gastrectomy for stage II/III gastric cancer: a propensity score matching analysis. Journal of Gastrointestinal Surgery 22:1489–1500 DOI 10.1007/s11605-018-3808-8. Yan G, Mcnamara JA, Sigler LM, Tiziano B. 2011. Comparison of craniofacial char- acteristics of typical Chinese and Caucasian young adults. European Journal of Orthodontics 33:205–211 DOI 10.1093/ejo/cjq054. Yang X, Zeng X. 1998. A study on the teeth shape and position of normal occlusion in Chinese. Journal of Peking University (Health Sciences) 30:528–531. Yassir YA, El-Angbawi AM, McIntyre GT, Revie GF, Bearn DR. 2018. A randomized clinical trial of the effectiveness of 0.018-inch and 0.022-inch slot orthodontic bracket systems: part 2-quality of treatment. European Journal of Orthodontics DOI 10.1093/ejo/cjy038. Zeng C, Dubreuil M, LaRochelle MR, Lu N, Wei J, Choi HK, Lei G, Zhang Y. 2019. Association of tramadol with all-cause mortality among patients with osteoarthritis. Chu et al. (2021), PeerJ, DOI 10.7717/peerj.10605 REFERENCES Jama 321:969–982 DOI 10.1001/jama.2019.1347. Zeng X, Gao X. 2008. Z2 appliance: the pre-adjusted appliance based on Chinese normal occlusion. West China Journal of Stomatology 26:573–579. 14/14 14/14
https://openalex.org/W2774202754	https://www.pure.ed.ac.uk/ws/files/50906689/Semple_R_The_metabolic_syndrome....pdf	English	null	The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion	Scientific reports	2,017	cc-by	10,356	Publisher Rights Statement: ub s e g ts State e t Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Digital Object Identifier (DOI): 10.1038/s41598-017-17746-8 Link: Link to publication record in Edinburgh Research Explorer Document Version: Publisher's PDF, also known as Version of record Published In: Scientific Reports General rights C i h f h General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Published In: S i ifi R Publisher Rights Statement: Edinburgh Research Explorer The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion Citation for published version: Rocha, N, Payne, F, Huang-Doran, I, Sleigh, A, Fawcett, K, Adams, C, Stears, A, Saudek, V, O'Rahilly, S, Barroso, I & Semple, RK 2017, 'The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion', Scientific Reports, vol. 7, no. 1, pp. 17593. https://doi.org/10.1038/s41598-017-17746-8 Citation for published version: Rocha, N, Payne, F, Huang-Doran, I, Sleigh, A, Fawcett, K, Adams, C, Stears, A, Saudek, V, O'Rahilly, S, Barroso, I & Semple, RK 2017, 'The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion', Scientific Reports, vol. 7, no. 1, pp. 17593. https://doi.org/10.1038/s41598-017-17746-8 The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion Received: 22 June 2017 Accepted: 30 November 2017 Published: xx xx xxxx Received: 22 June 2017 Accepted: 30 November 2017 Published: xx xx xxxx Nuno Rocha1,2, Felicity Payne5, Isabel Huang-Doran1,2, Alison Sleigh 3,4, Katherine Fawcett5, Claire Adams1,2, Anna Stears6, Vladimir Saudek1,2, Stephen O’Rahilly 1,2, Inês Barroso 1,5 & Robert K. Semple 1,2,7 Common genetic variants at the ARL15 locus are associated with plasma adiponectin, insulin and HDL cholesterol concentrations, obesity, and coronary atherosclerosis. The ARL15 gene encodes a small GTP-binding protein whose function is currently unknown. In this study adipocyte-autonomous roles for ARL15 were investigated using conditional knockdown of Arl15 in murine 3T3-L1 (pre)adipocytes. Arl15 knockdown in differentiated adipocytes impaired adiponectin secretion but not adipsin secretion or insulin action, while in preadipocytes it impaired adipogenesis. In differentiated adipocytes GFP- tagged ARL15 localized predominantly to the Golgi with lower levels detected at the plasma membrane and intracellular vesicles, suggesting involvement in intracellular trafficking. Sequencing of ARL15 in 375 severely insulin resistant patients identified four rare heterozygous variants, including an early nonsense mutation in a proband with femorogluteal lipodystrophy and non classical congenital adrenal hyperplasia, and an essential splice site mutation in a proband with partial lipodystrophy and a history of childhood yolk sac tumour. No nonsense or essential splice site mutations were found in 2,479 controls, while five such variants were found in the ExAC database. These findings provide evidence that ARL15 plays a role in adipocyte differentiation and adiponectin secretion, and raise the possibility that human ARL15 haploinsufficiency predisposes to lipodystrophy. Several metabolic traits, including insulin resistance and hyperinsulinemia, low HDL cholesterol, and hypoadi- ponectinemia, are strongly associated with each other, and with major diseases including diabetes mellitus and atherosclerosis. However, unpicking the direction of causality in such associations is a major challenge in which genetic studies play an important role. Genome-wide association studies (GWAS) agnostically seek statistical associations among common genetic variation and traits of interest. If the pertinent change in gene(s) expression or function that drives such genetic association can be identified, which usually requires fine mapping of the initial GWAS signal, then insights into disease mechanism may in principle be garnered. Take down policy Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 24. Oct. 2024 www.nature.com/scientificreports www.nature.com/scientificreports Received: 22 June 2017 Accepted: 30 November 2017 Published: xx xx xxxx Methods All human studies were approved by the National Health Service Research Ethics Committee of the United Kingdom and were conducted in accordance with the principles of the Declaration of Helsinki. Participants gave informed consent both for study participation and for use of clinical images where relevant. Antibodies and Immunoassays. Antibodies are listed in Supplementary Table 1. The mouse adipsin ELISA kit was from Millipore. The murine adiponectin autoDELFIA immunoassay was previously described11. Rabbit polyclonal antibody to adiponectin used in immunofluorescence experiments was a gift from Dr. P.E. Scherer. Cell culture. Reagents were from SIGMA unless otherwise indicated. Cells were maintained in DMEM with 10% neonatal calf serum (3T3-L1s) or fetal bovine serum (FBS) (HEK293T). To induce adipocyte differentiation, 2-day postconfluent 3T3-L1 preadipocytes were transferred to DMEM-Tet-Approved FBS (Clontech), supple- mented with 500 µmol/L IBMX, 1 µmol/L dexamethasone, 10 µmol/L insulin, and 100 nmol/L rosiglitazone for 3 days. Cells were further cultured in DMEM supplemented with 10% Tet-Approved FBS, 10 µmol/L insulin (Actrapid, Novo Nordisk) for 3 days and thereafter in DMEM with Tet-Approved FBS for 8 days unless otherwise indicated. Differentiation was assessed by Oil-Red-O staining12. Plasmid construction. Cloning of murine Arl15 into the mammalian pEGFP-N3 and the lentiviral pSLIK vectors, and shArl15 pSLIK is described in detail in Supplementary Information. The CFP-tagged fusions of the Golgin-245/Furin Golgi markers have been described previously13. Generation of Stable Cell Lines Conditionally Expressing shArl15. Lentiviral particles were gener- ated as previously described14. Low-passage 3T3-L1 preadipocytes were transduced at MOI ≤1 in the presence of polybrene and selected in G418. For RNAi experiments clones were selected using cloning cylinders. Clones with optimal RNAi inducibility, determined by GFP fluorescence and RT-qPCR, were pooled prior to differentiation. For subcellular localization studies, polyclonal 3T3-L1 lines with doxycycline-inducible Arl15-GFP expression were generated by G418 selection prior to adipocyte differentiation. All pSLIK-transduced cell lines were main- tained in G418-supplemented DMEM-NCS, with antibiotics withdrawn for adipocyte differentiation and doxy- cycline induction. The pSLIK lentiviral system has been described previously15. Gene Expression Analysis. Total RNA extracted with RNeasy Mini Kit (Qiagen) was used for MMLV reverse transcription (Promega). RT-qPCR used either custom designed or ABI TaqMan Gene Expression Assays run with ABI TaqMan Mastermix and analyzed on ABI 7900 HT system. Primer/probes were designed using Primer-Express (ABI). Catalog numbers and primer/probe sequences are listed in Supplementary Table 2. The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion However most genetic signals identified by GWAS do not lie in protein-coding regions of genes, and uncertainty about the protein or transcript-level perturbation directly associated with the trait of interest presents a bottleneck to full realisation of the power of GWAS as a tool to shed light on disease mechanism. 1The University of Cambridge Metabolic Research Laboratories, Wellcome Trust-MRC Institute of Metabolic Science, Cambridge, UK. 2The National Institute for Health Research Cambridge Biomedical Research Centre, Cambridge, UK. 3Wolfson Brain Imaging Centre, University of Cambridge School of Clinical Medicine, Cambridge Biomedical Campus, Cambridge, UK. 4National Institute for Health Research/Wellcome Trust Clinical Research Facility, Cambridge University Hospitals NHS Foundation Trust, Cambridge Biomedical Campus, Cambridge, UK. 5Wellcome Trust Sanger Institute, Wellcome Genome Campus, Hinxton, UK. 6Wolfson Diabetes and Endocrine Clinic, Institute of Metabolic Science, Addenbrooke’s Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK. 7Centre for Cardiovascular Sciences, University of Edinburgh, Queen’s Medical Research Institute, Edinburgh, UK. Correspondence and requests for materials should be addressed to R.K.S. (email: rks16@cam.ac.uk) SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 1 www.nature.com/scientificreports/ This study set out to investigate ARL15, a gene that encodes a small GTP binding protein of unknown function, and that has been implicated in metabolic traits and diseases by several different GWAS and other population-wide genetic studies. Genetic variation at a locus including only ARL15 was first associated with HDL cholesterol concentration and also plasma concentration of adiponectin1, a highly abundant and stable plasma protein almost exclusively secreted by adipocytes, and widely believed to act as an insulin-sensitising adipokine2. The same common genetic variants at the ARL15 locus are associated with indices of insulin resistance and cor- onary heart disease1, with other studies confirming variation at the ARL15 locus to be associated with fasting insulin, HDL cholesterol and triglyceride concentrations3–5. g y Based on similarities between the pattern of metabolic traits associated with ARL15 locus variants, and the pattern of metabolic derangement seen in Mendelian lipodystrophies, it has been hypothesized that the genetic variation at the ARL15 locus may affect metabolic traits such as fasting insulin and HDL concentrations through a primary action on adipose tissue development or function6,7. The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion Indeed, SNPs at the ARL15 locus have been asso- ciated with body shape8 and waist circumference9, and increased gene dosage has been associated with childhood obesity10.i y We thus set out firstly to assess whether ARL15 plays a cell autonomous role in the biosynthesis or secretion of adiponectin and/or preadipocyte differentiation using conditional Arl15 knockdown in the insulin-responsive, adiponectin-producing 3T3-L1 cell line, and secondly, to seek rare coding sequence variants in ARL15 in a cohort of volunteers with lipodystrophic or non lipodystrophic severe insulin resistance. Results A l15 k Arl15 knockdown impairs adiponectin expression and secretion. The murine 3T3-L1 preadipocyte cell line can be differentiated with high efficiency into lipid-laden adipocytes that exhibit many characteristics of primary adipocytes, including adiponectin secretion and insulin-stimulated glucose uptake. To investigate cell-autonomous consequences of Arl15 deficiency in mature adipocytes without potential confounding by any effect of the knockdown on differentiation, 3T3-L1 adipocytes conditionally expressing microRNA-like shRNA targeting murine Arl15 (shArl15) under the control of a doxycycline (DOX)-responsive promoter were gener- ated (Fig. 1A–C). DOX-induced RNAi acutely reduced Arl15 transcript levels by >75% in differentiated adi- pocytes relative to non-induced shArl15 adipocytes, with no discernible effect on adipocyte differentiation, as assessed by Oil-Red-O staining of lipid droplets (Fig. 1C). Acute knockdown of Arl15 moderately but significantly reduced AdipoQ mRNA levels, but had no effect on the mRNA of peroxisome proliferator-activated receptor gamma-2 (Pparg2) or Glut4 (Fig. 1D). Arl15 knockdown reduced total adiponectin secretion into cell culture medium, as assessed by either immunoblotting or immunoassay (Fig. 1E), and immunoblotting under non-re- ducing, non-heat-denaturing conditions revealed that all oligomeric forms of secreted adiponectin were affected (Fig. 1E). Despite the reduction in AdipoQ mRNA and secreted adiponectin, intracellular adiponectin levels were not affected by Arl15 knockdown, however (Fig. 1F).f f y g Given the effect of Arl15 downregulation on adiponectin secretion and the key role that some ARF and ARL (ARF-like) small G proteins have in intracellular transport pathways20–23, we assessed whether Arl15 suppression caused generalised inhibition of intracellular protein trafficking and secretion. However we found that secretion of adipsin, an adipokine whose trafficking route to the cell surface of 3T3-L1 adipocytes partly overlaps with that of adiponectin24, was not affected by suppression of Arl15, as assessed by either immunoblotting or ELISA (Fig. 1G). Arl15 knockdown has no impact on adipocyte-autonomous insulin action. We next tested whether Arl15 is involved in insulin signal transduction and/or insulin-responsive glucose uptake. Arl15 knockdown had no effect on peak insulin-stimulated phosphorylation of either AKT or ERK (extracellular signal-regulated kinase) (Fig. 2A), or on basal or insulin-stimulated glucose uptake (Fig. 2B), suggesting that Arl15 is not involved in insulin-stimulated mobilisation and fusion of Glut4-containing vesicles with the plasma membrane. Arl15 knockdown impairs 3T3-L1 pre-adipocyte differentiation. Most monogenic insulin resistance is caused not by direct perturbation of cell autonomous insulin signalling, but by defects in adipocyte differen- tiation or function. www.nature.com/scientificreports/ www.nature.com/scientificreports/ serum samples were diluted 1:50 and processed as described above for adipocyte culture medium. Adipocyte medium samples were diluted 1:50 for adiponectin DELFIA analysis and 1:10 for adipsin ELISA. Human serum samples were diluted 1:50 for adiponectin DELFIA. serum samples were diluted 1:50 and processed as described above for adipocyte culture medium. Adipocyte medium samples were diluted 1:50 for adiponectin DELFIA analysis and 1:10 for adipsin ELISA. Human serum samples were diluted 1:50 for adiponectin DELFIA. Deoxyglucose uptake assay. Insulin-stimulated deoxyglucose uptake was assayed as previously described16. Briefly, DOX-treated and serum-starved adipocytes were insulin-stimulated and incubated with [3H] deoxyglucose, washed and lysed. Radioactivity was normalized to protein concentration. Immunofluorescence microscopy. For detailed methods on (pre)-adipocytes and NMuMG immunoflu- orescence studies see Supplementary Information. Genetic studies. ARL15 (NM_019087.2) was screened in a total of 375 individuals ascertained by severe insulin resistance and 2,479 controls, using several different sequencing approaches (see Supplementary Information for details of cohorts studied and sequencing techniques). All rare ARL15 alleles reported were verified in non-amplified genomic DNA using Sanger sequencing. For assessment of ARL15 exon 4 splicing from whole-blood RNA see Supplementary Information. Human imaging and physiological studies. Body composition was assessed using Lunar Prodigy dual-energy X-ray absorptiometry (DEXA, GE Lunar). Hepatic triglyceride was assessed either using proton magnetic resonance spectroscopy on a Siemens 3T-Verio scanner17 (Patient 1) or using MRI proton density fat fractionation with IDEAL-IQ18 (Patient 2). Adipose distribution was determined by magnetic resonance imaging, using T1-weighted turbo spin echo, water-suppressed, transaxial images. g g p pp g Insulin, leptin and adiponectin were assayed using two-step time-resolved AutoDELFIA immunoassays11. Normative adiponectin, leptin and oral glucose tolerance testing data were derived from the MRC Ely Study cohort19. All other biochemical analyses were undertaken in an accredited clinical diagnostic laboratory in Addenbrooke’s Hospital, Cambridge, UK. Methods p g p p q pp y For immunoblotting, adipocytes were washed twice in ice-cold PBS and lysed on ice in 50 mmol/L HEPES, pH 7.5, 150 mmol/L NaCl, 30 mmol/L NaF, 10 mmol/L Na4P2O7, 1 mmol/L Na2VO4, 1% Triton X-100, 10 mmol/L EDTA and Protease Complete Cocktail (Roche). Extracts were cleared by centrifugation and protein con- centrations determined using Bio-Rad DC Protein Assay. Proteins were mixed with sample buffer (3% SDS, 50 mM Tris-HCl pH 6.8, 5% 2-mercaptoethanol, 12% glycerol) and heat-denatured at 95 °C. For complete reduction of adiponectin samples this buffer was supplemented with 10 mM dithiothreitol. For non-reducing non-heat-denaturing conditions, samples were incubated with reducing agent-free sample buffer for 1 h at room temperature. Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes using iBlot (Invitrogen). Membranes were blocked in 5% (w/v) skimmed milk or 3% (w/v) BSA (for phospho-protein immu- nodetection) TBS-T. Analysis of adiponectin from lysates and medium of 3T3-L1 adipocytes and human serum. 3T3-L1 adipocytes were washed twice with PBS and transferred to serum-free medium containing 3% BSA for 24 h. Medium (10 μL) was harvested, centrifuged and analyzed by immunoblotting under non-reducing and non-heat-denaturing or reducing/denaturing conditions. For analysis of adiponectin in lysates, protein sam- ples were obtained as described for immunoblotting of cell lysates. For analysis of human serum adiponectin, SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 2 Results A l15 k Left, Arl15 mRNA expression in shArl15 adipocytes, or adipocytes conditionally expressing shRNA against firefly luciferase (shLuc), treated with DOX or PBS. The same conditions were used in. Right, representative image of Oil-Red-O stained lipid droplets in shArl15 3T3-L1 adipocytes treated with DOX (RNAi) or PBS (control), as indicated. (D) Adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), and Glut4 mRNA expression in shArl15 adipocytes treated with PBS (black bars) or DOX (hatched bars) determined by real-time quantitative PCR and normalised to murine CypA (PBS, n = 3; DOX, n = 3). Data are expressed as mean ± SEM; differences between means analyzed by Student’s t test; *P > 0.001. (E) Total adiponectin (bottom left) or its oligomeric profile (top left) in shArl15 adipocyte culture medium assessed by SDS-PAGE under reducing, heat-denaturing or non-reducing, non-heat-denaturing SDS-PAGE respectively prior to immunoblotting. HMW (high-molecular weight, ≥9mer), hexamer, LMW (low molecular weight, trimer), and monomer are indicated. Right panel, quantification adiponectin concentrations in media from DOX- or PBS-treated shArl15 cultures measured by DELFIA immunoassay. (F) Left panel, total adiponectin (lower immunoblot) and its oligomeric profile (top immunoblot) in lysates of shArl15 adipocytes treated with DOX or PBS (control) revealed by Western blotting. Right panel, quantification of intracellular adiponectin determined by immunoassay (DELFIA). (G) Left panel, immunoblotting of adipsin in shArl15 adipocyte medium. Right panel, relative amounts of adipsin in culture medium determined using a mouse adipsin enzyme-linked immunoassay (ELISA). Means ± SEM are shown (n = 3). Figure 1. Arl15 knockdown impairs adiponectin, but not adipsin, secretion from 3T3-L1 adipocytes. (A) schematic of the pSLIK lentiviral vector used for doxycycline (DOX)-induced expression of a short hairpin RNA targeting Arl15 (shArl15). (B) Protocols for adipocyte differentiation and doxycycline (DOX)-induction of RNAi in 3T3-L1 adipocytes. (C) DOX-induced suppression of Arl15 by RNAi in adipocytes. Left, Arl15 mRNA expression in shArl15 adipocytes, or adipocytes conditionally expressing shRNA against firefly luciferase (shLuc), treated with DOX or PBS. The same conditions were used in. Right, representative image of Oil-Red-O stained lipid droplets in shArl15 3T3-L1 adipocytes treated with DOX (RNAi) or PBS (control), as indicated. (D) Adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), and Glut4 mRNA expression in shArl15 adipocytes treated with PBS (black bars) or DOX (hatched bars) determined by real-time quantitative PCR and normalised to murine CypA (PBS, n = 3; DOX, n = 3). Results A l15 k Data are expressed as mean ± SEM; differences between means analyzed by Student’s t test; P > 0.001. (E) Total adiponectin (bottom left) or its oligomeric profile (top left) in shArl15 adipocyte culture medium assessed by SDS-PAGE under reducing, heat-denaturing or non-reducing, non-heat-denaturing SDS-PAGE respectively prior to immunoblotting. HMW (high-molecular weight, ≥9mer), hexamer, LMW (low molecular weight, trimer), and monomer are indicated. Right panel, quantification adiponectin concentrations in media from DOX- or PBS-treated shArl15 cultures measured by DELFIA immunoassay. (F) Left panel, total adiponectin (lower immunoblot) and its oligomeric profile (top immunoblot) in lysates of shArl15 adipocytes treated with DOX or PBS (control) revealed by Western blotting. Right panel, quantification of intracellular adiponectin determined by immunoassay (DELFIA). (G) Left panel, immunoblotting of adipsin in shArl15 adipocyte medium. Right panel, relative amounts of adipsin in culture medium determined using a mouse adipsin enzyme-linked immunoassay (ELISA). Means ± SEM are shown (n = 3). Figure 1. Arl15 knockdown impairs adiponectin, but not adipsin, secretion from 3T3-L1 adipocytes. (A) schematic of the pSLIK lentiviral vector used for doxycycline (DOX)-induced expression of a short hairpin RNA targeting Arl15 (shArl15). (B) Protocols for adipocyte differentiation and doxycycline (DOX)-induction of RNAi in 3T3-L1 adipocytes. (C) DOX-induced suppression of Arl15 by RNAi in adipocytes. Left, Arl15 mRNA expression in shArl15 adipocytes, or adipocytes conditionally expressing shRNA against firefly luciferase (shLuc), treated with DOX or PBS. The same conditions were used in. Right, representative image of Oil-Red-O stained lipid droplets in shArl15 3T3-L1 adipocytes treated with DOX (RNAi) or PBS (control), as indicated. (D) Adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), and Glut4 mRNA expression in shArl15 adipocytes treated with PBS (black bars) or DOX (hatched bars) determined by real-time quantitative PCR and normalised to murine CypA (PBS, n = 3; DOX, n = 3). Data are expressed as mean ± SEM; differences between means analyzed by Student’s t test; P > 0.001. (E) Total adiponectin (bottom left) or its oligomeric profile (top left) in shArl15 adipocyte culture medium assessed by SDS-PAGE under reducing, heat-denaturing or non-reducing, non-heat-denaturing SDS-PAGE respectively prior to immunoblotting. HMW (high-molecular weight, ≥9mer), hexamer, LMW (low molecular weight, trimer), and monomer are indicated. Right panel, quantification adiponectin concentrations in media from DOX- or PBS-treated shArl15 cultures measured by DELFIA immunoassay. Results A l15 k Given the pattern of associations of the common ARL15 locus variant, which are sugges- tive of adipose dysfunction, a role of ARl15 in adipocyte differentiation was next assessed. Expression of Arl15 mRNA was significantly increased by day 6 of 3T3-L1 adipocyte differentiation and remained unchanged until day 10 (Fig. 3A). Knockdown of Arl15 prior to adipocyte differentiation was then undertaken (Fig. 3B). Arl15 mRNA expression at day 0 in knockdown cells was reduced to < 25% that of shArl15 3T3-L1s treated with saline (Fig. 3C), and the vast majority of DOX-treated cells failed to acquire the characteristic morphological features of differentiated adipocytes, including accumulation of cytoplasmic lipid droplets. Adipocyte differentiation was highly sensitive to Arl15 knockdown, with impaired lipid droplet formation seen at DOX concentrations as low as 10 ng/mL (Fig. 3D). Reduced transcript levels of adipocyte markers further corroborated impaired differentiation of Arl15 knockdown cells (Fig. 3E). SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 3 www.nature.com/scientificreports/ entificreports/ Figure 1. Arl15 knockdown impairs adiponectin, but not adipsin, secretion from 3T3-L1 adipocytes. (A) schematic of the pSLIK lentiviral vector used for doxycycline (DOX)-induced expression of a short hairpin RNA targeting Arl15 (shArl15). (B) Protocols for adipocyte differentiation and doxycycline (DOX)-induction of RNAi in 3T3-L1 adipocytes. (C) DOX-induced suppression of Arl15 by RNAi in adipocytes. Left, Arl15 mRNA expression in shArl15 adipocytes, or adipocytes conditionally expressing shRNA against firefly luciferase (shLuc), treated with DOX or PBS. The same conditions were used in. Right, representative image o Oil-Red-O stained lipid droplets in shArl15 3T3-L1 adipocytes treated with DOX (RNAi) or PBS (control), as indicated. (D) Adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), and Gl mRNA expression in shArl15 adipocytes treated with PBS (black bars) or DOX (hatched bars) determined by real-time quantitative PCR and normalised to murine CypA (PBS, n = 3; DOX, n = 3). Data are expressed as mean ± SEM; differences between means analyzed by Student’s t test; P > 0.001. (E) Total adiponectin (bottom left) or its oligomeric profile (top left) in shArl15 adipocyte culture medium assessed by SDS-PAGE Figure 1. Arl15 knockdown impairs adiponectin, but not adipsin, secretion from 3T3-L1 adipocytes. (A) schematic of the pSLIK lentiviral vector used for doxycycline (DOX)-induced expression of a short hairpin RNA targeting Arl15 (shArl15). (B) Protocols for adipocyte differentiation and doxycycline (DOX)-induction of RNAi in 3T3-L1 adipocytes. (C) DOX-induced suppression of Arl15 by RNAi in adipocytes. Results A l15 k (F) Left panel, total adiponectin (lower immunoblot) and its oligomeric profile (top immunoblot) in lysates of shArl15 adipocytes treated with DOX or PBS (control) revealed by Western blotting. Right panel, quantification of intracellular adiponectin determined by immunoassay (DELFIA). (G) Left panel, immunoblotting of adipsin in shArl15 adipocyte medium. Right panel, relative amounts of adipsin in culture medium determined using a mouse adipsin enzyme-linked immunoassay (ELISA). Means ± SEM are shown (n = 3). Arl15 predominantly localizes to the Golgi complex. To investigate the subcellular localization of Arl15, GFP-tagged Arl15 was expressed in 3T3-L1 cells and visualised by fluorescence confocal microscopy. Arl15-GFP was found in a predominantly perinuclear distribution characteristic of the Golgi apparatus, with SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 4 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 2. Arl15 knockdown has no impact on adipocyte-autonomous insulin action. (A) Insulin-stimulated phosphorylation of Akt on Threonine 308 (pAkt (T308) and Serine 473 (pAKT (S473)) or extracellular signal- regulated kinase 1/2 (pERK1/2) after 5-min treatment of serum-starved shArl15 adipocytes with 100 nmol/L insulin, assessed by Western blotting with GAPDH (GAPDH) and Akt shown as a loading controls (Akt). Arl15 knockdown was induced by DOX treatment. (B) Relative [3H]deoxyglucose uptake by Arl15 knockdown (+DOX) or control (+PBS) serum-starved shArl15 adipocytes treated with 100 nmol/L insulin or saline for 30 min. Means ± SEM are shown (n = 6); A.U., arbitrary units. Significance by ANOVA followed by Bonferroni Multiple Comparison tests was set at P < 0.05. ns = P > 0.05; P < 0.001. Figure 2. Arl15 knockdown has no impact on adipocyte-autonomous insulin action. (A) Insulin-stimulated phosphorylation of Akt on Threonine 308 (pAkt (T308) and Serine 473 (pAKT (S473)) or extracellular signal- regulated kinase 1/2 (pERK1/2) after 5-min treatment of serum-starved shArl15 adipocytes with 100 nmol/L insulin, assessed by Western blotting with GAPDH (GAPDH) and Akt shown as a loading controls (Akt). Arl15 knockdown was induced by DOX treatment. (B) Relative [3H]deoxyglucose uptake by Arl15 knockdown (+DOX) or control (+PBS) serum-starved shArl15 adipocytes treated with 100 nmol/L insulin or saline for 30 min. Means ± SEM are shown (n = 6); A.U., arbitrary units. Significance by ANOVA followed by Bonferroni Multiple Comparison tests was set at P < 0.05. ns = P > 0.05; P < 0.001. some signal also in intracellular vesicles. 3T3-L1 preadipocytes were then co-transfected with Arl15-GFP and two CFP-tagged Golgi markers (Golgin-245 or Furin) (Fig. 4A). Results A l15 k (C) Arl15 mRNA expression in 3T3-L1 preadipocytes harvested at day 0 of the RNAi and differentiation protocol described above. (D) Oil-Red-O staining of triglycerides in shArl15 adipocytes treated with either increasing concentrations of doxycycline (+DOX) or saline solution (+PBS), as indicated. Far right, representative bright field micrographs of shArl15 adipocytes treated with PBS (top) or 1 mg/mL DOX (bottom). (E) Arl15, adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), Glut4, and adipocyte protein 2 (aP2) mRNA expression in shArl15 adipocytes treated with DOX (hatched) or PBS (black), determined by real-time quantitative PCR. The same experiment was repeated using shLuc adipocytes, as control (chart on the right). mRNA levels were normalized to those of mouse CypA (cyclophilin A) (PBS, n = 3; DOX, n = 3). All data are shown as means ± SEM. A.U., arbitrary units. Differences between means were analysed by Student’s t test; P < 0.05, P < 0.01, P < 0.001. Figure 3. Arl15 knockdown impairs 3T3-L1 adipocyte differentiation. (A) Arl15 mRNA expression during differentiation of wild-type 3T3-L1 adipocytes, determined by real-time quantitative PCR. P < 0.01 vs. day 0 (n = 4). (B) DOX-induced RNAi and differentiation protocols used to assess the effects of Arl15 knockdown on 3T3-L1 adipocyte differentiation. (C) Arl15 mRNA expression in 3T3-L1 preadipocytes harvested at day 0 of the RNAi and differentiation protocol described above. (D) Oil-Red-O staining of triglycerides in shArl15 adipocytes treated with either increasing concentrations of doxycycline (+DOX) or saline solution (+PBS), as indicated. Far right, representative bright field micrographs of shArl15 adipocytes treated with PBS (top) or 1 mg/mL DOX (bottom). (E) Arl15, adiponectin (AdipoQ), peroxisome proliferator-activated receptor gamma-2 (Pparg2), Glut4, and adipocyte protein 2 (aP2) mRNA expression in shArl15 adipocytes treated with DOX (hatched) or PBS (black), determined by real-time quantitative PCR. The same experiment was repeated using shLuc adipocytes, as control (chart on the right). mRNA levels were normalized to those of mouse CypA (cyclophilin A) (PBS, n = 3; DOX, n = 3). All data are shown as means ± SEM. A.U., arbitrary units. Differences between means were analysed by Student’s t test; P < 0.05, P < 0.01, P < 0.001. the gene concerned is the source of the GWAS signal. Results A l15 k Arl15-GFP localized predominantly to a Furin- and Golgin-245-positive perinuclear region, as well as to intracellular vesicles. In some cells fluorescence was also detected at the plasma membrane. Brefeldin A (BFA), a fungal inhibitor that interferes with the GTPase cycle of Golgi ARFs and causes the collapse of the Golgi25–27, rapidly induced redistribution of Arl15-GFP throughout the cytoplasm, supporting association of Ar1l5 with the Golgi (Supplementary Figure S1). Day-3 3T3-L1 adipocytes showed Arl15-GFP in the perinuclear region, vesicular structures and, albeit with varying relative intensities, the plasma membrane (Fig. 4B). This plasma membrane localisation was particularly visible in some cell lines, including NMuMG (mammary gland epithelial cell line) (Fig. 4C). The presence of Arl15 in these intracellular compartments is consistent with a role for this small G protein in trafficking to the plasma membrane. 24 28 fi Treatment of adipocytes with BFA has been shown to block adiponectin secretion24,28 suggesting that this requires an intact Golgi. To investigate if collapse of the Golgi also played a role in inhibition of adiponectin secre- tion in DOX-treated shArl15 cells (Fig. 1E), we used antibodies against the Golgi marker Syntaxin 6. No evidence of compromise of Golgi structure was seen upon Arl15 knockdown (Supplementary Figure S2), consistent with lack of effect of Arl15 knockdown on adipsin secretion (Fig. 1G), which also depends on integrity of the Golgi and post-Golgi trafficking24.fi fi To test the possibility that Arl15 acts directly on compartments utilised in adiponectin trafficking we exam- ined co-localization of Arl15-GFP with endogenous adiponectin in day-5 adipocytes using confocal micros- copy. Despite the cytoplasm-wide presence of adiponectin-positive vesicular compartments, no extensive overlap (non-thresholded average Pearson’s correlation coefficients (r) of 0.39) of the two proteins was observed, however (Fig. 4D). Identification of two lipodystrophic patients with ARL15 haploinsufficiency. Identification of rare and deleterious coding variants in a gene at a locus implicated by GWAS in a trait of interest, and demonstra- tion that these variants are associated with a severe form of the relevant trait or disease, provides evidence that SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 5 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 3. Arl15 knockdown impairs 3T3-L1 adipocyte differentiation. (A) Arl15 mRNA expression during differentiation of wild-type 3T3-L1 adipocytes, determined by real-time quantitative PCR. P < 0.01 vs. day 0 (n = 4). (B) DOX-induced RNAi and differentiation protocols used to assess the effects of Arl15 knockdown on 3T3-L1 adipocyte differentiation. Results A l15 k (B) Day-3 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green), fixed, and stained with neutral lipid stain LipidTOX (red) and DAPI (blue). A representative confocal micrograph of adipocytes expressing Arl15-GFP at the perinuclear region, intracellular vesicles, and the plasma membrane. (C) NMuMG cell transiently expressing Arl15-GFP (green) at the plasma membrane, internal vesicles, and the perinuclear region. (D) Day-5 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green) were fixed and immunostained with anti-adiponectin antibodies and Alexa Fluor 594 (AdipoQ, red). Panel on the right shows the two channels merged. The average Pearson’s correlation coefficient (r) for the GFP and Alexa channels in hand-drawn “regions of interest” (ROI) encompassing entire cells was: 0.39 (without Costes automatic thresholding), 0.02 (below Costes Auto thresholding), and 0.11 (above Costes automatic thresholding). Scale bars, 10 μm. Figure 4. Arl15 predominantly localizes to the Golgi complex. (A) Wild-type 3T3-L1 preadipocytes were co-transfected with GFP-tagged Arl15 and either CFP-Golgin or CFP-Furin, two Golgi markers, fixed, and imaged by confocal microscopy. Arl15-GFP and CFP-Furin (top) or CFP-Golgin (bottom) are show in grayscale and overlayed on the right (merge) in green and red, respectively. Insets on the far right (zoom-in) show in detail regions 1 or 2, as indicated in the corresponding merge panel. (B) Day-3 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green), fixed, and stained with neutral lipid stain LipidTOX (red) and DAPI (blue). A representative confocal micrograph of adipocytes expressing Arl15-GFP at the perinuclear region, intracellular vesicles, and the plasma membrane. (C) NMuMG cell transiently expressing Arl15-GFP (green) at the plasma membrane, internal vesicles, and the perinuclear region. (D) Day-5 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green) were fixed and immunostained with anti-adiponectin antibodies and Alexa Fluor 594 (AdipoQ, red). Panel on the right shows the two channels merged. The average Pearson’s correlation coefficient (r) for the GFP and Alexa channels in hand-drawn “regions of interest” (ROI) encompassing entire cells was: 0.39 (without Costes automatic thresholding), 0.02 (below Costes Auto thresholding), and 0.11 (above Costes automatic thresholding). Scale bars, 10 μm. Figure 4. Arl15 predominantly localizes to the Golgi complex. (A) Wild-type 3T3-L1 preadipocytes were co-transfected with GFP-tagged Arl15 and either CFP-Golgin or CFP-Furin, two Golgi markers, fixed, and imaged by confocal microscopy. Arl15-GFP and CFP-Furin (top) or CFP-Golgin (bottom) are show in grayscale and overlayed on the right (merge) in green and red, respectively. Results A l15 k We thus sequenced coding exons and flanking sequences of ARL15 using Sanger sequencing in 189 probands with insulin resistance that was disproportionately severe to their total degree of adiposity, in a further 186 such probands using either exome sequencing (n = 62) or tar- geted next generation sequencing (n = 124), and 2,479 controls (Supplementary Information). Four variants in ARL15 were thus identified in the severely insulin resistant cohort (Table 1). Two of these variants were missense variants also identified in the ExAC database with low allele frequencies (0.01–0.02%) but not in the control groups in this study. One variant, p.Cys133Arg, was found in a Bengali woman with late onset generalised lipo- dystrophy in addition to insulin resistance, and the other, p.Ala110Thr, was found in an obese Italian woman. The scaled Combined Annotation Dependent Depletion Score (32) was 18 for p.Cys133Arg and 27 for p.Ala- 110Thr. A score of 20 or above indicates that a variant falls in the top 0.1% of all possible variants when ranked by potential deleteriousness. Scaled CADD scores are given for all variants identified in ExAC and controls in Supplementary Information (Table 1, Supplementary Table 3). Neither patient was available for further study. The other variants were an early nonsense mutation, p.Pro26Thrfs9, and an essential splice-acceptor site muta- tion in intron 3 (c.254–2 A > G; g.54113409 A > G), both in patients with partial lipodystrophy (Fig. 5). Neither mutation was identified in the ExAC database nor in 2,479 controls sequenced in this study (Supplementary Table 3). Furthermore no other essential splice site or nonsense mutations were identified in the control group in this study, although five patients in the ExAC database (which includes around 20,000 patients ascertained through cardiometabolic traits such as type 2 diabetes) did have heterozygous nonsense variants (p.Glu160(1 SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 6 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 4. Arl15 predominantly localizes to the Golgi complex. (A) Wild-type 3T3-L1 preadipocytes were co-transfected with GFP-tagged Arl15 and either CFP-Golgin or CFP-Furin, two Golgi markers, fixed, and imaged by confocal microscopy. Arl15-GFP and CFP-Furin (top) or CFP-Golgin (bottom) are show in grayscale and overlayed on the right (merge) in green and red, respectively. Insets on the far right (zoom-in) show in detail regions 1 or 2, as indicated in the corresponding merge panel. Results A l15 k P1 P2 Reference Range Age, years 53 22 ARL15 variant p.Pro26ThrfsTer9 c.254-2 A > G (g.54113409 A > G) B.M.I., kg/m2 30.9 20.2 19–25 % Fat Mass: Arms: 40.2 20.2 35.7–42.9§ Legs: 16.4 25.6 38.4–46.0§ Trunk: 39.6 46.2 44.1–48.8§ Liver triglyceride content, % 24.7 (<5.5) 8 (<5) ¥ HbA1c, mmol/mol 60 35 <42 Glucose, mg/dL 94 67 <100 Insulin, pmol/L 76 395 0–60 Free fatty acids, mg/dL 8.5 9.1 8–25 Adiponectin, mg/L 3.5 (2.6–14.9) 4.5 (4.4–17.7) * Leptin, μg/L 40.1 (14.9–60.2) 38.1 (2.4–24.4) Triglyceride, mg/dL 399 301 <175 HDL-cholesterol, mg/dL 37 35 >42 LDL-cholesterol, mg/dL # 100 <85 ALT, mU/L 36 133 9–52 Testosterone, ng/dL 98 187 <60 Cortisol, μg/dL 18.6 N/A 5–25 Medication Metformin Insulin (228U/day) Prednisolone (5 mg/day) Acipimox Ciprofibrate Spironolactone Flutamide Metformin Thyroxine Fenofibrate Orlistat Table 2. Clinical and metabolic profile of Patients P1 and P2 with loss-of-function ARL15 variants. ¥Method specific references ranges are given in brackets; sex and B.M.I.-adjusted normal range shown in brackets; #could not be calculated. §Controls were 37 adult female volunteers with BMI >30 kg/m2. P1 P2 Reference Range Age, years 53 22 ARL15 variant p.Pro26ThrfsTer9 c.254-2 A > G (g.54113409 A > G) B.M.I., kg/m2 30.9 20.2 19–25 % Fat Mass: Arms: 40.2 20.2 35.7–42.9§ Legs: 16.4 25.6 38.4–46.0§ Trunk: 39.6 46.2 44.1–48.8§ Liver triglyceride content, % 24.7 (<5.5) 8 (<5) ¥ HbA1c, mmol/mol 60 35 <42 Glucose, mg/dL 94 67 <100 Insulin, pmol/L 76 395 0–60 Free fatty acids, mg/dL 8.5 9.1 8–25 Adiponectin, mg/L 3.5 (2.6–14.9) 4.5 (4.4–17.7) Leptin, μg/L 40.1 (14.9–60.2) 38.1 (2.4–24.4) Triglyceride, mg/dL 399 301 <175 HDL-cholesterol, mg/dL 37 35 >42 LDL-cholesterol, mg/dL # 100 <85 ALT, mU/L 36 133 9–52 Testosterone, ng/dL 98 187 <60 Cortisol, μg/dL 18.6 N/A 5–25 Medication Metformin Insulin (228U/day) Prednisolone (5 mg/day) Acipimox Ciprofibrate Spironolactone Flutamide Metformin Thyroxine Fenofibrate Orlistat Table 2. Clinical and metabolic profile of Patients P1 and P2 with loss-of-function ARL15 variants. ¥Method specific references ranges are given in brackets; sex and B.M.I.-adjusted normal range shown in brackets; #could not be calculated. §Controls were 37 adult female volunteers with BMI >30 kg/m2. Table 2. Clinical and metabolic profile of Patients P1 and P2 with loss-of-function ARL15 variants. ¥Method specific references ranges are given in brackets; sex and B.M.I.-adjusted normal range shown in brackets; #could not be calculated. §Controls were 37 adult female volunteers with BMI >30 kg/m2. Results A l15 k Insets on the far right (zoom-in) show in detail regions 1 or 2, as indicated in the corresponding merge panel. (B) Day-3 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green), fixed, and stained with neutral lipid stain LipidTOX (red) and DAPI (blue). A representative confocal micrograph of adipocytes expressing Arl15-GFP at the perinuclear region, intracellular vesicles, and the plasma membrane. (C) NMuMG cell transiently expressing Arl15-GFP (green) at the plasma membrane, internal vesicles, and the perinuclear region. (D) Day-5 3T3-L1 adipocytes expressing GFP-tagged Arl15 (green) were fixed and immunostained with anti-adiponectin antibodies and Alexa Fluor 594 (AdipoQ, red). Panel on the right shows the two channels merged. The average Pearson’s correlation coefficient (r) for the GFP and Alexa channels in hand-drawn “regions of interest” (ROI) encompassing entire cells was: 0.39 (without Costes automatic thresholding), 0.02 (below Costes Auto thresholding), and 0.11 (above Costes automatic thresholding). Scale bars, 10 μm. SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 7 www.nature.com/scientificreports/ ID Sex Age years Ethnicity B.M.I. kg/m2 Clinical Diagnoses ARL15 Variant Detected ARL15 variant ExAC allele frequency CADD Score P1 F 53 White British 30.9 Partial Lipodystrophy Congenital Adrenal Hyperplasia p.Pro26ThrfsTer9 0 P2 F 22 White British 20.2 Partial Lipodystrophy Yolk Sac Tumour c.254-2 A > G (g.54113409 A > G) 0 25 P3 F 24 Bengali 14.9 (16 years old) Late onset generalised lipodystrophy p.Cys133Arg 25/120366 (0.00021) 18 P4 F 19 Italian 31.2 Obesity PCOS p.Ala110Thr 16/120,448 (0.00013) 27 Table 1. Characteristics of four severely insulin resistant patients with ARL15 variants. B.M.I. = Body Mass Index; ExAC = Exome Aggregation Consortium32. All ancestries are included; CADD Score = Scaled Combined Annotation Dependent Depletion Score33 [CADDv1.3 accessed January 2017]. Table 1. Characteristics of four severely insulin resistant patients with ARL15 variants. B.M.I. = Body Mass Index; ExAC = Exome Aggregation Consortium32. All ancestries are included; CADD Score = Scaled Combined Annotation Dependent Depletion Score33 [CADDv1.3 accessed January 2017]. Results A l15 k individual), p.Arg30(1), and p.Tyr92(3)), among a total of 57 protein-altering variants affecting 187 partici- pants (Supplementary Table 3). No parents were available to study for the first proband (proband 1 (P1)), while the mother of the second proband (proband 2 (P2)) did not carry the splice site mutation. Her father was not available for study. Detailed clinical histories of P1 and P2 are given in the Supplementary Information. In brief, P1, who was heterozygous for the p.Pro26ThrfsTer9 mutation (Fig. 5D,J), is a 53 year old woman who developed obesity with severe clinical features of elevated androgenic hormones at 20 years old, leading to diagnosis of non-classical con- genital adrenal hyperplasia due to compound heterozygous mutations in the CYP21A2 gene, and treatment with a mean daily dose of prednisolone of 7.5 mg per day from 25 years old. Extreme elevation of serum triglyceride was noted at 28 years old, and at 30 years old diabetic ketoacidosis developed and insulin therapy was begun. Subsequent metabolic control remained suboptimal despite daily insulin doses up to 250 units per day, with HbA1c levels between 7 and 10%, and fasting plasma triglyceride concentrations from 190–2,200 mg/dL. At 38 years old peripheral lipodystrophy was recorded. P1’s mother, who had diabetes mellitus, and who was said to have had a similar body habitus, died at 68 years of a myocardial infarction. Her father died at 89 years old of heart failure. Neither were available for study. SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 8 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 5. Identification of two lipodystrophic patients with ARL15 haploinsufficiency. (A) Axillary acanthosis nigricans in P1 with (B) loss of femorogluteal and accumulation of truncal fat. (C) Coronal section of whole body MRI of P1 showing femorogluteal lipodystrophy. (D) Sequencing chromatograms showing frameshift mutation in ARL15 in P1 but not a control. (E) ARL15 transcript levels in immortalized lymphoblastoid cells from P1 compared to cells from healthy controls. RT-qPCR data are normalized to 36B4 expression. Means ± SEM are shown (n = 3). A.U., arbitrary units. Differences between means were analysed by Student’s t test; *P < 0.001. (F) Axillary acanthosis nigricans in P2. (G) Femorogluteal lipodystrophy in P2 with mammary hypoplasia. (H) Sequencing of ARL15 in P2, her mother, and a healthy control. The red arrow indicates the heterozygous 3′ splice site mutation in P2. (I) Detection of ARL15 exon 4 (209 bp) in lymphoblastoid cell mRNA. Results A l15 k Serum adiponectin concentrations as determined by immunoassay (DELFIA) are shown above the blots in red typeface (mg/L). Further information about control samples used is found in Supplementary Table 4. non-reducing and non-heat denaturing conditions, as indicated. The lower panel shows the monomeric form of adiponectin under heat-denaturing and reducing conditions. Serum adiponectin concentrations as determined by immunoassay (DELFIA) are shown above the blots in red typeface (mg/L). Further information about control samples used is found in Supplementary Table 4. On clinical assessment at 52 years old her Body Mass Index (B.M.I.) was 30.9 kg/m2 and she had centripetal obesity but peripheral lipodystrophy, with calf hypertrophy and little subcutaneous adipose tissue in the femoro- gluteal region (Fig. 5A–C, Table 2). Head and neck fat was increased, but there was no easy bruising, thin skin nor muscle wasting. There was persisting axillary acanthosis nigricans and marked clinical hyperandrogenism, with androgenetic alopecia and moderate hirsutism. Fasting biochemical evaluation and imaging revealed metabolic dyslipidemia, fatty liver, raised plasma testosterone but normal cortisol. Adiponectin was mildly low only, and not disproportionate to the overall clinical and biochemical profile (Table 2). Epstein Barr virus-transformed lymph- oblastoid cells derived from P1 with the p.Pro26ThrfsTer9 mutation expressed ARL15 mRNA at less than 50% of the level seen in healthy controls (Fig. 5E), but no mutant allele was seen on sequencing of cDNA, consistent with instability of the mutant mRNA (data not shown). P2, who carries an essential splice-acceptor site mutation in intron 3 (c.254–2 A > G; g.54113409 A > G) (Fig. 5H), is a 22 year old woman with subcutaneous lipodystrophy involving limbs and trunk including the mammary region. At 2 years old she was treated for a metastatic primary yolk sac tumour with JEB chemotherapy (carboplatin, etoposide, and bleomycin29) followed by surgical resection at 2.5 years old. No local or systemic radiotherapy was used. Progressive obesity was noted from around 4 years old. At 11 years old a predominantly centripetal pattern of adipose deposition was noted, as well as acanthosis nigricans. Marked “metabolic” dys- lipidaemia (high serum triglyceride and low HDL cholesterol) and polycystic ovary syndrome, consistent with severe insulin resistance, were recorded shortly afterwards. Discussion Several intronic single nucleotide polymorphisms (SNPs) at the ARL15 locus have been associated with components of the metabolic syndrome and its sequelae including plasma adiponectin, insulin, HDL cholesterol, and triglyceride concentrations, and coronary artery disease. The pattern of these associations, allied to associations with anthropo- metric traits including body shape and obesity, suggests that SNPs at this locus may primarily influence adipose tissue development or function, and give rise to metabolic traits indirectly by a lipodystrophy-like mechanism. Data from the Genotype-Tissue Expression (GTEx) Project indicate that ARL15 is relatively highly expressed in visceral and subcu- taneous adipose tissues, however eQTL data do not show significant association between the GWAS SNPs and ARL15 expression in the adipose tissue tested. Nevertheless adipose tissue is complex, and whether mature subcutaneous adi- pose tissue reflects the depot and developmental stage at which the ARL15 SNPs exert their influence is unclear.f ll We show that Arl15 expression is upregulated during 3T3-L1 adipocyte differentiation, and that its knock- down impairs 3T3-L1 adipocyte differentiation. We also provide evidence for a cell autonomous role for ARL15 in adiponectin expression and secretion from adipocytes, suggesting that adiponectin concentrations in the blood may be associated with ARL15 SNPs by more than one mechanism. The effects observed are consistent with, but do not prove, the case that the metabolic GWAS signals are mediated through effects on ARL15 itself.fi f We also report two haploinsufficient individuals with peripheral lipodystrophy, and centripetal obesity with dyslipidemic severe insulin resistance, and two less well studied individuals with severe insulin resistance and rare missense variants in ARL15. Neither haploinsufficient patient had adiponectin levels or oligomeric profiles that were distinct from control volunteers with primary lipodystrophy or hyperandrogenemic, insulin resistant PCOS. In both haploinsufficient patients we describe there are cofactors, namely lifelong androgen excess and childhood malignancy treated with systemic chemotherapy, each of which may contribute to adipose tissue dam- age or remodelling. Moreover, although no frameshift, essential splice site or nonsense mutations were identified in 2,479 controls studied, the ExAC database shows three nonsense mutations in five participants. On the other hand more than 20,000 patients in the ExAC component cohorts were ascertained through cardiometabolic traits potentially related to ARL15, so this cannot be construed as a healthy control dataset. Results A l15 k Two PCR products are visible for P2: spliced exon 4-containing and exon 4-less. For P2’s mother and healthy control only the exon 4-containing PCR amplicon is detected. (J) Schematic of ARL15 with the predicted Small GTPase superfamily ARF/SAR type functional domain shown in blue. The frameshift caused by the single nucleotide insertion in P1 results in eight aberrant amino acid residues Figure 5. Identification of two lipodystrophic patients with ARL15 haploinsufficiency. (A) Axillary acanthosis nigricans in P1 with (B) loss of femorogluteal and accumulation of truncal fat. (C) Coronal section of whole body MRI of P1 showing femorogluteal lipodystrophy. (D) Sequencing chromatograms showing frameshift mutation in ARL15 in P1 but not a control. (E) ARL15 transcript levels in immortalized lymphoblastoid cells from P1 compared to cells from healthy controls. RT-qPCR data are normalized to 36B4 expression. Means ± SEM are shown (n = 3). A.U., arbitrary units. Differences between means were analysed by Student’s t test; *P < 0.001. (F) Axillary acanthosis nigricans in P2. (G) Femorogluteal lipodystrophy in P2 with mammary hypoplasia. (H) Sequencing of ARL15 in P2, her mother, and a healthy control. The red arrow indicates the heterozygous 3′ splice site mutation in P2. (I) Detection of ARL15 exon 4 (209 bp) in lymphoblastoid cell mRNA. Two PCR products are visible for P2: spliced exon 4-containing and exon 4-less. For P2’s mother and healthy control only the exon 4-containing PCR amplicon is detected. (J) Schematic of ARL15 with the predicted Small GTPase superfamily ARF/SAR type functional domain shown in blue. The frameshift caused by the single nucleotide insertion in P1 results in eight aberrant amino acid residues immediately followed by a translation termination codon. The portion of ARL15 deleted in P2 as a result of mis- splicing is indicated by the two dashed lines flanking the box labelled “Exon 4”. (K) Immunoblotting analysis of serum adiponectin from the affected patients and controls with either polycystic ovary syndrome (PCOS), idiopathic familial partial lipodystrophy type 1 (FPLD1), or familial partial lipodystrophy type 2 due to LMNA p.Arg482Trp heterozygosity. Top, high-molecular weight (HMW), middle-molecular weight (MMW), and low-molecular weight (LMW), indicate the different oligomeric states of serum adiponectin resolved under SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 9 www.nature.com/scientificreports/ non-reducing and non-heat denaturing conditions, as indicated. The lower panel shows the monomeric form of adiponectin under heat-denaturing and reducing conditions. Results A l15 k A marked reduction in body mass index ensued, and on detailed evaluation at 16.5 years old, although there was residual centripetal adiposity, there was striking, relative paucity of adipose tissue from limbs and trunk including the mammary region, where there was breast hypoplasia. Head and neck adipose tissue was preserved, and moderate acanthosis nigricans could be seen in skin folds, but there were no clinical signs of dyslipidaemia, enlarged liver or hyperandrogenism.h On assessment at 22 years old height was 1.63 m and B.M.I. 20.2 kgm−2. There was subcutaneous lipodystro- phy affecting limbs and trunk including the mammary region, however head and neck adipose tissue was largely spared. Widespread moderate acanthosis nigricans was present in flexural regions (Fig. 5F,G). Fasting biochemi- cal evaluation and liver imaging revealed persisting metabolic dyslipidemia, fatty liver, raised plasma testosterone and insulin, and relatively normal adiponectin and leptin (Table 2).i y PCR amplification of cDNA derived from whole blood of P2 with the intron 3 splice-acceptor site mutation and her mother showed the presence of a truncated mRNA species which was absent from controls (Fig. 5I). Bidirectional Sanger sequencing of the PCR product showed this to be the result of skipping of exon 4, which is predicted to remove amino acids 85 to 154, including significant parts of the canonical GTP-binding motif in the event of the truncated protein being stable (Fig. 5I,J). Finally we compared plasma adiponectin concentrations and oligomeric profiles among P1, P2 and patients with familial partial lipodystrophy or insulin resistant polycys- tic ovary syndrome to assess whether P1 and P2 had disproportionately low adiponectin, however both P1 and P2 showed quantitative and qualitative adiponectin profiles sitting within the control group (Fig. 5K). SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 References 1. Richards, J. B. et al. A genome-wide association study reveals variants in ARL15 that influence adiponectin levels. PLoS Genet. 5 e1000768 (2009). 2. Stern, J. H., Rutkowski, J. M. & Scherer, P. E. Adiponectin, Leptin, and Fatty Acids in the Maintenance of Metabolic Homeostasi through Adipose Tissue Crosstalk. Cell Metab. 23, 770–84 (2016). g p 3. Teslovich, T. M. et al. Biological, clinical and population relevance of 95 loci for blood lipids. Nature 466, 707–13 (2010). 4. Global Lipids, G. C. et al. Discovery and refinement of loci associated with lipid levels. Nat. Genet. 45, 1274–83 (2013). g p 3. Teslovich, T. M. et al. Biological, clinical and population relevance of 95 loci for blood lipids. Nature 466, 707–13 (2010).i ation relevance of 95 loci for blood lipids. Nature 466, 707–13 (2010 3. Teslovich, T. M. et al. Biological, clinical and population relevance of 95 loci for blood lipids. Nature 466, 707 13 (201 4. Global Lipids, G. C. et al. Discovery and refinement of loci associated with lipid levels. Nat. Genet. 45, 1274–83 (2013 p yi p 5. Scott, R. A. et al. Large-scale association analyses identify new loci influencing glycemic traits and provide insight into the underlying biological pathways. Nat. Genet. 44, 991–1005 (2012). y g g p y 6. Scott, R. A. et al. Common genetic variants highlight the role of insulin resistance and body fat distribution in type 2 diabetes independent of obesity. Diabetes 63, 4378–87 (2014). p y 7. Yaghootkar, H. et al. Genetic evidence for a normal-weight ‘metabolically obese’ phenotype linking insulin resistance, hypertension, coronary artery disease, and type 2 diabetes. Diabetes 63, 4369–77 (2014).i y y yp 8. Ried, J. S. et al. A principal component meta-analysis on multiple anthropometric traits identifies novel loci for body shape. Nat Commun. 7, 13357 (2016). 9. Shungin, D. et al. New genetic loci link adipose and insulin biology to body fat distribution. Nature 518, 187–196 (2015). g g p gy y 10. Glessner, J. T. et al. A genome-wide study reveals copy number variants exclusive to childhood obesity cases. Am. J. Hum. Genet. 87, 661–6 (2010). 1. Semple, R. K. et al. Elevated plasma adiponectin in humans with genetically defective insulin receptors. J. Clin. Endocrinol. Metab 91, 3219–3223 (2006). 2. Rochford, J. J. et al. ETO/MTG8 Is an Inhibitor of C/EBP Activity and a Regulator of Early Adipogenesis. Mol. Cell. Biol. 24 9863–9872 (2004).hl 3. Discussion Thus, while the human genetic studies we describe do not prove that ARL15 haploinsufficiency gives rise to or predisposes to lipod- ystrophic insulin resistance, we suggest that this possibility warrants further study. SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 10 www.nature.com/scientificreports/ To date there have been no cellular functions ascribed to ARL15, and given that the term ARL merely indicates that the protein is structurally related to ARFs, no functional information can necessarily be inferred from it23. Nevertheless as small G proteins from the Rab and ARF families are critical regulators of vesicle trafficking21–23,28,30, we assessed whether ARL15 may play a role in adiponectin routing to the cell surface. We report Arl15 to be local- ised to the Golgi apparatus, and also to vesicular compartments and to the plasma membrane, which is consistent with this. However although adiponectin is also found in the Golgi, from where it is transported in vesicles to the surface of adipocytes28, and although Arl15 knockdown reduced adiponectin secretion, co-localization studies revealed no extensive overlap between the two molecules. Arl15’s action may thus be restricted to a subset of adiponectin transport compartments, including the subset which are destined for secretion. Emerging evidence has revealed a broad range of protein trafficking regulatory mechanisms governed by ARFs, encompassing mod- ulation of signalling inputs, actin dynamics and vesicle formation, organelle structure, recycling of specialized membrane compartments, or lipid composition of compartment membranes20,24,31. Whichever mechanism is at play in the case of Arl15, it shows some degree of cargo specificity, as adipsin secretion via the Golgi24 and insulin-stimulated Glut4 translocation appear to not have been affected by its knockdown. Unpicking which of these possibilities explains our observations will require substantial further study. p p q y In summary, we have studied the function of Arl15 in a murine preadipocyte cell line, and provide evidence that it is associated with the Golgi apparatus, secretory vesicles and the plasma membrane. Knockdown impairs both preadipocyte differentiation and adiponectin secretion from differentiated adipocytes. We also report two haploinsufficient individuals with partial lipodystrophy, increased centripetal adiposity, and severe insulin resist- ance and dyslipidemia, although in both cases major additional stressors on adipose function were present. We suggest that a possible role for heterozygous ARL15 loss-of-function variants in pathological adipose remodelling and insulin resistance-related traits merits further study. References Zavzavadjian, J. R. et al. The alliance for cellular signaling plasmid collection: a flexible resource for protein localization studies and signaling pathway analysis. Mol. Cell. Proteomics 6, 413–24 (2007).i 13. Zavzavadjian, J. R. et al. The alliance for cellular signaling plasmid collection: a flexible resource for protein localization studies and signaling pathway analysis. Mol. Cell. Proteomics 6, 413–24 (2007). 14. Payne, F. et al. Hypomorphism in human NSMCE2 linked to primordial dwarfism and insulin resistance. J. Clin. Invest. 124, signaling pathway analysis. Mol. Cell. Proteomics 6, 413–24 (2007). 14. Payne, F. et al. Hypomorphism in human NSMCE2 linked to primordial dwarfism and insulin resistance. J. Clin. Invest. 124, 4028–38 (2014). 14. Payne, F. et al. Hypomorphism in human NSMCE2 linked to primordial dwarfism and insulin resistance. J. Clin. Invest. 4028–38 (2014). 5. Shin, K.-J. et al. A single lentiviral vector platform for microRNA-based conditional RNA interference and coordinated transgene expression. Proc. Natl. Acad. Sci. USA 103, 13759–64 (2006). p 6. Huang-Doran, I. et al. Genetic defects in human pericentrin are associated with severe insulin resistance and diabetes. Diabetes 60 925–35 (2011). 17. Semple, R. K. et al. Postreceptor insulin resistance contributes to human dyslipidemia and hepatic steatosis. J. Clin. Invest. 119, 315–22 (2009). ( ) 18. Reeder, S. B., Cruite, I., Hamilton, G. & Sirlin, C. B. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J. Magn. Reson. Imaging 34, 729–49 (2011). p py g g g 9. Williams, D. R. R. et al. Undiagnosed Glucose Intolerance in the Community: the Isle of Ely Diabetes Project. Diabet. Med. 12 30–35 (1995). 20. Donaldson, J. G. Arfs and membrane lipids: sensing, generating and responding to membrane curvature. Biochem. J. 414, e1–e2 (2008).h 21. Chavrier, P. & Goud, B. The role of ARF and Rab GTPases in membrane transport. Current Opinion in Cell Biology 11, 466–475 (1999).fi ( ) 22. D’Souza-Schorey, C. & Chavrier, P. ARF proteins: roles in membrane traffic and beyond. Nat. Rev. Mol. Cell Biol. 7, 347–358 (2 22. D’Souza-Schorey, C. & Chavrier, P. ARF proteins: roles in membrane traffic and beyond. Nat. Rev. Mol. Cell Biol. 7, 347–358 (2006). 23. Gillingham, A. K. & Munro, S. The small G proteins of the Arf family and their regulators. Annu. Rev. Cell Dev. Biol. 23, 579–611 (2007). y pfi y ( ) 23. Gillingham, A. K. & Munro, S. SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 References The small G proteins of the Arf family and their regulators. Annu. Rev. Cell Dev. Biol. 23, 579–611 (2007). ( ) 4. Clarke, M., Ewart, M. A., Santy, L. C., Prekeris, R. & Gould, G. W. ACRP30 is secreted from 3T3-L1 adipocytes via a Rab11 dependent pathway. Biochem. Biophys. Res. Commun. 342, 1361–1367 (2006).fi 25. Klausner, R. D., Donaldson, J. G. & Lippincott-Schwartz, J. Brefeldin A: insights into the control of membrane traffic and organelle structure. J. Cell Biol. 116, 1071–80 (1992). S l T l f b d f h A l d h f G T ffi ( ) ( ) 26. Szul, T. et al. Dissection of membrane dynamics of the ARF-guanine nucleotide exchange factor GBF1. Traffic 6, 374–85 (2005 27. Peyroche, A. et al. Brefeldin A acts to stabilize an abortive A fi 27. Peyroche, A. et al. Brefeldin A acts to stabilize an abortive ARF-GDP-Sec. 7 domain protein complex: involvement of specific residues of the Sec. 7 domain. Mol. Cell 3, 275–85 (1999).fi 28. Xie, L., O’Reilly, C. P., Chapes, S. K. & Mora, S. Adiponectin and leptin are secreted through distinct trafficking pathways in adipocytes. Biochim. Biophys. Acta - Mol. Basis Dis. 1782, 99–108 (2008). y p p adipocytes. Biochim. Biophys. Acta - Mol. Basis Dis. 1782, 99–108 p y p y , ( ) 29. Pinkerton, C. R. et al. ‘JEB’–a carboplatin based regimen for malignant germ cell tumours in children. Br. J. Cancer 62, 25 (1990). SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 11 www.nature.com/scientificreports/ 0. Xie, L. et al. Intracellular trafficking and secretion of adiponectin is dependent on GGA-coated vesicles. J. Biol. Chem. 281 7253–7259 (2006). ( ) 31. Casanova, J. E. et al. Meeting report - Arf and Rab family G proteins. J. Cell Sci. 126, 5313–6 (2013). y p g g 33. Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet 46, 310 (2014). y p g g 33. Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet 46, 310–315 (2014). y g g 33. Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet 46, 310–315 (2014). Author Contributions I.B. and R.K.S. conceived the study. N.R., F.P., I.H.-D., C.A., A.Sl., and K.F. researched data. R.K.S. and N.R. wrote the manuscript. A.St., A.Sl., V.S., C.A., F.P., S.O’R. contributed to discussion and reviewed/edited the manuscript. Acknowledgements g We are grateful to the patients for their participation in these studies, to Keith Burling and the UK NIHR Clinical Biochemistry Assay Laboratory for immunoassays, to Laura Watson of the NIHR/Wellcome Trust Clinical Research Facility for providing body composition analysis and control data, and to Prof David Lomas of the University of Cambridge Department of Radiology for liver fat quantification in Patient 2. A full list of the investigators who contributed to the generation of the UK10K data is available from www.UK10K.org. The authors would like to thank the Exome Aggregation Consortium and the groups that provided exome variant data for comparison. A full list of contributing groups can be found at http://exac.broadinstitute.org/about. We also thank the staff of the DNA Pipelines team at the Wellcome Trust Sanger Institute. This work was supported by the Wellcome Trust [grant numbers WT098498, WT098051, with data from the UK10K Consortium [WT091310] Rare Diseases Arm] the Medical Research Council [MRC_MC_UU_12012/5]; the United Kingdom National Institute for Health Research (NIHR) Cambridge Biomedical Research Centre; and the EU/EFPIA Innovative Medicines Initiative Joint Undertaking (EMIF grant no. 115372). SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-017-17746-8. Competing Interests: The authors declare that they have no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2017 SCIENTIFIC RePOrTS \| 7: 17593 \| DOI:10.1038/s41598-017-17746-8 12
https://openalex.org/W4391780204	https://www.biorxiv.org/content/biorxiv/early/2024/02/13/2024.02.12.579951.full.pdf	English	null	Short-term memory capacity predicts willingness to expend cognitive effort for reward	bioRxiv (Cold Spring Harbor Laboratory)	2,024	cc-by	21,960	1 1 PREDICTING COGNITIVE EFFORT Short-term memory capacity predicts willingness to expend cognitive effort for 1 reward 2 . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 3 Abstract We must often decide whether the effort required for a task is worth the reward. Past 25 rodent work suggests that willingness to deploy cognitive effort can be driven by individual 26 differences in perceived reward value, depression, or chronic stress. However, many factors 27 driving cognitive effort deployment - such as short-term memory ability - cannot easily be 28 captured in rodents. Furthermore, we do not fully understand how individual differences in 29 short-term memory ability, depression, chronic stress, and reward sensitivity impact 30 cognitive effort deployment for reward. Here, we examined whether these factors predict 31 cognitive effort deployment for higher reward in an online visual short-term memory task. 32 Undergraduate participants were grouped into high and low effort groups (nHighEffort = 348, 33 nLowEffort = 81; nFemale = 332, nMale = 92, MAge = 20.37, RangeAge = 16-42) based on 34 decisions in this task. After completing a monetary incentive task to measure reward 35 sensitivity, participants completed short-term memory task trials where they could choose 36 to encode either fewer (low effort/reward) or more (high effort/reward) squares before 37 reporting whether or not the colour of a target square matched the square previously in 38 that location. We found that only greater short-term memory ability predicted whether 39 participants chose a much higher proportion of high vs. low effort trials. Drift diffusion 40 modeling showed that high effort group participants were more biased than low effort group 41 participants towards selecting high effort trials. Our findings highlight the role of individual 42 differences in cognitive effort ability in explaining cognitive effort deployment choices. 43 ff We must often decide whether the effort required for a task is worth the reward. Past 25 rodent work suggests that willingness to deploy cognitive effort can be driven by individual 26 differences in perceived reward value, depression, or chronic stress. However, many factors 27 driving cognitive effort deployment - such as short-term memory ability - cannot easily be 28 captured in rodents. Furthermore, we do not fully understand how individual differences in 29 short-term memory ability, depression, chronic stress, and reward sensitivity impact 30 cognitive effort deployment for reward. Here, we examined whether these factors predict 31 cognitive effort deployment for higher reward in an online visual short-term memory task. Short-term memory capacity predicts willingness to expend cognitive effort for 1 reward 2 Brandon J. Forys1, Catharine A. Winstanley1, 2, Alan Kingstone1, and Rebecca M. Todd1, 2 3 1Department of Psychology, University of British Columbia, Vancouver, BC, Canada 4 2Djavad Mowafaghian Centre for Brain Health, University of British Columbia, Vancouver, 5 BC, Canada 6 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 2 Author Note The authors declare no competing financial interests. We would like to acknowledge 9 Alex Terpstra for creating the delayed monetary incentive task, Rita Jin for assisting in 10 data analyses, and Claire Hales for her input on the drift diffusion model results. This 11 research was supported by a Natural Sciences and Engineering Research Council of Canada 12 (NSERC) grant (#F19-05182) to R. M. Todd and the UBC Djavad Mowafaghian Centre 13 for Brain Health Innovation Fund Kickstart Research Grant (#F19-05932), as well as an 14 NSERC Canada Graduate Scholarship - Doctoral (CGS D) award to B.J. Forys and a 15 Michael Smith Foundation for Health Research Scholar Award to R.M. Todd. 16 The authors made the following contributions. Brandon J. Forys: 17 Conceptualization, Investigation, Methodology, Writing - Original Draft Preparation, 18 Writing - Review & Editing; Catharine A. Winstanley: Conceptualization; Alan Kingstone: 19 Conceptualization; Rebecca M. Todd: Conceptualization, Funding acquisition, Project 20 Administration, Resources, Supervision, Writing - Review & Editing. 21 Correspondence concerning this article should be addressed to Brandon J. Forys, 22 2136 West Mall, Vancouver, BC Canada V6T 1Z4. E-mail: brandon.forys@psych.ubc.ca 23 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 4 PREDICTING COGNITIVE EFFORT Significance statement We must often make decisions about when cognitive effort is worth the potential 49 reward. Reward value, depression, and chronic stress in rodents can impact cognitive effort 50 deployment decisions for reward, but factors like short-term memory ability can only be 51 easily characterized in humans. We examined whether short-term memory ability, 52 depression, chronic stress, and reward sensitivity predict cognitive effort decisions for 53 reward. In a short-term visual memory task with a choice of easier or harder trials for low 54 vs. high reward, we found that only short-term memory ability predicted more choices of 55 high vs. low effort trials. This research suggests that cognitive effort decisions could be 56 driven by cognitive effort ability more than motivational factors like depression or chronic 57 stress. 58 Abstract 32 Undergraduate participants were grouped into high and low effort groups (nHighEffort = 348, 33 nLowEffort = 81; nFemale = 332, nMale = 92, MAge = 20.37, RangeAge = 16-42) based on 34 decisions in this task. After completing a monetary incentive task to measure reward 35 sensitivity, participants completed short-term memory task trials where they could choose 36 to encode either fewer (low effort/reward) or more (high effort/reward) squares before 37 reporting whether or not the colour of a target square matched the square previously in 38 that location. We found that only greater short-term memory ability predicted whether 39 participants chose a much higher proportion of high vs. low effort trials. Drift diffusion 40 modeling showed that high effort group participants were more biased than low effort group 41 participants towards selecting high effort trials. Our findings highlight the role of individual 42 differences in cognitive effort ability in explaining cognitive effort deployment choices. 43 Keywords: cognitive effort, working memory, chronic stress 44 Word count: 7374 45 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Introduction It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Introduction We are often faced with diﬀicult choices about work and life. For example, we may 60 choose to spend time thoroughly studying for an exam to gain a few extra percent points 61 on a grade; alternatively, we may trade off this reward to work less and be able to do other 62 tasks. These choices require trading off more or less effort for a larger or smaller reward, 63 and thus involve deciding how much cognitive effort to deploy. In general, the motivation 64 to deploy cognitive effort can be influenced by the potential reward to be gained (Shenhav, 65 Botvinick, & Cohen, 2013; Shenhav, Cohen, & Botvinick, 2016; Yee, Crawford, 66 We are often faced with diﬀicult choices about work and life. For example, we may 60 choose to spend time thoroughly studying for an exam to gain a few extra percent points 61 on a grade; alternatively, we may trade off this reward to work less and be able to do other 62 tasks. These choices require trading off more or less effort for a larger or smaller reward, 63 and thus involve deciding how much cognitive effort to deploy. In general, the motivation 64 to deploy cognitive effort can be influenced by the potential reward to be gained (Shenhav, 65 Botvinick, & Cohen, 2013; Shenhav, Cohen, & Botvinick, 2016; Yee, Crawford, 66 Lamichhane, & Braver, 2021). Beyond this, an individual’s willingness to expend cognitive 67 effort can also be linked to individual differences in factors that influence overall 68 motivation, such as mood disorder levels (Grahek, Shenhav, Musslick, Krebs, & Koster, 69 2019; Pruessner, Barnow, Holt, Joormann, & Schulze, 2020; Yee, Adams, Beck, & Braver, 70 2019). Research in rodents has revealed patterns of individual differences in motivation to 71 Lamichhane, & Braver, 2021). Beyond this, an individual’s willingness to expend cognitive 67 effort can also be linked to individual differences in factors that influence overall 68 motivation, such as mood disorder levels (Grahek, Shenhav, Musslick, Krebs, & Koster, 69 2019; Pruessner, Barnow, Holt, Joormann, & Schulze, 2020; Yee, Adams, Beck, & Braver, 70 2019). Research in rodents has revealed patterns of individual differences in motivation to 71 PREDICTING COGNITIVE EFFORT 5 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 6 & Zald, 2009) and anticipation of anhedonia (Sherdell, Waugh, & Gotlib, 2012) generally 98 predicts reduced motivation to deploy physical effort for reward (Culbreth, Moran, & 99 Barch, 2018). However, many tasks in life are cognitive as opposed to physical in nature; 100 cognitive effort tasks may offer more naturalistic value as a result. 101 & Zald, 2009) and anticipation of anhedonia (Sherdell, Waugh, & Gotlib, 2012) generally 98 predicts reduced motivation to deploy physical effort for reward (Culbreth, Moran, & 99 Barch, 2018). However, many tasks in life are cognitive as opposed to physical in nature; 100 cognitive effort tasks may offer more naturalistic value as a result. 101 Cognitive factors influencing willingness to deploy cognitive effort in humans are 102 more complex than those that we can evaluate with rodents (Stephan, Volkmann, & 103 Rossner, 2019). Humans require less training than rodents on cognitive effort tasks, and 104 can be evaluated on a greater variety of behavioural measures that may reveal factors 105 influencing individual differences in willingness and ability to deploy cognitive effort for 106 reward. For example, even if we are highly motivated to exert cognitive effort, reduced 107 executive function capability could drive increased reliance on reward incentives - leading 108 us to deploy cognitive effort beyond our capabilities and increasing the risk of failure (Kool, 109 McGuire, Rosen, & Botvinick, 2010; Sandra & Otto, 2018). One executive process that we 110 use in everyday life is visuospatial working memory, or the ability to hold in mind and 111 manipulate object locations in space. Increased working memory ability can in turn drive 112 improved attentional control - an aspect of cognitive effort - while reduced working memory 113 ability is associated with poorer control (Unsworth & Robison, 2020). Furthermore, during 114 cognitively effortful tasks, reward sensitivity modulates activation in neural circuitry that 115 supports working memory (Fuentes-Claramonte et al., 2015). 116 The Expected Value of Control theory (Shenhav et al., 2013) predicts that, in 117 general, people will deploy more cognitive effort when they know that they can effectively 118 obtain greater rewards while doing so (Frömer, Lin, Dean Wolf, Inzlicht, & Shenhav, 2021). 119 Additionally, the theory predicts that greater reward anticipation will lead to higher 120 expected value of deploying cognitive effort to obtain a reward (Grahek, Musslick, & 121 Shenhav, 2020). PREDICTING COGNITIVE EFFORT CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 5 exert cognitive effort, where rats are classified as “workers” (high effort group) or “slackers” 72 (low effort group) depending on whether they are willing to work more or less for reward, 73 respectively (Hosking, Cocker, & Winstanley, 2016; Silveira, Wittekindt, Ebsary, & 74 Winstanley, 2021; Silveira, Wittekindt, Mortazavi, Hathaway, & Winstanley, 2020). 75 However, factors influencing individual differences in willingness to deploy cognitive effort 76 for reward have yet to be fully characterized in humans. 77 However, factors influencing individual differences in willingness to deploy cognitive effort 76 for reward have yet to be fully characterized in humans. 77 In both rodents and humans, chronic stress (Birn, Roeber, & Pollak, 2017; Watt, 78 Weber, Davies, & Forster, 2017) and depressive traits (Grahek et al., 2019; Silveira et al., 79 2020) have been shown to negatively impact cognitive effort deployment for reward. In 80 rodents, chronic stress dampens reward sensitivity even as acute stress heightens it 81 (Ironside, Kumar, Kang, & Pizzagalli, 2018; Kúkel’ová et al., 2018). Here, chronic stress is 82 typically induced by prolonged social defeat (Kúkel’ová et al., 2018) or non-social chronic 83 mild stress, such as modifications to housing (Slattery & Cryan, 2017). In humans, chronic 84 stress - unlike acute stress - cannot ethically be induced. Research into the impact of 85 chronic stress on reward motivation primarily focuses on reports early childhood stress 86 (Birn et al., 2017; Watt et al., 2017), and not shorter-term chronic stress of the kind that 87 university students may experience (Towbes & Cohen, 1996). This latter form of chronic 88 stress is more widely experienced than early childhood stress, and yet its impact on our 89 willingness and ability to complete everyday cognitive tasks is not well understood. 90 Alongside chronic stress, depression is another motivational factor that has been 91 shown to dampen reward sensitivity (Terpstra et al., 2023; A. Westbrook et al., 2020; S. R. 92 Westbrook, Hankosky, Dwyer, & Gulley, 2018). In particular, anhedonia - a key symptom 93 of depression impacting interpretations of reward (Slattery & Cryan, 2017) - can drive 94 reduced reward sensitivity and negatively impact willingness to deploy effort in rodents 95 (Scheggi, De Montis, & Gambarana, 2018). Similarly, in humans, trait-level anhedonia 96 (Treadway, Bossaller, Shelton, & Zald, 2012; Treadway, Buckholtz, Schwartzman, Lambert, 97 . cognitive effort for higher reward is not known. 124 In rodents, different levels of motivation between ‘workers’ and ‘slackers’ have also 125 been associated with differences in learning rate and bias towards choosing high effort 126 trials, and that this may be linked to differences in cognitive capacity. Drift diffusion 127 modeling in rodents completing a cognitive effort task has revealed that ‘workers’ with high 128 accuracy may accumulate evidence more quickly towards selecting high effort, high reward 129 trials than ‘slackers’ or ‘workers’ with low accuracy (Hales et al., 2024). Overall, the above 130 findings suggest that effort cost computations interact with reward sensitivity as factors we 131 rely on to make judgments about the value of deploying more or less cognitive effort for 132 reward. In turn, individual differences in reward sensitivity may be linked to depressive 133 traits or chronic stress, and rodent research indicates that executive capacity may also 134 influence choice behavior (Eichenbaum, 2017). The goal of the present study was to 135 examine the degree to which individual differences in reward sensitivity, chronic stress 136 levels, and depressive traits on the one hand, and visuospatial short-term memory as an 137 index of executive function on the other, predict human choice behaviour independently or 138 in interaction. 139 PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 7 PREDICTING COGNITIVE EFFORT However, the degree to which individual differences in reward sensitivity 122 and executive function capacity influence the likelihood of choosing to deploy more 123 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Materials and Methods 165 PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 8 given the choice between a low effort, low reward (LR) trial or a high effort, high reward 150 (HR) trial. Then, in a basic memory task, they must poke their noses in a hole that lights 151 up for either 1000 ms (LR) or 200 ms (HR) (Cocker et al., 2012). Rodents must rapidly 152 encode the location of the illuminated light to succeed. Such a task can be scaled up to 153 human working memory capabilities by including more stimuli and features that must be 154 encoded to obtain a reward. For our study, we used harder and easier conditions (smaller 155 vs. larger set size) from a visual short-term memory task (Luck & Vogel, 1997) to serve as 156 high and low cognitive effort choices within the same kind of choice paradigm offered in the 157 rCET. In each trial, participants could choose to perform an easier visual short-term 158 memory task for a lower reward or a harder task for a higher reward. We examined overall 159 visuospatial short-term memory capacity as well as indices of depressive traits (Beck, Steer, 160 & Brown, 1996), chronic stress (Levenstein et al., 1993), anhedonia (Rizvi et al., 2015), and 161 reward sensitivity (Terpstra et al., 2023) as predictors of the tendency to choose lower or 162 higher effort tasks. Furthermore, we used drift diffusion modeling (Hales et al., 2024; 163 Ratcliff, 1978) to examine factors that may contribute to choice biases. 164 The present study 140 One way of examining factors that motivate humans to deploy more or less effort for 141 reward is to offer participants the choice of completing an easier or harder cognitive effort 142 task for a low or high reward, respectively (Treadway et al., 2009). In the current study, we 143 built on an existing rodent choice task to investigate whether visual short-term memory 144 capacity, chronic stress, depressive traits, and reward sensitivity predict one’s choices in 145 deploying cognitive effort for reward. We adapted the rodent Cognitive Effort Task (rCET) 146 (Cocker, Hosking, Benoit, & Winstanley, 2012; Hosking et al., 2016; Silveira et al., 2020) 147 for use in humans. In the rodent paradigm, rodents first learn to associate illuminated 148 lights with the opportunity to obtain a reward. In the main phase of the task, rodents are 149 . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 8 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 9 evaluating the probability of a participant selecting 70% or more high effort trials (being in 169 the high effort group in the study). This analysis gave us a target sample size of N = 487. 170 We recruited N = 570 participants from the Human Subject Pool of psychology 171 undergraduate students at the University of British Columbia; they were each compensated 172 with bonus points for their courses as well as a CAD $5.00 Starbucks gift card. Of these, n 173 = 42 participants did not complete the initial survey and n = 7 participants did not 174 complete the reward anticipation task. Furthermore, n = 11 participants completed the 175 change detection task more than once, n = 2 participants spent more than 30 seconds 176 choosing any one trial - as this could indicate disengagement with task demands, and n = 177 79 participants performed at or below chance (50%) during the choice phase of the task. In 178 total, we analyzed data from N = 429 participants (n = 92 male, n = 332 female, n = 5 179 other; Table 1). The study was approved by the Behavioural Research Ethics Board at the 180 University of British Columbia, ethics code H20-01388. 181 Table 1 Demographic information for all participants, by sex and effort deployment group. BDI prop = Beck Depression Inventory II proportion score (score divided by max score). PSS = Perceived Stress Scale score. DARS = Dimensional Anhedonia Rating Scale. Ant. = mean reward anticipation in the behavioural monetary incentive delay task. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. 𝐺𝑟𝑜𝑢𝑝 𝑆𝑒𝑥 𝑛 𝑀𝑎𝑔𝑒 𝑆𝐷𝑎𝑔𝑒 𝑀𝑖𝑛𝑎𝑔𝑒 𝑀𝑎𝑥𝑎𝑔𝑒 𝑀𝐵𝐷𝐼𝑝𝑟𝑜𝑝 𝑆𝐷𝐵𝐷𝐼𝑝𝑟𝑜𝑝 𝑀𝑃𝑆𝑆 𝑆𝐷𝑃𝑆𝑆 𝑀𝐷𝐴𝑅𝑆 𝑆𝐷𝐷𝐴𝑅𝑆 𝑀𝑎𝑛𝑡. 𝑆𝐷𝑎𝑛𝑡. High effort Female 263 20.34 2.37 16 42 0.27 0.18 20.83 6.21 35.61 11.71 3.94 1.81 Male 80 20.70 2.34 17 32 0.19 0.14 17.69 5.30 37.69 9.62 4.29 1.75 Other 5 20.40 0.89 19 21 0.17 0.10 19.60 3.65 43.00 10.07 4.80 1.64 Low effort Female 69 20.01 1.94 16 26 0.28 0.16 20.68 6.54 38.04 8.94 4.00 1.79 Male 12 20.92 2.97 18 28 0.19 0.14 21.08 4.38 39.17 8.61 3.96 1.51 We powered our study using a power analysis through the pwrss package (Bulus, 167 167 2023) to achieve an expected power of 80% at an odds ratio of 0.70 for a logistic regression 168 2023) to achieve an expected power of 80% at an odds ratio of 0.70 for a logistic regression 168 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 10 asked to view another scale where they would indicate how excited they felt. They 194 completed this task for a series of eight trials. 195 asked to view another scale where they would indicate how excited they felt. They 194 completed this task for a series of eight trials. 195 In the second phase of the study, as a measure of each individual’s visuospatial 196 short-term memory capacity we presented a visuospatial short-term memory task (Fig. 1B, 197 C) modified from Luck and Vogel (1997) and adapted from an open-source version of the 198 task on Pavlovia (de Oliveira, João Roberto Ventura, 2023). In this task, we presented a 199 series of trials with between 2 and 8 coloured squares that were presented for 500 ms. Each 200 square subtended a visual angle of approximately 0.05∘on the screen. A mask with 201 multi-coloured squares would then appear at each of the original squares’ locations for 202 another 500 ms, followed by a single coloured square appearing in one of the positions of 203 the original squares. This final square had a 50% chance of being the same or a different 204 colour from the square appearing in the same position in the initial part of the trial. After 205 indicating with a keyboard press whether the square was the same or a different colour 206 from the initial square, they would see whether or not they gained points towards a 207 monetary reward. 208 Stimulus Presentation 182 All participants completed the study online on their own devices, via the Pavlovia 183 online study platform using PsychoPy 2021.2.3 (RRID: SCR_006571) (Peirce et al., 2019). 184 Participants were not allowed to complete the study on mobile phones or tablets. 185 All stimuli used in the study were generated by and implemented in PsychoPy 187 (Peirce et al., 2019). In the first phase of the study, participants performed an online 188 behavioural monetary incentive delay task as a measure of reward sensitivity (Fig. 1A) as 189 outlined in Terpstra et al. (2023). In brief, during this initial task phase, participants saw 190 a series of scales that they could click to indicate how excited they were to receive a reward. 191 After a brief wait, participants would see a small happy face appear on one side of the 192 screen. Afterwards, they were shown whether they received a reward or not, and were 193 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Questionnaires and monetary incentive delay task 210 Participants began the study by completing an online questionnaire. After giving 211 consent and demographic information, they were asked how about their history of 212 depression and anxiety; COVID-related stress; the Perceived Stress Scale (PSS; Levenstein 213 et al. (1993)); the Beck Depression Inventory II (BDI; Beck et al. (1996)); and the 214 Dimensional Anhedonia Rating Scale (DARS; Rizvi et al. (2015)). Afterwards, they were 215 redirected to the first phase of our study, the online behavioural monetary incentive delay 216 task (Fig. 1A; Terpstra et al. (2023)). This task measured participants’ reward sensitivity 217 via the mean of the excitement ratings given before each trial. 218 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint 11 PREDICTING COGNITIVE EFFORT You are playing for 1 ticket. How excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. A B + + Cue 500 ms Mask 500 ms Probe Until response + Cue 500 ms Mask 500 ms Probe Until response C Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select Low effort, low reward trial High effort, high reward trial Success Failure +1 point! Score: 11 +0 points! Score: 10 Success Failure +10 points! Score: 21 +0 points! Score: 10 Probe Until response Cue 500 ms Mask 500 ms Fixation 500 ms Fixation 500 ms Fixation 500 ms Choice Time Time Time Time Figure 1 Layout of the experimental tasks. Figure 1 Figure 1 Layout of the experimental tasks. (A) The Monetary Incentive Delay Task, in which participants indicate their excitement in playing for tickets towards a monetary reward.(B) The calibration phase of the change detection task, where participants saw an array of either 2, 4, 6, or 8 squares and indicated whether the final square that appeared (the probe) was of the same or a different colour from the square that appeared in the same location in the previously show array. (C) The reward phase of the change detection task, where participants saw an array of either 2 or 4, or 6 or 8, shapes on screen depending on their performance in the calibration phase of the task. Once again, participants indicated with a keyboard press whether the final square that appeared (the probe) is the same or a different colour from the square that appeared in the same location in the array. Here, if they made the correct decision, they would receive 1 point (low effort trial) or 10 points (high effort trial); they would receive 0 points for an incorrect decision. Questionnaires and monetary incentive delay task 210 (A) The Monetary Incentive Delay Task, in which participants indicate their excitement in playing for tickets towards a monetary reward.(B) The calibration phase of the change detection task, where participants saw an array of either 2, 4, 6, or 8 squares and indicated whether the final square that appeared (the probe) was of the same or a different colour from the square that appeared in the same location in the previously show array. (C) The reward phase of the change detection task, where participants saw an array of either 2 or 4, or 6 or 8, shapes on screen depending on their performance in the calibration phase of the task. Once again, participants indicated with a keyboard press whether the final square that appeared (the probe) is the same or a different colour from the square that appeared in the same location in the array. Here, if they made the correct decision, they would receive 1 point (low effort trial) or 10 points (high effort trial); they would receive 0 points for an incorrect decision. You are playing for 1 ticket. How excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. A Time A + Cue 500 ms Mask 500 ms Probe Until response + Probe Until response C Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select Low effort, low reward trial High effort, high reward trial Success Failure +1 point! Score: 11 +0 points! Score: 10 Success Failure +10 points! Score: 21 +0 points! Score: 10 Cue 500 ms Mask 500 ms Fixation 500 ms Fixation 500 ms Choice Time Time C B + Cue 500 ms Mask 500 ms Probe Until response Fixation 500 ms Time B Visuospatial short-term memory task 219 Versions of this task served both as a means to measure individual differences in 220 visuospatial short-term memory as well as the tasks of varied cognitive effort to be 221 subsequently chosen for high/low effort trials. After completing the monetary incentive 222 delay task, participants were redirected to the second phase of our study, the visuospatial 223 short-term memory task (Luck & Vogel, 1997). Here, this task served as a means to 224 measure individual differences in visuospatial short-term memory as an index of working 225 memory. After receiving instructions on which stimuli would appear and how to respond to 226 them, participants first completed a series of ten practice trials presented in a random 227 order. On half of these trials, the probe square was the same colour as that of the cue 228 square (congruent/change trial); on the other half, the probe square was a different colour 229 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 12 from that of the cue square (incongruent/no change trial). Of these practice trials, four had 230 a set size = 2, another four had a set size = 4, and another two had a set size = 6. In each 231 trial the probe square could be anywhere in the array, and participants were required to 232 hold the whole array in visual short-term memory in order to successfully indicate whether 233 the probe color had changed. Participants did not receive feedback on whether their 234 from that of the cue square (incongruent/no change trial). Of these practice trials, four had 230 a set size = 2, another four had a set size = 4, and another two had a set size = 6. In each 231 trial the probe square could be anywhere in the array, and participants were required to 232 hold the whole array in visual short-term memory in order to successfully indicate whether 233 the probe color had changed. Participants did not receive feedback on whether their 234 responses were correct or not in the practice block, in keeping with Luck and Vogel (1997). 235 Following these practice trials, participants completed (Fig. 1B) a series of 120 trials of the 236 visuospatial short-term memory task with 60 change and 60 no change trials. In total, 30 237 trials of each set size (2, 4, 6, and 8 squares onscreen) were presented in a randomized 238 order. Although the original task by Luck and Vogel (1997) contained trials with up to 10 239 squares onscreen, this largest set size was determined to be too diﬀicult for participants to 240 reliably complete correctly following piloting; as such, the maximum set size was 8. 241 responses were correct or not in the practice block, in keeping with Luck and Vogel (1997). 235 Following these practice trials, participants completed (Fig. 1B) a series of 120 trials of the 236 visuospatial short-term memory task with 60 change and 60 no change trials. In total, 30 237 trials of each set size (2, 4, 6, and 8 squares onscreen) were presented in a randomized 238 order. PREDICTING COGNITIVE EFFORT Although the original task by Luck and Vogel (1997) contained trials with up to 10 239 squares onscreen, this largest set size was determined to be too diﬀicult for participants to 240 reliably complete correctly following piloting; as such, the maximum set size was 8. 241 Based on participants’ performance in this phase of the task, a K estimate of their 242 visuospatial working memory capability (Rouder, Morey, Morey, & Cowan, 2011) was 243 calculated for each set size, as follows: 244 ̂𝑘𝑝= 𝑁( ̂ℎ− ̂𝑓 1 − ̂𝑓 ) ̂𝑘𝑝= 𝑁( ̂ℎ− ̂𝑓 1 − ̂𝑓 ) where 𝑁is the set size, ̂ℎis the hit rate, and ̂𝑓is the false alarm rate. Although our 245 task was a single-probe recognition task, we used the initially proposed whole-display K 246 estimate measure (Pashler, 1988) to ensure a stricter criterion than the single-probe K 247 estimate measure. The whole-display K estimate measure has also previously been used in 248 single-probe change detection tasks (Rouder et al., 2011). 249 where 𝑁is the set size, ̂ℎis the hit rate, and ̂𝑓is the false alarm rate. Although our 245 task was a single-probe recognition task, we used the initially proposed whole-display K 246 estimate measure (Pashler, 1988) to ensure a stricter criterion than the single-probe K 247 estimate measure. The whole-display K estimate measure has also previously been used in 248 single-probe change detection tasks (Rouder et al., 2011). 249 245 In addition to serving as indices of individual differences in visuospatial short-term 250 memory, K estimate scores were used to calibrate the tasks used in the effort choice task 251 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 13 (Fig. 1C), which was based on the rodent cognitive effort task (Hosking et al., 2016; Silveira 252 et al., 2021). Here, participants completed a series of 30 trials. They would receive rewards 253 for making a correct response, and they could choose a high effort, high reward (HR) trial 254 or a low effort, low reward (LR) trial at each trial. Importantly in this task, trials with 255 small and large arrays from the visuospatial short-term memory task served as the LR and 256 HR trials. HR trials used a larger set size, but gave a reward of 10 points towards a 257 monetary reward. LR trials had a smaller set size than the high effort/high reward trials, 258 but yielded a reward of 1 point towards a monetary reward. In order to ensure that the 259 task diﬀiculty in this phase was balanced by participants’ working memory capability, we 260 used participants’ K estimate score at set size = 4 from the initially-presented visuospatial 261 short-term memory task - the last set size before performance began to drop off - to set a 262 criterion for the available set sizes in HR and LR trials. Specifically, if the K estimate at 263 set size 4 was <= 3, participants could choose an LR trial with a set size of 2 or an HR 264 trial with a set size of 6. If the K estimate at set size 4 was > 3, participants could choose 265 an LR trial with a set size of 4 or an HR trial with a set size of 8. Although participants 266 were told that the the number of points they gained in this phase was proportion to the 267 monetary reward they would receive, all participants received the same reward (a $5 CAD 268 Starbucks gift card) at the end of the study. Afterwards, participants were redirected to a 269 debriefing survey and received their course credit and monetary reward. 270 All analyses were conducted using R 4.3.1 “Beagle Scouts” (R Development Core 272 Team, 2011) through RStudio (Booth et al., 2018). 273 All analyses were conducted using R 4.3.1 “Beagle Scouts” (R Development Core 272 Team, 2011) through RStudio (Booth et al., 2018). PREDICTING COGNITIVE EFFORT 273 The primary predictor variables in our study were 1) working memory ability, 274 operationalized as a participant’s K estimate at a set size of 4; 2) depressive traits, 275 operationalized as a participant’s BDI proportion score; 3) chronic stress levels, 276 operationalized as a participant’s PSS score; and 4) reward sensitivity, operationalized as a 277 operationalized as a participant’s PSS score; and 4) reward sensitivity, operationalized as a 277 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 14 participant’s mean excitement before playing for tickets in the monetary incentive delay 278 task (Terpstra et al., 2023). As our overall depression score measures were of more 279 translational interest than anhedonia, given the role of depression in downweighing effort 280 deployment and reward anticipation in both rodents (Slattery & Cryan, 2017) and humans 281 (Grahek et al., 2019), we focus on reporting this measure as opposed to the anhedonia 282 measure we also collected. The primary dependent variable in our study was 1) proportion 283 of HR trials chosen in the reward phase of the task; operationalized as whether participants 284 chose the HR option for more or less than 70% of all trials in the reward phase. We further 285 examined 2) accuracy, operationalized as the proportion of correct responses (hits and 286 correct rejections) in the effort choice (reward) phase of the task and 3) choice latency, 287 operationalized as the time in seconds until participants selected a diﬀiculty level on each 288 trial in the effort choice phase of the task. Finally, drift diffusion model parameters of drift 289 rate, starting point, boundary separation, and non-decision time (Hales et al., 2024) during 290 choices in the effort choice (reward) phase of the task were additional outcome variables. 291 To evaluate whether participant motivation to choose high effort trials for high reward 292 significantly explained performance (accuracy in the reward phase of the task) and choice 293 latency (time until participants selected a diﬀiculty level in the reward phase), we classified 294 participants into two categories according to the criteria discussed in Silveira et al. (2021): 295 participants who chose the HR option for more than 70% of all trials in the reward phase 296 were in the high effort preference group, while participants who chose the HR option for 297 less than or equal to 70% of all trials in the reward phase were in the low effort preference 298 group. This grouping was chosen to ensure continuity with comparable cognitive effort 299 studies in rodents (Hales et al., 2024; Hosking et al., 2016; Silveira et al., 2021). 300 For our analyses, we first conducted a series of t-tests to evaluate sex differences in 301f ( participant’s mean excitement before playing for tickets in the monetary incentive delay 278 task (Terpstra et al., 2023). PREDICTING COGNITIVE EFFORT As our overall depression score measures were of more 279 translational interest than anhedonia, given the role of depression in downweighing effort 280 deployment and reward anticipation in both rodents (Slattery & Cryan, 2017) and humans 281 (Grahek et al., 2019), we focus on reporting this measure as opposed to the anhedonia 282 measure we also collected. The primary dependent variable in our study was 1) proportion 283 of HR trials chosen in the reward phase of the task; operationalized as whether participants 284 chose the HR option for more or less than 70% of all trials in the reward phase. We further 285 examined 2) accuracy, operationalized as the proportion of correct responses (hits and 286 correct rejections) in the effort choice (reward) phase of the task and 3) choice latency, 287 operationalized as the time in seconds until participants selected a diﬀiculty level on each 288 trial in the effort choice phase of the task. Finally, drift diffusion model parameters of drift 289 rate, starting point, boundary separation, and non-decision time (Hales et al., 2024) during 290 choices in the effort choice (reward) phase of the task were additional outcome variables. 291 To evaluate whether participant motivation to choose high effort trials for high reward 292 significantly explained performance (accuracy in the reward phase of the task) and choice 293 latency (time until participants selected a diﬀiculty level in the reward phase), we classified 294 participants into two categories according to the criteria discussed in Silveira et al. (2021): 295 participants who chose the HR option for more than 70% of all trials in the reward phase 296 were in the high effort preference group, while participants who chose the HR option for 297 less than or equal to 70% of all trials in the reward phase were in the low effort preference 298 group. This grouping was chosen to ensure continuity with comparable cognitive effort 299 studies in rodents (Hales et al., 2024; Hosking et al., 2016; Silveira et al., 2021). 300 For our analyses, we first conducted a series of t-tests to evaluate sex differences in 301 participant’s mean excitement before playing for tickets in the monetary incentive delay 278 task (Terpstra et al., 2023). PREDICTING COGNITIVE EFFORT CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 15 For sex difference comparisons, only male and female participants were included as we were 305 under-powered to report sex differences from those reporting sex as “other”. Next, we 306 divided participants into high and low effort groups to ensure translational comparability 307 with existing rodent work. As in Cocker et al. (2012), participants who chose the HR 308 option for more than 70% of all trials in the reward phase of the task were in the high 309 effort preference group, while those who chose the HR option for 70% or less of all trials in 310 this phase were in the low effort preference group. We first conducted a binomial linear 311 regression to determine whether sex, working memory ability, depressive traits, chronic 312 stress levels, and reward anticipation significantly predicted whether a participant was in 313 the high or low effort group in terms of their effort choices. We then conducted two 2x2 314 within-between ANOVAs (Trial type x Group) using anova_test through the rstatix 315 package (Kassambara, 2023) to evaluate whether accuracy or choice latency significantly 316 differed by trial effort level (HR vs. LR) and motivation group (high effort vs. low effort 317 groups). We examined accuracy and choice latency to evaluate whether participants were 318 matched for performance and time spent selecting a trial (Cocker et al., 2012), regardless of 319 how many high vs. low effort trials they selected. We then ran two multi-level models 320 through the lmerTest package (Kuznetsova, Brockhoff, & Christensen, 2017) to evaluate 321 whether sex, working memory ability, depressive traits, chronic stress levels, reward 322 anticipation, and effort level significantly predict choice latency or accuracy. Lastly, in 323 order to evaluate whether the high and low effort groups differed in response strategies and 324 biases towards selecting HR vs. LR trials, we fit a hierarchical Bayesian drift diffusion 325 model, adapted from Ratcliff (1978) and run for 2000 iterations with 1000 warmup 326 iterations and 4 Markov chains for Monte Carlo sampling, to data from all participants 327 using the hBayesDM package in R (Ahn, Haines, & Zhang, 2017). We used this model to 328 compare drift rate, starting point, boundary separation, and non-decision time between 329 participants who chose high levels of effort vs. those who chose low levels of effort. PREDICTING COGNITIVE EFFORT As our overall depression score measures were of more 279 translational interest than anhedonia, given the role of depression in downweighing effort 280 deployment and reward anticipation in both rodents (Slattery & Cryan, 2017) and humans 281 (Grahek et al., 2019), we focus on reporting this measure as opposed to the anhedonia 282 measure we also collected. The primary dependent variable in our study was 1) proportion 283 of HR trials chosen in the reward phase of the task; operationalized as whether participants 284 chose the HR option for more or less than 70% of all trials in the reward phase. We further 285 examined 2) accuracy, operationalized as the proportion of correct responses (hits and 286 correct rejections) in the effort choice (reward) phase of the task and 3) choice latency, 287i operationalized as the time in seconds until participants selected a diﬀiculty level on each 288 trial in the effort choice phase of the task. Finally, drift diffusion model parameters of drift 289 rate, starting point, boundary separation, and non-decision time (Hales et al., 2024) during 290 choices in the effort choice (reward) phase of the task were additional outcome variables. 291 To evaluate whether participant motivation to choose high effort trials for high reward 292 significantly explained performance (accuracy in the reward phase of the task) and choice 293 latency (time until participants selected a diﬀiculty level in the reward phase), we classified 294 participants into two categories according to the criteria discussed in Silveira et al. (2021): 295 participants who chose the HR option for more than 70% of all trials in the reward phase 296 were in the high effort preference group, while participants who chose the HR option for 297 less than or equal to 70% of all trials in the reward phase were in the low effort preference 298 group. This grouping was chosen to ensure continuity with comparable cognitive effort 299 studies in rodents (Hales et al., 2024; Hosking et al., 2016; Silveira et al., 2021). 300 For our analyses, we first conducted a series of t-tests to evaluate sex differences in 301 BDI and PSS scores. Past work suggests sex differences in depressive traits (Altemus, 302 Sarvaiya, & Neill Epperson, 2014; Forys et al., 2023) and chronic stress levels (Watt et al., 303 2017), with women presenting with higher depression and chronic stress scores than men. 304 . Depressive traits and chronic stress levels 339 Depressive traits and chronic stress levels 339 Summary statistics for depressive traits (BDI) and chronic stress levels (PSS) can 340 be found in Table 1. After correcting for multiple comparisons using the Bonferroni 341 method, women had significantly higher depressive trait scores (t(261.23) = 5.35, p = 342 <0.001, d = 0.45) (Fig. 2A) and chronic stress scores (t(270.03) = 4.93, p <0.001, d = 343 0.40) than men (Fig. 2B). However, men and women did not significantly differ on 344 anhedonia scores (t(253.79) = -2.05, p = 0.373, d = -0.17) (Fig. 2C) or levels of mean 345 reward anticipation (t(233.53) = -2.50, p = 0.118, d = -0.23) (Fig. 2D). 346 PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 16 (selecting an HR or LR trial). The starting point captures the initial bias, at the beginning 332 of each trial, that participants have towards selecting an HR over an LR trial. The 333 boundary separation captures a trade off between speed and selecting the HR trial as 334 opposed to the LR trial. Lastly, non-decision time captures the part of choice latency that 335 is not related to cognitive effort decisions, such as the time required to execute a motor 336 response upon making a decision. 337 Results 338 PREDICTING COGNITIVE EFFORT CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT The 330 drift rate captures the rate at which participants drift towards a decision boundary 331 For sex difference comparisons, only male and female participants were included as we were 305 under-powered to report sex differences from those reporting sex as “other”. Next, we 306 divided participants into high and low effort groups to ensure translational comparability 307 with existing rodent work. As in Cocker et al. (2012), participants who chose the HR 308 option for more than 70% of all trials in the reward phase of the task were in the high 309 effort preference group, while those who chose the HR option for 70% or less of all trials in 310 this phase were in the low effort preference group. We first conducted a binomial linear 311 regression to determine whether sex, working memory ability, depressive traits, chronic 312 stress levels, and reward anticipation significantly predicted whether a participant was in 313 the high or low effort group in terms of their effort choices. We then conducted two 2x2 314 within-between ANOVAs (Trial type x Group) using anova_test through the rstatix 315 package (Kassambara, 2023) to evaluate whether accuracy or choice latency significantly 316 differed by trial effort level (HR vs. LR) and motivation group (high effort vs. low effort 317 groups). We examined accuracy and choice latency to evaluate whether participants were 318 matched for performance and time spent selecting a trial (Cocker et al., 2012), regardless of 319 how many high vs. low effort trials they selected. We then ran two multi-level models 320 through the lmerTest package (Kuznetsova, Brockhoff, & Christensen, 2017) to evaluate 321 whether sex, working memory ability, depressive traits, chronic stress levels, reward 322 anticipation, and effort level significantly predict choice latency or accuracy. Lastly, in 323 order to evaluate whether the high and low effort groups differed in response strategies and 324 biases towards selecting HR vs. LR trials, we fit a hierarchical Bayesian drift diffusion 325 model, adapted from Ratcliff (1978) and run for 2000 iterations with 1000 warmup 326 iterations and 4 Markov chains for Monte Carlo sampling, to data from all participants 327 using the hBayesDM package in R (Ahn, Haines, & Zhang, 2017). We used this model to 328 compare drift rate starting point boundary separation and non-decision time between 329 . Predictors of overall high vs. low effort choices 347 Our primary research question focused on factors that predict the likelihood of 348 choosing high vs low effort/reward options (Table 2; Fig. 3). Here, results of the binomial 349 regression showed that only visuospatial working memory (K estimate) score significantly 350 predicted whether a participant was in the high effort or the low effort group (z = 2.26, p 351 = 0.024, 𝑒𝛽= 1.24). Participants with higher working memory ability had a higher 352 probability of being in the high effort group. Sex, depressive traits, chronic stress levels, 353 and reward anticipation did not significantly predict whether participants were in the low 354 vs. high effort groups. 355 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 17 High Effort Low Effort 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0 1 2 BDI proportion score Density A High Effort Low Effor 0 10 20 30 0 10 20 30 0.000 0.025 0.050 0.075 PSS score B 0 10 20 30 40 50 0 10 20 30 40 50 0.00 0.01 0.02 0.03 0.04 DARS score Density C 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D Sex Female Male Figure 2 Distribution of (A) depressive trait (BDI) proportion scores (score divided by total possible score), (B) perceived stress (PSS) scores, (C) anhedonia (DARS) scores, and (D) mean reward anticipa- tion scores by sex. BDI = Beck Depression Inventory, PSS = Perceived Stress Scale, DARS = Dimensional Anhedonia Rating Scale. Predictors of overall high vs. low effort choices 347 0 High Effort Low Effor 0 10 20 30 0 10 20 30 0.000 0.025 0.050 0.075 PSS B High Effort Low Effort 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0 1 2 BDI ti Density A High Effort Low Effor 0 10 20 30 0 10 20 30 0.000 0.025 0.050 0.075 PSS B High Effort Low Effort 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 0 1 2 BDI proportion score Density A B A BDI proportion score PSS score 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D p p 0 10 20 30 40 50 0 10 20 30 40 50 0.00 0.01 0.02 0.03 0.04 DARS score Density C D C Figure 2 Distribution of (A) depressive trait (BDI) proportion scores (score divided by total possible score), (B) perceived stress (PSS) scores, (C) anhedonia (DARS) scores, and (D) mean reward anticipa- tion scores by sex. BDI = Beck Depression Inventory, PSS = Perceived Stress Scale, DARS = Dimensional Anhedonia Rating Scale. g Distribution of (A) depressive trait (BDI) proportion scores (score divided by total possible score), (B) perceived stress (PSS) scores, (C) anhedonia (DARS) scores, and (D) mean reward anticipa- tion scores by sex. BDI = Beck Depression Inventory, PSS = Perceived Stress Scale, DARS = Dimensional Anhedonia Rating Scale. Accuracy 356 In all ANOVAs reported below all p-values were corrected for multiple comparisons 357 using Greenhouse-Geiser correction. A 2x2 within-between ANOVA (Group x Trial Type; 358 Fig. 4A) revealed a main effect of Trial Type on accuracy such that participants showed 359 poorer performance on the more diﬀicult HR trials compared to the easier LR trials (F(1, 360 273) = 724.10, p <0.001, 𝛾= 0.515). This was qualified by a Trial Type x Group 361 interaction such that participants in the high effort group performed better than those in 362 the low effort group on LR but not HR trials (F(1, 273) = 11.81, p <0.001, 𝛾= 0.017). 363 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. Predictors of overall high vs. low effort choices 347 ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 18 0.00 0.25 0.50 0.75 1.00 -2 0 2 4 Working memory score (K estimate) Probability of being in High Effort group Figure 3 Binomial logistic regression predicting whether participants were in the high effort vs. low effort group. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. 0.00 0.25 0.50 0.75 1.00 -2 0 2 4 Working memory score (K estimate) Probability of being in High Effort group Working memory score (K estimate) Working memory score (K estimate) Figure 3 Binomial logistic regression predicting whether participants were in the high effort vs. low effort group. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. N.S. * * 0.6 0.8 1.0 High Low Selected effort level Accuracy A N.S. * * 0.1 1.0 10.0 High Low Selected effort level Log choice latency (s) B Motivation group High Effort Low Effort Figure 4 Accuracy and choice latency in the choice phase of the change detection task by motivation group (high effort vs. low effort group). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. N.S. * * 0.6 0.8 1.0 High Low Selected effort level Accuracy A Motivation group N.S. * * 0.1 1.0 10.0 High Low Selected effort level Log choice latency (s) B Hi h Eff L Eff A . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Figure 4 g Accuracy and choice latency in the choice phase of the change detection task by motivation group (high effort vs. low effort group). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. g Accuracy and choice latency in the choice phase of the change detection task by motivation group (high effort vs. low effort group). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 19 PREDICTING COGNITIVE EFFORT Table 2 Binomial logistic regression predicting whether participants were in the high effort vs. low effort group. High effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Binomial logistic regression: High vs. low effort group (Intercept) 1.14+ (0.60) Working memory ability (K estimate) 0.21* (0.09) BDI (prop. score) −0.78 (0.96) PSS 0.00 (0.03) Reward anticipation 0.00 (0.07) Num.Obs. 429 AIC 419.7 BIC 440.0 Log.Lik. −204.833 RMSE 0.39 + p < 0.1, * p < 0.05, p < 0.01, * p < 0.001 Binomial logistic regression: High vs. low effort group PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 20 Additional predictors of accuracy and choice latency 373 We additionally evaluated predictors of accuracy and choice latency from individual 374 differences in participant visuospatial short-term memory capacity, mood disorder score, 375 and reward sensitivity by sex. A multi-level model analysis revealed that the effort level 376 required in the selected trial significantly predicted accuracy across all trials such that 377 accuracy was higher on low effort (LR) trials compared to high effort (HR) trials. 378 Importantly, of the predictors of interest, only visuospatial short-term memory score 379 predicted acccuracy, such that higher short-term memory ability was associated with 380 higher accuracy (Table 3; Fig. 5A). Trial effort level also predicted choice latency, with 381 slower choices on low relative to high effort trials. Depression scores also predicted choice 382 latency such that higher depression scores predicted faster choices across all trials (Table 4; 383 Fig. 5B) Sex, chronic stress scores, and reward anticipation did not significantly predict 384 accuracy or choice latency. 385 We additionally evaluated predictors of accuracy and choice latency from individual 374 differences in participant visuospatial short-term memory capacity, mood disorder score, 375 and reward sensitivity by sex. A multi-level model analysis revealed that the effort level 376 required in the selected trial significantly predicted accuracy across all trials such that 377 accuracy was higher on low effort (LR) trials compared to high effort (HR) trials. 378 Importantly, of the predictors of interest, only visuospatial short-term memory score 379 predicted acccuracy, such that higher short-term memory ability was associated with 380 higher accuracy (Table 3; Fig. 5A). Trial effort level also predicted choice latency, with 381 slower choices on low relative to high effort trials. Depression scores also predicted choice 382 latency such that higher depression scores predicted faster choices across all trials (Table 4; 383 Fig. 5B) Sex, chronic stress scores, and reward anticipation did not significantly predict 384 accuracy or choice latency. 385 PSS Choice latency 364 A 2x2 within-between ANOVA (Group x Trial Type; Fig. 4B) revealed a main effect 365 of Group such that participants were slower to choose HR trials compared to LR trials 366 (F(1, 273) = 37.50), p <0.001, 𝛾= 0.063). Furthermore, high effort participants were 367 slower than low effort participants to make choices across all trial types (F(1, 273) = 368 10.10), p = 0.002, 𝛾= 0.019). There was also an Trial Type x Group interaction, such that 369 participants in the high effort group took longer to select LR trials than HR trials, while 370 choice latency between high effort and low effort groups did not differ on HR trials (F(1, 371 273) = 37.50), p <0.001, 𝛾= 0.021). 372 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Drift diffusion model 386 To evaluate group differences in evidence accumulation and bias towards deploying 387 higher amounts of cognitive effort in decision making when choosing HR or LR trials, we fit 388 a hierarchical Bayesian drift diffusion model (Ahn et al., 2017) to choice (HR vs. LR trial 389 selected) and corresponding choice latency data on each trial from all participants who had 390 at least one correct response on an HR and an LR trial (N = 136, 𝑛𝐻𝑖𝑔ℎ𝑒𝑓𝑓𝑜𝑟𝑡= 94, 391 𝑛𝐿𝑜𝑤𝑒𝑓𝑓𝑜𝑟𝑡= 42). We then evaluated this overall model fitting data from these participants 392 to explore group-level differences in whether participants selected HR or LR trials more 393 often. For each participant in the high effort and low effort groups, we evaluated four 394 parameters: drift rate (speed of evidence accumulation in deciding on an HR vs. LR trial), 395 starting point (bias towards HR vs. LR trial), boundary separation (extent to which speed 396 and effort choice are traded off between HR and LR trials), and non-decision time (portion 397 of choice latency time unrelated to trial choice, such as time for motor response) (Fig. 6). 398 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. Drift diffusion model 386 ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 21 High Effort -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1.0 Accuracy A High Effort Low Effort -2 0 2 4 -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1.0 Accuracy A -2 0 2 4 -2 0 2 4 0.1 1.0 10.0 Working memory score (K estimate) Log choice latency (s) B Figure 5 (A) Accuracy and (B) choice latency in the reward phase of the change detection task by working memory ability (K estimate). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. A -2 0 2 4 0.1 1.0 10.0 Working memor Log choice latency (s) B Figure 5 B (A) Accuracy and (B) choice latency in the reward phase of the change detection task by working memory ability (K estimate). The choice latency plot has been log-transformed on the y-axis to more clearly show small choice latency values. These parameters capture individual differences in the speed and motivation to make a 399 decision about deploying more or less cognitive effort for reward. After Bonferroni 400 correction for multiple comparisons, we found that high effort group participants had a 401 starting point closer to the HR trial decision boundary (t(71.58) = 3.81, p = 0.003, d = 402 0.74) and had wider decision boundaries (t(68.92) = 4.40, p <0.001, d = 0.87) than low 403 effort group participants. However, high and low effort group participants did not 404 significantly differ in drift rate (t(78) = 1.88, p = 0.58, d = 0.35) or in non-decision time 405 (t(106.11) = 2.17, p = 0.29, d = 0.36). These findings suggest that participants who were 406 in the high effort group required more evidence to select high effort trials, and were more 407 These parameters capture individual differences in the speed and motivation to make a 399 decision about deploying more or less cognitive effort for reward. After Bonferroni 400 correction for multiple comparisons, we found that high effort group participants had a 401 starting point closer to the HR trial decision boundary (t(71.58) = 3.81, p = 0.003, d = 402 0.74) and had wider decision boundaries (t(68.92) = 4.40, p <0.001, d = 0.87) than low 403 effort group participants. Drift diffusion model 386 However, high and low effort group participants did not 404 significantly differ in drift rate (t(78) = 1.88, p = 0.58, d = 0.35) or in non-decision time 405 significantly differ in drift rate (t(78) = 1.88, p = 0.58, d = 0.35) or in non-decision time 405 (t(106.11) = 2.17, p = 0.29, d = 0.36). These findings suggest that participants who were 406 in the high effort group required more evidence to select high effort trials, and were more 407 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 22 strongly biased towards selecting high effort/high reward trials, than those in the low effort 08 group. 09 strongly biased towards selecting high effort/high reward trials, than those in the low effort 408 group. 409 group. 9 NS. -1 0 1 2 3 High Effort Low Effort Drift rate A * 0.0 0.5 1.0 1.5 High Effort Low Effort Starting point B * 0 1 2 3 High Effort Low Effort Motivation group Boundary separation C NS. 0.00 0.25 0.50 0.75 1.00 High Effort Low Effort Motivation group Non-decision time (s) D Figure 6 Outputs of a hierarchical Bayesian drift diffusion model fit to all participants in the effort choice task, divided by effort group. High effort participants had a significantly higher (B) starting point and (C) boundary separation compared to low-effort participants, but did not significantly differ on (A) drift rate or (D) non-decision time compared to low-effort participants. NS. -1 0 1 2 3 High Effort Low Effort Drift rate A * 0.5 1.0 1.5 Starting point B B A 1 2 Drift rate D Outputs of a hierarchical Bayesian drift diffusion model fit to all participants in the effort choice task, divided by effort group. Drift diffusion model 386 High effort participants had a significantly higher (B) starting point and (C) boundary separation compared to low-effort participants, but did not significantly differ on (A) drift rate or (D) non-decision time compared to low-effort participants. Discussion In this study, we investigated whether working memory performance, depressive 411 traits, and chronic stress levels influence people’s motivation and ability to consistently 412 deploy cognitive effort for reward. Results showed that greater working memory ability 413 significantly predicted the likelihood of systematically choosing high effort/high reward 414 trials (whether the participant was classified in the high vs. low effort group), whereas 415 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 24 deploying more effort or wanting higher rewards. Furthermore, they could indicate that 442 impulsiveness to select a HR or LR trial could differ between humans and rodents. This 443 could be explained by differences in the strength of reinforcement provided by the reward, 444 which was a secondary reinforcement (points towards a monetary reward) in our task but a 445 primary reinforcement (sugar pellets) in the rCET (Cocker et al., 2012; Hales et al., 2024), 446 or by other differences between our task and the rCET as well as between humans and 447 rodents in terms of how reward motivation is processed. 448 Additionally, previous studies with the rCET have shown no overall accuracy 449 differences between high and low effort group participants across either LR or HR trials 450 (Hosking et al., 2016; Silveira et al., 2020). However, in a recent, larger analysis, Hales et al. 451 (2024) showed that rodents choosing more HR trials were slightly more accurate on HR 452 trials, while rodents choosing more LR trials were slightly more accurate on LR trials. 453 Additionally, previous studies with the rCET have shown no overall accuracy 449 differences between high and low effort group participants across either LR or HR trials 450 (Hosking et al., 2016; Silveira et al., 2020). However, in a recent, larger analysis, Hales et al. 451 (2024) showed that rodents choosing more HR trials were slightly more accurate on HR 452 trials, while rodents choosing more LR trials were slightly more accurate on LR trials. 453 Although participants practiced task contingencies ahead of the main choice phase and the 454 task diﬀiculty was calibrated to working memory ability, individual differences in working 455 memory - which predicted whether a participant selected high or low effort trials more 456 often - still predicted these performance differences in humans. Furthermore, chronic stress 457 - which can negatively impact availability of executive resources - did not explain accuracy 458 or choice latency in this task. Thus, participants’ willingness to deploy cognitive effort in 459 our task was explained more by individual differences in executive function (working 460 memory capability) (Sandra & Otto, 2018) rather than affective factors or the effects of 461 affective factors on executive function (Slattery & Cryan, 2017). PREDICTING COGNITIVE EFFORT It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 23 chronic stress, depression trait levels, and reward sensitivity were not predictive of effort 416 deployment choices. Drift diffusion modeling further indicated that those who were 417 categorized as high-effort participants required less evidence to select high effort/high 418 reward trials, and were more strongly biased towards selecting these trials, than low-effort 419 participants. Furthermore, working memory ability significantly predicted accuracy, while 420 depressive traits predicted choice latency; both factors significantly differed depending on 421 the effort required in the trial. 422 These results partially recapitulate findings from rodent research using the rCET 423 regarding differences in performance and decision-making biases between those who select 424 high effort trials more often, and those who select them less often. As in Cocker et al. 425 (2012), participants who chose high effort trials over 70% of the time performed better than 426 those who chose high effort trials 70% or less of the time, on LR - although not HR - trials. 427 This finding could be explained by high effort group participants being less sensitive to 428 rewards and punishments than low effort group participants, leading them to act 429 sub-optimally and continue deploying effort even when they are not gaining more rewards 430 from doing so. 431 Furthermore, our findings align with past work in rodents (Hales et al., 2024) in 432 that our drift diffusion model results showed wider decision boundaries between HR and 433 LR trials for high effort group compared to low effort group participants, as well as a 434 starting point closer to HR trials for high effort group participants. Additionally, as with 435 rodents, participants in the low effort group made faster choices than those in the high 436 effort group on LR trials. However, participants in our study did not differ in drift rate 437 between groups - while rodents in the equivalent high effort group had steeper drift rates 438 on correct responses (Hales et al., 2024). Taken together, these findings could suggest 439 similarities between humans and rodents in the influence of cognitive capabilities on 440 performance when deploying more cognitive effort, as well as similarities in biases towards 441 PREDICTING COGNITIVE EFFORT 24 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. PREDICTING COGNITIVE EFFORT Note that although we did 462 not measure executive control as a global construct in this study, visuospatial working 463 memory ability is typically described as a process that falls withing the suite of executive 464 functions (Miyake & Friedman, 2012). 465 Although participants practiced task contingencies ahead of the main choice phase and the 454 task diﬀiculty was calibrated to working memory ability, individual differences in working 455 memory - which predicted whether a participant selected high or low effort trials more 456 often - still predicted these performance differences in humans. Furthermore, chronic stress 457 - which can negatively impact availability of executive resources - did not explain accuracy 458 or choice latency in this task. Thus, participants’ willingness to deploy cognitive effort in 459 our task was explained more by individual differences in executive function (working 460 memory capability) (Sandra & Otto, 2018) rather than affective factors or the effects of 461 affective factors on executive function (Slattery & Cryan, 2017). Note that although we did 462 not measure executive control as a global construct in this study, visuospatial working 463 memory ability is typically described as a process that falls withing the suite of executive 464 functions (Miyake & Friedman, 2012). 465 Our findings suggest that choices to deploy cognitive effort for reward are primarily 466 driven by participants’ cognitive capacity (as measured by visuospatial working memory 467 PREDICTING COGNITIVE EFFORT 25 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 26 as reflected in increased choice latency among high effort group participants for LR trials - 495 whereas selecting a high proportion of low effort trials could entail more model-free, rapid, 496 and random decisions. Furthermore, participants with higher levels of depressive traits 497 responded significantly faster across all trials, a result that runs counter to past work 498 suggesting increased choice latency in reward tasks in rodents (Hales et al., 2023) or 499 humans (Di Schiena, Luminet, Chang, & Philippot, 2013) with depressive-like traits. 500 as reflected in increased choice latency among high effort group participants for LR trials - 495 whereas selecting a high proportion of low effort trials could entail more model-free, rapid, 496 and random decisions. Furthermore, participants with higher levels of depressive traits 497 responded significantly faster across all trials, a result that runs counter to past work 498 suggesting increased choice latency in reward tasks in rodents (Hales et al., 2023) or 499 humans (Di Schiena, Luminet, Chang, & Philippot, 2013) with depressive-like traits. 500 However, after Bonferroni correction, depressive traits and choice latency were not 501 significantly correlated across all participants and trials (t(702) = -1.21, p = 1, r = -0.05). 502 As depressive traits did not impact accuracy, these findings could instead suggest a lack of 503 engagement with decision-making in the task characterized by higher levels of automatic 504 responding given high levels of depressive traits (Teachman, Joormann, Steinman, & 505 Gotlib, 2012). 506 We found no significant sex differences in performance or choice latency in our 507 study. Women generally present more with depression than men (Parker & Brotchie, 2010) 508 and are also more impacted by chronic stress across the lifespan (Hodes & Epperson, 2019). 509 However, the results of our study were primarily driven by visuospatial short-term memory 510 ability, where sex differences are overall smaller and women may be better than men at 511 memory for location (Voyer, Voyer, & Saint-Aubin, 2017) - a relevant measure in our task. 512 Our undergraduate psychology student sample strongly skewed female; in future studies, it 513 would be important to obtain a more balanced sample in order to further explore potential 514 sex differences. 515 There are a number of caveats we must consider when interpreting these results. 516 First, the study was conducted fully online on participants’ own devices. PREDICTING COGNITIVE EFFORT CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 25 ability) and not by depressive traits, chronic stress, or reward sensitivity, in participants 468 within a non-clinical range of depression traits and chronic stress. Potentially, the 469 aversiveness of expending cognitive effort for particpants with lower working memory 470 capacity may have overridden overall sensitivity for reward. Additionally, although our 471 population exhibited a wide range of depression scores - including those above clinical 472 thresholds (Wang & Gorenstein, 2013) - this population was non-clinical. In contrast, 473 much work on the impact of reward sensitivity on cognitive flexibility and related 474 constructs has focused on participants who were clinically diagnosed with depression (Alloy, 475 Olino, Freed, & Nusslock, 2016; Terpstra et al., 2023) or on rodents in which chronic stress 476 was directly induced (Watt et al., 2017). In order to obtain the highest accuracy and 477 reward possible, participants may have titrated the task’s diﬀiculty - through choosing a 478 higher proportion of LR trials - to a level at which they can consistently complete the task 479 and at which the value of their effort was greatest (Shenhav et al., 2013). 480 Choice latency indicates how long participants spend when deciding which effort 481 level to choose. High effort group participants, who generally selected HR rewards more 482 often and who were more biased towards selecting HR trials than low effort group 483 participants, appear to have spent more time deliberating before selecting a LR trial than 484 low effort group participants. They may have been trading off the benefits and costs of 485 selecting a LR trial more consciously than low effort group participants were (Pruessner et 486 al., 2020; Sandra & Otto, 2018), even as they performed significantly better than low effort 487 group participants on LR but not HR trials. Indeed, our drift diffusion model results 488 suggest that high effort group participants had to overcome a bias towards HR trials in 489 order to select LR trials. This process could be driven by adjustments to effort deployment 490 strategies based on performance, as suggested by recent interpretations of the Expected 491 Value of Control theory (Shenhav, Prater Fahey, & Grahek, 2021). Through this 492 framework, choosing a high proportion of high effort trials could be seen as optimizing 493 effort deployment and making considered, model-based, decisions based on performance - 494 PREDICTING COGNITIVE EFFORT 26 . PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 27 than originally intended for some participants. However, participant performance was still 521 near ceiling on LR trials and very high on HR trials, so this was unlikely to have a large 522 effect. Second, the set sizes used on the change detection task were modified from those 523 originally used in Luck and Vogel (1997). The set size of 10 was removed from the pool of 524 trials as pilot participants had very low performance at this set size. Although this change 525 reduced the possible variety of trials used, participants still reported a maximum set size of 526 8 to be highly challenging. Third, the method used to calculate the K estimate slightly 527 differed between the task - used to set the criterion level - and our analysis - used to 528 establish the K estimate for each participant. In the task, the K estimate was calculated 529 based on the current hit and false alarm rate after every trial, with the final K estimate 530 being used to set the criterion for the choice phase of the task. In our analysis, we 531 calculated the K estimate based on the full hit and false alarm rate at each set size. 532 Because of rounding differences in these calculations, the K estimate values used in our 533 analysis differ from those in our task by M = 0.01, SD = 0.63. We consider this difference 534 to be small enough as to not have an effect on the diﬀiculty of the task. Last, differences in 535 performance that are explained by the K estimate could also be explained by the 536 differences in set sizes available associated with being above or below the K criterion of 3 537 on set size = 4. However, the set sizes at each level of criterion were established with 538 piloting such that the diﬀiculty of each criterion level was matched as closely as possible. 539 Future studies could make further use of information about effort deployment 540 choices that can only be captured in human studies. In particular, an expanding field of 541 research uses qualitative approaches to investigate why participants make the judgments 542 they do about effort deployment, as well as ask about participant experiences of task 543 diﬀiculty and the value of rewards vs. PREDICTING COGNITIVE EFFORT While this 517 allowed us to obtain a larger and better powered sample, differences in screen size and 518 background distractions could have impacted participants’ ability to encode the positions 519 and colours shapes on each trial. These differences may have made the task more diﬀicult 520 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 28 traditional self-report measures. This approach would also allow us to further explore how 548 rodent-based constructs of cognitive effort deployment are reflected in human 549 decision-making. Additionally, exploring how the brain represents information about effort 550 task diﬀiculty - especially in regions relevant to cognitive control like the dorsal anterior 551 cingulate cortex (Shenhav et al., 2016; Yee et al., 2021) - through fMRI would help us 552 better understand how motivational states in cognitive effort decision-making are reflected 553 in neural circuitry. 554 traditional self-report measures. This approach would also allow us to further explore how 548 rodent-based constructs of cognitive effort deployment are reflected in human 549 Our study suggests that, as in rodents, significant individual differences exist in the 555 human tendency to choose harder but more rewarding options in cognitive effort tasks. 556 Our study suggests that, as in rodents, significant individual differences exist in the 555 human tendency to choose harder but more rewarding options in cognitive effort tasks. 556 Using a visuospatial working memory task, we show that individual differences in accuracy 557 and choice latency are primarily driven not by depressive traits or chronic stress - as has 558 previously been shown in the rodent literature - but instead by working memory capability. 559 These findings may help to inform clinical interventions aimed at increasing motivation to 560 seek rewards and engage in work in everyday life, and illustrate the importance of 561 translating and extending rodent work with human-specific measures. 562 PREDICTING COGNITIVE EFFORT the effort required to obtain them (Vásquez-Rosati, 544 Montefusco-Siegmund, López, & Cosmelli, 2019). Understanding the phenomenology of 545 cognitive effort deployment choices would help us extend our understanding of how and 546 why we make decisions to deploy more or less cognitive effort for reward, beyond 547 . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT Data and code availability 563 All task code, stimuli, data and code used to generate this manuscript and the 564 figures are available at https://osf.io/c4h7s/. 565 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint 29 PREDICTING COGNITIVE EFFORT . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint References 566 Booth, D. S., Szmidt-Middleton, H., King, N., Westbrook, M. J., Young, S. L., Kuo, 583 A., … Rstudio Team. (2018). RStudio: Integrated development for r. Nature. 584 https://doi.org/10.1108/eb003648 585 Bulus, M. (2023). {Pwrss}: Statistical power and sample size calculation tools. 586 Retrieved from https://CRAN.R-project.org/package=pwrss 587 Cocker, P. J., Hosking, J. G., Benoit, J., & Winstanley, C. A. (2012). Sensitivity to 588 Bulus, M. (2023). {Pwrss}: Statistical power and sample size calculation tools. 586 Cocker, P. J., Hosking, J. G., Benoit, J., & Winstanley, C. A. (2012). Sensitivity to 588 Cognitive Effort Mediates Psychostimulant Effects on a Novel Rodent 589 Cost/Benefit Decision-Making Task. Neuropsychopharmacology, 37(8), 590 1825–1837. https://doi.org/10.1038/npp.2012.30 591 Culbreth A J Moran E K & Barch D M (2018) Effort-cost decision-making 592 Cocker, P. J., Hosking, J. G., Benoit, J., & Winstanley, C. A. (2012). Sensitivity to 588 Cognitive Effort Mediates Psychostimulant Effects on a Novel Rodent 589 Cost/Benefit Decision-Making Task. Neuropsychopharmacology, 37(8), 590 1825–1837. https://doi.org/10.1038/npp.2012.30 591 Culbreth, A. J., Moran, E. K., & Barch, D. M. (2018). Effort-cost decision-making 592 Culbreth, A. J., Moran, E. K., & Barch, D. M. (2018). Effort-cost decision-making 592 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 30 PREDICTING COGNITIVE EFFORT 30 in psychosis and depression: Could a similar behavioral deficit arise from 3 disparate psychological and neural mechanisms? Psychological Medicine, 48(6), 4 889–904. https://doi.org/10.1017/S0033291717002525 5 de Oliveira, João Roberto Ventura. (2023). Luck and Vogel Change Detection Task. 6 Retrieved from https://gitlab.pavlovia.org/joaorobertoventura/luck-and-vogel- 7 change-detection-task 8 de Oliveira, João Roberto Ventura. (2023). Luck and Vogel Change Detection Task. Retrieved from https://gitlab.pavlovia.org/joaorobertoventura/luck-and-vogel- change-detection-task Di Schiena, R., Luminet, O., Chang, B., & Philippot, P. (2013). Why are Depressive Individuals Indecisive? Different Modes of Rumination Account for Indecision in Non-clinical Depression. Cognitive Therapy and Research, 37(4), 713–724. https://doi.org/10.1007/s10608-012-9517-9 Forys, B. J., Tomm, R. J., Stamboliyska, D., Terpstra, A. R., Clark, L., 606 Chakrabarty, T., … Todd, R. M. (2023). Gender Impacts the Relationship 607 between Mood Disorder Symptoms and Effortful Avoidance Performance. 608 Eneuro, 10(2), ENEURO.0239–22.2023. 609 Eneuro, 10(2), ENEURO.0239–22.2023. 609 https://doi.org/10.1523/ENEURO.0239-22.2023 610 https://doi.org/10.1523/ENEURO.0239-22.2023 Frömer, R., Lin, H., Dean Wolf, C. K., Inzlicht, M., & Shenhav, A. (2021). 611 Expectations of reward and eﬀicacy guide cognitive control allocation. Nature 612 Communications, 12(1), 1030. https://doi.org/10.1038/s41467-021-21315-z 613 Fuentes-Claramonte, P., Ávila, C., Rodríguez-Pujadas, A., Ventura-Campos, N., 614 Bustamante, J. C., Costumero, V., … Barrós-Loscertales, A. (2015). Reward 615 Sensitivity Modulates Brain Activity in the Prefrontal Cortex, ACC and 616 Striatum during Task Switching. PLOS ONE, 10(4), e0123073. 617 https://doi.org/10.1371/journal.pone.0123073 618 Grahek, I., Musslick, S., & Shenhav, A. (2020). A computational perspective on the 619 Sensitivity Modulates Brain Activity in the Prefrontal Cortex, ACC and Striatum during Task Switching. PLOS ONE, 10(4), e0123073. Grahek, I., Musslick, S., & Shenhav, A. (2020). A computational perspective on the 619 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 31 Psychoneuroendocrinology, 147, 105953. Hodes, G. E., & Epperson, C. N. (2019). Sex differences in vulnerability and resilience to stress across the life span. Biological Psychiatry, 86(6), 421–432. https://doi.org/10.1016/j.biopsych.2019.04.028 Hosking, J. G., Cocker, P. J., & Winstanley, C. A. (2016). Prefrontal Cortical 637 Inactivations Decrease Willingness to Expend Cognitive Effort on a Rodent 638 Cost/Benefit Decision-Making Task. Cerebral Cortex, 26(4), 1529–1538. 639 https://doi.org/10.1093/cercor/bhu321 640 Ironside, M., Kumar, P., Kang, M.-S., & Pizzagalli, D. A. (2018). Brain mechanisms 641 mediating effects of stress on reward sensitivity. Current Opinion in Behavioral 642 Sciences, 22, 106–113. https://doi.org/10.1016/j.cobeha.2018.01.016 643 Kassambara, A. (2023). Rstatix: Pipe-friendly framework for basic statistical tests. 644 Retrieved from https://CRAN.R-project.org/package=rstatix 645 Kool, W., McGuire, J. T., Rosen, Z. B., & Botvinick, M. M. (2010). Decision 646 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 32 Levenstein, S., Prantera, C., Varvo, V., Scribano, M. L., Berto, E., Luzi, C., & 656 Andreoli, A. (1993). Development of the perceived stress questionnaire: A new 657 tool for psychosomatic research. Journal of Psychosomatic Research, 37(1), 658 19–32. https://doi.org/10.1016/0022-3999(93)90120-5 659 Luck, S. J., & Vogel, E. K. (1997). The capacity of visual working memory for 660 features and conjunctions. Nature, 390(6657), 279–281. 661 https://doi.org/10.1038/36846 662 Miyake, A., & Friedman, N. P. (2012). The Nature and Organization of Individual 663 Differences in Executive Functions: Four General Conclusions. Current 664 Directions in Psychological Science, 21(1), 8–14. 665 https://doi.org/10.1177/0963721411429458 666 Parker, G., & Brotchie, H. (2010). Gender differences in depression. International 667 Review of Psychiatry, 22(5), 429–436. 668 https://doi.org/10.3109/09540261.2010.492391 669 Pashler, H. (1988). Familiarity and visual change detection. Perception & 670 Psychophysics, 44(4), 369–378. https://doi.org/10.3758/BF03210419 671 Pashler, H. (1988). Familiarity and visual change detection. Perception & 670 Psychophysics, 44(4), 369–378. https://doi.org/10.3758/BF03210419 671 Peirce, J., Gray, J. R., Simpson, S., MacAskill, M., Höchenberger, R., Sogo, H., … 672 Lindeløv, J. K. (2019). PsychoPy2: Experiments in behavior made easy. 673 Psychophysics, 44(4), 369 378. https://doi.org/10.3758/BF03210419 671 Peirce, J., Gray, J. R., Simpson, S., MacAskill, M., Höchenberger, R., Sogo, H., … 672 Lindeløv, J. K. (2019). PsychoPy2: Experiments in behavior made easy. 673 Lindeløv, J. K. (2019). PsychoPy2: Experiments in behavior made easy. 673 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint 33 PREDICTING COGNITIVE EFFORT Behavior Research Methods, 51(1), 195–203. 674 https://doi.org/10.3758/s13428-018-01193-y 675 Pruessner, L., Barnow, S., Holt, D. V., Joormann, J., & Schulze, K. (2020). A 676 Cognitive Control Framework for Understanding Emotion Regulation Flexibility 677 Emotion, 20(1), 21–29. 678 R Development Core Team, R. (2011). R: A language and environment for 679 statistical computing. 680 Ratcliff, R. (1978). A Theory of Memory Retrieval. Psychological Review, 85(2), 681 59–108. 682 Rizvi, S. J., Quilty, L. . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT C., Sproule, B. A., Cyriac, A., Michael Bagby, R., & 683 Kennedy, S. H. (2015). Development and validation of the dimensional 684 anhedonia rating scale (DARS) in a community sample and individuals with 685 major depression. Psychiatry Research, 229(1-2), 109–119. 686 https://doi.org/10.1016/j.psychres.2015.07.062 687 Rouder, J., Morey, R., Morey, C., & Cowan, N. (2011). How to measure working 688 memory capacity in the change detection paradigm. Psychonomic Bulletin & 689 Review, 18, 324–330. https://doi.org/10.3758/s13423-011-0055-3 690 Sandra, D. A., & Otto, A. R. (2018). Cognitive capacity limitations and Need for 691 Cognition differentially predict reward-induced cognitive effort expenditure. 692 Cognition, 172, 101–106. https://doi.org/10.1016/j.cognition.2017.12.004 693 Scheggi, S., De Montis, M. G., & Gambarana, C. (2018). Making sense of rodent 694 models of anhedonia. International Journal of Neuropsychopharmacology, 695 21(11), 1049–1065. https://doi.org/10.1093/ijnp/pyy083 696 Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013). The expected value of 697 control: An integrative theory of anterior cingulate cortex function. Neuron, 698 https://doi.org/10.3758/s13428-018-01193-y Pruessner, L., Barnow, S., Holt, D. V., Joormann, J., & Schulze, K. (2020). A 676 Cognitive Control Framework for Understanding Emotion Regulation Flexibility. 677 Emotion, 20(1), 21–29. 678 R Development Core Team, R. (2011). R: A language and environment for statistical computing. Ratcliff, R. (1978). A Theory of Memory Retrieval. Psychological Review, 85(2), 59–108. Rizvi, S. J., Quilty, L. C., Sproule, B. A., Cyriac, A., Michael Bagby, R., & Rizvi, S. J., Quilty, L. C., Sproule, B. A., Cyriac, A., Michael Bagby, R., & Kennedy, S. H. (2015). Development and validation of the dimensional anhedonia rating scale (DARS) in a community sample and individuals with major depression. Psychiatry Research, 229(1-2), 109–119. https://doi.org/10.1016/j.psychres.2015.07.062 ( ) anhedonia rating scale (DARS) in a community sample and individuals with major depression. Psychiatry Research, 229(1-2), 109–119. https://doi.org/10.1016/j.psychres.2015.07.062 Rouder, J., Morey, R., Morey, C., & Cowan, N. (2011). How to measure working memory capacity in the change detection paradigm. Psychonomic Bulletin & Review, 18, 324–330. https://doi.org/10.3758/s13423-011-0055-3 Sandra, D. A., & Otto, A. R. (2018). Cognitive capacity limitations and Need for 691 Cognition differentially predict reward-induced cognitive effort expenditure. 692 Cognition, 172, 101–106. https://doi.org/10.1016/j.cognition.2017.12.004 693 Scheggi, S., De Montis, M. G., & Gambarana, C. (2018). Making sense of rodent 694 models of anhedonia. International Journal of Neuropsychopharmacology, 695 21(11), 1049–1065. https://doi.org/10.1093/ijnp/pyy083 696 Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013). The expected value of 697 control: An integrative theory of anterior cingulate cortex function. Neuron, 698 79(2), 217–240. PREDICTING COGNITIVE EFFORT https://doi.org/10.1016/j.neuron.2013.07.007 699 Shenhav, A., Cohen, J. D., & Botvinick, M. M. (2016). Dorsal anterior cingulate 700 Sandra, D. A., & Otto, A. R. (2018). Cognitive capacity limitations and Need for 691 Cognition differentially predict reward-induced cognitive effort expenditure. 692 Cognition, 172, 101–106. https://doi.org/10.1016/j.cognition.2017.12.004 693 Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013). The expected value of 697 control: An integrative theory of anterior cingulate cortex function. Neuron, 698 79(2), 217–240. https://doi.org/10.1016/j.neuron.2013.07.007 699 Shenhav, A., Cohen, J. D., & Botvinick, M. M. (2016). Dorsal anterior cingulate 700 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT 34 cortex and the value of control. Nature Neuroscience, 19(10), 1286–1291. 701 https://doi.org/10.1038/nn.4384 702 Shenhav, A., Prater Fahey, M., & Grahek, I. (2021). Decomposing the Motivation 703 to Exert Mental Effort. Current Directions in Psychological Science, 30(4), 704 307–314. https://doi.org/10.1177/09637214211009510 705 Sherdell, L., Waugh, C. E., & Gotlib, I. H. (2012). Anticipatory pleasure predicts 706 motivation for reward in major depression. Journal of Abnormal Psychology, 707 121(1), 51–60. https://doi.org/10.1037/a0024945 708 Silveira, M. M., Wittekindt, S. N., Ebsary, S., & Winstanley, C. A. (2021). 709 Evaluation of cognitive effort in rats is not critically dependent on ventrolateral 710 orbitofrontal cortex. European Journal of Neuroscience, 53(3), 852–860. 711 https://doi.org/10.1111/ejn.14940 712 attention, and impulsive action in female rats. Journal of Psychopharmacology, 715 34(4), 452–466. https://doi.org/10.1177/0269881119896043 716 D. A., & Cryan, J. F. (2017). Modelling depression in animals: at the ce of reward and stress pathways. Psychopharmacology, 234(9), 1451–1465. //doi.org/10.1007/s00213-017-4552-6 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 36 PREDICTING COGNITIVE EFFORT depression inventory-II: A comprehensive review. Revista Brasileira de 755 Psiquiatria, 35(4), 416–431. https://doi.org/10.1590/1516-4446-2012-1048 756 Watt, M. J., Weber, M. A., Davies, S. R., & Forster, G. L. (2017). Impact of 757 juvenile chronic stress on adult cortico-accumbal function: Implications for 758 cognition and addiction. Progress in Neuro-Psychopharmacology and Biological 759 Psychiatry, 79, 136–154. https://doi.org/10.1016/j.pnpbp.2017.06.015 760 Westbrook, A., Bosch, R. van den, Määttä, J. I., Hofmans, L., Papadopetraki, D., 761 Cools, R., & Frank, M. J. (2020). Dopamine promotes cognitive effort by biasing 762 the benefits versus costs of cognitive work. Science, 367(6484), 1362–1366. 763 https://doi.org/10.1126/science.aaz5891 764 depression inventory-II: A comprehensive review. Revista Brasileira de 755 Psiquiatria, 35(4), 416–431. https://doi.org/10.1590/1516-4446-2012-1048 756 Watt, M. J., Weber, M. A., Davies, S. R., & Forster, G. L. (2017). Impact of 757 juvenile chronic stress on adult cortico-accumbal function: Implications for 758 cognition and addiction. Progress in Neuro-Psychopharmacology and Biological 759 Psychiatry, 79, 136–154. https://doi.org/10.1016/j.pnpbp.2017.06.015 760 Westbrook, A., Bosch, R. van den, Määttä, J. I., Hofmans, L., Papadopetraki, D., 761 Cools, R., & Frank, M. J. (2020). Dopamine promotes cognitive effort by biasing 762 the benefits versus costs of cognitive work. Science, 367(6484), 1362–1366. 763 https://doi.org/10.1126/science.aaz5891 764 Westbrook, S. R., Hankosky, E. R., Dwyer, M. R., & Gulley, J. M. (2018). Age and 765 sex differences in behavioral flexibility, sensitivity to reward value, and risky 766 decision-making. Behavioral Neuroscience, 132(2), 75–87. 767 https://doi.org/10.1037/bne0000235 768 Yee, D. M., Adams, S., Beck, A., & Braver, T. S. (2019). Age-Related Differences in 769 Motivational Integration and Cognitive Control. Cognitive, Affective, & 770 Behavioral Neuroscience, 19(3), 692–714. 771 https://doi.org/10.3758/s13415-019-00713-3 772 Yee, D. M., Crawford, J. L., Lamichhane, B., & Braver, T. S. (2021). Dorsal 773 Anterior Cingulate Cortex Encodes the Integrated Incentive Motivational Value 774 of Cognitive Task Performance. The Journal of Neuroscience, 41(16), 3707–3720. 775 https://doi.org/10.1523/JNEUROSCI.2550-20.2021 776 PREDICTING COGNITIVE EFFORT 35 PREDICTING COGNITIVE EFFORT 35 Humaira, A., … Todd, R. M. (2023). Cognitive-affective processes and suicidality 728 in response to repetitive transcranial magnetic stimulation for treatment 729 resistant depression. Journal of Affective Disorders, 321, 182–190. 730 https://doi.org/10.1016/j.jad.2022.10.041 731 Towbes, L. C., & Cohen, L. H. (1996). Chronic stress in the lives of college students: 732 Scale development and prospective prediction of distress. Journal of Youth and 733 Adolescence, 25(2), 199–217. https://doi.org/10.1007/BF01537344 734 Treadway, M. T., Bossaller, N. A., Shelton, R. C., & Zald, D. H. (2012). 735 Effort-based decision-making in major depressive disorder: A translational 736 model of motivational anhedonia. Journal of Abnormal Psychology, 121(3), 737 553–558. https://doi.org/10.1037/a0028813 738 Treadway, M. T., Buckholtz, J. W., Schwartzman, A. N., Lambert, W. E., & Zald, 739 D. H. (2009). Worth the ’EEfRT’? The effort expenditure for rewards task as an 740 objective measure of motivation and anhedonia. PLoS ONE, 4(8), 1–9. 741 https://doi.org/10.1371/journal.pone.0006598 742 Unsworth, N., & Robison, M. K. (2020). Working memory capacity and sustained 743 attention: A cognitive-energetic perspective. Journal of Experimental 744 Psychology: Learning, Memory, and Cognition, 46(1), 77–103. 745 https://doi.org/10.1037/xlm0000712 746 Vásquez-Rosati, A., Montefusco-Siegmund, R., López, V., & Cosmelli, D. (2019). 747 Emotional Influences on Cognitive Flexibility Depend on Individual Differences: 748 A Combined Micro-Phenomenological and Psychophysiological Study. Frontiers 749 in Psychology, 10, 1138. https://doi.org/10.3389/fpsyg.2019.01138 750 Voyer, D., Voyer, S. D., & Saint-Aubin, J. (2017). Sex differences in visual-spatial 751 working memory: A meta-analysis. Psychonomic Bulletin & Review, 24(2), 752 307–334. https://doi.org/10.3758/s13423-016-1085-7 753 Wang, Y. P., & Gorenstein, C. (2013). Psychometric properties of the beck 754 Wang, Y. P., & Gorenstein, C. (2013). Psychometric properties of the beck 754 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Legend Figure 1. Layout of the experimental tasks. (A) The Monetary Incentive Delay 778 Task, in which participants indicate their excitement in playing for tickets towards a 779 monetary reward.(B) The calibration phase of the change detection task, where 780 Figure 1. Layout of the experimental tasks. (A) The Monetary Incentive Delay 778 Task, in which participants indicate their excitement in playing for tickets towards a 779 monetary reward.(B) The calibration phase of the change detection task, where 780 C G COG O . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 37 participants saw an array of either 2, 4, 6, or 8 squares and indicated whether the final 781 square that appeared (the probe) was of the same or a different colour from the square that 782 appeared in the same location in the previously show array. (C) The reward phase of the 783 change detection task, where participants saw an array of either 2 or 4, or 6 or 8, shapes on 784 screen depending on their performance in the calibration phase of the task. Once again, 785 participants indicated with a keyboard press whether the final square that appeared (the 786 probe) is the same or a different colour from the square that appeared in the same location 787 in the array. Here, if they made the correct decision, they would receive 1 point (low effort 788 trial) or 10 points (high effort trial); they would receive 0 points for an incorrect decision. 789 Figure 2. Demographic distributions. Distribution of (A) depressive trait (BDI) 790 proportion scores (score divided by total possible score), (B) perceived stress (PSS) scores, 791 (C) anhedonia (DARS) scores, and (D) mean reward anticipation scores by sex. BDI = 792 Beck Depression Inventory, PSS = Perceived Stress Scale, DARS = Dimensional 793 Anhedonia Rating Scale. 794 Figure 3. Plot of binomial logistic regression predicting whether participants were in 795 the high effort vs. low effort group. Binomial logistic regression predicting whether 796 participants were in the high effort vs. low effort group. High effort group: > 70% HR 797 trials selected; low effort group: <= 70% HR trials selected. 798 Figure 4. Accuracy and choice latency by motivation group. Accuracy and choice 799 latency in the choice phase of the change detection task by motivation group (high effort 800 vs. low effort group). The choice latency plot has been log-transformed on the y-axis to 801 more clearly show small choice latency values. High effort group: > 70% HR trials selected; 802 low effort group: <= 70% HR trials selected. 803 Figure 4. Accuracy and choice latency by motivation group. Accuracy and choice 799 latency in the choice phase of the change detection task by motivation group (high effort 800 vs. low effort group). The choice latency plot has been log-transformed on the y-axis to 801 more clearly show small choice latency values. PREDICTING COGNITIVE EFFORT High effort group: > 70% HR trials selected; 802 low effort group: <= 70% HR trials selected. 803 Figure 5. Accuracy and choice latency by working memory ability. (A) Accuracy 804 and (B) choice latency in the reward phase of the change detection task by working 805 Figure 5. Accuracy and choice latency by working memory ability. (A) Accuracy 804 and (B) choice latency in the reward phase of the change detection task by working 805 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 38 memory ability (K estimate). The choice latency plot has been log-transformed on the 806 y-axis to more clearly show small choice latency values. 807 memory ability (K estimate). The choice latency plot has been log-transformed on the 806 y-axis to more clearly show small choice latency values. 807 Figure 6. Drift diffusion model outputs for high and low effort participants. Figure 6. Drift diffusion model outputs for high and low effort participants. 808 Outputs of a hierarchical Bayesian drift diffusion model fit to all participants in the effort 809 choice task, divided by effort group. High effort participants had a significantly higher (B) 810 starting point and (C) boundary separation compared to low-effort participants, but did 811 not significantly differ on (A) drift rate or (D) non-decision time compared to low-effort 812 participants. 813 Table 1. Demographic information for all participants, by sex and effort deployment 814 group. BDIprop = Beck Depression Inventory II proportion score (score divided by max 815 score). PSS = Perceived Stress Scale score. DARS = Dimensional Anhedonia Rating Scale. 816 Ant. = mean reward anticipation in the behavioural monetary incentive delay task. High 817 effort group: > 70% HR trials selected; low effort group: <= 70% HR trials selected. 818 814 Table 2. Binomial logistic regression predicting whether participants were in the 819 high effort vs. low effort group. High effort group: > 70% HR trials selected; low effort 820 group: <= 70% HR trials selected. BDI (prop. score) = depression score on the Beck 821 Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided 822 by total possible score. AIC = Akaike information criterion, BIC = Bayesian information 823 criterion, ICC = intraclass correlation, RMSE = root mean squared error. 824 Table 3. Multi-level model analysis coeﬀicients and standard errors for accuracy in 825 the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression 826 Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total 827 possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, 828 ICC = intraclass correlation, RMSE = root mean squared error. 829 Table 4. Multi-level model analysis coeﬀicients and standard errors for choice 830 Table 4. Multi-level model analysis coeﬀicients and standard errors for choic 830 . . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint PREDICTING COGNITIVE EFFORT PREDICTING COGNITIVE EFFORT 39 latency in the reward phase of the task. BDI (prop. score) = depression score on the Beck 831 Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided 832 by total possible score. AIC = Akaike information criterion, BIC = Bayesian information 833 criterion, ICC = intraclass correlation, RMSE = root mean squared error. 834 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint 40 PREDICTING COGNITIVE EFFORT Table 3 Table 3 Multi-level model analysis coeﬀicients and standard errors for accuracy in the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Table 3 Multi-level model analysis coeﬀicients and standard errors for accuracy in the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Multi-level model: Accuracy (Intercept) 0.64* (0.02) Sex 0.01 (0.01) Working memory ability (K estimate) 0.01 (0.00) BDI (prop. score) −0.01 (0.04) PSS 0.00 (0.00) Reward anticipation 0.00 (0.00) Effort level 0.25* (0.01) SD (Intercept participant) 0.04 SD (Observations) 0.10 Num.Obs. 542 R2 Marg. 0.592 R2 Cond. 0.656 AIC −840.0 BIC −801.3 ICC 0.2 RMSE 0.09 + p < 0.1, * p < 0.05, p < 0.01, * p < 0.001 Multi-level model: Accuracy 0 64* . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint 41 PREDICTING COGNITIVE EFFORT . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted February 13, 2024. ; https://doi.org/10.1101/2024.02.12.579951 doi: bioRxiv preprint Table 4 Multi-level model analysis coeﬀicients and standard errors for choice latency in the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Table 4 Table 4 Multi-level model analysis coeﬀicients and standard errors for choice latency in the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Multi-level model analysis coeﬀicients and standard errors for choice latency in the reward phase of the task. BDI (prop. score) = depression score on the Beck Depression Inventory, PSS = Perceived Stress Scale. Proportion scores are scores divided by total possible score. AIC = Akaike information criterion, BIC = Bayesian information criterion, ICC = intraclass correlation, RMSE = root mean squared error. Multi-level model: Choice latency (Intercept) 1.25* (0.34) Sex −0.07 (0.17) Working memory ability (K estimate) 0.05 (0.05) BDI (prop. score) −1.11* (0.54) PSS 0.01 (0.01) Reward anticipation −0.04 (0.04) Effort level 1.06*** (0.13) SD (Intercept participant) 0.18 SD (Observations) 1.53 Num.Obs. 542 R2 Marg. 0.114 R2 Cond. 0.127 AIC 2042.9 BIC 2081.6 ICC 0.0 RMSE 1.51 + p < 0.1, * p < 0.05, p < 0.01, * p < 0.001 Multi-level model: Choice latency ying for 1 et. ow excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. + Cue 500 ms Mask 500 ms Probe Until response + Cue 500 ms Mask 500 ms Probe Until response C Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select Low effort, low reward trial High effort, high reward trial Success Failure +1 point! Score: 11 +0 points! Score: 10 Success Failure +10 points! Score: 21 +0 points! Score: 10 Probe Until response Cue 500 ms Mask 500 ms Fixation 500 ms Fixation 500 ms Choice Time Time Time Time You are playing for 1 ticket. How excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. A B + Cue 500 ms Mask 500 ms Probe Until response Fixation 500 ms Time Time You are playing for 1 ticket. How excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. Table 4 A B + + Cue 500 ms Mask 500 ms Probe Until response + Cue 500 ms Mask 500 ms Probe Until response C Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select Low effort, low reward trial High effort, high reward trial S S Probe Until response Cue 500 ms Mask 500 ms Fixation 500 ms Fixation 500 ms Fixation 500 ms Choice Time Time Time Time You are playing for 1 ticket. How excited do you feel? 0 - 50 GET READY! (Press the spacebar!) Tickets banked: 1 Congratulations! You were fast enough. A Time A C + Cue 500 ms Mask 500 ms Probe Until response + Probe Until response C Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select Low effort, low reward trial High effort, high reward trial Success Failure +1 point! Score: 11 +0 points! Score: 10 Success Failure +10 points! Score: 21 +0 points! Score: 10 Cue 500 ms Mask 500 ms Fixation 500 ms Fixation 500 ms Choice Time Time You are playing for 1 ticket. Table 4 High effort, high reward trial Choose your task: Easy Task Hard Task 1 point 10 points press A to select press L to select 10 points press L to select 1 point press A to select Low effort, low reward trial Success B + Cue 500 ms Mask 500 ms Probe Until response Fixation 500 ms Time B High Effort Low Effort 25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 BDI proportion score High Effort Low Effor 0 10 20 30 0 10 20 30 0.000 0.025 0.050 0.075 PSS score B 10 20 30 40 50 0 10 20 30 40 50 DARS score 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D Sex Female Male Effort Low Effort 50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 BDI proportion score High Effort Low Effor 0 10 20 30 0 10 20 30 0.000 0.025 0.050 0.075 PSS score B B A High Effort Low Effort High Effort 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 BDI proportion score 0 10 20 30 0 10 20 30 0.000 0.025 PSS score 0 10 20 30 40 50 0 10 20 30 40 50 0.00 0.01 0.02 0.03 0.04 DARS score 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D Sex Female Male BDI proportion score PSS score 0 10 20 30 40 50 0 10 20 30 40 50 DARS score 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D Sex Female Male rtion score PSS score 0 10 20 30 40 50 S score 2 4 6 2 4 6 0.00 0.05 0.10 0.15 0.20 Mean reward anticipation D D C -2 0 2 4 Working memory score (K estimate) N.S. * * 0.6 0.8 1.0 High Low Selected effort level Accuracy A N.S. * * 0.1 1.0 10.0 High Low Selected effort level Log choice latency (s) B Motivation group High Effort Low Effort 1 Log choice latency (s) B 0.1 1.0 10.0 Log choice latency (s) B N.S. * * 0.1 1.0 10.0 High Low Selected effort level Log choice latency (s) B Log choice latency (s) B A N.S. * * 0.6 0.8 1.0 High Low Selected effort level Accuracy A N.S. 0.1 1.0 10.0 High Selec Log choice latency (s) B * * A High Effort -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1.0 Accuracy A High Effort Low Effort -2 0 2 4 -2 0 2 4 -2 0 2 4 -2 0 2 4 Working memory score (K estimate) Low Effort High Effort B NS. 1 0 1 2 3 High Effort Low Effort * 0.0 0.5 1.0 1.5 High Effort Low Effort Starting point B * 0 2 3 High Effort Low Effort Motivation group NS. 0.00 0.25 0.50 0.75 1.00 High Effort Low Effort Motivation group Non-decision time (s) D B A 1.5 3 NS. -1 0 1 2 3 High Effort Low Effort Drift rate 0.0 0.5 1.0 High Eff Starting point * 0 1 2 3 High Effort Low Effort Motivation group Boundary separation C 0.00 0.25 0.50 0.75 1.00 High Ef M Non-decision time (s) D * 0.5 1.0 Starting point * Drift rate NS. *
https://openalex.org/W2267023236	https://breast-cancer-research.biomedcentral.com/counter/pdf/10.1186/bcr3795	English	null	Breast biopsy in patients on anti-coagulants: is new guidance needed?	Breast cancer research	2,015	cc-by	16,896	O1 Accuracy of GE digital breast tomosynthesis versus supplementary mammographic views for diagnosis of screen-detected soft tissue breast lesions Accuracy of GE digital breast tomosynthesis versus supplementary mammographic views for diagnosis of screen-detected soft tissue breast lesions Introduction: There is little evidence regarding the optimum interval between mammograms in a population breast cancer screening programme. The UK NHS Breast Screening Programme employs a 3-year interval, unlike other countries which screen more frequently. This study uses variations in the screening interval within a single large UK screening service to examine possible relationships between screening interval and screen-detected cancer characteristics. Eleanor Cornford1, Anne Turnbull2, Jonathan James1, Rachel Tsang1, Tayeba Akram2, Helen Burrell1, Lisa Hamilton1, Sarah Tennant1, Mark Bagnall2, Shama Puri2, Graham Balls3, Yan Chen4, Vivienne Jones5 1Nottingham Breast Institute, Nottingham, UK; 2Royal Derby Hospital, Derby, UK; 3Nottingham Trent University, Nottingham, UK; 4Loughborough University, Loughborough, UK; 5Northampton General Hospital, Northampton, UK Breast Cancer Research 2015, 17(Suppl 1):O1 Eleanor Cornford1, Anne Turnbull2, Jonathan James1, Rachel Tsang1, Tayeba Akram2, Helen Burrell1, Lisa Hamilton1, Sarah Tennant1, Mark Bagnall2, Shama Puri2, Graham Balls3, Yan Chen4, Vivienne Jones5 1Nottingham Breast Institute, Nottingham, UK; 2Royal Derby Hospital, Derby, UK; 3Nottingham Trent University, Nottingham, UK; 4Loughborough University, Loughborough UK; 5Northampton General Hospital Northampton UK 1Nottingham Breast Institute, Nottingham, UK; 2Royal Derby Hospital, Derby, UK; 3Nottingham Trent University, Nottingham, UK; 4Loughborough University, Loughborough UK; 5Northampton General Hospital Northampton UK Methods: A total of 1107 women diagnosed with breast cancer on incident screening over a 5-year period were included. Age, time since last mammogram and surgical histopathology data (tumour type, size, grade, nodal stage, receptor status) were recorded. The Nottingham prognostic index (NPI) was calculated for invasive cancers. Analysis with Spearman’s rho and Pearson’s correlation was performed. Introduction: The aim was to compare the accuracy of standard supplementary views and GE digital breast tomosynthesis (DBT) for assessment of soft tissue mammographic abnormalities. Introduction: The aim was to compare the accuracy of standard supplementary views and GE digital breast tomosynthesis (DBT) for assessment of soft tissue mammographic abnormalities. Methods: Women recalled for further assessment of soft tissue abnormalities were recruited and received standard supplementary views (typically spot compression views) and two-view GE DBT. The added value of DBT in the assessment process was determined by analysing data collected prospectively by unblinded radiologists working up the cases. Results: The median patient age was 63. Most screening intervals were between 800 and 1200 days. Open Access MEETING ABSTRACTS Open Access Breast Cancer Research 2015, Volume 17 Suppl 1 Breast Cancer Research 2015, Volume 17 Suppl 1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 These abstracts are available online at http://www.breast-cancer-research.com/supplements/17/S1 ese abstracts are available online at http://www.breast-cancer-research.com/supplements/17/S Manchester, UK; 3Centre for Imaging Sciences, Institute of Population Health, University of Manchester, UK; 4Medical Statistics Department, Education and Research Centre, University Hospital of South Manchester, UK; 5Centre for Biostatistics, Institute of Population Health, University of Manchester, UK Breast Cancer Research 2015, 17(Suppl 1):O2 © 2015 various authors. All articles published in this supplement are distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 O1 Accuracy of GE digital breast tomosynthesis versus supplementary mammographic views for diagnosis of screen-detected soft tissue breast lesions The proportion of women with ductal carcinoma in situ (DCIS) decreased significantly with increasing screening interval, from 24/94 (25.5%) for < 2.5 years to 35/240 (14.6%) for ≥3 years (p = 0.032). No significant associations were found between other tumour variables and screening interval. The average NPI score was 3.49 with a Pearson correlation coefficient of −0.032 (p = 0.389). Following anonymisation of cases, there was also a retrospective multireader review. The readers first read bilateral standard two-view digital mammography (DM) together with the supplementary mammographic views and gave a combined score for suspicion of malignancy on a five- point scale. The same readers then read bilateral standard two-view DM together with two-view DBT. Pathology data were obtained. Differences were assessed using ROC analysis. p Conclusion: The findings suggest that there is a significant rate of progression of DCIS to invasive disease within the current 3-year screening interval. This, together with the rate and characteristics of interval cancers (which were not examined in this study), appears to be more important in determining the optimum screening interval than the characteristics of the invasive cancers detected by screening. Results: The study population was 342 lesions in 322 patients. Final diagnosis was malignant in 113 cases (33%) and benign/normal in 229 cases (67%). In the prospective analysis, the performance of two-view DM plus DBT was at least equivalent to the performance of two-view DM and standard mammographic supplementary views–area under the curve (AUC) was 0.946 and 0.922 respectively, which did not reach statistical significance. Similar results were obtained for the retrospective review– AUC was 0.900 (DBT) and 0.873 (supplementary views), which did not reach statistical significance. O3 Breast density measurements with ultrasound tomography: a comparison with non-contrast MRI Elizabeth O’Flynn1, Jeremie Fromageau1, Minty Ledger1, Alessandro Messa1, Ashley D’Aquino1, Minouk Schoemaker1, Maria Schmidt1, Neb Duric1,2, Anthony Swerdlow1, Jeff Bamber1 1Insititute of Cancer Research and Royal Marsden Hospital, Sutton, UK; 2Karmanos Institute, Detroit, MI, USA Breast Cancer Research 2015, 17(Suppl 1):O3 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Phys Med Biol 2011, 56(18):N195-N205. g q y Patsy Whelehan1,2, Andrew Evans1, Gozde Ozakinci2 1University of Dundee, Dundee, UK; 2University of St Andrews, St Andrews, UK Breast Cancer Research 2015, 17(Suppl 1):O6 O4 Breast tumour localisation using Iodine seeds in the UK: the first 100 patients Nidhi Sibal, Nerys Forester Newcastle Teaching Hospitals, Newcastle, UK Breast Cancer Research 2015, 17(Suppl 1):O4 Introduction: While the concept of overdiagnosis can be difficult to understand, it has been shown that women wish to be informed about it. The latest breast screening information leaflet offers considerable detail about potential benefits and harms of screening, including overdiagnosis. However, it is unknown how much use women attending for screening make of the leaflet. We report qualitative findings on informed choice and consent within the UK breast screening programme. Introduction: Wire localization techniques for impalpable breast tumours require wire placement ideally on the day of surgery. Tumour localization using iodine-125 seeds allows tumour localization to occur prior to surgery, improving both work flow dynamics and the patient experience. Newcastle Hospitals Trust is the first centre in the UK to adopt this technique. Here we present our initial experience of the first 100 patients to undergo wire-free surgery. Methods: Participants were clients and mammographers from breast screening units in Scotland and London. Semi-structured, in-depth, individual interviews were conducted and thematic analysis performed. y p Results: Twenty-two clients were interviewed, aged 50−72: seven first- attenders and 15 subsequent, from a range of deprivation categories. Eighteen mammographer-participants included assistant, registered, and advanced practitioners, with a wide range of ages and lengths of experience. Most clients understood that screening aims to detect breast cancer early to improve the chances of survival. Several were aware of the possibility of false positive results and the risk of mammography inducing a cancer. Others could not name any risks of screening. Women had mostly either skimmed the information leaflet or not read it at all. Several mammographers recounted experiences where women had appeared to attend under pressure from others and where severe challenges existed in ascertaining consent. Methods: From September 2014, data were prospectively collected on all patients undergoing iodine seed tumour localization. Seeds were placed under ultrasound guidance into tumours identifiable on ultrasound between 7 and 14 days preoperatively. Seeds were removed with the tumour after intraoperative localization using a gamma probe. Results: Our first 100 patients are included in this initial analysis. Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Introduction: As part of diagnostic work for radiological abnormalities seen in the breast, there has been an increase in use of vacuum-assisted biopsies for diagnosis. This allows more tissue to be sampled and therefore leads to a greater degree of diagnostic accuracy. In addition to diagnosis, in some centres the same procedure has also been used for removal of the entire lesion–vacuum-assisted excision (VAE). This is sometimes offered in place of a diagnostic surgical excision in cases of B3 lesions. We wanted to examine whether VAE can be a safe alternative for B3 lesion that show atypia. Introduction: As part of diagnostic work for radiological abnormalities seen in the breast, there has been an increase in use of vacuum-assisted biopsies for diagnosis. This allows more tissue to be sampled and therefore leads to a greater degree of diagnostic accuracy. In addition to diagnosis, in some centres the same procedure has also been used for removal of the entire lesion–vacuum-assisted excision (VAE). This is sometimes offered in place of a diagnostic surgical excision in cases of B3 lesions. We wanted to examine whether VAE can be a safe alternative for B3 lesion that show atypia. density. This study was to evaluate the technique of UST and compare VASS with percentage water density from non-contrast MRI. Methods: This single-centre cross-sectional trial had research ethics committee approval. Fifty healthy volunteers from the Generations study [3] (median age 40 years, range 30−64 years) underwent bilateral breast UST. Forty-six underwent MRI using a 2-point Dixon technique [4]. VASS and percentage water density measurements were evaluated in both beasts and compared with Pearson’s correlation coefficient. Results: Mean VASS measurements for the cohort were 1446 ± 148 ms−1 (range 1434−1541 ms). There was high similarity between measurements from the right and left breasts (1463 ± 29 ms−1, 1459 ± 29 ms−1 respectively (p = 0.516)) (ICC = 0.98). Mean percentage water density for the cohort was 34.6 ± 14.5% (range 13.5−74.4%) with good right-to-left consistency (35.7 ± 15.3%, 34.4 ± 14.6% respectively (p = 0.55)). There was excellent correlation between VASS and percentage water density (r2 = 0.97, p < 0.0001). Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 The majority of patients had a wide local excision, with 10 undergoing therapeutic mastectomy. Thirteen patients returned to theatre for positive margins or completion mastectomy, depending on the final pathology. No seeds were lost during use. One patient had a second tumour identified at the time of seed placement which required wire localization. No radiological complications occurred. Introduction of iodine seeds improved radiological workflow, with creation of a planned outpatient ‘seed list’, remote from the day of surgery and radiological high demand times. g g Conclusion: These qualitative findings that some women attend for breast screening with little knowledge of the balance of risks and benefits, and in some cases may encounter coercion, require further investigation. New methods of communication may be indicated. Conclusion: Iodine seed tumour localization in the UK is achievable, patient friendly and has great benefits for radiologists in terms of department workflow. Noticeably, patients (and surgeons) appear much more relaxed since the introduction of this technique and initial patient satisfaction surveys have been positive. Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Methods: We identified all patients, at Leeds Teaching Hospital NHS Trust, who had undergone a surgical diagnostic excision following a core biopsy which had revealed the following lesions: fibroadenoma, papilloma or radial scar with atypia (FEA, AIDP or ISLN) during the period between 2009 and 2013. We reviewed the slides of the core biopsy and the subsequent excision biopsy to confirm the histological diagnosis. Results: Twenty-nine cases in total satisfied our inclusion criteria. There were nine cases of fibroadenomas with ISLN and/or AIDP. None of the cases showed upgrading of the atypia. There were eight cases of radial scar that had either ISLN, LCIS, epithelial atypia or AIDP, of which two showed DCIS in the surgical excision. There were 12 cases of papilloma with either ISLN or AIDP; of these, five had DCIS on surgical excision. Conclusion: UST holds promise as a robust, reliable and accurate technique to evaluate breast density without the use of ionising radiation and has additional benefits of lower cost and greater patient acceptability. References 1. Glide C, Duric N, Littrup P: Novel approach to evaluating breast density utilising ultrasound tomography. Medical Physics 2007, 34(2):744-753. Conclusion: VAE is safe for fibroadenomas with atypia and radial scars with atypia provided the periphery can be adequately sampled, to help diagnose DCIS. Papilloma with atypia requires surgical excision due to complex histological architecture. 2. Duric N, Boyd N, Littrup P, Sak M, Myc L, Li C, et al: Breast density measurmement with ultrasound tomography: a comparison with film and digital mammography. Medical Physics 2013, 40(1):013501. 2. Duric N, Boyd N, Littrup P, Sak M, Myc L, Li C, et al: Breast density measurmement with ultrasound tomography: a comparison with film and digital mammography. Medical Physics 2013, 40(1):013501. 3. Swerlow AJ, Jones ME, Schoemaker MJ, Hemming J, Thomas D, Williamson J, et al: The Breakthrough Generations Study: design of a long-term UK cohort study to investigate breast cancer aetiology. Br J Cancer 2011, 105(7):911-917. O6 Informed choice and consent among women attending for breast screening in the UK: data from a qualitative study Patsy Whelehan1,2, Andrew Evans1, Gozde Ozakinci2 1University of Dundee, Dundee, UK; 2University of St Andrews, St Andrews, UK Breast Cancer Research 2015, 17(Suppl 1):O6 4. Schmidt MA: Phase-uncertainty quality map for two-point Dixon fat- water separation. Phys Med Biol 2011, 56(18):N195-N205. 4. Schmidt MA: Phase-uncertainty quality map for two-point Dixon fat- water separation. O3 Conclusion: The accuracy of GE DBT in the assessment of screen- detected soft tissue abnormalities is equivalent to the use of standard supplementary mammographic views. Introduction: Ultrasound tomography (UST) is an emerging whole breast 3D imaging technique that obtains quantitative tomograms of speed of sound (as well as other properties) of the entire breast. The measured parameter is the volume averaged speed of sound (VASS) [1,2]. It improves on mammography by measuring density at each voxel and holds promise as a cheap, patient-acceptable, non-ionising radiation method to evaluate O2 Breast screening interval and the characteristics of screen-detected cancers Carl A Tupper1, Anthony J Maxwell2,3, Susan Astley2,3, Megan Bydder2, Soujanya Gadde2, Elaine Harkness2,3, Yit Y Lim2,3, Mary Wilson2, Julie Morris4,5 1University of Manchester Medical School, Manchester, UK; 2Nightingale Centre and Genesis Prevention Centre, University Hospital of South © 2015 various authors. All articles published in this supplement are distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Page 2 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Accuracy of axillary nodal ultrasound and ultrasound fine needle aspiration/ core biopsy in the preoperative staging of patients with invasive breast cancer William Cheung, Suzanne McLenachan, Gauripriya Babu, Melanie Smith Breast Unit, Edinburgh, UK Breast Cancer Research 2015 17(Suppl 1):P2 Alice Leaver, Simon Lowes, Peter Newton, Linsley Lunt, Anuradha Anand, Linda Hutchinson, Sally Athey, Amanda Jane Potterton, Sheetal Sharma, Alan Redman Introduction: Patients with invasive breast cancer undergo axillary ultrasound ± ultrasound fine needle aspiration (US-FNA)/core biopsy for preoperative staging depending on the ultrasound appearance. At our institution, abnormal axillary lymph node assessment includes: a cortical thickness >3 mm, focal or eccentric cortical thickening, nodal shape (spherical) and replaced appearance with loss of echogenic nodal hilum. Our aims were to evaluate the accuracy of preoperative US + US- FNA/core biopsy for detecting axillary metastatic disease. Introduction: In our Trust, all breast cancer patients undergo preoperative axillary staging with ultrasound. Over the past year we have introduced intradermal sulphur hexafluoride microbubbles ultrasound contrast injection to help identify sentinel lymph nodes for a preoperative needle biopsy in each patient. Only patients with malignant node morphology on grey-scale ultrasound undergo biopsy without microbubbles injection. Methods: Excluding those patients who underwent neoadjuvant chemotherapy, we identified 120 patients with invasive breast cancer between January and December 2013, which yielded axillary node metastases on final surgical pathology. We performed a retrospective analysis of the clinical records and used descriptive statistics. j Methods: Prospective audit of data collated at the time of the microbubbles procedure together with multidisciplinary meeting records identified relevant screening and symptomatic patients with primary breast cancer treatment including axillary node surgery between 1 July 2014 and 1 July 2015. Descriptive statistics were performed. y p Results: Preoperative US correctly identified 60/120 (50%) patients with axillary metastatic disease, 42/60 (70%) had subsequent true positive US biopsies. Of the cases where a biopsy was not performed, 88% (53/60) had one or two positive nodes confirmed after surgery and 12% (7/60) had at least three nodes. Thirty-four of 60 (57%) were from the symptomatic population. Of the total 21 false negative US biopsies from the 18 patients, 81% (17/21) were performed via FNA and 19% (4/21) via core biopsy. Eleven of 18 (61%) were from the symptomatic population. Twenty-nine of 42 (69%) true positive US biopsies were from the symptomatic population. Conclusion: The results highlight the need for a review of our biopsy criteria, which may result in a decrease in our biopsy threshold. POSTER PRESENTATIONS P1 Sensitivity of US and FNAC for staging of the axilla in patients presenting with symptomatic breast cancer Diane Lister, Miaad Al-Attar, Elizabeth Denton, Lisa Grosvenor, Gayle McDonald, Dave Purnell Breast Care Centre, UHL NHS Trust, Leicester, UK Breast Cancer Research 2015, 17(Suppl 1):P1 P1 Sensitivity of US and FNAC for staging of the axilla in patients presenting with symptomatic breast cancer Diane Lister, Miaad Al-Attar, Elizabeth Denton, Lisa Grosvenor, Gayle McDonald, Dave Purnell Breast Care Centre, UHL NHS Trust, Leicester, UK Breast Cancer Research 2015, 17(Suppl 1):P1 O5 Is surgical diagnostic excision always necessary for solid lesions with atypia? Nisha Sharma, Rebecca Millican-Slater, Eldo Verghese Leeds Teaching Hospital NHS Trust, Leeds, UK Breast Cancer Research 2015, 17(Suppl 1):O5 O5 Is surgical diagnostic excision always necessary for solid lesions with atypia? Nisha Sharma, Rebecca Millican-Slater, Eldo Verghese Leeds Teaching Hospital NHS Trust, Leeds, UK Breast Cancer Research 2015, 17(Suppl 1):O5 Page 3 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Introduction: Contrast-enhanced ultrasound of the axilla can be used to identify the axillary sentinel lymph node. We introduced this into our practice in 2013. During the study period there was an upgrade of our US equipment. The purpose of our audit was to see the negative predictive value of CEUS biopsy of the SLN. Introduction: Contrast-enhanced ultrasound of the axilla can be used to identify the axillary sentinel lymph node. We introduced this into our practice in 2013. During the study period there was an upgrade of our US equipment. The purpose of our audit was to see the negative predictive value of CEUS biopsy of the SLN. Introduction: We investigated our sensitivity for axillary node staging, in patients presenting with symptomatic breast cancer from January to December 2012. Methods: Of 430 patients identified, 288 had first-line surgical treatment, 63 had neoadjuvant therapy first. Seventy-nine women were unfit for surgery, had less aggressive evaluation of the axilla and were excluded from sensitivity calculations. US axilla ± FNA were performed at presentation. Nodal disease prevalence, sensitivity for diagnosis and the NPV of our tests were calculated. In the neoadjuvant cases, pretreatment nodal status was not accurately known. Methods: This was a retrospective audit. In total, 110 patients with invasive breast cancer were identified at the breast MDT. P4 Use of sulphur hexafluoride microbubbles injection to identify the sentinel lymph node in breast cancer patients: initial experience in a UK breast unit P2 Accuracy of axillary nodal ultrasound and ultrasound fine needle aspiration/ core biopsy in the preoperative staging of patients with invasive breast cancer William Cheung, Suzanne McLenachan, Gauripriya Babu, Melanie Smith Breast Unit, Edinburgh, UK Breast Cancer Research 2015, 17(Suppl 1):P2 P2 P4 Use of sulphur hexafluoride microbubbles injection to identify the sentinel lymph node in breast cancer patients: initial experience in a UK breast unit Alice Leaver, Simon Lowes, Peter Newton, Linsley Lunt, Anuradha Anand, Linda Hutchinson, Sally Athey, Amanda Jane Potterton, Sheetal Sharma, Alan Redman Gateshead Hospitals NHS Trust, Gateshead, UK Breast Cancer Research 2015, 17(Suppl 1):P4 POSTER PRESENTATIONS The US core biopsy, surgical sentinel node biopsy and subsequent axillary histology were documented. Results: CEUS was successful in identifying the first draining lymph node in 88.1% (97/111). Eighty-three of 97 cases (86%) had a definitive biopsy (B2−B5) result with 13 being malignant and 69 were benign. Fifteen were non-diagnostic with B1 core biopsy. The prevalence of axillary metastases at surgery was 31% (30/97) (22 macrometastases, six micrometastases and two isolated tumour cells) of which 42% were detected by CEUS, with 100% specificity. Two of the 30 cases were in palpable, non-sentinel nodes. The negative predictive value of CEUS with core biopsy is 80% but 90% if only macrometastases are included. Conclusion: CEUS and biopsy is a promising technique for reducing the false negative rate of imaging at the time of SLNB. Our numbers are small and we had a transition to different equipment during the study, but it is felt that reproducible data comparable with Cox et al. [1] is achievable. y Results: The prevalence of nodal metastases in our surgery first cases was 43% (123/288). Twenty-four per cent of cases were micrometastases (29/123). US sensitivity for macrometastases was 51% (48/94); 41% including micrometastases (50/123). FNA sensitivity for macrometastases was 38% (36/94; 35 results C5); 30% (37/123) including micrometastases. Combining all groups, FNA was definitive (C5 or C2) in 90% (134/149) of cases. The NPV of imaging was 65% (137/210); 75% (137/183) with micrometastases excluded. The NPV of a C1/2 result was 72% (28/39) giving a false negative FNA rate of 28%. Of neoadjuvant cases, FNA was positive in 60% (38/63; 35 results C5), giving a minimum disease prevalence and diagnostic sensitivity of 60%. Combining both groups, nodal disease prevalence lies between 46% (161/351) and 53% (186/351). FNA sensitivity is between 48 and 57% for macrometastases (75/157; 75/132); and 40−46% including micrometastases (75/186; 75/161). Conclusion: Axillary staging depends on both US sensitivity and FNAC technique. US sensitivity is adversely affected by micrometastases. In our symptomatic patients, FNA sensitivity for macrometastases lies between 48 and 57%. Reference 1. Cox K, et al: Contrast -enhanced ultrasound biopsy of sentinel lymph nodes in patients with breast cancer. Implications for axillary metastases and conservation 2015, o(1):1-7. P3 Evaluation of the use of microbubbles in the ultrasound assessment of the axilla in breast cancer patients Nisha Sharma, Isobel Haigh, Rebecca Millican-Slater, Benjamin Dessauvagie St James’s Hospital, Leeds, UK Breast Cancer Research 2015, 17(Suppl 1):P3 P5 Preoperative identification and biopsy of sentinel lymph nodes using contrast-enhanced ultrasound: the Tunbridge Wells experience Use of preoperative breast MRI to determine disease extent Nicky Dineen, Jennifer Weeks, Ruxandra Pietrosanu, Karina Cox, Pippa Mills Maidstone General Hospital, Maidstone, Kent, UK Breast Cancer Research 2015, 17(Suppl 1):P7 Introduction: Breast MRI can be performed in the preoperative workup of patients with biopsy-proven breast cancer to size lesions, if there is discrepancy regarding the extent of disease from clinical, mammography or ultrasound assessment, and to identify multicentric or multifocal disease. The purpose of breast MRI is to plan the optimum surgical procedure, thus reducing the local tumour recurrence rate and the need for the patient to undergo additional surgery. Introduction: At Maidstone and Tunbridge Wells NHS Trust, all newly diagnosed breast cancer patients with a normal grey-scale axillary ultrasound have a procedure to identify and biopsy sentinel lymph nodes (SLN) using contrast-enhanced ultrasound (CEUS). Methods: Retrospective data were collected on patients undergoing a CEUS guided biopsy over a 42-month period at Tunbridge Wells Breast Clinic (TWBC). We compared the results of the first group of patients with the most recent to determine the performance of the test over time. Methods: In this poster we have reviewed breast MRI examinations from patients with a new diagnosis of breast cancer, who had a preoperative MRI. Results: There were 75 scans in total. Patients who had MRI occult disease or neoadjuvant chemotherapy were excluded, leaving a total of 51 breast MRI scans. Invasive tumour size and total tumour size (invasive tumour + DCIS) as seen on MRI were compared with the size reported in the surgical pathology specimen. There was accurate correlation in invasive tumour size in 81% and significantly discordant sizing in 19%. Correlation in overall tumour size including DCIS was 70% and significantly discordant in 30%. In three patients in whom the total tumour size was overestimated, the patients consequently had complete local excision with wide excision margins. In another patient, in whom the total disease extent was underestimated on MRI, following complete local excision, repeat surgery was required for positive margins. In these four patients the MRI was misleading for guiding the optimum surgical procedure. Results: Between February 2011 and June 2012, 94 patients had a CEUS guided biopsy of SLN. Twenty patients were excluded; five had neoadjuvant therapy, five were unfit for surgery, one had an abnormal grey-scale ultrasound and nine had incomplete data. SLN were visualised in 92% and 83% had a successful SLN core biopsy. P6 Correlation of post-neoadjuvant chemotherapy response on MRI with final histology in breast cancer patients Introduction: Invasive lobular cancer has been associated with an increased risk of multifocal and contralateral disease. The literature suggests an incidence of contralateral disease as high as 15%. Current national (NICE) recommendations are that all patients with lobular carcinoma being considered for breast-conserving surgery have a preoperative breast MRI. The objective was to identify the rate of additional MRI-detected multifocal and contralateral disease in patients with a newly diagnosed lobular cancer to determine whether it is as high as the literature suggests. Based on our findings we hope to further explore whether another imaging alternative should be considered. Introduction: We used MRI breasts to assess neoadjuvant chemotherapy response in line with departmental protocol. The aims were to see the correlation of findings on MRI with final histology in patients with breast cancer receiving neoadjuvant chemotherapy, and to assess the accuracy of our reporting and to evaluate the cause for any discordancy. p g y y Methods: Retrospective data collection from March 2012 to May 2014. Data on 50 consecutive patients, who had both pre and post neoadjuvant MRIs, were collected. Use of CRIS, NBT PACS and UHB PACS for reports and image visualisation. Use of Ultra inquires on the intranet for histopathology reports. Methods: A retrospective search was done on PACS to identify all those patients in Northern Ireland investigated with bilateral breast MRI for a newly diagnosed cancer during a 15-month period. MRI findings were correlated with histopathology records from all regional labs and the data analysed. p gy p Results: Discrepancy in size of residual tumour between MRI and histology within 10 mm was considered concordant. Concordant size between MR and histology = 31/50 (62%). Discordant size between MR and histology = 19/50 (38%). For complete response: sensitivity = 46%, specificity = 86%, positive predictive value = 71% (95% CI 45−88%), negative predictive value = 70% (95% CI 53−82%), accuracy = 70%. Conclusion: Good specificity but low sensitivity in line with published literature. The time interval between second MRI and time to surgery did not affect the ability of MRI to predict response. Presence of DCIS and LCIS (42%) influenced MRI-histology discrepancy. Hormonal-positive, Her2- positive and triple-positive were more likely to show size discrepancy compared with other hormone profiles. Recommendation: We plan to integrate DWI into our reports to increase y Results: Discrepancy in size of residual tumour between MRI and histology within 10 mm was considered concordant. Accuracy of axillary nodal ultrasound and ultrasound fine needle aspiration/ core biopsy in the preoperative staging of patients with invasive breast cancer * An increase in the use of core biopsies may yield greater accuracy in correctly identifying axillary nodal disease. Results: Sixty-four female patients underwent microbubbles injection and axillary node surgery. Overall combined sensitivity and specificity of microbubbles ultrasound/biopsy procedure were 67% (8/12) and 100% (52/52) respectively. Seventy-five per cent of operative sentinel node biopsies (45/60) showed evidence of previous needle biopsy (four axillary clearance specimens excluded). Needle biopsy detection of micrometastatic disease only, shortly after commencing microbubbles use, led to multidisciplinary meeting consideration of size of needle biopsy metastasis and ultrasound appearance of sentinel and surrounding nodes in triage of patients to type of axillary surgery. Results represent the combined learning curve of seven radiologists. The procedure was well tolerated by patients and technically easy to perform. The greatest challenges were optimising ultrasound machines for microbubbles visualisation, and finding time within busy clinics to perform the procedure. Conclusion: In this small patient cohort, introduction of microbubbles has facilitated reliable and effective identification of the sentinel lymph node for assessment of morphology on ultrasound and also biopsy. Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Page 4 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 P7 P5 Preoperative identification and biopsy of sentinel lymph nodes using contrast-enhanced ultrasound: the Tunbridge Wells experience Tania De’Silva, Mohamed Hashem, Nick Wakeham, Sarah Kirwan, Neal Chayya, Pippa Mills, Karina Cox Maidstone and Tunbridge Wells NHS Trust, Tunbridge Wells, UK Breast Cancer Research 2015, 17(Suppl 1):P5 P5 Preoperative identification and biopsy of sentinel lymph nodes using contrast-enhanced ultrasound: the Tunbridge Wells experience The sensitivity of the test to detect SLN metastases was 56%, specificity 100%, negative predictive value 86% and prevalence of lymph node (LN) metastases 27%. Between October 2013 and September 2014, 99 patients had the test. Thirty patients were excluded from the final analysis. SLN were visualised in 86% and 74% had a successful SLN core biopsy. The sensitivity was 69%, specificity 100%, negative predictive value 90% and prevalence of LN metastases 25%. p Conclusion: The percentage of SLN visualised and successfully biopsied in TWBC decreased between the two study periods. These findings may represent a normal fluctuation of the test’s performance, be a function of missing data in retrospective collection or the cumulative ‘learning- curves’ of newly appointed radiologists. This warrants further analysis. Conclusion: MRI tumour size assessment particularly for DCIS should be interpreted with caution. P8 MRI evaluation of multifocality and contralateral disease in lobular breast cancer: what can we learn from one region’s experience? Claire Magee1,2, Keith Lowry1 1Belfast City Hospital, Belfast Health and Social Care Trust, Belfast, UK; 2NIMDTA, Belfast, UK Breast Cancer Research 2015, 17(Suppl 1):P8 P6 Correlation of post-neoadjuvant chemotherapy response on MRI with final histology in breast cancer patients Anjum Mahatma, Lyn Jones, Alexandra Valencia Breast care centre, North Bristol NHS Trust, Bristol, UK Breast Cancer Research 2015, 17(Suppl 1):P6 P11 Newcastle Teaching Hospitals, Newcastle, UK Breast Cancer Research 2015, 17(Suppl 1):P9 Introduction: Breast MRI monitors tumour response to neoadjuvant chemotherapy (NAC) and guides breast-conserving-therapy (BCT). It is unclear how accurately MRI predicts pathological response. This audit investigates concordance between MRI findings and final pathology following NAC. Methods: Patients undergoing NAC between January 2011 and December 2014 were retrospectively identified. MRI was performed before, during and after NAC. At final MRI, response was graded as radiological complete response (CR no/<5 mm enhancement), partial response (PR <90% original enhancement), or no response (NR <10% reduction in enhancement). After surgery, pathological outcomes were no residual cancer (NRC), <5 mm invasive cancer/DCIS present (PRC), or >5 mm residual invasive cancer (RC). Radiological and pathological responses were either concordant or discordant. Introduction: Preoperative assessment of tumour extent is crucial in the management of breast cancer. MRI is currently indicated in cases of invasive lobular carcinoma on histology, a dense breast parenchymal pattern on 2D digital mammography (2DDM) or if there is a discrepancy between the clinical and radiological extent of disease. We compared the imaging characteristics of multifocal breast cancers on MRI, digital breast tomosynthesis (DBT), ultrasound and 2DDM to demonstrate the accuracy of each modality in the assessment of multifocal cancers. Results: Forty-six patients had NAC over 4 years (mean age 52 years), 43 IDC; three inflammatory carcinoma (not analysed). Radiological CR was seen in 19, PR in 18 and NR in six. Pathological outcome was NRC in 10, PRC in nine, and RC in 24. Responses were concordant in 30/43. BCT was attempted in 22 patients. Three required mastectomy for margins (despite two demonstrating radiological CR). MRI correctly predicted complete pathological response in 7/19 patients. In 12/19 there was residual disease despite MRI appearances. All six patients with no MRI response had residual invasive disease. Three patients with a partial MRI response demonstrated pathological complete response. Methods: A retrospective review of 74 cases over a 4-year period was conducted. We included all cases whereby MRI or DBT identified two or more lesions that were considered suspicious or highly suggestive for malignancy. We compared the sign on MRI (including morphology and enhancement characteristics) against the lesion detectability on DBT. The final histology of these lesions obtained following ultrasound- guided core biopsy, vacuum-assisted MR-guided biopsy or surgical excision was considered. P11 Results: There were 142 proven malignancies on histology out of the 74 cases, all of which were detected on MRI. The results of the MRI led to a change in surgical management in approximately 50% of cases but overstaged 16% of cases. Conclusion: MRI during NAC is useful, particularly when the MRI response is PR or NR. However, a complete radiological response predicts a complete pathological response in less than 50% of cases. Patients undergoing BCT following NAC should be aware of the risks of subsequent surgery. Conclusion: MRI is more sensitive than the other three imaging modalities combined in accurately identifying multifocal breast cancer; however, DBT is still a useful adjunct in the evaluation of multifocal disease. There was no correlation between the pathological subtype and the non-detectability of multifocal cancer on the combined imaging modalities. P9 Use of MRI to predict response following neoadjuvant chemotherapy for breast cancer: how accurately can it guide surgical choice? Sue Tan, Simon Lowes, Carol Ellen Holmes, Nidhi Sibal, Lesley McLean, Nerys Forester 9 Use of MRI to predict response following neoadjuvant chemotherapy for breast cancer: how accurately can it guide surgical choice? Sue Tan, Simon Lowes, Carol Ellen Holmes, Nidhi Sibal, Lesley McLean, Nerys Forester P11 Comparison of MRI and digital breast tomosynthesis in the preoperative evaluation of multifocal breast cancer Bhavna Batohi, Valeria Vinci, Clare Peacock, Michael Michell, Asif Iqbal, David Evans, Juliet Morel, Keshthra Satchithananda, Reema Wasan, Rumana Rahim King’s College Hospital, London, UK Breast Cancer Research 2015, 17(Suppl 1):P11 P10 Is pretreatment assessment of the contralateral breast with MRI useful following a new diagnosis of invasive lobular cancer? Harriet Russell1,2 , Rebecca Geach1,2, Iain Lyburn1,2, Helen Massey1,2 1Thirlestaine Breast Centre, Cheltenham, Gloucestershire, UK; 2Cheltenham Imaging Centre, Cheltenham, Gloucestershire, UK Breast Cancer Research 2015, 17(Suppl 1):P10 P12 MRI guided breast biopsy: initial experience of service expansion in West Midlands Muthyala Sreenivas, Vandana Gaur UHCW NHS Trust, Coventry, UK Breast Cancer Research 2015, 17(Suppl 1):P12 Introduction: There is a reported increased incidence of contralateral disease at presentation of invasive lobular cancer (ILC). In our unit breast MRI is undertaken to assess the extent of all newly diagnosed ILC. If mastectomy is planned MRI is still carried out to assess the contralateral breast–we set out to evaluate this. Introduction: At UHCW Hospitals NHS Trust (which is the sole provider of diagnostic breast MRI for UHCW, South Warwickshire and George Elliot Hospitals NHS Trusts) the MRI guided breast biopsy service has been running since June 2011. Initially, the service was offered to patients imaged at UHCW NHS Trust. With increased experience and confidence the service is now offered to all the eight NHSBSP screening sites in the West Midlands. Here we present our experience regarding the outcomes. Methods: We reviewed 160 reports of consecutive dynamic contrast- enhanced breast MRIs of newly diagnosed ILC (January 2010−June 2015). All cases had been double reported according to the BI-RADS lexicon by two trained readers. We looked at the number of cases of BI-RADS MRM scores of 3 or above in the contralateral breast, second-look ultrasound findings, biopsy rate (U/S or MRI guided) and resultant contralateral cancer detection. Methods: Since June 2011, 50 cases were referred for MRI guided breast biopsy of which 10 cases were referred from outside the UHCW NHS Trust diagnostic imaging cohort. There were three high-risk screening cases whilst the remaining cases already had diagnosis of cancer but MRI identified further lesions. All cases were graded B3 or above on diagnostic imaging. All images were reviewed (obtained via IEP) and biopsy performed within 10 days of the initial request. Biopsy samples were sent to local hospitals for pathological analysis. Methods: Since June 2011, 50 cases were referred for MRI guided breast biopsy of which 10 cases were referred from outside the UHCW NHS Trust diagnostic imaging cohort. There were three high-risk screening cases whilst the remaining cases already had diagnosis of cancer but MRI identified further lesions. All cases were graded B3 or above on diagnostic imaging. P6 Correlation of post-neoadjuvant chemotherapy response on MRI with final histology in breast cancer patients Concordant size between MR and histology = 31/50 (62%). Discordant size between MR and histology = 19/50 (38%). For complete response: sensitivity = 46%, specificity = 86%, positive predictive value = 71% (95% CI 45−88%), negative predictive value = 70% (95% CI 53−82%), accuracy = 70%. Results: A total of 141 patients had an MRI for biopsy-proven lobular carcinoma. Within this regional cohort the incidence of additional contralateral and multifocal disease was 2.13% and 13.4% respectively. Conclusions: The incidence of contralateral lobular disease is 2.13%, within our reasonably large study population, significantly less than the currently cited 15%. Our study does show a significant increase in detection of multifocal disease in the same breast by MRI. Based on our results consideration should be given to exploring the use of tomography or contrast-enhanced mammography prior to MRI to attempt to detect further disease. This could potentially expedite patient care. Results: A total of 141 patients had an MRI for biopsy-proven lobular carcinoma. Within this regional cohort the incidence of additional contralateral and multifocal disease was 2.13% and 13.4% respectively. Conclusion: Good specificity but low sensitivity in line with published literature. The time interval between second MRI and time to surgery did not affect the ability of MRI to predict response. Presence of DCIS and LCIS (42%) influenced MRI-histology discrepancy. Hormonal-positive, Her2- positive and triple-positive were more likely to show size discrepancy compared with other hormone profiles. Recommendation: We plan to integrate DWI into our reports to increase sensitivity. Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Page 5 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Conclusion: The incidence of ‘conventional imaging occult’ contralateral disease in ILC may be lower than initially reported. The routine use of MRI to assess the contralateral breast is potentially questionable. P9 Use of MRI to predict response following neoadjuvant chemotherapy for breast cancer: how accurately can it guide surgical choice? Sue Tan, Simon Lowes, Carol Ellen Holmes, Nidhi Sibal, Lesley McLean, Nerys Forester Newcastle Teaching Hospitals, Newcastle, UK Breast Cancer Research 2015, 17(Suppl 1):P9 P10 P10 Is pretreatment assessment of the contralateral breast with MRI useful following a new diagnosis of invasive lobular cancer? Harriet Russell1,2, Rebecca Geach1,2, Iain Lyburn1,2, Helen Massey1,2 1Thirlestaine Breast Centre, Cheltenham, Gloucestershire, UK; 2Cheltenham Imaging Centre, Cheltenham, Gloucestershire, UK Breast Cancer Research 2015, 17(Suppl 1):P10 P15 Comparative study of radiation dose between tomosynthesis and standard compression views in mammography * P13 Can public consultation effectively optimise the design of a patient information leaflet about breast magnetic resonance imaging? Lyn Jones, Anna Mankelow, Joanne Robson, Anjum Mahatma, Alice Pocklington, Alexandra Valencia Bristol Breast Care Centre at North Bristol NHS Trust, Bristol, UK Breast Cancer Research 2015, 17(Suppl 1):P13 P15 Comparative study of radiation dose between tomosynthesis and standard compression views in mammography Introduction: Objectives were a comparative study of the radiation dose of two-view digital breast tomosynthesis (DBT) and two-view spot compression views in a symptomatic breast service. Introduction: Breast magnetic resonance imaging (MRI) involves multiple aspects that are unique to a medical environment and may seem frightening and strange to a person from a non-medical background (the tunnel, no credit cards, keys or watches, loud noises, intravenous pump injector). The purpose of an information leaflet is to inform people about what they should expect, and to prepare them for the experience. During public consultation about breast MRI, we discovered that women considered the current information provided by the NHS (from several different hospitals) to be inadequate. They told us that their experience of the process of breast MRI had been more distressing that it would have been had they been better informed. We decided to ask their advice on the design of an information leaflet to see if it could be optimised to better prepare women for the experience. Methods: Two hundred patients were included in the study, 100 who had undergone two-view spot compression views and 100 two-view DBT. DBT was carried out using GE Seno Claire with an iso-dose setting and grid system. A retrospective computer-based search of patients in the two categories was undertaken and the accumulative dose for each technique was identified and recorded, as was the thickness of the breast from the original cc mammogram projection. The percentage variance of dose between DBT and spot compression views was calculated according to breast thickness. Results: The mean accumulative glandular dose for the whole group regardless of breast thickness was 2.84 for DBT compared with 3.50 for spot compression views. In this patient population, the AGD was lower for DBT than for FFDM in 64 % of the patients. When patients were categorized according to breast thickness, the accumulative glandular dose of DBT was on average 22 % less than spot compression mammography with a reduction ranging from 53 to 1 %. P14 Contrast-enhanced spectral mammography: what is the ‘added value’ in a symptomatic setting? Initial findings from a UK centre Sarah Tennant1, Eleanor Cornford1, Jonathan James1, Helen Burrell1, Lisa Hamilton1, Yan Chen2 1Nottingham University Hospitals NHS Trust, Nottingham, UK; 2Loughborough University, Leicester, UK Breast Cancer Research 2015, 17(Suppl 1):P14 Bhavna Batohi, Juliet Morel, Reema Wasan, Asif Iqbal, David Evans, Jane Goligher, Michael Michell, Clare Peacock, Rumana Rahim, Keshthra Satchithananda King’s College Hospital, London, UK Breast Cancer Research 2015, 17(Suppl 1):P16 Introduction: Digital breast tomosynthesis (DBT) is increasingly used for the further assessment of mammographically detected abnormalities due to its superior specificity compared with 2D digital mammography (2DDM). In this study we evaluate the positive predictive value (PPV) of mammographic features on DBT and assessment categories as per the Royal College of Radiologists (RCR) breast group classification system. Introduction: Contrast-enhanced spectral mammography (CESM) is a new technology. Dual energy acquisitions during one exposure yield two sets of images: a low energy (LE) set, equivalent to standard full field digital mammography (FFDM); and a recombined set displaying contrast uptake. In our symptomatic breast service, specific patients, including those with a P4/5 clinical abnormality are offered CESM instead of FFDM. Despite encouraging data from Europe and the USA, there are, until now, no UK data to support its use in this setting. y g g g y Methods: Women recalled following routine screening mammograms underwent bilateral 2DDM and DBT over an 18-month period. Experienced screening radiologists prospectively evaluated each case, documenting mammographic sign, size and classification according to the RCR breast group guidelines. DBT findings and final pathology were then correlated. Methods: Retrospective multi-reader review of 50 consecutive patients undergoing CESM. Anonymised LE images were reviewed and given a score for suspicion of malignancy. At least 3 weeks later, the entire examination (LE and recombined images) was reviewed. Pathology data were obtained for all cases. Differences in performance were assessed using receiver-operative characteristic (ROC) analysis. Sensitivity, specificity and lesion size (vs. MRI or histopathology) were analysed using a two-way independent t test. Results: A total of 759 abnormalities were included. On DBT, 221 (29 %) were normal. Of the remaining 538, there were 207 circumscribed masses, 89 spiculate masses, 156 microcalcifications, 35 distortions and 51 asymmetric densities. Final histology revealed 204 malignant and 334 benign lesions. P15 Comparative study of radiation dose between tomosynthesis and standard compression views in mammography * There was no evidence in this study that dose reduction with DBT significantly increased with increasing breast thickness. Methods: Public consultation was used to identify aspects of breast MRI that required explanation in an information leaflet and how they would like the information presented. We incorporated their suggestions into our new design and asked for comments at a second public consultation session. Results: The need for intravenous access, the tunnel, the nature of the loud noises that changed during the scan and knowledge that the radiographers could see and hear them throughout the scan were all emphasised as requiring explanation. The public suggested the use of multimedia including links to videos, sounds and personal accounts of experience. Conclusion: The radiation dose of patients undergoing two-view DBT on average showed a significant reduction compared to two-view spot compression views. The dose reduction may be attributed to the grid and iso-dose technology used in the selected DBT system. Conclusion: Our new leaflet has been approved by the public and patients. P10 Is pretreatment assessment of the contralateral breast with MRI useful following a new diagnosis of invasive lobular cancer? * All images were reviewed (obtained via IEP) and biopsy performed within 10 days of the initial request. Biopsy samples were sent to local hospitals for pathological analysis. Results: Forty-eight out of 50 cases were considered for biopsy. Two cases were deemed benign (one on review of diagnostic MRI and one case on second-look US). There were 24 malignancies (50 % of all cases). A follow-up Results: Of the 160 cases, 23 (14.4%) had an indeterminate or suspicious lesion reported in the contralateral breast. Three of these were contralateral cancers that had already been diagnosed by conventional imaging prior to MRI examination. Seventeen (10.6%) had second-look ultrasound of the contralateral breast: 15 lesions were subsequently biopsied in 11 women. Following negative second-look ultrasounds, two women had MRI-guided biopsy. MRI and subsequent work-up identified three women (1.9%) with previously undiagnosed contralateral malignancies. These were a 5 mm invasive ductal cancer, a 16 mm DCIS and a multicentric ILC. Results: Forty-eight out of 50 cases were considered for biopsy. Two cases were deemed benign (one on review of diagnostic MRI and one case on second-look US). There were 24 malignancies (50 % of all cases). A follow-up Page 6 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 23 % vs. 31 %, p <0.025). ROC analysis showed CESM performance was better than LE alone (AUC 0.933 vs. 0.848, p <0.05). MRI or surgical excision recommendation was made as necessary for non- malignant cases in the biopsy report. g p y p Conclusion: MRI guided breast biopsy has been successfully established at UHCW NHS Trust which currently serves as the regional referral centre in the West Midlands. Conclusion: This preliminary study demonstrates the additional clinical utility of CESM in symptomatic patients. Its potential use in other clinical settings (e.g. screening of high-risk women) requires evaluation. P15 Comparative study of radiation dose between tomosynthesis and standard compression views in mammography Alice Bannister, Julie Scudder Guys and St Thomas’ NHS Foundation Trust, London, UK Breast Cancer Research 2015, 17(Suppl 1):P15 P15 Comparative study of radiation dose between tomosynthesis and standard compression views in mammography Alice Bannister, Julie Scudder Guys and St Thomas’ NHS Foundation Trust, London, UK Breast Cancer Research 2015, 17(Suppl 1):P15 P18 P18 Initial outcome of contrast-enhanced spectral mammography in lesion characterisation in the mammographically dense breast Zeiad Hussain, Peter Duggan, Julie Scudder, Sarah McWilliams Guys and St Thomas NHS Trust, London, UK Breast Cancer Research 2015, 17(Suppl 1):P18 Introduction: The accuracy of mammography is limited in the dense breast. Contrast-enhanced spectral mammography (CESM) is a relatively new alternative with promising results. The aim of this study is to assess the sensitivity, specificity, positive predictive value (PPV) and negative predicative value (NPV) of CESM in the detection and characterization of breast lesions. Introduction: The purpose of this study was to assess the diagnostic performance/utility of digital breast tomosynthesis (DBT) in symptomatic patients in a multidisciplinary clinical setting. Methods: The study was registered as a Cambridge University Hospital (CUH) service evaluation audit. Patients from the CUH symptomatic breast clinic from October 2014 to February 2015 were included in the study. Patients were included as clinic workflow permitted and DBT and full field digital mammography (FFDM) (SenoClaire, GE) were interpreted prospectively. Image quality of the DBT and 2D synthetic images were rated on a 5-point scale compared to FFDM. The imaging findings were graded on both FFDM and DBT as normal, benign, indeterminate, suspicious or malignant. Patients were clinically examined and additional ultrasound was carried out as appropriate. FFDM and DBT findings were correlated with the ultrasound findings and when performed to histopathology. Methods: Retrospective review of the prospectively maintained database of patients who underwent CESM over a period of 6 months. The sensitivity, specificity, PPV and NPV of CESM were assessed against the histopathology result. In a subgroup of eight patients, the CESM outcome was also compared to MRI scan results. In addition, patients’ demographics and correlation with mammography and ultrasound outcomes were obtained. Results: Twenty-four patients (23 female) underwent CESM over a 6-month period. CESM was found to have a sensitivity of 76 %, specificity of 66 %, PPV of 93 % and NPV of 66 %. The median maximal lesion dimension (MMLD) on CESM was 29 mm. In a subgroup of eight patients MRI was also performed where the MMLD on MRI was 22 mm and on CEM was 20 mm. MRI accurately diagnosed all malignant lesions (8/8) while CEM demonstrated a false negative results in 2/8 patients. P21 P21 A review of BRCA gene carrier demographics in Wales Tom Evans1, Jenny Long1, Georgina Devenish1,2, Mike Lewis3, Damian Bailey3, Kate Gower Thomas2,3, Alexandra Murray1,4 1Cardiff and Vale University Health Board, Cardiff, UK; 2Breast Test Wales, Cardiff, UK; 3University of South Wales, Pontypridd, UK; 4All Wales Genetics Service, Cardiff, UK Breast Cancer Research 2015, 17(Suppl 1):P21 Arul Selvan1, Yan Chen2, Alastair Gale2 1Sheffield Hallam University, Sheffield, UK; 2Loughborough University, Loughborough, UK Breast Cancer Research 2015, 17(Suppl 1):P19 Poster presentation: Being able to accurately determine the extent of a possible malignancy on a mammogram is an important task as this can affect the potential follow-up surgical treatment that a woman receives after breast screening. It is known that this can be a difficult task, particularly where the lesion has diffuse abnormalities. A potential computer-aided approach is to employ hierarchical clustering-based segmentation (HCS) and this interactive educational exhibit dynamically demonstrates this technique. HCS is an unsupervised segmentation process that when applied to an image yields a hierarchy of segmentations based on image pixel dissimilarities and so can be used to highlight areas in the mammographic image to aid interpretation. Introduction: Women who inherit a mutated copy of the BRCA-1 or BRCA-2 genes have a higher lifetime risk of developing breast cancer. There have been no large epidemiological studies looking at BRCA- positive patients in the UK. p p Methods: Across the All Wales Genetics Service, individuals with confirmed BRCA mutation, since formal testing began (1995) to 1 January 2015, were included–identified from a prospectively gathered database. Genetics case notes were obtained and retrospective analysis carried out. Results: A total of 419 females with mean age 47 (19−81) were included in the study. Of these, 206 were identified using diagnostic testing with the remaining 213 undergoing predictive testing. Of the predictive group who subsequently had cancer, 18 (78 %) developed breast cancer. Seven (39 %) had wide local excision (WLE), six (33 %) had single mastectomy while the remaining five (28 %) had bilateral mastectomies as their primary operation. Five of the predictive group (22 %) had ovarian cancer. Of these, four (80 %) went on to have prophylactic breast surgery too. Of the 13 patients who underwent WLE or single mastectomy, four (31 %) went on to have completion risk reduction mastectomies (RRM). P18 In five of the 11 patients who initially had mammography and four of the 19 patients who initially had ultrasound scan demonstrating indeterminate lesions, CEM correctly characterised the lesions. Results: A total of 134 patients were included. Eighty-five per cent of the synthetic images were considered qualitatively similar or better than the FFDM images. Nineteen lesions were graded as indeterminate, 10 lesions were graded as suspicious and six lesions were graded as malignant on FFDM, whereas three lesions were graded as indeterminate, four lesions were graded as suspicious and 14 lesions were graded as malignant on DBT. Seventy-nine per cent of the indeterminate lesions on FFDM were downgraded accurately by DBT and 16 % were upgraded accurately by DBT. There was one false negative and three false positives with FFDM and DBT. y Conclusion: CESM is a promising and affordable technique in the assessment of suspicious lesions in the mammographically dense breast. CEM has a good sensitivity, specificity, PPV and NPV. Conclusion: DBT helps characterise indeterminate lesions more accurately compared with FFDM in the symptomatic setting. P19 A perceptual aid to delineating the extent of potential mammographic abnormalities Arul Selvan1, Yan Chen2, Alastair Gale2 1Sheffield Hallam University, Sheffield, UK; 2Loughborough University, Loughborough, UK Breast Cancer Research 2015, 17(Suppl 1):P19 P14 The PPVs were 97.7 % for spiculate masses, 65.7 % for distortions, 35.8 % for microcalcifications, 16.9 % for circumscribed masses and 5.8 % for asymmetric densities. Results: Fifty females, mean age 49. Thirty-four (68 %) patients had biopsy- proven malignancy, 16 (32 %) were benign. CESM was more sensitive than LE alone (94 % vs. 86 %, p <0.025), more specific than LE alone (84 % vs. 63 %, p <0.025) and showed better size estimation (mean size difference Conclusion: DBT allows more accurate assessment of mammographic lesions without the impedance of overlying tissues. Spiculate masses have the highest PPV on both 2DDM and DBT. Although the PPV for Page 7 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 this approach in increasing the perceptual recognition of abnormal appearances. asymmetric densities appears relatively low on DBT, this is still nearly twice that of 2DDM. Better understanding of the likelihood of tomographic signs indicating malignancy will increase the value of DBT. P20 Diagnostic implications of digital breast tomosynthesis in symptomatic patients Sanjeeva Ramasundara1,2, Lorraine Tucker1,2, Matthew Wallis1, Peter Britton1, Penny Moyle1, Kethryn Taylor1, Ruchi Sinnatamby1,2, Alan Freeman1, Matthew Gaskarth1, Fiona Gilbert1,2 1Cambridge University Hospital, Cambridge, UK; 2University of Cambridge, UK Breast Cancer Research 2015, 17(Suppl 1):P20 P22 Breast Cancer Research 2015, 17(Suppl 1):P24 Audit of high prevalent breast screening recall rates: Torbay Hospital Alexander Crowther, Rebecca Green Torbay DGH, Torquay, UK Breast Cancer Research 2015, 17(Suppl 1):P22 Introduction: The aim was to investigate the impact of index of multiple deprivation (IMD) and ethnicity on breast cancer screening uptake in the North West of England. Introduction: The aim was to investigate the impact of index of multiple deprivation (IMD) and ethnicity on breast cancer screening uptake in the North West of England. Methods: Data for screening uptake rates were collected from 2005 to 2014 using data from the North West Breast Screening Units and the annual breast screening statistics reports. These were correlated with the IMD published in 2007 and 2010. The uptake rates were also correlated with ethnicity data obtained from the Census 2011. Then, the results for ethnicity were adjusted for IMD. Introduction: The target percentage of women recalled after prevalent round breast screening is <7 % with minimum standards <10 %. Torbay Hospital’s prevalent round recall is high at 11.4 %. We plan to assess patterns of recall by category to see if any particular reason for recall could be decreased. Methods: Retrospective audit of 12 months of prevalent round recalls March 2013-February 2014. All age groups were included. Each recall was grouped into one/more of the following categories: calcification, well- defined mass, ill-defined mass, asymmetric density, distortion, clinical, other. We will calculate the proportion of recalls per group that proved to be malignancy and assess to see if any category was a poor predictor of malignancy. All histology proven malignancies from 2012/13 and 2014/15 will also be categorised by group. Results: Both prevalent and incident uptake rates have declined from 2005/06 to 2013/14. Deprivation was shown to negatively correlate with breast screening uptake in all rounds, the strongest correlation being with prevalent screening rounds (IMD 2007 p = 0.005 and 2010 p = 0.016). The incident round negative correlation was IMD 2007 p = 0.002 (significant) and IMD 2010 p = 0.163 (not significant). For ethnicity, the Caucasian population showed a positive correlation while Asian, a negative correlation. This was more significant in the Pakistani and Bangladeshi groups. Interestingly, when the results were adjusted for deprivation, ethnicity did not show a significant correlation with uptake rates. Results: There were 215 recalls for ages 49–69, 15 proven malignancies. P23 P23 Outcome of ultrasound of the mammographically normal contralateral breast in patients recalled to the screening assessment clinic Preet Hamilton, Simon Lowes, Sheetal Sharma, Alicia Wright, Alice Leaver Gateshead Hospitals NHS Trust, Gateshead, UK Breast Cancer Research 2015, 17(Suppl 1):P23 Introduction: Tomosynthesis is a new technology that is being used increasingly to evaluate the breast for assessment in the UK. It has, however, been approved as a screening tool in the United States, Canada and several European countries. We implemented tomosynthesis in the Assessment Clinic at Avon Breast Screening Unit (ABSU) last year as recommended by the NHSBSP. A retrospective audit of 134 consecutive cancers diagnosed from 9 June 2014 to 31 December 2014 was performed. The aim was to evaluate whether tomosynthesis gives additional information to increase the grading of mammographic features of a lesion seen on initial screening mammography and increase the assessor’s confidence. Introduction: Women diagnosed with breast cancer are at increased risk of contralateral breast cancer; some of these cancers will be synchronous and mammographically occult (M1). Ultrasound may detect M1 breast cancers but also benign lesions that necessitate needle testing, conferring additional patient morbidity that could be termed ‘over investigation’. Local guidelines for ultrasound of the M1 contralateral breast vary between units. We present a retrospective audit of contralateral M1 breast ultrasound within our screening assessment clinics. Methods: Screening and pathology hospital databases of 2013 and 2014 identified records of 331 women with screen-detected breast cancer. Descriptive statistics were performed. Result: A total of 134 cancers were reviewed. Sixty-six lesions were graded the same on screening mammography and the assessment tomosynthesis. Thirty-six were M5 lesions at screening and assessment. Thirty M3 or M4 lesions remained unchanged. One patient had an M3 lesion that was downgraded. Three patients had incidental cancers found on ultrasound. Sixty-four lesions were upgraded with tomosynthesis. Forty-four of 64 M3 or M4 lesions were upgraded to tomosynthesis 5. Twenty of 64 were upgraded from M3 to tomosynthesis 4. The morphology of the lesions upgraded was spiculated 30/64, 7/64 distortions and 7/64 ill-defined densities. Thirty-one of 44 tomosynthesis 5 lesions measured 10 mm or less. Results: All 331 women underwent ipsilateral breast ultrasound; 288 (87 %) underwent ultrasound of their contralateral mammographically normal (M1) breast. Six contralateral breast lesions were needle sampled: four B2 lesions, two B3 without atypia. No subsequent breast cancer has been detected in any of these patients to date. P22 77% of Ill-defined mass, 22% of distortion and 10% of calcifications recalled proved to be malignant and are the strongest predictors of malignancy. Well-defined mass and asymmetric density had 0% malignancy rates and accounted for 129 (59.4 %) of prevalent recalls. Thirteen clinical recalls (1.4 %) were also 0 % for malignancy but beyond the control of the screening service. Further audit was performed looking at the proven malignancies from 2012/13 and 2014/15, which showed a total of 33 malignancies with 13 calcifications, 17 ill-defined masses, one asymmetry, one distortion and one clinical recall. Conclusions: Our results clearly show that the more deprived an area, the lower the breast screening uptake rate. Moreover, the higher the proportion of Asian in a population, the lower the uptake rates and this is more significant in the Pakistani and Bangladeshi group compared to the Indian and Chinese. Overall the impact is most marked in the prevalent round. y y Conclusion: A high proportion of recalls (60 %) are for well-defined mass and asymmetric density which have poor predictive outcome. These groups are potential areas to decrease recall rates. A total 1.4 % of clinical recalls are beyond the control of the screening service, which would bring prevalent recalls to a compliant level of 10 %. P21 From the remaining 190 individuals in the predictive group with no cancer diagnosis, 102 (54 %) have had no risk reduction surgery, Methods: Across the All Wales Genetics Service, individuals with confirmed BRCA mutation, since formal testing began (1995) to 1 January 2015, were included–identified from a prospectively gathered database. Genetics case notes were obtained and retrospective analysis carried out. g g g p g p A set of 15 known difficult FFDM mammographic cases were selected from PERFORMS case sets where expert radiologists had previously delineated the extent of various abnormalities. Regions of interest (ROI) around these abnormalities were extracted from the DICOM images and processed using HCS algorithms resulting in a set of paired original mammographic ROI images and related HCS processed ROI images. In the exhibit these paired images are presented and delegates interactively select which of the pair they think best identifies abnormality extent. The original expert delineated abnormality is then provided as feedback. Over the course of the conference, data will be collected on how useful the HCS approach has been found and this information fed back to participants. The learning objectives are to demonstrate the potential of Page 8 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 32 (17 %) RRM only, 31 (16 %) BSO only and 25 (13 %) underwent both procedures. 32 (17 %) RRM only, 31 (16 %) BSO only and 25 (13 %) underwent both procedures. Impact of index of multiple deprivation and ethnicity on breast screening uptake in the North West of England 1 1 2 1 y , ; g g Prevention Centre, University Hospital of South Manchester NHS Foundation Trust, Manchester, UK y , ; g g Prevention Centre, University Hospital of South Manchester NHS Foundation Trust, Manchester, UK Does tomosynthesis increase confidence in grading the suspicious appearance of a lesion? An audit of cancers diagnosed in the assessment clinic using tomosynthesis: initial experience at Avon Breast Screening Unit Does tomosynthesis increase confidence in grading the suspicious appearance of a lesion? An audit of cancers diagnosed in the assessment clinic using tomosynthesis: initial experience at Avon Breast Screening Unit * Gillian Clark, Alexandra Valencia Avon Breast Screening Unit, Bristol Breast Care Centre, North Bristol NHS Trust, Bristol, UK Breast Cancer Research 2015, 17(Suppl 1):P25 Gillian Clark, Alexandra Valencia Avon Breast Screening Unit, Bristol Breast Care Centre, North Bristol NHS Trust, Bristol, UK Breast Cancer Research 2015, 17(Suppl 1):P25 Gillian Clark, Alexandra Valencia Avon Breast Screening Unit, Bristol Breast Care Centre, North Bristol NHS Trust, Bristol, UK Breast Cancer Research 2015, 17(Suppl 1):P25 P24 Impact of index of multiple deprivation and ethnicity on breast screening uptake in the North West of England Jayeshwaraj Bhola1, Anil Jain1,2, Philip Foden1 1The University of Manchester, UK; 2The Nightingale Centre and Genesis Prevention Centre, University Hospital of South Manchester NHS Foundation Trust, Manchester, UK Breast Cancer Research 2015, 17(Suppl 1):P24 p Conclusion: There is variation in the surgical management of BRCA positive patients in Wales. This has implications for service allocation and demands for screening for these high-risk patients. P26 P26 An audit of marker placement in stereotactic guided biopsy Jane Prady1,2, Lucy Hill1,3, Alison Gray1, Alison Gilchrist1 1South East of Scotland Breast Screening Service, Edinburgh, UK; 2Queen Margaret University, Edinburgh, UK; 3British Society of Breast Radiology, Edinburgh, UK Breast Cancer Research 2015, 17(Suppl 1):P26 The 3.5-year to 13.5-year follow up of 137 lesions of uncertain malignant potential (B3 lesions) diagnosed by vacuum-assisted biopsy alone Karen Mullin, Anne Turnbull, Sharma Puri, Mark Bagnall Royal Derby Hospital, Derby, UK Breast Cancer Research 2015, 17(Suppl 1):P28 An audit of marker placement in stereotactic guided biopsy Jane Prady1,2, Lucy Hill1,3, Alison Gray1, Alison Gilchrist1 1South East of Scotland Breast Screening Service, Edinburgh, UK; 2Queen Margaret University, Edinburgh, UK; 3British Society of Breast Radiology, Edinburgh, UK Breast Cancer Research 2015, 17(Suppl 1):P26 Introduction: Vacuum-assisted biopsy (VAB) was introduced in Derby in 2001, as the second procedure after 14g core biopsy. We present 3.5-year to 13.5-year follow up of cases where B3 lesions have been managed with VAB alone. Introduction: Anecdotal evidence suggests that there is a greater incidence of marker migration using large volume sampling techniques in stereotactic guided breast biopsies. Methods: The NBSS and local BASO databases were searched from January 2002 to December 2011 for all cases with B3 histopathology, a VAB procedure and no surgery. Screening and symptomatic women were included. Methods: Prospective study of 130 biopsies with markers done between June and December 2014. Markers more than 10 mm from the target lesion were considered migrated. The aim of the audit was to quantify the number of markers migrating, distance and direction of migration and conditions under which markers migrate. Results: There were 137 women who met the criteria. The pathologies found are presented in Table 1. The cases where atypia was found were individually discussed at MDT to ensure that the abnormal feature had either been excised or very well sampled. Only one breast cancer has developed at the same site in a woman who had 5 mm calcification excised at VAB. This lobular cancer was identified 4 years later at recall from annual surveillance. Five other cancers have developed in the 137 cases, one contralaterally and four different lesions in different sites in the same breast. Conclusion: This study provides further evidence for the safety of the use of VAB alone in the diagnosis of B3 lesions in the longer term. P29 P29 Quantitative study: should vacuum-assisted biopsy be the first biopsy approach in breast interventional techniques in stereotactic guided microcalcifications rather than 14 gauge core needle biopsy? Anuma Shrestha, Louise Wilkinson, Rosalind Given-Wilson, Judi Curtis St George’s Healthcare NHS Trust, London, UK Breast Cancer Research 2015, 17(Suppl 1):P29 Quantitative study: should vacuum-assisted biopsy be the first biopsy approach in breast interventional techniques in stereotactic guided microcalcifications rather than 14 gauge core needle biopsy? Anuma Shrestha, Louise Wilkinson, Rosalind Given-Wilson, Judi Curtis St George’s Healthcare NHS Trust, London, UK Breast Cancer Research 2015, 17(Suppl 1):P29 Conclusion: This audit found that marker migration occurred predominantly within lucent breast tissue and using the latero-medial approach when using the Bard Encor system. g , , Breast Cancer Research 2015, 17(Suppl 1):P29 P26 g Results: A total of 12.3 % had migrated markers: 10.7 % from use of the Bard Encor system and 1.5 % from use of the Bard Vacora system. The greatest marker migration occurred using a latero-medial approach. The majority of migrated markers were deeper than the target lesion. Marker migration was significantly greater using the Encor system within lucent breast tissue. Firstly, further audit is required incorporating lesion size, routine vacuuming of the cavity before deployment of marker, specific sequencing of marker films, correlation of compressed breast thickness and target depth, clinical impact of marker migration and possible development of expanding marker. Secondly, the breast screening service should provide guidelines regarding distances, thresholds and targets for marker migration. P23 Conclusion: Two years of routine contralateral ultrasound has yielded no cancers but also very few benign biopsies. Ongoing audit and discussion of risk/benefit to patients is indicated. Page 9 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 but in six cases a second Intact biopsy and/or a VACB was required to complete that excision, with extra cost implications. In two patients with M3 microcalcification the Intact pathology demonstrated ductal carcinoma in situ, leading to surgical wide local excision. Conclusion: Tomosynthesis is excellent at showing the spiculate nature of lesions, upgrading the appearance of a lesion from M3 and M4 to tomosynthesis 5 which increases the assessor’s confidence during the assessment clinic. It is also excellent in helping identify small suspicious lesions of 10 mm or less. However, ultrasound should always be performed in addition to tomosynthesis as lesions may rarely be downgraded. Conclusion: Our early experience shows Intact as a reliable and effective tool for handheld diagnostic and/or therapeutic excision of selected breast lesions. P32 Use of WHO checklist in interventional breast radiological procedures Trupti Kulkarni1,2, Andrew O’Connor1 Use of WHO checklist in interventional breast radiological procedures Trupti Kulkarni1,2, Andrew O’Connor1 1University Hospital Aintree, Liverpool, UK; 2Clatterbridge Breast Unit, Wirral, UK P30 Safety of vacuum-assisted biopsy/mammatome guided, non-operative management of B3 lesions without atypia: a 7-year follow-up study Alex Wilkins1,2, Peter Kneeshaw1, Penelope McManus1, Kartikae Grover1, Caroline Bradley1, Ayesha Rahman1, Anne Hubbard1 1Castle Hill Hospital, Cottingham, UK; 2Hull York Medical School, Hull, UK Breast Cancer Research 2015, 17(Suppl 1):P30 Introduction: Increasing awareness of safety in healthcare provision has resulted in incorporation of risk-reducing strategies. The WHO checklist is now increasingly being used by interventional radiology. Is it relevant for the interventional breast radiologist? Methods: A questionnaire for assessing awareness and use of the WHO checklist used for surgical procedures (or a modified checklist) was devised. Responses were collected and analysed via Survey Monkey. Introduction: B3 management balances safe treatment of potential malignancy against the morbidity of surgical excision of benign lesions. Vacuum-assisted biopsy (VAB) increases diagnostic accuracy, removing some lesions entirely without surgery. Few follow-up data are available to assess the safety and effectiveness of this approach. Results: Eighty-one complete responses were received and analysed. In total, 93.83 % were aware of the WHO checklist; 83.95 % worked in departments where this was used by IR; and 46.91 % used the checklist individually or as a department. The list was locally devised in 43.21 %. Of those who did not employ use of the list, 27.16 % had considered its use. A total of 24.69 % had never considered using it. Fifty-four per cent opined it was relevant to a therapeutic vacuum-assisted procedure with various individual procedures having scores ranging from 12 to 47 %. Adherence to CQC standards was cited as the reason for use of the checklist. Naysayers quoted increase in time required and poor work flow as reasons for not using it. Methods: A total of 215 patients with B3 biopsies without atypia were identified using Labcentre histopathology codes at a single centre. Hospital and NBSS records were analysed to identify patients who were treated with VAB and mammographic surveillance alone and to determine outcome over a follow-up period of 52−149 months (median 85). Local Labcentre and regional Pathlinks histopathology records were independently checked. Mammograms of ipsilateral re-presentations were reviewed by a consultant radiographer and consultant radiologist to determine whether lesions developed at the site of B3 biopsy. P27 P27 Handheld ultrasound-guided 20 mm basket Intact breast lesion excision system biopsy for excision of benign breast lesions Simon Lowes, Alice Leaver, Alice Townend, Jackie Westgarth, Peter Newton, Dianne Hemming, Alan Redman Gateshead Hospitals NHS Trust, Gateshead, UK Breast Cancer Research 2015, 17(Suppl 1):P27 Introduction: Stereotactic guided 14 gauge core needle biopsy (14GCNB) and vacuum-assisted biopsy (VAB) are the two commonly used biopsy methods for obtaining an accurate diagnosis for microcalcifications. Table 1(abstract bcr3790) Table 1(abstract bcr3790) Pathology Initial biopsy Final pathology CSL 39 37 Papillary lesion, no atypia 51 50 Papillary lesion, atypia 2 0 CCC atypia/flat epithelial atypia 12 0 ADH/apocrine atypia/AIDEP 9 1 Lobular neoplasia 6 5 Mucocoele-like lesion 5 0 Other B3 atypia 5 0 Fibroepithelial lesion 2 0 B2 4 40 Fibroadenoma 0 3 B1 2 0 VAB only 0 1 Total 137 137 Table 1(abstract bcr3790) Introduction: In selected patients, our Unit has recently moved from handheld ultrasound-guided vacuum-assisted core biopsy (VACB) piecemeal acquisition of tissue to the handheld Intact Breast Lesion Excision System (Intact). Intact removes a single piece of tissue, potentially allowing radiologists to excise the entire lesion as well as allowing pathologists to visualise lesion architecture more easily and to calculate margins. We evaluated our early experience of excising benign or likely benign breast lesions using the 20 mm Intact. y g g Methods: Prospective data collection was performed on all patients undergoing handheld ultrasound-guided Intact excision under local anaesthetic in 2014 and 2015, which comprised 19 lesions in 18 female patients aged 29−73. Results: The device was technically straightforward to operate and well- tolerated by patients with no significant complications. Handheld needle orientation was difficult within dense glandular tissue (only one acquisition is possible per needle), but improved with increased operator experience. Achieving adequate analgesia required higher quantities of local anaesthetic than for the equivalent VACB. Pathologists found specimens easier to interpret than VACB samples. In all cases adequate excision was completed sonographically at one outpatient appointment, Page 10 of 13 Page 10 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 some lesions entirely without surgery. Few follow-up data are available to assess the safety and effectiveness of this approach. Retrospective review of 399 patients who underwent biopsy for breast microcalcification during screening assessment from April 2012 to March 2013 was used to evaluate the performance and cost-effectiveness of both methods. P27 Methods: A total of 129 patients with B3 VAB with atypia were identified using Labcentre histopathology codes at a single centre. Hospital and NBSS records were analysed to identify patients treated with VAB and mammographic surveillance alone and to determine outcome over a follow-up period of 52−142 months (median 85). Methods: The repeat biopsy rate, diagnostic accuracy, time taken and cost of both methods was calculated. Microsoft Excel (2010) and SPSS 22 were used for statistical analysis. Results: The repeat biopsy rate for 14GCNB was 10 % and VAB was 6 %. Specificity, PPV and NPV were all 90 % or higher when compared against post-surgical final diagnosis in both methods. The sensitivity of VAB was 93.75 % vs. 71.88 % for 14GCNB for first biopsy. There was no significant difference in procedure time between two methods (p = 0.291). VAB necessitated almost double the rate of clip deployment compared with 14GCNB. The cost of VAB would be £69,922 greater than 14GCNB if used as the first-line biopsy method in this series. Results: Ten per cent progressed directly to surgery (13/129). A total of 116 were followed mammographically after VAB with no surgical intervention (49 ADH, 2 ALH, 21 LCIS, 44 atypia (not otherwise specified)). Nine patients re-presented to the service with invasive carcinoma (six ipsilateral) and two with DCIS (both ipsilateral) between 12 and 80 months. The ipsilateral re-presentation rate was highest for ADH (5/49) and LCIS (2/21). In the absence of ADH or LCIS, the only ipsilateral representation was one low-grade DCIS, 62 months after VAB. Conclusion: Re-presentation with ipsilateral carcinoma following VAB excision for ADH and LCIS is comparable to surgical excision for ADH and LCIS. National guidance is required. Conclusion: This study found VAB to have higher sensitivity than 14GCNB. There was also a trend for lower repeat biopsy rate, higher diagnostic accuracy and lower surgical upgrade with VAB. If VAB had been used as the first biopsy method for microcalcifications, the cost would have been significantly higher. 14GCNB is a cost-effective but less sensitive first biopsy method for selected microcalcifications. P32 P32 Use of WHO checklist in interventional breast radiological procedures Trupti Kulkarni1,2, Andrew O’Connor1 1University Hospital Aintree, Liverpool, UK; 2Clatterbridge Breast Unit, Wirral, UK B t C R h 2015 17(S l 1) P32 P33 Breast biopsy in patients on anti-coagulants: is new guidance needed? Trupti Kulkarni1,2, Andrew O’Connor1 1University Hospital Aintree, Liverpool, UK; 2Clatterbridge Breast Unit, Wirral, UK Breast Cancer Research 2015, 17(Suppl 1):P33 Breast Cancer Research 2015, 17(Suppl 1):P33 P32 Use of WHO checklist in interventional breast radiological procedures Trupti Kulkarni1,2, Andrew O’Connor1 Conclusion: Breast radiological intervention procedures, although low risk and with low complications, remain health interventions. An adverse event should not be a necessary trigger for change of practise. Opinion on use of additional safeguards such as an intervention checklist, while divided, suggested that a modified checklist is called for in complex procedures involving recall of patient at a different date, multiple radiologists involved and therapeutic procedures under vacuum guidance. Results: Twenty per cent had excision biopsy (42/215) of which <5 % (2/42) contained carcinoma. A total of 144 patients had VAB which identified 30 high-risk cases analysed separately (DCIS, B4 or atypia). In total, 114 B3 lesions without atypia (on either core biopsy or VAB) were followed mammographically after VAB with no surgical intervention. Four patients re-presented to the service with malignancy; 37, 38, 41 and 67 months after VAB. Sixty-one per cent (69/114) of individuals were screened locally 2012−2015. y Conclusion: VAB of B3 biopsies without atypia appears to be safe with no representations in the first 3 years and overall carcinoma and DCIS incidence of 3.5 % over 7 years (4/114). National guidance on B3 lesion management is required. P34 Role of mammographic templates in managing ever increasing workloads * Anuradha Anand1, Lekha Potti2, Geoff Naisby1, Sheetal Sharma2, Alan Redman2 1James Cook University Hospital, Middlesbrough, UK; 2Queen Elizabeth Hospital, Gateshead, UK Breast Cancer Research 2015, 17(Suppl 1):P36 Introduction: Recent studies have questioned the value of bone scans (BS) in staging breast cancer when a CT chest, abdomen and pelvis is also performed. We retrospectively reviewed breast cancer staging CTs and BS performed within 2 months of each other, to see if BS identified more skeletal metastases than CT. Introduction: A breast imaging report is a key component of the breast cancer diagnostic process. The report must be clear and concise to avoid ambiguity and confusion. However, substantial variation in the information provided in a breast imaging report is not uncommon to see. We sought to develop a report template containing a summary of all essential information pertinent to the surgeons and the radiologists. Methods: Our study was performed at the breast screening unit at Queen Elizabeth Hospital, Gateshead (QE) and the symptomatic breast unit at James Cook University Hospital (JCUH), Middlesbrough. Experienced radiologists blinded to primary BS reports retrospectively assessed CTs performed for primary breast cancer, known recurrence or to explain symptoms of pain. They then reviewed the same patient’s BS. CT and BS were marked positive, negative or indeterminate for skeletal metastases. Methods: Breast surgeons and radiologists were consulted as to what was required in a report and they stated breast density, correlation with clinical findings, lesion characteristics, R1−R5 category, site and size of lesion, and is clinical area of concern biopsied. A retrospective audit of 30 breast imaging reports of recently diagnosed carcinomas between October 2014 and January 2015 were reviewed to see if these were recorded. Results: Combined data from both units yielded 253 cases in total. CT and BS concurred in 217 cases. Of the remaining 36, CT identified skeletal metastases in five where BS was negative and two where BS was indeterminate. CT excluded metastases in 23 which were indeterminate on BS. BS confirmed or excluded metastases in five cases where CT was indeterminate and identified metastases in only one case which was negative on CT. This lesion proved to be benign and hence BS was false positive in this case. Results: Ten per cent of reports did not mention breast density. The most frequent information provided is lesion size (ultrasound 100 %, mammography 73 %). P37 P35 Local experience of referral for breast assessment resulting from incidental findings on CT and PET-CT studies over a 5-year period Richard Sidebottom, Jean Lee, Shaheel Bhuva, Vaishali Parulekar Oxford Breast Imaging Centre, Oxford University Hospitals NHS Trust, Oxford, UK Breast Cancer Research 2015 17(Suppl 1):P35 Breast Cancer Research 2015, 17(Suppl 1):P35 Breast Cancer Research 2015, 17(Suppl 1):P35 Breast Cancer Research 2015, 17(Suppl 1):P35 Introduction: Incidental findings of breast abnormalities from cross- sectional imaging (CT and PET-CT) are a relatively common source of referral for breast assessment at our unit. We sought to describe and quantify our local experience of these referrals and to determine which cross-sectional imaging findings were more predictive of malignancy. Introduction: The aim was to compare the histopathological and prognostic differences in DCIS between age-matched Asian and Caucasian female patients. Introduction: The aim was to compare the histopathological and prognostic differences in DCIS between age-matched Asian and Caucasian female patients. Methods: Data related to presentation, histopathology, prognosis and treatment of DCIS were gathered from 48 women from the Asian Breast Cancer Database at the Nightingale Centre, all of whom had begun with an initial diagnosis (at biopsy) of DCIS. These were compared with age- matched Caucasian patients, also diagnosed with DCIS at the time of biopsy. The total study included 96 patients. Methods: Retrospective review using radiology information system searches for mammography referrals resulting from CT and PET-CT scan findings performed over a 5-year period (July 2010−July 2015) in Oxford University Hospitals NHS Trust. Studies in patients with known active breast malignancy were excluded. Cross-sectional imaging characteristics of the abnormalities were collected including CT enhancement, PET avidity, size and shape. Assessment imaging features, subsequent biopsy and clinical outcomes were recorded. Results: Out of 48 Asian women more presented symptomatically (25, 52.1 %) compared to Caucasian women (13, 27.1 %), p = 0.012. Asian women had a larger mean value in regards to tumour size (28.48 mm) compared to Caucasian women (21.59 mm), and more progressed from an initial diagnosis of DCIS, to a final diagnosis of DCIS with an invasive component (12.5 % compared to 2.1 %), p = 0.05. However, differences in the average Van Nuys Prognostic Index score were not statistically Results: A total of 126 patients were assessed as a result of incidental breast abnormalities. P31 Introduction: Patients on anti-coagulation requiring breast biopsies are more at risk of bleeding. Newer anti-coagulants may not have a method for quantifying coagulation unlike the INR for warfarin. Also, some of these such as dabigatran do not have antidotes and rely on the body’s ability to excrete the drug which may be altered by renal function. There are no up-to-date national guidelines on breast biopsy in such patients. Introduction: Patients on anti-coagulation requiring breast biopsies are more at risk of bleeding. Newer anti-coagulants may not have a method for quantifying coagulation unlike the INR for warfarin. Also, some of these such as dabigatran do not have antidotes and rely on the body’s ability to excrete the drug which may be altered by renal function. There are no up-to-date national guidelines on breast biopsy in such patients. Methods: A questionnaire for assessing practise of breast biopsy in patients on various anti-coagulants was devised. Responses were collected and analysed via Survey Monkey. p g p y p Methods: A questionnaire for assessing practise of breast biopsy in patients on various anti-coagulants was devised. Responses were collected and analysed via Survey Monkey. Breast Cancer Research 2015, 17(Suppl 1):P31 Introduction: B3 management balances safe treatment of potential malignancy against the morbidity of surgical excision of benign lesions. Vacuum-assisted biopsy (VAB) increases diagnostic accuracy, removing Results: Seventy-eight complete responses were received and analysed. Thirty-eight per cent of respondents said they had local guidelines while Page 11 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 shape and CT enhancement features will be presented. Lesions with high avidity on PET-CT scans were more likely to be primary breast cancer on biopsy (83 % SUVmax >2.5). Of 36 breast malignancies identified, three patients underwent mastectomy surgery, 10 had wide local excision and 20 had non-surgical management. Three patient outcomes are unknown at the time of writing. 45 % used BSBR guidelines 2012. Sixty-three per cent would refer back to the GP/specialist in cases of warfarinised patients,14 % in cases of patients on clopidogrel and only 1 % of those on aspirin. Eighty-eight per cent of respondents did not have a policy for dabigatran and rivaroxaban. Practise was different in screening and symptomatic groups in 7 % due to the site of screening units away from A/E. P31 Unit policy in warfarinised patients requiring vacuum-assisted biopsy (VAB) was not available to 38 %. Anecdotally, a number of radiologists reported that they would not perform VAB in patients on clopidogrel. 45 % used BSBR guidelines 2012. Sixty-three per cent would refer back to the GP/specialist in cases of warfarinised patients,14 % in cases of patients on clopidogrel and only 1 % of those on aspirin. Eighty-eight per cent of respondents did not have a policy for dabigatran and rivaroxaban. Practise was different in screening and symptomatic groups in 7 % due to the site of screening units away from A/E. Unit policy in warfarinised patients requiring vacuum-assisted biopsy (VAB) was not available to 38 %. Anecdotally, a number of radiologists reported that they would not perform VAB in patients on clopidogrel. Conclusion: Referrals arising from incidental abnormalities identified on cross-sectional imaging have a high yield for breast malignancy (29 %). Incidental PET findings, while less often a route of referral, have the highest likelihood of identifying a malignant lesion. y g Conclusion: There is a wide variation in practise while performing biopsies in patients on anti-coagulation including the newer anti- coagulants. P34 Role of mammographic templates in managing ever increasing workloads * Correlation with referral was unclear in 10 %, R1 −R5 category not given in 3 %. Site of lesion was not provided in 3 %. Seven per cent of the reports were 3−4 pages long, described as confusing and difficult to read by the two data extractors. Thirty per cent of reports were not separated into mammography/ultrasound/biopsy sections. There were 23 different ways of characterising lesions on mammography and 24 on ultrasound. Conclusion: BS does not detect more skeletal deposits than CT in the initial assessment or follow-up of breast cancer. CT should be used as the first-line investigation for skeletal and visceral metastasis and BS reserved for problem-solving. Conclusion: The audit highlighted the need for a breast reporting template that met the needs of the clinicians to ensure the relevant facts were included to further improve the patient pathway. P36 P36 Do bone scans add to CT in detecting skeletal metastases in breast cancer staging? * P40 Breast pain in the over 40s: impact on imaging Trupti Kulkarni, Andrew O’Connor Aintree University Hospital, Liverpool, UK Breast Cancer Research 2015, 17(Suppl 1):P40 Breast pain in the over 40s: impact on imaging Trupti Kulkarni, Andrew O’Connor Andrew Steele1, Anil Jain1,2, Navneet Randhawa1, Philip Foden1, Julie Morris1 1The University of Manchester, UK; 2The Nightingale Centre and Genesis Prevention Centre, Manchester, UK Breast Cancer Research 2015, 17(Suppl 1):P38 Introduction: Current practice in our unit as agreed with the local Cancer Network Group is for women over 40 years presenting with breast pain and with a normal clinical examination to have a mammogram. NICE recommends no imaging in this group of patients. The aim was to measure workload impact from current practice, and assess diagnostic yield. Introduction: In the UK, ethnic minority groups have reported lower awareness of breast screening and have presented with breast cancer symptoms later than Caucasian women. Our study compared prognostic indices in symptomatic and screen-detected breast cancer between Asian and Caucasian patients. p Methods: Of the 310 Asian women diagnosed with breast cancer between 1999 and 2014 in the Asian Breast Cancer Database, 217 with invasive cancer were selected (57 screen-detected and 160 symptomatic). Data on invasive tumour size, grade, lymph node status and NPI were compared with age and mode-matched Caucasian breast cancer patients. Results: Asian symptomatic women had larger invasive tumours (median 25.0 mm, IQR 17.1−35.8 mm), compared with Caucasian patients (median 17.0 mm, IQR 12.0−26.4 mm) (p <0.001); higher proportions of grade 3 tumours (64.4 %) (p = 0.007) and with more than one lymph node involved (46.2 %) (p = 0.004), compared with Caucasian patients (48.8 % and 30.0 % respectively); worse NPI scores (median 4.6, IQR 4.3−5.6), compared with Caucasian patients (median 4.3, IQR 3.3−4.7) (p <0.001); and higher proportions with poor prognosis (33.8 %), compared with Caucasian patients (11.9 %) (p <0.001). Multivariable analysis showed invasive grade and tumour size were statistically significant independent discriminators with lymph node status as borderline significant. However, there was no statistically significant difference between the ethnic groups for screen-detected invasive tumours. y Methods: Retrospective audit of imaging and biopsy in female patients over 40 years, presenting with breast pain, and who had normal clinical examination. Results: A total of 100 patients, aged 40−65, from 30 clinics over 3 months, 2014. Eighty normal mammograms. P37 Thirty-six of 126 (29 %) were subsequently found to have breast malignancy (CT 28/110, 25 % and PET CT 8/16, 50 %). Size, Page 12 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Results: Forty patients were referred for a CT staging, four patients did not proceed. Indications: 26 (65 %) had four or more metastatic nodes, six (15 %) T3, eight (20 %) T4. 20/36 (56 %) had no evidence of metastatic disease; 8/36 (22 %) had definite metastases identified (four, >4 nodes, three T4 and one T3); 8/36 (22 %) had indeterminate findings. In three cases the diagnosis of metastatic disease contributed to the decision not to proceed with surgery. No negative impact on treatment was reported in the indeterminate cases. significant in Asian (7.13) and Caucasian (7.51) patients, p = 0.236. Interestingly, significantly more Asian women were treated with mastectomy (47.9 %) compared to Caucasian women (22.9 %), p = 0.015. Conclusion: Asian women presented with a larger tumour size, more progressed to a diagnosis of invasive carcinoma, and more had mastectomies compared to Caucasian women. Since fewer Asian women are presenting via the screening programme, education and awareness of breast cancer and screening needs to be increased in Asian women to increase their screening uptake rates. Conclusion: The new proforma for guiding staging CT scans has reduced the number of overall scans performed with a relatively high pick-up rate of 22 %. P40 Seven of these had ultrasound for focal tenderness or probable glandular tissue, all of which were normal. Twenty abnormal mammograms: eight calcifications, six asymmetry, five discrete masses, one implant rupture. Total imaging workload: nine requests for previous imaging from elsewhere, eight further mammographic views, 11 ultrasounds, two stereo core biopsies (benign), one ultrasound-guided FNA followed by core biopsy (malignant). Yield: one cancer (25 mm grade 2 invasive ductal, negative sentinel lymph node). Conclusion: Workload is appreciably impacted by breast pain investigations. The final diagnosis was often delayed because of the wait for pathology results and previous imaging, increasing patient anxiety. The cancer detection rate number is too low for significance, but nevertheless compares favourably to screening. After discussion with clinicians it was decided to keep to our current practice as a means of opportunistic screening, particularly as our unit is in an area of poor screening uptake. Conclusion: Prognostic indices in Asian women were worse in symptomatic breast cancer, but similar in screen-detected invasive cancer, compared with age-matched Caucasian women. Greater initiatives need to be implemented to promote breast cancer awareness, education and screening among the Asian ethnic minorities. P41 P39 Prospective study looking at CT staging for metastases in early breast cancer Michelle McMahon, Nisha Sharma, David Dodwell The Leeds Teaching Hospitals, NHS Trust, Leeds, UK Breast Cancer Research 2015, 17(Suppl 1):P39 P41 Breast cancer in women under 35 years Kate Hunter, Deirdre Boyle, John Kavanagh, Yvonne Hanhauser, Elizabeth Connolly, Sylvia O’Keeffe St James’s Hospital, Dublin, Ireland Breast Cancer Research 2015, 17(Suppl 1):P41 P44 Follow-up imaging of breast symptomatic patients: a waste of radiologist time? Follow-up imaging of breast symptomatic patients: a waste of radiologist time? * Diana Dalgliesh, Alison Smith, Anjum Mahatma Royal United Hospital, Bath, UK Diana Dalgliesh, Alison Smith, Anjum Mahatma Royal United Hospital, Bath, UK Breast Cancer Research 2015, 17(Suppl 1):P44 Introduction: The NHSBSP does not recommend early recall following assessment of screen-detected abnormalities. Symptomatic patients in our breast clinic may be invited to return for repeat imaging. A survey of repeat imaging in our symptomatic breast clinic was undertaken to understand whether we can justify reducing the number of patients recalled and to gauge associated anxiety levels. Results: Pre consultant involvement, 576 patients were seen and 175 referred with P = 1 (30 %). A total of five biopsies (3 % of referred) were performed retrieving two malignancies (1 % of referred). Post reorganisation, 771 patients were seen and 308 referred to imaging with P = 1 (40 %). A total of 32 biopsies were performed (10 % of referred) with three malignancies (<1 % of referred). In this group only one patient was <40 years old. All cancers were invasive ductal (B5b). All malignancies were in areas of presenting concern. There was a significant increase in workload with decreasing sensitivity of radiology and clinical examination post reorganisation. The background incidence of malignancy was low and stable. Methods: We identified 71 consecutive patients attending an imaging appointment from 1 February 2013 who had a repeat imaging recommendation. Patients were asked to complete a questionnaire. We recorded reason for recall, imaging interval, imaging outcome, and feedback from questionnaires. Results: One patient did not attend. Mean interval between initial and repeat imaging: 4−16 weeks. Fifty-five episodes classified R1/R2 at initial imaging; 11 R3; four R4. Outcomes: 68 % were discharged; 11 % were invited for a third imaging appointment and all were then discharged; 13 % had a benign biopsy; 7 % returned to the surgical clinic for management of their benign symptom. Twenty-three questionnaires were completed −one patient was ‘very anxious’ about repeat imaging, seven patients were ‘mildly anxious’, 10 were ‘relieved’, six were ‘not bothered’. Conclusion: There is increasing demand on imaging for patients with normal clinical examination unrelated to clinical seniority. Cancers are present even with normal clinical examination and patient’s initial clinical concern proved to be an important predictor. Careful scrutiny of the patient’s presenting symptom may allow detection of all cancers. P39 Prospective study looking at CT staging for metastases in early breast cancer Michelle McMahon, Nisha Sharma, David Dodwell The Leeds Teaching Hospitals, NHS Trust, Leeds, UK Breast Cancer Research 2015, 17(Suppl 1):P39 Introduction: Breast cancer is rare in young women under 35 years; however, it can present a diagnostic challenge. This study was undertaken to determine the presentation of breast cancer in young women and the role of imaging including the predictive value in determining tumour size. Introduction: Practice is variable nationally with no agreed guidelines for performing CT staging of asymptomatic patients with a new diagnosis of breast cancer. We have devised a new proforma for performing staging CT in asymptomatic women with high-risk early breast cancer. In our unit, 600 cancers are diagnosed/year. Methods: All breast cancer diagnoses in women aged ≤35 years from 2006 to 2014 were identified. Data were then extracted from PACS and EPR. Analysis was performed on Microsoft Excel. Paired t tests were used to assess the accuracy of imaging in predicting final pathological tumour size. Methods: Prospective audit identifying patients eligible for CT staging based on our proforma over a 12-month period were identified at the breast cancer MDT. A staging scan of the chest, abdomen and pelvis was performed. CT results and clinic letters were reviewed. Criteria: asymptomatic patients diagnosed with new breast cancer requiring staging (T4, inflammatory breast cancer or tumour which extends into the chest wall, skin, or both, fixed nodal disease, arm oedema, nodal disease in SCF; T3, tumour >50 mm clinically, radiologically or pathologically; ≥4 positive nodes at surgery; part of clinical trial involvement or extensive residual disease at surgery after NACT). Results: Seventy patients with 74 presentations of carcinoma were included. Mean age 31 years (SD = 3.7). Of 73 examination scores (E): 7 % (5/73) were screen detected, 52 % (38/73) were E2−3 and 38 % (28/73) were suspected (E4−5). At ultrasound, 16 % (12/74) were U3 and 82 % (61/74) were suspected to be malignant (U4−5). Seventy-four per cent (51/69) had a mammogram score M4−M5. Seventy-five per cent (50/67) Page 13 of 13 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 Breast Cancer Research 2015, Volume 17 Suppl 1 http://www.breast-cancer-research.com/supplements/17/S1 us to review the surgical letters to GPs to see if the radiologist’s contribution was acknowledged. us to review the surgical letters to GPs to see if the radiologist’s contribution was acknowledged. of patients were ACR density of 3−4. P39 Prospective study looking at CT staging for metastases in early breast cancer At MRI (42/70), tumour size correlated with final tumour size on pathology (N = 24, Pearson R 0.45). There was no significant difference between MR estimates of size and final tumour size (t = −0.88, p = 0.39). In contrast, there was a significant difference between US size estimates and final pathology (N = 43, t = −2.56, p <0.05). g Methods: We reviewed 20 surgical letters from the initial fast-track appointment to GPs about patients with a proven cancer. All imaging and biopsies were performed by a consultant radiologist or radiographer. Results: Seventeen (85 %) letters were written by a consultant breast surgeon and three (15 %) by breast registrars. Only one (5 %) letter mentioned radiological involvement and two described the biopsies as ‘we performed’ giving the impression that the biopsy had been performed by the surgical team. In 17 the description of the imaging and biopsy was neutral. Conclusion: Clinical examination has a low PPV in young women with ultrasound demonstrating a superior performance. However, 16 % of cancers were unsuspected at ultrasound. An important finding is the usefulness of MR in defining tumour size, suggesting it should be performed in all young patients. All of the letters, however, were judged to be excellent in terms of information to the GP. Conclusion: Many medical professionals outside of the breast team are unaware of the role of the radiologist. The radiologist, despite doing all the imaging and biopsies in our clinic (which is the case in many units around the country), was essentially invisible in 19 of 20 of the letters we reviewed. We need to debate how we promote the contribution of the radiologist. This could be by reviewing the GP letter template with our surgical colleagues or by promoting the role of radiology to GPs, clinicians, medical students and the general public. ’Peace of mind’? Demand for breast imaging investigation following normal clinical examination: establishing the patient benefits and service implications for a symptomatic service ’Peace of mind’? Demand for breast imaging investigation following normal clinical examination: establishing the patient benefits and service implications for a symptomatic service * Miaad Al-Attar, Lubna Bashir, Lisa Grosvenor, Diane Lister, Elizabeth Denton, Moin Hoosein, Lakshmi Sundaram, Gayle McDonald, Niki Hartley Breast Care Centre, Glenfield Hospital, UHL, Leicester, UK Breast Cancer Research 2015, 17(Suppl 1):P42 Introduction: In the symptomatic service, we noted that requests for imaging after normal examination appeared to be significantly increasing but not improving cancer detection. We aimed to identify demand for imaging following normal clinical examination and their incidence of cancer. Methods: Our unit underwent clinic reorganisation, with the consultant surgeon as primary clinical assessor in February 2014. We carried out a retrospective audit, choosing a month prior and post service reorganisation. All patients referred to imaging with normal clinical examination (P = 1) were included. We recorded demographics, presenting complaint, requestor, findings and biopsies outcomes. P42 P42 ’Peace of mind’? Demand for breast imaging investigation following normal clinical examination: establishing the patient benefits and service implications for a symptomatic service Miaad Al-Attar, Lubna Bashir, Lisa Grosvenor, Diane Lister, Elizabeth Denton, Moin Hoosein, Lakshmi Sundaram, Gayle McDonald, Niki Hartley Breast Care Centre, Glenfield Hospital, UHL, Leicester, UK Breast Cancer Research 2015, 17(Suppl 1):P42 P44 Follow-up imaging of breast symptomatic patients: a waste of radiologist time? * Conclusion: Repeat imaging did not yield any diagnoses of malignancy. All patients were eventually discharged with a benign outcome. We can justify reducing follow-up imaging of our symptomatic patients in line with guidelines for screening assessments. Radiologist time may be better directed towards meeting the symptomatic breast 2-week wait standard. P43 The radiologist in the fast track breast clinic: the invisible man (woman) Nicholas Ridley1*, Charlotte Jones2, Nathan Coombs2, Sarah Taylor1 1Breast Unit, Great Western Hospital, Swindon, UK; 2Department of Surgery, Great Western Hospital, Swindon, UK Breast Cancer Research 2015, 17(Suppl 1):P43 Cite abstracts in this supplement using the relevant abstract number, e.g.: Dalgliesh et al.: Follow-up imaging of breast symptomatic patients: a waste of radiologist time?. Breast Cancer Research 2015, 17(Suppl 1):P44 Cite abstracts in this supplement using the relevant abstract number, e.g.: Dalgliesh et al.: Follow-up imaging of breast symptomatic patients: a waste of radiologist time?. Breast Cancer Research 2015, 17(Suppl 1):P44 Introduction: During a visit to radiology a GP expressed surprise when she discovered that all the breast imaging and biopsies were done by radiologists. She assumed these were done by the surgical team. This led
https://openalex.org/W3117713598	https://strathprints.strath.ac.uk/75185/1/Shittu_etal_MT_2020_systematic_review_of_structural_reliability_methods_for_deformatio_fatigue.pdf	English	null	A Systematic Review of Structural Reliability Methods for Deformation and Fatigue Analysis of Offshore Jacket Structures	Metals	2,020	cc-by	24,699	Abdulhakim Adeoye Shittu 1,* , Athanasios Kolios 2 and Ali Mehmanparast 1 Abdulhakim Adeoye Shittu 1 Energy and Power Theme, Cranﬁeld University, Cranﬁeld MK43 0AL, UK; a.mehmanparast@cranﬁeld.ac.uk 2 Department of Naval Architecture, Ocean & Marine Engineering, University of Strathclyde, Glasgow G1 1XQ, UK; athanasios.kolios@strath.ac.uk * Correspondence: a.a.shittu@cranﬁeld.ac.uk; Tel.: +44-798-116-9719 1 Energy and Power Theme, Cranﬁeld University, Cranﬁeld MK43 0AL, UK; a.mehmanparast@cranﬁeld.ac.uk 2 Department of Naval Architecture, Ocean & Marine Engineering, University of Strathclyde, Glasgow G1 1XQ, UK; athanasios.kolios@strath.ac.uk * Correspondence: a.a.shittu@cranﬁeld.ac.uk; Tel.: +44-798-116-9719 Abstract: This paper presents the state of the art in Structural Reliability Analysis (SRA) methods with a view of identifying key applications of each method and its proposed variations, qualifying characteristics, advantages, and limitations. Due to the increasing complexity and scale of modern offshore jacket structures, it becomes increasingly necessary to propose an accurate and efﬁcient approach for the assessment of uncertainties in their material properties, geometric dimensions, and operating environments. SRA, as a form of uncertainty analysis, has been demonstrated to be a useful tool in the design of structures because it can directly quantify how uncertainty about input parameters can affect structural performance. Herein, attention was focused speciﬁcally on the probabilistic fracture mechanics approach because this accounts accurately for fatigue reliability mostly encountered as being dominant in the design of such structures. The well-established analyti- cal/approximate methods such as the First- and Second-Order Reliability Methods (FORM/SORM) are widely used as they offer a good balance between accuracy and efﬁciency for realistic problems. They are, however, inaccurate in cases of highly non-linear systems. As a result, they have been modiﬁed using methods such as conjugate search direction approach, saddle point approximation, subset simulation, evidence theory, etc. in order to improve accuracy. Initially, direct simulations methods such as the Monte Carlo Simulation Method (MCS) with its various variance reduction techniques such as the Importance Sampling (IS), Latin Hypercube Sampling (LHS), etc. are ideal for structures having non-linear limit states but perform poorly for problems that calculate very low probabilities of failure. Overall, each method has its own merits and limitation, with FORM/SORM being the most commonly used, but recently, simulation methods have increasingly been used due to continuous advances in computation powers. Other relevant methods include the Response Surface Methods (RSM) and the Surrogate Models/Meta-models (SM/MM), which are advanced approxi- mation methods and are ideal for structures with implicit limit state functions and high-reliability indices. Citation: Shittu, A.A.; Kolios, A.; Mehmanparast, A. A Systematic Review of Structural Reliability Methods for Deformation and Fatigue Analysis of Offshore Jacket Structures. Metals 2021, 11, 50. https://doi.org/10.3390/met11010050 Received: 29 November 2020 Accepted: 24 December 2020 Published: 28 December 2020 Keywords: probabilistic fracture mechanics; SRA; FORM; SORM; MCS; RSM Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional claims in published maps and institutional afﬁliations. Publisher’s Note: MDPI stays neu- tral with regard to jurisdictional claims in published maps and institutional afﬁliations. metals metals metals metals Keywords: probabilistic fracture mechanics; SRA; FORM; SORM; MCS; RSM Abdulhakim Adeoye Shittu 1,* , Athanasios Kolios 2 and Ali Mehmanparast 1 Combinations of advanced approximation methods and reliability analysis methods are also found in literature as they can be suitable for complex, highly non-linear problems. Citation: Shittu, A.A.; Kolios, A.; Mehmanparast, A. A Systematic Review of Structural Reliability Methods for Deformation and Fatigue Analysis of Offshore Jacket Structures. Metals 2021, 11, 50. https://doi.org/10.3390/met11010050 1. Introduction Modern offshore jacket structures such as those supporting wind turbines are often exposed to severe environmental conditions. Besides environmental impacts, failures occurring will result in signiﬁcant ﬁnancial losses. This moves the point of focus toward structural reliability assessment of such structures [1]. Most of the existing offshore wind turbines (OWTs) use monopile foundations and are installed in water depths less than 50 m. However, for larger turbines in deeper waters, monopiles become very large and increasingly uneconomical due to the difﬁculty of fabricating and installing such systems, as well as the consideration of modal requirements. Space frame structures, such as jackets derived from the petroleum industry, offer a lighter and yet stiff alternative to monopiles. Copyright: © 2020 by the authors. Li- censee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). https://www.mdpi.com/journal/metals Metals 2021, 11, 50. https://doi.org/10.3390/met11010050 Metals 2021, 11, 50 2 of 37 2 of 37 Jackets can be cheaper in deep-water circumstances and, hence, they can contribute to the offshore wind industry’s quest for the reduction of Levelized Cost of Energy (LCoE). However, an effective design of these structures is resource-intensive, especially when designing to withstand the wide set of dynamic loading mechanisms. Thus, research is still required to improve the design and analysis of modern jacket-type support structures, with due consideration of the requirement for manufacturing optimization if ever these structures are to be mass produced [2–4]. Jackets can be cheaper in deep-water circumstances and, hence, they can contribute to the offshore wind industry’s quest for the reduction of Levelized Cost of Energy (LCoE). However, an effective design of these structures is resource-intensive, especially when designing to withstand the wide set of dynamic loading mechanisms. Thus, research is still required to improve the design and analysis of modern jacket-type support structures, with due consideration of the requirement for manufacturing optimization if ever these structures are to be mass produced [2–4]. p Structural reliability (SR) can be deﬁned as ‘the ability that a structure complies with given requirements under speciﬁc conditions during its intended design life’ [4–7]. Suitable SR levels will, therefore, avoid intolerable damage to a structure over a speciﬁed period of time [8,9]. 1. Introduction Use of probabilistic methods in structural design is increasing, and many standards have made allowance for reliability analysis (RA) either in the calibration of partial safety factors or in design and analysis [10]. ISO 2394 gives information on the principles of reliability methods. Safety margins and factors in design are determined considering T-D (Time-dependent) deterioration mechanisms [11] (see Figure 1). RA allows for methods that analyze the service life variation of Failure Probability for speciﬁed modes of failure, treating uncertainties systematically. The cause of structural failure is the exceedance of limit state. When g(x) < 0 (Limit State Function) they are deﬁned as a failure domain, when g(x) > 0 a safe domain, and a failure surface when g(x) = 0. Structural reliability analysis (SRA) revolves around modeling uncertainties emanating from poor knowledge of design quantities, such as likelihood of events, variability, lack of knowledge, degree of belief, inaccuracy, etc. [6]. These terms are considered basic variables that consist of quantities of material properties and structural dimensions, yield stress, and other limitations in ultimate response, operational conditions and degradation, loading and environmental factors, etc. Uncertainties for SRA can be mainly grouped into two categories: aleatory and epistemic [12]. Aleatory uncertainties are the unavoidable natural randomness associated with an uncertain quantity [6]. Epistemic uncertainties result from inadequate knowledge or information about a quantity. Since epistemic uncertainties stem from inadequate knowledge, they can be minimized by gathering data for a longer period, taking more measurements, or carrying out further tests, among others. The present methodologies available for characterization of uncertainties can be classiﬁed into three groups, namely: the non-probabilistic approach, the precise probability approach, and the imprecise probability approach. Further information on these can be found in [13]. p p y pp Reliability methods have four levels (levels I to IV) of classiﬁcation, according to the way that uncertainty is considered in the analysis [5]. Level I methods are deterministic reliability methods, which only apply one characteristic value to deﬁne each uncertain vari- able. Popular design standard formats, allowable stress, and load resistance belong to this category. They can be applied in combination with more advanced, higher-order methods in the case of calibration of partial safety factors, which can lead to their optimization [7]. 1. Introduction Level II applies two values for the description of each random variable (e.g., the mean and the variance) and a supplementary measure of the correlation between the variables (e.g., covariance). Reliability can be interpreted geometrically as a relative distance from the mean values of loads and their effects. In Level III, the joint probability distribution of the sum of the uncertain variables is introduced, directly computing the POF (Probability of Failure) for a performance function (PF). This category consists of the approximate analytical methods, including the First- and Second-Order Reliability Methods, advanced mathematical techniques such as numerical integration (NI), and simulation methods, such as the Monte Carlo Simulation (MCS) and the Directional Sampling methods. Level IV reliability methods introduce elements of target cost to the engineering principles, so as to derive a technically feasible and, as such, an economically optimized solution. These methods are the most advanced and can set a target reliability level, which is acceptable such that Level III methods can be applied. Metals 2021, 11, 50 3 of 37 which Figure 1. Schematic representation of load process and degradation of resistance. See [11] for more details. Reprinted with permission from [11]. Copyright 2020, Elsevier. Figure 1. Schematic representation of load process and degradation of resistance. See [11] for more details. Reprinted with permission from [11]. Copyright 2020, Elsevier. Figure 1. Schematic representation of load process and degradation of resistance. See [11] for more details. Reprinted with permission from [11]. Copyright 2020, Elsevier. Figure 1. Schematic representation of load process and degradation of resistance. See [11] for more details. Reprinted with permission from [11]. Copyright 2020, Elsevier. This paper aimed to present a state-of-the-art review of Level II and III reliability methods, identifying key applications of each method, qualifying characteristics, ad- vantages and limitations. To the best of the authors’ knowledge, the work performed herein is the first study to review research works conducted on the reliability/probabilistic assessment/evaluation of structures focusing specifically on applicability to design of modern offshore jacket structures against deformation and fatigue. The findings of this research can provide invaluable insights to researchers about the method used for such analyses. 1. Introduction In order to identify relevant sources, a systematic review approach was fol- owed, focusing the search mainly on works published from 2005 to 2020, using prede- fined words and combinations of words, which included “first-order reliability method” or “FORM” or “second-order reliability method” or “SORM” or “Monte Carlo” or “prob- ability of failure (POF)” or “reliability index (RI)” or “safety” or “probabilistic safety” or “safety level” or “reliability safety” or “stochastic” or “probabilistic” or “structural relia- bility” or “reliability” and “analysis” or “assessment” or “evaluation”, etc. on Scopus, sci- ence direct, web of science, ASME (American Society of Mechanical Engineers) digital col- ection, ASCE (American Society of Civil Engineers) library, ICE (Institution of Civil En- gineers)virtual library, one petro, etc. Figure 2 illustrates a classification of the various SRA h d h d i d f This paper aimed to present a state-of-the-art review of Level II and III reliability methods, identifying key applications of each method, qualifying characteristics, advan- tages and limitations. To the best of the authors’ knowledge, the work performed herein is the ﬁrst study to review research works conducted on the reliability/probabilistic assess- ment/evaluation of structures focusing speciﬁcally on applicability to design of modern offshore jacket structures against deformation and fatigue. The ﬁndings of this research can provide invaluable insights to researchers about the method used for such analyses. In order to identify relevant sources, a systematic review approach was followed, focusing the search mainly on works published from 2005 to 2020, using predeﬁned words and combinations of words, which included “ﬁrst-order reliability method” or “FORM” or “second-order reliability method” or “SORM” or “Monte Carlo” or “probability of failure (POF)” or “reliability index (RI)” or “safety” or “probabilistic safety” or “safety level” or “reliability safety” or “stochastic” or “probabilistic” or “structural reliability” or “reliabil- ity” and “analysis” or “assessment” or “evaluation”, etc. on Scopus, science direct, web of science, ASME (American Society of Mechanical Engineers) digital collection, ASCE (American Society of Civil Engineers) library, ICE (Institution of Civil Engineers)virtual library, one petro, etc. Figure 2 illustrates a classiﬁcation of the various SRA methods that are presented in a structured form. Metals 2021, 11, 50 als 2021, 11, x FOR PEE 4 of 37 38 Figure 2. Methods for Structural Reliability Analysis [7,8,14,15], LSF: Limit State Function. Figure 2. Methods for Structural Reliability Analysis [7,8,14,15], LSF: Limit State Function. 2. Level III Analytical Structural Reliability Analysis Methods 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) Theoretical background about FORM can be found in [5,17]. In the FORM, the random variables are represented by means and variances and assume normal distributions. The reliability levels are measured based on the RI. Iterative correlation procedures are used to predict the POF of a structural system or structure. In general, this involves an idealization of the failure domain and it is usually assumed to be a basic representation of the joint probability distribution of a variable. This method of RA is computationally efﬁcient and produces results with desired degree of accuracy (see Figure 3). Theoretical background about FORM can be found in [5,17]. In the FORM, the ran- dom variables are represented by means and variances and assume normal distributions. The reliability levels are measured based on the RI. Iterative correlation procedures are used to predict the POF of a structural system or structure. In general, this involves an idealization of the failure domain and it is usually assumed to be a basic representation of the joint probability distribution of a variable. This method of RA is computationally effi- cient and produces results with desired degree of accuracy (see Figure 3). Figure 3. Illustration of the First-Order Reliability Methods (FORM) approximation. See [18] for more details. Adapted with permission from [18]. Copyright 2020, Elsevier. Figure 3. Illustration of the First-Order Reliability Methods (FORM) approximation. See [18] for more details. Adapted with permission from [18]. Copyright 2020, Elsevier. Figure 3. Illustration of the First-Order Reliability Methods (FORM) approximation. See [18] for more details. Adapted with permission from [18]. Copyright 2020, Elsevier. Figure 3. Illustration of the First-Order Reliability Methods (FORM) approximation. See [18] for more details. Adapted with permission from [18]. Copyright 2020, Elsevier. The Mean Value FOSM (MVFOSM) simplifies the process of calculating the POF of an LSF. Following this method, the LSF is estimated by the first-order Taylor series ex- pansion at the mean value point. In the Hasofer and Lind (HL) method, the RI is expressed as the minimum geometrical distance of Most Probable failure Point (MPP) on the limit state surface from the origin of a u-dimensional space. The HL RI method changes the expansion point from the point of mean value to the MPP, thereby optimizing the ap- proach. 2. Level III Analytical Structural Reliability Analysis Methods The probability of failure (POF) of a structural element with regards to a single mode of failure can be formally expressed as [16]: Pf = Z g(x)≤0 fX(x)dx. (1) (1) The complexity in directly calculating the failure probability Pf from the integral expressed in Equation (1) resulted in the establishment of approximate reliability meth- ods. Such methods involve approximating the failure surface to some simple forms, e.g., hyperplane or quadratic surfaces at certain locations, referred to as design points. This procedure is known as forward SRA. The method used for this computation algorithm is a level II technique, whereby the multidimensional integral expressed in Equation (1) is estimated after the operations: (1) The basic uncertain variables are transferred onto a set of independent Gaussian random variables represented by the U vector. The transformation operation is denoted by T such that U = T(X). (2) The LSF in the U space Z = g(u) is Metals 2021, 11, 50 5 of 37 5 of 37 approximated to a linear or second-order (quadratic) function at the limit state surface, which forms a hyperplane or a quadratic failure surface. Methods based on linear ap- proximation are referred to as the “First-Order Reliability Methods (FORM)” and those based on quadratic approximation are known as the “Second-Order Reliability Methods (SORM)” [4,5]. W 5 of 37 on quadratic approximation are known as the “Second-Order Reliability Methods (SORM)” [4,5]. 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) µe g(µX) β = µeg σeg = g(µX) ∑n i=1 ∂g(µX) ∂xi 2 ·σ2 xi 1 2 (2) αi = − ∂g(X∗) ∂xi ·σxi r ∑n i=1 ∂g(X∗) ∂xi σxi 2 (3) (2) (3) (4) Calculate a new design point Xk and Uk function value, as well as gradients at this new design point. β (4) (4) Calculate a new design point Xk and Uk function value, as well as gradients at this new design point. (4) Calculate a new design point Xk and Uk function value, as well as gradients at this new design point. xi,k = µxi + βσxiαi (4) ui,k = xi,k −µxi σxi (5) (4) (5) (5) Compute the RI β and direction cosine αi using Equations (4) and (5) respectively. (5) Compute the RI β and direction cosine αi using Equations (4) and (5) respectively. β = g(U∗) −∑n i=1 ∂g(U) ∂xi σxiu∗ i r ∑n i=1 ∂g(U∗) ∂xi σxi 2 (6) αi = − ∂g(X∗) ∂xi ·σxi r ∑n i=1 ∂g(X∗) ∂xi σxi 2 (7) (6) αi = − ∂g(X∗) ∂xi ·σxi r ∑n i=1 ∂g(X∗) ∂xi σxi 2 (7) (7) Iterate steps (4) to (5) until the values of the RI β converge. The failure probability Pf can then be calculated from the following: Pf = Φ(−β) = 1 −Φ(β) (8) (8) where Φ(·) is the standard normal cumulative distribution function (CDF) given as where Φ(·) is the standard normal cumulative distribution function (CDF) given as Φ(β) = Z β −∞ 1 √ 2π e−( 1 2 )z2dz. (9) (9) The Hasofer Lind–Rackwitz Fiessler (HL–RF) method is an improvement to the HL method such that its accuracy is close to level III method. This method is ideal for cases of non-Gaussian variables, and it further enables information on the distribution of the random variables to be incorporated (i.e., apart from the mean and standard deviation included in the HL algorithm, information on distributions such as Weibull, Lognormal, Normal, etc. of the random variables are also incorporated in the HL–RF (Hasofer Lind- Rackwitz Fiessler) algorithm). This method involves transformation to the normalized space, and an example is the application of the Rosenblatt transformation, a transformation method for dependent non-normal design variables [5]. An alternative matrix procedure for the HL–RF algorithm is presented in [16]. 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) For instance, once the PF is ascertained, the HL algorithm is employed in calcu- lating the RI 𝛽This recursive algorithm is summarized below [2 3 16 19] The Mean Value FOSM (MVFOSM) simpliﬁes the process of calculating the POF of an LSF. Following this method, the LSF is estimated by the ﬁrst-order Taylor series expansion at the mean value point. In the Hasofer and Lind (HL) method, the RI is expressed as the minimum geometrical distance of Most Probable failure Point (MPP) on the limit state surface from the origin of a u-dimensional space. The HL RI method changes the expansion point from the point of mean value to the MPP, thereby optimizing the approach. For instance, once the PF is ascertained, the HL algorithm is employed in calculating the RI, β. This recursive algorithm is summarized below [2,3,16,19]. lating the RI, 𝛽. This recursive algorithm is summarized below [2,3,16,19]. 1) Define the PF for the corresponding LS, e.g., ultimate limit state (ULS), serviceability li it t t (SLS) f ti li it t t (FLS) t (1) Deﬁne the PF for the corresponding LS, e.g., ultimate limit state (ULS), serviceability limit state (SLS), fatigue limits state (FLS), etc. limit state (SLS), fatigue limits state (FLS), etc. (2) Let the mean value point be the initial design point, i.e., 𝒙௜,௞= 𝜇𝒙೔ 𝑖= 1,2, … , 𝑛, and evaluate the gradients ∇𝑔(𝑿௞) of the LSF at this design point, where 𝒙௜,௞ represents the 𝑖௧௛ element in the vector 𝑿௞ of the 𝑘௧௛ iteration and 𝜇𝒙೔ is the mean value of the (2) Let the mean value point be the initial design point, i.e., xi,k = µxi i = 1, 2, . . . , n, and evaluate the gradients ∇g(Xk) of the LSF at this design point, where xi,k represents the ith element in the vector Xk of the kth iteration and µxi is the mean value of the ith element. Metals 2021, 11, 50 6 of 37 (3) Compute the initial RI, β adopting the mean-value approach, i.e., β = µeg σeg and its direction cosine α. µeg g(µX) (3) Compute the initial RI, β adopting the mean-value approach, i.e., β = µeg σeg and its direction cosine α. 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) For the case of correlated random variables, the HL–RF algorithm is modiﬁed by introducing a correlation matrix [ρ]. The correlation matrix [ρ] is the matrix of correlation coefﬁcients for the uncertain variables involved in the LSF. Further information on this can be found in [4,5,20]. To enhance the precision of the original HL–RF method, speciﬁc modiﬁcations were suggested in several sources. Keshtegar and Chakraborty [21] presented a conjugate search direction approach that overcomes the unstable solutions resulting from periodic nature and chaos for RA problems that involve highly non-linear PFs. In the study, two recursive FORM schemes were examined based on the conjugate descent direction applying hybrid self-adaptive conjugate (HSAC) and the self-adaptive conjugate (SAC) search directions for Metals 2021, 11, 50 7 of 37 l the estimation of RI. Keshtegar and Meng [22] developed a relaxed HL–RF (RHL–RF) based on a relaxed factor, which is calculated dynamically by the second-order interpolation between zero and one. They proposed a hybrid relaxed HL–RF (HRHL–RF) method whereby the HL–RF and RHL–RF are implemented adaptively by using an angle criterion to enhance the efﬁciency and robustness of the FORM formula. In [23], an efﬁcient and robust iterative algorithm, referred to as ﬁnite-based Armijo search direction (FAL) method, was proposed for FORM-based SRA. To achieve the stabilization of the FORM algorithm, a ﬁnite step size was developed applying the Armijo rule and sufﬁcient descent condition. The FAL is adjusted adaptively based on the information acquired from the recursive algorithm at each iteration and Armijo rule. directions for the estimation of RI. Keshtegar and Meng [22] developed a relaxed HL–RF (RHL–RF) based on a relaxed factor, which is calculated dynamically by the second-order interpolation between zero and one. They proposed a hybrid relaxed HL–RF (HRHL–RF) method whereby the HL–RF and RHL–RF are implemented adaptively by using an angle criterion to enhance the efficiency and robustness of the FORM formula. In [23], an effi- cient and robust iterative algorithm, referred to as finite-based Armijo search direction (FAL) method, was proposed for FORM-based SRA. To achieve the stabilization of the FORM algorithm, a finite step size was developed applying the Armijo rule and sufficient descent condition. The FAL is adjusted adaptively based on the information acquired from the recursive algorithm at each iteration and Armijo rule. 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) A di t [24] th HL RF l ith f th FORM h d b k f th h According to [24], the HL–RF algorithm of the FORM has a drawback of the phe- nomena of convergence failure, for example, chaos, periodic oscillation, and bifurcation for some non-linear problems. The essential reasons for numerical instabilities due to chaotic dynamics that include chaos and periodic oscillation common to recursive FORM solutions were revealed. According to Keshtegar [25], the STM with chaos feedback control is inefﬁcient for both concave and convex reliability issues. The author proposed the STM with chaotic conjugate search direction to enhance both the efﬁciency and robustness of the FORM algorithm, by developing a chaos control factor based on a logistic map and following a recursive procedure involving the RI and the logistic map to adaptively deﬁne a transformation matrix. See Figure 4 for an illustration. Pedroso [26] presented a solution to reliability problems using a parallel evolutionary algorithm with accuracy and repeatability of results. The stochastic nature of evolutionary algorithms prevents it from generating identical results. Consideration was given to an optimization problem resulting from the FORM with an implicit LSF that can include a call to Finite Element Analysis (FEA). A detailed explanation was given on a strategy to handle failure from the transformation of random variables or from the ﬁnite element call during the evolution process. According to [24], the HL–RF algorithm of the FORM has a drawback of the phenom- ena of convergence failure, for example, chaos, periodic oscillation, and bifurcation for some non-linear problems. The essential reasons for numerical instabilities due to chaotic dynamics that include chaos and periodic oscillation common to recursive FORM solu- ions were revealed. According to Keshtegar [25], the STM with chaos feedback control is nefficient for both concave and convex reliability issues. The author proposed the STM with chaotic conjugate search direction to enhance both the efficiency and robustness of he FORM algorithm, by developing a chaos control factor based on a logistic map and ollowing a recursive procedure involving the RI and the logistic map to adaptively define a transformation matrix. See Figure 4 for an illustration. Pedroso [26] presented a solution o reliability problems using a parallel evolutionary algorithm with accuracy and repeat- ability of results. The stochastic nature of evolutionary algorithms prevents it from gener- ating identical results. 2.2. Second-Order Reliability Method (SORM) 2.2. Second-Order Reliability Method (SORM) Theoretical background on SORM can be found in [5,28]. RI estimates predicted through FORM produce adequate results when the LSF is nearly linear close to the design point and the LSS (Limit State Surface) has only one minimal distance point. For other conditions, the POF predicted by FORM, using the RI β, may produce inaccurate and unreliable results [5]. To overcome this drawback, second-order Taylor series expansions (or other polynomials) may be introduced [28]. Where a second-order (parabolic) LSS is attached to the non-linear LSF at the design point, it is then regarded as a SORM. According to this approach, the LSF is represented in terms of two independent random functions, whereby one is linear while the other is quadratic in the u-dimensional space. y q p According to [29], in the SORM, the LSF in arbitrarily distributed random variables is approximated by a quadratic polynomial of standard normal variables where the ﬁt- ted quadratic polynomial is then applied in calculating the POF of the LS. Due to the unavailability of a closed-form solution for the failure probability of a general quadratic polynomial surface, a new SORM for RA using saddle point approximation (SPA) was developed. In their study, a new SORM was presented where the failure probability was derived directly from the ﬁtted quadratic polynomial surface because parabolic approxima- tion consideration was removed and the accompanying errors were eliminated [29]. In the new SORM approach, ﬁrst, a Nataf transformation to a standard normal u-space is carried out. Then the design point is located and a quadratic ﬁt at the design point is carried out, after which the cumulant generating function (CGF) of the ﬁtted quadratic polynomial surface is analytically derived. Finally, the Probability Density Function (PDF), Cumulative Density Function (CDF), and failure probability of the LS are established by applying SPA. 2.1. First-Order Reliability Method (FORM) 2.1. First-Order Reliability Method (FORM) Consideration was given to an optimization problem resulting rom the FORM with an implicit LSF that can include a call to Finite Element Analysis FEA). A detailed explanation was given on a strategy to handle failure from the transfor- mation of random variables or from the finite element call during the evolution process. Figure 4. Schematic iterative procedure of FORM-based conjugate search direction. See [25] for more details. Reprinted with permission from [25]. Copyright 2020, Elsevier. Shi et al. [27] investigated the efficiency of different SRA approaches with implicit LSF. Response data sets determined via non-linear FEA are then applied in RA. They pro- posed an efficient response variability method and RA method. The approach is a combi- nation of the Maximum Entropy Fitting Method (MEM) used to model the responses sto- chastically and the FORM for RA. Figure 4. Schematic iterative procedure of FORM-based conjugate search direction. See [25] for more details. Reprinted with permission from [25]. Copyright 2020, Elsevier. Shi et al. [27] investigated the efﬁciency of different SRA approaches with implicit LSF. Response data sets determined via non-linear FEA are then applied in RA. They proposed an efﬁcient response variability method and RA method. The approach is a combination of the Maximum Entropy Fitting Method (MEM) used to model the responses stochastically and the FORM for RA. Figure 4. Schematic iterative procedure of FORM-based conjugate search direction. See [25] for more details. Reprinted with permission from [25]. Copyright 2020, Elsevier. Figure 4. Schematic iterative procedure of FORM-based conjugate search direction. See [25] for more details. Reprinted with permission from [25]. Copyright 2020, Elsevier. Shi et al. [27] investigated the efficiency of different SRA approaches with implicit LSF. Response data sets determined via non-linear FEA are then applied in RA. They pro- posed an efficient response variability method and RA method. The approach is a combi- nation of the Maximum Entropy Fitting Method (MEM) used to model the responses sto- chastically and the FORM for RA. Shi et al. [27] investigated the efﬁciency of different SRA approaches with implicit LSF. Response data sets determined via non-linear FEA are then applied in RA. They proposed an efﬁcient response variability method and RA method. The approach is a combination of the Maximum Entropy Fitting Method (MEM) used to model the responses stochastically and the FORM for RA. Metals 2021, 11, 50 8 of 37 3. Level III (Direct) Reliability Methods The POF of a structural element in relation to a single failure mode is derived from the probability integration expression in Equation (1). Calculating this integration expects that the LSS, g(x) = 0, and the joint PDF of X, fx(x), are known. The methods that calculate the POF directly from this integration are known as Level III reliability methods, which are perceived to be most accurate. The most common relevant level III methods are: (1) Monte Carlo Methods (MC), (2) NI, (3) Analytical Integration (AI), (4) Surrogate Models (SMs), and (5) Stochastic Finite Element Method (SFEM) [1,8,9]. Theoretical background on AI, NI, and MCS can be found in [30]. Computation of the exact failure probability from Equation (1) by using analytical and NIs involves approximations in the solution process and may become inefﬁcient for certain problems. This drawback can be overcome by applying the MCS technique, which avoids the extra layer of approximation; however, it becomes inefﬁcient when the LS needs a continuous calculation procedure such as FEA [5,30], in which case it will be impractical. The concept behind MCS is that the number of samples falling into the failure domain Nf is determined by sampling the basic random variables/vector X, each possessing different probability distributions. p y An effective sampling method is the Latin Hypercube sampling (LHS). LHS is a method whereby multiple variables can be represented such that overlapping data sets are avoided. First, the application divides each stochastic variable distribution in n, non- overlapping intervals with equal probability. This involves associating the analysis point derived from each data set by randomly selecting one value of each variable from each interval. In comparison with the crude Monte Carlo (MC) sampling, this method yields small variance in the response because of the homogeneous allocation of intervals on the PDF [5]. Besides the LHS, other available variance reduction techniques include (1) Adaptive Sampling, which updates sampling density dynamically as the simulation proceeds [5], and (2) Conditional Expectation techniques, which consist of Axis-orthogonal Simulation techniques for convex failure sets, and Directional Simulation for convex safe sets. The IS (Importance Sampling) method is a derivation from the MCS whereby to achieve greater Metals 2021, 11, 50 9 of 37 9 of 37 efﬁciency the simulation is biased; the sampling is basically carried out in the tail of the distribution to ensure the occurrence of adequate simulation failures. 3. Level III (Direct) Reliability Methods An alternative variance reduction technique is the subset sampling, where the failure event is expressed as a sequence of partial failure events (subsets). In the design-point simulation method, the MC sampling is applied around the design point. Initially, having approximated the MPP in the u-dimensional space, MCS is used here instead of performing simulations in the wider range of each distribution. In [31], it was reported that MCS is a powerful tool, straightforward in implementation and capable of solving a wide range of SRA problems. In the same study, neural networks were combined with MCS to address the issues of low POF for highly reliable structures at low computational cost, and the validity of the methodology was demonstrated. Zhang et al. [32] considered SR when statistical parameters of distribution functions could not be ascertained accurately as a result of epistemic uncertainties. Interval bounds were used to model uncertainties in estimates constructed from conﬁdence intervals. They developed an interval MC method that combines the simulation process with the interval analysis. Naess et al. [33] proposed a new MC-based method for approximating system reliabilities, which aims to reduce the computational burden involved in the brute force MCS methods for complicated systems. It harnesses the regularity of tail probabilities to establish an approximation procedure for the calculation of far tail POFs based on the POFs estimated from MCS at more moderate levels. Gaspar et al. [34] claimed the reliability of complex structural systems could be predicted accurately by MCS. They proposed an MCS method for evaluating system reliabilities, which aims to reduce the computational cost associated with highly reliable structural systems. This method applied non-linear FEA combined with response surface (RS) modeling. Jahani et al. [35] presented SRA such that stochastic variables are modeled as fuzzy random variables, and Interval MCS (IMC) combined with Interval Finite Element Method (IFEM) was applied in approximating the POF. IMC-IFEM and Genetic Algorithm (GA) were compared to ascertain the most efﬁcient. It was concluded that the IMC-IFEM provided higher efﬁciency compared to the GA method. Dai et al. [36] developed an IS method based on support vector density estimation and adaptive Markov chain simulation. According to the methodology, samples that can adaptively populate the importance region by the adaptive metropolis algorithm were generated and IS density by support vector density was constructed. In [37] directional simulation was merged with IS. 4.1. Response Surface Method RSMs revolve around creating a polynomial closed-form approximation, ˆg(x), for the exact LSF, g(x), which is usually recognized through an algorithmic procedure, via (1) a few select deterministic analyses and (2) regression analysis of these results. Quadratic functions are commonly employed in practice, as: g(x) ≈ˆg(x) = a0 + N ∑ i=1 aixi + N ∑ i=1 aiixi 2 + N ∑ i=1 N ∑ j=1,j̸=i aijxixj = VT(x)a (10) (10) where the coefﬁcients aT = a0, ai, aii, aij are to be determined and the vector V(x) is described as VT(x) = 1, xi, xi2, xixj . To develop the RS, a ﬁnite number of evaluations of the LSF, for instance, by applying ﬁnite element runs, is needed. Then the RA can be carried out analytically by means of the expression given by Equation (10). This approach is especially desirable when simulation methods are used to determine reliability. The unknown coefﬁcients, aT = a0, ai, aii, aij , are determined via the least-square method. Having identiﬁed a set of ﬁtting points, {xk, k = 1, . . . , NF}, where NF denotes the number of ﬁtting points, referred to as the experimental design, the exact values yk = g xk are calculated and the error is calculated as: error(a) = ∑NF k=1 yk −VT xk ·a 2 . (11) (11) The error expression is minimized with respect to the vector a to calculate the unknown coefﬁcients. After determining a response surface failure function from Equation (10), a standard reliability technique is used to compute the POF [8,9]. According to [7], the response of a structure subjected to certain loading conditions can be evaluated through FEA modeling. The work used a methodology, based on a generalized SRSM (Stochastic Response Surface Method) for the SRA of an offshore jacket, selected as a reference application. According to the methodology, FEA simulation results were incorporated with numerical reliability techniques through multivariate (quadratic) polynomial regression (MPR), so as to predict the reliability levels of components. This ap- proach is particularly useful as it enables enhanced analysis of elements under a stochastic perspective, accounting for design uncertainties efﬁciently [46]. In [2,3,47–49], the efﬁciency of the SRSM for advanced RA of Offshore Wind Turbine (OWT) jacket support structures was demonstrated. According to the authors, a parametric FEA model was developed, and then stochastic FEA simulations were performed. 3. Level III (Direct) Reliability Methods This involves deﬁning a sampling function on the unit hyper-sphere, which samples random directions aiming toward the MPP. Random directions are generated from the sampling function (made adaptive by a closed-form rule to renew the sampling parameters) using spherical coordinates. Zhang [38] developed a new interval IS method by applying the IS technique to an imprecise probability. This methodology has the advantage that an expensive interval analysis is not necessary. Recently, [3] reported the use of direct simulations applying the LHS to predict POF of a complex frame-type structure in the presence of stochastic loads based on studies previously conducted in [39]. p y According to [40], Subset Simulation (SS) is an adaptive simulation method, which solves SRA problems having numerous random variables efﬁciently. In the study, a novel approach for Markov Chain MC (MCMC) sampling was introduced in the standard normal space. This developed two algorithm variants: a basic variant that is simpler than existing formulations with equal efﬁciency and accuracy and a superior variant with adaptive scaling. The accuracy of the SS method was reportedly improved. In [41], an approach to SRA of deteriorating systems that accounts for stochastic dependence among element deterioration was presented. Bayesian updating of the system deterioration model was applied. The updated system reliability is then derived via coupling a probabilistic structural model with the updated deterioration model. SS was reported as a robust and efﬁcient sampling-based scheme ideal for such analyses as solving the underlying high-dimensional SR problems. Metals 2021, 11, 50 10 of 37 10 of 37 4. Advanced Approximation Modeling Methods RS method modeling is applied in deriving equation(s) to express one or more inde- pendent variables in terms of a dependent variable. The fundamental idea is to replace the true LSF by an approximation whose function values can be computed more easily [42]. RSM (Response Surface Model) and SM (Surrogate Modelling) methods are classiﬁcations of approximation methods usually applied in modeling complex engineering systems. RSM derives an analytical expression of the system (deﬁned as polynomial equations) while the latter enters the system data sets normally in a matrix closed form for post- processing [43,44]. The expression derived from either category of methods can further be incorporated with RA methods, such as FORM/SORM and MCS, and applied in multidis- ciplinary optimization problems [1,45]. 4.1. Response Surface Method The results obtained from the FEA were post-processed through MPR in order to obtain the PFs expressed in terms of the stochastic variables, and the RI was then computed through FORM. In the presence of Structural Health Monitoring/Condition Monitoring (SHM/CM) data, the structure may be reassessed and updated. The updated safety index gives vital information for deci- sion making for the inspection and maintenance of OWT support structures. According to [50], the SRSM is a technique employed for RA of complex structural systems with Metals 2021, 11, 50 11 of 37 11 of 37 time-consuming, implicit, or computational costly LSFs. The collection of sample points, the approximation of RS, and the estimation of the POF are the main aspects of the SRSM. Sample points were selected close to the MPP and the actual LS surface (LSS). They used the weighted regression technique to ﬁt the RS, which enables the ﬁtting points to be weighted based on their distance from the failure surface. They analytically derived the cumulant generating function (CGF) of the RS. The POF of the structural system was computed by utilizing the Saddle point Approximation (SPA) method. g p pp RSM is a popular method for RA, especially when the LSF is meant to be highly non- linear or closed-form mechanical models to deﬁne complex structural systems that are not available [6]. According to [51], the SRSM is used for RA of complex systems with low POF, for which approximate methods are inaccurate and MCS is too computationally intensive. The SRSM approximates by ﬁtting the polynomial to a number of sampling points from the LSF with a multidimensional quadratic polynomial. To address the problem related to ill-conditioned systems and an approximated LS (which is very imprecise outside the domain of the stochastic parameters) an algorithm using orthogonal polynomials was proposed. According to [52], RA of a jacket-type OWT support structure under extreme ocean environment loads was performed. The use of the RSM methodology was demonstrated to express the LSF and RI was calculated using the FORM. Rücker and Faber [53] presented the RA of an OWT support structure building upon structural, loading, LS, and uncertainty models comprising design, production, and erection data by applying the SFEM. An algorithm that accounts for complexity in the individual models dictates an efﬁcient solution scheme for reliability. This consists of an adaptive RS and an IS-MC algorithm. 4.2. Surrogate Models (SMs) Theoretical background on SMs can be found in [54–56]. According to [55], SMs, or meta-models (MM) or RSM [55,56], are used for time-consuming implicit LSFs in the context of SRA. To reduce the computational burden of the direct MCS approach, the SMs can be applied to estimate implicitly the LSFs involving FEA. Artiﬁcial neural network (ANN), Kriging, adaptive Kriging, support vector regression (SVR), etc. are applied to estimate the failure probability based on the prediction of the probabilistic model with uncertainties [21]. According to [2], the SRA of a complex OWT support structure subjected to pitting- corrosion fatigue was assessed based on the damage tolerance modeling approach. A non-intrusive formulation incorporating ANN-SM and FORM was proposed to perform the SRA of the crack propagation regime following a sequence of steps. Figure 5 depicts a ﬂowchart of a non-intrusive formulation for probabilistic fracture mechanics in the presence of pitting-corrosion fatigue employing the stochastic parametric FEA. 4.1. Response Surface Method The RS algorithm is based on predetermined DoE (Design of Experiments), which facilitates the adjustment of design parameters for an optimized prediction variance in the domain of the design points. 4.2.1. Kriging Incorporated with FORM Metals 2021, 11, x FOR PEER REVIEW 13 of 37 Flowchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [2]. t 2020, Wiley. wchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [ 20, Wiley. FOR PEER REVIEW 13 4.2.1. Kriging Incorporated w Kriging or Gaussian pro defined over a basis function function. The methodology of is as follows: After obtaining a basis function that projects to take a prior distribution an the Gaussian Process is define defines the smoothness and c dependent variables, 𝒙, Gaus covariance, 𝑘(𝒙, 𝒙’) [57]. Gaspar et al. [55] propos of input basic random variabl lems associated with non-line model applies an adaptive, kr 4.2.1. Kriging Incorporated with FORM Kriging or Gaussian process regression is a stochastic process where distribution is defined over a basis function that can take any form, for example, a squared exponent function. The methodology of earlier instances of Gaussian Process used for curve fitting is as follows: After obtaining a predictive distribution, the regression is then applied over a basis function that projects the input onto the feature space, using a Bayesian approach to take a prior distribution and updating it to form a posterior distribution. The output of the Gaussian Process is defined as the mean and covariance matrix. The covariance matrix defines the smoothness and can be represented by the kernel function. Given a set of in- dependent variables, 𝒙, Gaussian process can be fully defined with a mean, 𝑚(𝒙), and a covariance, 𝑘(𝒙, 𝒙’) [57]. ℱ(𝒙)~𝐺𝑃൫𝑚(𝒙), 𝑘(𝒙, 𝒙’)൯ (12) Gaspar et al. [55] proposed an adaptive kriging SM that applies a moderate number of input basic random variables with active refinement, to overcome component RA prob- lems associated with non-linear computationally intensive implicit LSFs. The presented model applies an adaptive, kriging-based, trust-region approach to search for the design point in the standard normal space and calculates an initial POF based on the FORM and sensitivity factors for the input stochastic variables. In the second stage, the initial estimate is then verified or enhanced using MCS with IS based on Kriging SM defined recursively around the MPP by employing an active refinement algorithm. In addition, they devel- oped a convergence criterion, which detects the stabilization of the POF approximation during the active refinement process, per Figure 6. Figure 6. 4.2.1. Kriging Incorporated with FORM Reliability assessment problem in the standard Gaussian space with the two stages involved in a proposed adap- tive Kriging SM: (a) stage 1 for design point search and POF prediction based on FORM with the adaptive Kriging-based, trust-region method; (b) stage 2 for the POF prediction based on MCS-IS with the Kriging model defined iteratively around the FORM design point with the active refinement algorithm of the AK-MCS method [55]. Reprinted with permission from [55]. Copyright 2020, Elsevier. 4.2.2. Kriging Incorporated with IS/SS Figure 6. Reliability assessment problem in the standard Gaussian space with the two stages involved in a proposed adaptive Kriging SM: (a) stage 1 for design point search and POF prediction based on FORM with the adaptive Kriging-based, trust-region method; (b) stage 2 for the POF prediction based on MCS-IS with the Kriging model deﬁned iteratively around the FORM design point with the active reﬁnement algorithm of the AK-MCS method [55]. Reprinted with permission from [55]. Copyright 2020, Elsevier. 1. Krigin sion is a stochastic process where distributi ake any form, for example, a squared expo tances of Gaussian Process used for curve fi e distribution, the regression is then applied nto the feature space, using a Bayesian appr it to form a posterior distribution. The outp an and covariance matrix. The covariance m esented by the kernel function. Given a set o ss can be fully defined with a mean, 𝑚(𝒙), a 𝐺𝑃൫𝑚(𝒙), 𝑘(𝒙, 𝒙’)൯ tive kriging SM that applies a moderate num ve refinement, to overcome component RA p ationally intensive implicit LSFs. The prese d, trust-region approach to search for the de , ensive im n appro lem mod point in the standard normal space and calculates an initial POF based on the FORM and sensitivity factors for the input stochastic variables. In the second stage, the initial estimat is then verified or enhanced using MCS with IS based on Kriging SM defined recursively around the MPP by employing an active refinement algorithm. In addition, they devel oped a convergence criterion, which detects the stabilization of the POF approximation during the active refinement process, per Figure 6. Figure 6. 4.2.1. Kriging Incorporated with FORM Kriging or Gaussian process regression is a stochastic process where distribution is deﬁned over a basis function that can take any form, for example, a squared exponent function. The methodology of earlier instances of Gaussian Process used for curve ﬁtting is as follows: After obtaining a predictive distribution, the regression is then applied over a basis function that projects the input onto the feature space, using a Bayesian approach to take a prior distribution and updating it to form a posterior distribution. The output of the Gaussian Process is deﬁned as the mean and covariance matrix. The covariance matrix deﬁnes the smoothness and can be represented by the kernel function. Given a set of independent variables, x, Gaussian process can be fully deﬁned with a mean, m(x), and a covariance, k(x, x′) [57]. (12) F(x) ∼GP m(x), k x, x′ (12) F(x) ∼GP m(x), k x, x′ F(x) ∼GP m(x), k x, x′ Metals 2021, 11, 50 12 of 37 12 of 37 Gaspar et al. [55] proposed an adaptive kriging SM that applies a moderate number of input basic random variables with active reﬁnement, to overcome component RA problems associated with non-linear computationally intensive implicit LSFs. The presented model applies an adaptive, kriging-based, trust-region approach to search for the design point in the standard normal space and calculates an initial POF based on the FORM and sensitivity factors for the input stochastic variables. In the second stage, the initial estimate is then veriﬁed or enhanced using MCS with IS based on Kriging SM deﬁned recursively around the MPP by employing an active reﬁnement algorithm. In addition, they developed a convergence criterion, which detects the stabilization of the POF approximation during the active reﬁnement process, per Figure 6. EW 12 of 37 Figure 5. Flowchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [2]. Copyright 2020, Wiley. Figure 5. Flowchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [2]. Copyright 2020, Wiley. Metals 2021, 11, x FOR PEER REVIEW 13 of 37 Figure 5. Flowchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [2]. Copyright 2020, Wiley. Figure 5. Flowchart of a proposed non-intrusive formulation. See [2] for more details. Reprinted with permission from [2]. Copyright 2020, Wiley. 4.2.4. Sequential Kriging Reliability Analysis (SKRA) According to [54], the sequential Kriging reliability analysis (SKRA) method was pro- posed for non-linear implicit PFs, which were computationally unaffordable and included EGRA and AK-MCS. An adaptive sampling regions’ strategy was developed to avoid selecting samples in the areas where the probability density is so low, thereby having an insigniﬁcant impact on the results. The size of the sampling space was adapted based on the POF approximated by the last recursion. 4.2.1. Kriging Incorporated with FORM Reliability assessment problem in the standard Gaussian space with the two stages involved in a proposed adap- tive Kriging SM: (a) stage 1 for design point search and POF prediction based on FORM with the adaptive Kriging-based, trust-region method; (b) stage 2 for the POF prediction based on MCS-IS with the Kriging model defined iteratively around the FORM design point with the active refinement algorithm of the AK-MCS method [55]. Reprinted with permission from [55]. Copyright 2020, Elsevier. 4 2 2 Kriging Incorporated with IS/SS Figure 6. Reliability assessment problem in the standard Gaussian space with the two stages involved in a proposed adaptive Kriging SM: (a) stage 1 for design point search and POF prediction based on FORM with the adaptive Kriging-based, trust-region method; (b) stage 2 for the POF prediction based on MCS-IS with the Kriging model deﬁned iteratively around the FORM design point with the active reﬁnement algorithm of the AK-MCS method [55]. Reprinted with permission from [55]. Copyright 2020, Elsevier. Metals 2021, 11, 50 13 of 37 13 of 37 4.2.2. Kriging Incorporated with IS/SS 4.2.2. Kriging Incorporated with IS/SS In [58], an efﬁcient reliability method combining adaptive IS and KIM (Kriging Inter- polation Method) based on active learning mechanism was presented. It was claimed that the Kriging MM, adaptive IS, and active learning mechanism superiorities are inherited and only enable evaluating the interested samples in actual PF. In [59], a modiﬁcation to an algorithm for the efﬁcient estimation of small probabilities, combining FORM and an adaptive Kriging-based IS strategy (AK-IS), was proposed. It was reported that the modiﬁcation overcomes an important limitation of the original AK-IS, providing the algo- rithm with the ﬂexibility to deal with multiple failure regions characterized by complex, non-linear LS. In [60], the time-consuming task inherent in the SS and SM method when the PF needs to be numerically evaluated was addressed by an active learning method combining KIM and SS (AK-SS). It was asserted that the efﬁciency of the new method relies upon the advantages of SS in evaluating small POFs and the KIM with active learning for approximating the true PF. 4.2.3. Efﬁcient Global Reliability Analysis (EGRA) Bichon et al. [61] developed an efﬁcient RA method known as the Efﬁcient Global RA (EGRA) that characterizes the LS across the domain of random variables accurately. The technique begins by building a KIM from a small number of samples and then selecting adaptively where to generate subsequent samples to enhance the accuracy of the model near the LS. The KIM produced is then sampled by applying multimodal adaptive IS to compute the reliability level of interest. Highly non-linear complex LS can be modeled by locating multiple points on or around the LS, resulting in a more accurate probability integration. Few numbers of true function evaluations are required to generate a quality SM by concentrating the samples in the region where accuracy is essential. 4.2.5. Support Vector Approach Support vector machines (SVM) can be applied for regression, and when doing so are called Support Vector Regression (SVR). SVMs work via linear domain division where the division is made to be as large as possible. This can also be extended to higher-order domains and be used for regression through the use of kernels. In this method, a support vector is drawn such that the error is minimized by selecting a hyperplane that maximizes the margin. For linear SVR, an approximation can be derived from [57]: y = N ∑ i=1 (αi −α∗ i )xi, x + b (13) (13) where αi and α∗ i are Lagrange multipliers, there are N training variables, and b is a real number constant. For non-linear SVR, a kernel is applied to xi, x. For a Gaussian radial basis function kernel this becomes: K xi, xj = e(− xi−x2 j 2σ2 ) (14) (14) where σ is a free parameter. Support vector machines can be applied in either regression or classiﬁcation form and have been used in a wide range of applications where it was proven that it provides Metals 2021, 11, 50 14 of 37 14 of 37 predictions with high ﬁdelity and that hybrid method can enhance predictions. In other applications, SVM gives a good performance, but also relatively quick training times compared to other methods tested due to its simplicity. predictions with high ﬁdelity and that hybrid method can enhance predictions. In other applications, SVM gives a good performance, but also relatively quick training times compared to other methods tested due to its simplicity. According to [62], metamodeling has been widely adopted for RA to enhance compu- tational efﬁciency. They developed an efﬁcient reliability method that harnesses Adaptive Support Vector Machine (ASVM) and the MCS. This applied a pool-based ASVM for the construction of MM with the minimum number of training samples, whereby a learning function was proposed to sequentially select informative training samples. Then the MCS was used to calculate the POF based on the SVM classiﬁer obtained. Dai et al. [63] presented a multiwavelet linear programming SVR method for RA that mitigates the difﬁculties inher- ent in the standard quadratic programming SVR, such as being computationally expensive and possessing the inability to guarantee sufﬁcient model sparsity. 4.2.6. Artiﬁcial Neural Network Approach ANNs are established to be universal function approximators and have found ap- plications in Structural Reliability Assessment (SRA) by several researchers [4,64]. The neural structure of the brain has been employed in creating mathematical models deﬁning ANNs. ANNs establish a functional relationship between two spaces of data during a learning process and replicate that connection during a recall process. According to [31], ANNs can approximate highly non-linear functions accurately over the entire domain with very high ﬁdelity. Several studies [2,42,65] have also been carried out, proving the precision and efﬁciency of the response surface method based on ANN in comparison with the conventional response surface methods for reliability assessment. y Multilayer Perceptron Neural Networks use more than one layer of neurons: an input layer, one or more hidden layers, and an output layer. The neurons have activation func- tions to relate the input they receive to the output they send to the next layer. Furthermore, there are weights between different neurons and biases that are trained iteratively, conven- tionally using reinforcement learning via back propagation where the weights are adjusted between each neuron depending on how accurate the prediction is to the desired result. The model can be applied to an approximation problem by applying a suitable output layer function and minimizing a cost function [57]. To address the extreme number of FEA required to achieve accuracy, a directional approach was developed, which was reported to signiﬁcantly improve efﬁciency. This utilizes deterministic point sets to preserve the underlying joint probability distribution of the random vector describing the structure and adopts neural networks to focus the simulation efforts in the most crucial regions [66]. In [67], two adaptive stochastic search algorithms are used to locate and trace an implicitly deﬁned function to construct an SM for RA. An ANN-SM was applied in the implementation, and it was claimed that the method could, in principle, be applied with any form of SM. In both algorithms proposed, the SM evolved continuously with sample selection and was used in the choice of new samples such that convergence was achieved rapidly to an accurate representation of the limit surface. 4.2.5. Support Vector Approach The method involves developing a novel multiwavelet kernel by the construction of an autocorrelation function of multiwavelets and uses this kernel in context of linear programming SVR for predicting the LS. 4.2.7. Radial Basis Function Approach A typical Radial Basis Function (RBF) model is a form of a feed-forward neural network composed of single neurons using RBF transfer functions. The result of this approach is that a radial basis function ﬁts a surface through the measured sample points. The values between the sample points are calculated from functions based on the radial distance from the original point. The equation for a multi-quadratic basis function is expressed as [57] q ϕ(r) = q 1 + (εr)2 (15) ϕ(r) = q 1 + (εr)2 (15) Metals 2021, 11, 50 15 of 37 15 of 37 where ε is the shape parameter and r is the radius. The output is from the entire model, whose basis function is a function of the Euclidian norm, which is effectively the radius, s(x) = n ∑ i=1 λiϕ(x −xi) (16) (16) where s(x) is the output of the model and λi is the weight of the i-th node. RBF nodes can be employed in a variety of models; however, they are usually applied in a single-layer model with a single node for each sample point in the training set. where s(x) is the output of the model and λi is the weight of the i-th node. RBF nodes can be employed in a variety of models; however, they are usually applied in a single-layer model with a single node for each sample point in the training set. Bucher and Most [42] compared approximate response functions in terms of their capacities to reduce computational costs in SRA. The RS approaches are based on poly- nomial functions, radial basis functions (RBFs), and ANN. In the polynomial approach, the higher-order polynomial shows severe oscillations, requiring too many support points, and this drawback could be addressed via the application of smoothing techniques, e.g., the moving least-squares method (MLS). In the ANN method, the output is linked with the input parameters via simple, ﬂexible functions including linear, step, or sigmoid func- tions that are combined by adjustable weights. The RBF approach also allows for ﬂexible adjustment of the interpolation scheme. The availability of interpolating functions that can be augmented incrementally by allowing extra support points implies that it permits approximation sequences to be provided, ensuring quick convergence of its reliability estimates to the true value. Zhang et al. 4.2.7. Radial Basis Function Approach [68] proposed an efﬁcient RSM to evaluate structural reliability using evidence theory in order to overcome the associated high computational cost. They de- veloped a DoE technique whose key issue is the search of the important control points at intersections of the LS and uncertainty domain. These points have a signiﬁcant contribution to the accuracy of the subsequent RS. Based on these, a highly precise RBF to the actual LSS was introduced. 5. Probabilistic Fatigue and Fracture Mechanics Approaches Since limit state design is a common trend in modern design, details of this will be presented in this section [7]. The general design requirement is to provide structures with sufﬁcient safety margins taking into account all types of uncertainties having effects on its integrity (i.e., load and capacity variability, modeling idealizations, etc.). Limit state design can be deﬁned simply as that the load/demand of a structural system should under no circumstances exceed its resistance/capacity. For offshore and marine structures, various limit states are prescribed in design standards that should be assessed within a comprehensive design. According to the DNV (Det Norske Veritas) [69], the four main types of limit states that should be considered are: (1) serviceability limit state (SLS): deformation and vibration limit states; (2) ultimate limit state (ULS): buckling and stress limit states; (3) fatigue limit state (FLS); and (4) accidental limit state (ALS). Signiﬁcant cyclic loads induced by wind and wave imposed on OWT support struc- tures make their design to be generally dominated by FLS. The assessment of the FLS is performed using two types of methods, i.e., the S-N (Stress-Number of Cycles) curve method and fracture mechanics (FM) method. The S-N curve approach is based on experi- mental fatigue-test data. This approach is commonly used in practice and well described in standards for fatigue design of offshore structures [34,69]. Based on the S-N curve approach, the number of loading cycles to failure, N, can be determined from [3]: logN = A −mlog∆S (17) (17) where A and m represents the intercept and the slope of the S-N curve on the log-log plot, respectively, and ∆S is the stress range. Design standards, such as the DNVGL-ST-0126 (Det Norske Veritas Germanischer Lloyd Standard), prescribe values for the intercept A Metals 2021, 11, 50 16 of 37 16 of 37 and slope m in Equation (17). The Equation (17) can also be referred to as the Basquin’s formulation. The PF of FRA, according to the S-N curve method, is given as [2]: and slope m in Equation (17). The Equation (17) can also be referred to as the Basquin’s formulation. 5. Probabilistic Fatigue and Fracture Mechanics Approaches The PF of FRA, according to the S-N curve method, is given as [2]: (18) gf,SN = logN −logN(t) (18) where subscripts f and SN represent the FLS and S-N curve method, respectively, N is the number of loading cycles to failure, as given by Equation (17), and N(t) is the expected number of loading cycles during the given design life. FM method is applied at the design stage of offshore structures, as this provides a platform for predicting the fatigue life, and during the operational stage, to allow the decision-making process strategies for inspection scheduling and repair. Common among large and complex structures, such as offshore tubular structures, crack-like imperfections, notches, or other forms of discontinuities exist. The basic concept of the FM approach revolves around the characterization of the stress ﬁeld in terms of a single parameter, ∆K, which is known as the stress intensity factor (SIF) near the crack [5,30]. This parameter is a function of both the stress S and crack size a. The FM approach is based on the models created to predict the crack growth in a ma- terial when variable loads are applied. The crack growth is a process not fully understood at the atomic level. The engineering analysis of this is studied using relationships between SIF and crack-growth rates. The Paris law can be expressed as da dN = C(∆K)m (19) (19) where a is the crack length, N is the number of cycles, ∆K is the SIF range, and C and m are material constants. State-of-the-art numerical and experimental research on bilinear crack-growth law phenomena can be found in [70–79]. The safety margin for fatigue reliability analysis based on LEFM (Linear Elastic Fracture Mechanics) is given by: gf, FM = Z ac ao 1 Y(a)m√πa m da −C∆Sm(N(t) −No) (20) (20) where subscript f and FM represent the fatigue limit state and FM method, respectively; ao is the initial crack depth (or the crack depth at time t0); ac denotes the critical crack depth; Y(a) is the compliance function, which is related to crack depth a; m and C are Paris’ law constants; N(t) is the total number of stress cycles in the time period [t0, t]; and No is the initial number of stress cycles. The ﬁndings of the state-of-the-art-review studied herein will be presented later in Section 7. 6. Other Methods and Applications 6.1. Stochastic Finite Element Method (SFEM) 6.1. Stochastic Finite Element Method (SFEM) Component reliability methods are based on a failure function g(ϑ) in the space of basic random variables, which are collated in the vector θ (here θ is used for random ﬁeld instead of X to prevent confusion between spatial variation and random ﬁeld). It is easy to implement if g(ϑ) is explicitly deﬁned or known. In practical cases, the failure function is usually unavailable explicitly in the closed form, and the response is derived via FEA. Several computational techniques may be employed for the RA with implicit failure functions. The major techniques include Perturbation Techniques, Neumann expansion solution, RS approach, and branch and bound techniques [8,9]. A state-of-the-art review on SFEM is presented in [80]. The perturbation techniques are desirable owing to their efﬁciency in computation times and accuracy. Theoretical background on perturbation method can be found in [8,9]. Li et al. [81] proposed a new class of hybrid perturbation-Galerkin methods to establish the response function (surface) to overcome the major challenge in the RA of the complex structure of being unable to ﬁnd the response function from which the LS can be determined. Single-variable and double-variable approximations are the main methods incorporating a Metals 2021, 11, 50 17 of 37 17 of 37 combination of the perturbation technique and Bubnov–Gerlekin projection, where several orders of summation terms of polynomial expansions are adopted as the Galerkin trial functions or basis vectors. Feng et al. [82] presented a robust stochastic-free vibration analysis for engineering structures involving hybrid, yet spatially variant, uncertain system parameters. Both the stochastic and non-stochastic representations of the spatial dependencies of the uncertain- ties are simultaneously incorporated within the uniﬁed analysis as distinguished from the conventional hybrid uncertain eigenvalue problem. It was asserted that the applicability and effectiveness of the proposed computational framework were demonstrated by the numerical investigations on various engineering structures. In [83], a general FE (Finite Element)-based formulation was proposed for the assess- ment of the mean and mean-square response of stochastic structural systems with material properties deﬁned by random ﬁelds governed by a ﬂexibility-based formulation without involving approximations. Integral expressions of closed form for the mean and mean- square value of the displacement response of structures that are stochastic and statically indeterminate was introduced. 6.1. Stochastic Finite Element Method (SFEM) Two new quantities, variability response function (VRF) and the mean response function (MRF) for the mean-square response and the uncertain material properties, were modeled using the stochastic ﬁeld spectral density function, which can be referred to as integral expressions. Toward achieving an efﬁcient and accurate numerical evaluation of the VRF and MRF, a FEM-based fast MCS procedure (FEM-FMCS) was introduced. 7. Critical Discussion Various challenges, as well as their respective solutions, with respect to SRA methods were identiﬁed and presented in Sections 2–6 in this study. This section collates and discusses key aspects from the studies that stand out as particularly important. Tables 1–4 summarize the most popular SR methods and their capabilities and limitations [1,5,6,16,17, 19,22,28–30,55,64,80]. The developments over the last 40 years, with an emphasis on recent development in the FORM and SORM, were reviewed in great detail by Breitung [17] and show that the FORM and SORM are relevant and indispensable in the area of RA as applied to structural engineering. This is in agreement with previous studies carried out by Rackwitz [28]. The well-practiced quantitative approaches to reliability-based risk analysis of analytical nature, such as the concept of LS and FORM/SORM or MCS methods, are still common practice [1]. The conventional SR methods, such as FORM/SORM and MCS, still remain the de facto methods and serve as the basis for enhanced methods that overcome their inherent limitations. Recent research [80] shows that the FORM is the most efﬁcient SRA method ahead of the SORM. The concepts underlining the intricate FORM/SORM algorithms, including the application of numerical optimization methods that form the basis of advancements already discussed in Section 2, are examined in depth and elucidated below [17]. Having deﬁned the LSF g(u1, . . . , un) in the u-dimensional Euclidean space, Equation (7), the FORM aims to approximate the failure domain F = {u; g(u) < 0} by a halfspace by replacing the LSF by a linear tangent hyper-plane at the point u∗. The LSS G = {u; g(u) = 0} has the nearest geometric distance to the origin, which means that the PDF is maximal there since it is proportional to −\|u\|2. The HL algorithm revolves around ﬁnding a point u∗for an LSF g(u) at the normal standard space such that \|u∗\| = min g(u)=0\|u\| = min g(u)≤0\|u\|, (21) (21) i.e., with the shortest distance to the origin. Basically, this involves linearizing the LSF at an initial point, calculating the design point for the linearized LSF, and then proceeding recursively, always again linearizing until convergence is achieved. Line-search and trust-region methods are common deterministic minimization ap- proaches for differentiable functions [88]. In terms of saving computational costs, the former is usually employed. 6.3. Time-Variant Reliability of Systems 6.3. Time-Variant Reliability of Systems Theoretical background on the transfer into T-V (Time-variant) system method and on the outcrossing approach can be found in [8,9,30]. In many cases, reliability aspects are T-V, for instance, some kind of degrading mechanism on the resistance side or ﬂuctuating loads may be the reason. In [11], T-D RA of aging structures was presented, wherein uncertainties such as structural deterioration and non-stationarities in structural load process were taken into account. It was claimed that the improved approximate method, which requires only low-dimensional integration, reduces signiﬁcantly the cost of assessing T-V reliability over a service life extending to 50 years. In [86], a T-V reliability method was formulated as a large-scale series system consisting of T-V response functions obtained by discretizing T-V continuous response functions within the forecast time period. This is used instead of analyzing outcrossing rates, which have a limitation of being inaccurate for low boundary reliability with dependent outcrossing rates. In [87], an efﬁcient cross-entropy-based adaptive IS method was proposed to facilitate the wide application of stochastic process- based T-V reliability methods in complex problems. The LHS with proper correlation control is used to extend cross-entropy-based IS to T-V RA. 6.2. Reliability Analysis of Systems System RA could be conducted directly or as a follow-up of a set of a single element or mode analysis. The ﬁrst option is through NI or MC methods. It may be viable, in many cases, to begin with an analysis of individual components and process the outcomes afterwards to compute the POF for the system, and it may be evident that the methods of the series and parallel systems are capable of providing solutions to the combined system problems. For complicated structural systems, applying the FORM/SORM and MC techniques directly could either be too computationally expensive or the LSF could be unavailable explicitly in the closed form. For realistic structures, then, the response is obtained through a numerical technique such as FEA, whereby the derivatives are unavailable and each evaluation of the implicit LSF is time demanding. A wide range of computational procedures can be used for the RA with implicit failure functions in the form of the Stochastic Finite Element Method (SFEM) [8,9]. Detailed explanation on Parallel System can be found in [8,9,30]. y In the series system, failure of the weakest link results in failure of the entire system. Series systems are modeled commonly by assuming multiple failure modes of a component or multiple failure paths of a structure. A detailed explanation of this can be found in [8,9]. Gong and Zhou [84] evaluated the system reliability of series systems using a proposed improved equivalent component approach. The FORM context was considered for an analytical expression derived to estimate the unit normal vector related to the equivalent component. Hence, the computational efﬁciency for establishing the correlation coefﬁcients between the equivalent component approach and the system is enhanced. At each combining step, the two components with the utmost correlation coefﬁcient are combined in an adaptive combining process. The efﬁciency and accuracy of the enhanced equivalent component approach were demonstrated for a series system with unequally and equally correlated components. Zhang et al. [85] described a framework for developing reliability-based system resistance factors suitable for use with a Direct Design Method (DDM), which is a system- based, design-by-advanced analysis approach. A design-by-inelastic analysis method that relies on existing resistance factors originally developed from member reliability considerations was implemented for minimum system reliability requirements [68]. Metals 2021, 11, 50 18 of 37 18 of 37 7. Critical Discussion In the line-search approach, to ﬁnd the minimum of a function f (x), a sequence xk of points, of which convergence toward a minimum is sought, can be calculated iteratively via: (22) xk+1 = xk −αkH−1 k ∇f (xk) (22) Metals 2021, 11, 50 19 of 37 19 of 37 where xk is the iteration point at present, Hk a symmetric and positive deﬁnite matrix, ∇f (xk) the gradient, and αk the step length. 1 where xk is the iteration point at present, Hk a symmetric and positive deﬁnite matrix, ∇f (xk) the gradient, and αk the step length. f ( ) g p g The search direction H−1 k ∇f (xk) is assumed as a direction such that the target function decreases and αk in such a way that the decrease based on a certain criterion is sufﬁcient. Only a step length is chosen for a somehow sufﬁcient decrease as no exact line search is often made for αk—an inexact line search [88]. The inexact line search is carried out either by ﬁnding the minimum value after polynomial approximation is made to the function on the line or by calculating the function value for a decreasing sequence of step lengths until attaining a sufﬁcient decrease. g Hk is selected based on the information and storage space available. Newton methods, where Hk = ∇2 f (xk), or quasi-Newton methods, where Hk is an approximation of the Hessian, are applied if the information on second derivatives can be collected. However, if this information is unavailable, steepest descent approaches, which use the n-dimensional unity matrix Hk = In, or conjugate gradient approaches are applied. In the full Newton approaches, knowledge about Hessian of the target function is required, which implies that these are usually computationally costly for problems having high dimension. It seems the steepest descent method, −∇f (xk), should be the optimal search direction choice. However, this approach does not possess optimal convergence characteristics [88]. This has linear convergence speed and its application is usually discouraged according to literature. A superior approach that makes use of only gradient information is the conjugate gradient method [88]. Quasi-Newton methods involving reconstructing the Hessian from the change in the gradients are more efﬁcient. Calculation of second derivatives is avoided here but storing the approximate n × n −matrix Hessian is required. This sequence of steps is applied in modiﬁed form for ﬁnding constrained extrema. 7. Critical Discussion In exceptional cases, the HL–RF algorithm behaves like the Newton–Raphson method with a unique feature of possessing quadratic convergence speed, assuming the algorithm Metals 2021, 11, 50 20 of 37 20 of 37 starts on the connecting line of the origin with the beta point, such that the LSF gradient at a point u on this line is parallel always to the position vector u. For example, in a hyperparabolic LSF problem, having deﬁned h(v) = g(v∇g(u)), then the iterative HL–RF algorithm can be expressed as: starts on the connecting line of the origin with the beta point, such that the LSF gradient at a point u on this line is parallel always to the position vector u. For example, in a hyperparabolic LSF problem, having deﬁned h(v) = g(v∇g(u)), then the iterative HL–RF algorithm can be expressed as: vn+1 = vn −h(vn) h′(vn). (25) (25) This is the 1D Newton–Raphson method used to ﬁnd the zero of the function h(v) characterized by quadratic convergence speed. Since steepest descent procedure uses only local descent direction information and due to its inability to establish more efﬁcient search directions as quasi-Newton methods via varying the search direction, problems of considerable starting point-beta point distance may run a long time through valleys. In the case of steepest descent close to the MPP, the problem is that the Lagrangian in the tangential space there is approximately equal to its second-order Taylor series expansion L(u∗, λ∗) + (u −u∗)T∇2 uL(u∗, λ∗)(u −u∗)/2 at the design point u∗(wherein λ∗is the Lagrange multiplier), where the gradient is vanishing. Thus, these methods have slow convergence and zigzagging characteristics. Conversely, these methods need only one evaluation of the LSF gradient at each step and appear to be quite robust. Most studies performed in the last years proposing improvements have failed to address the inherent challenges, such as convergence speed in the original iteration scheme, because these approaches only apply not such a signiﬁcant step-length adjustment. Most optimization schemes revolve around arriving at super-linear convergence rate, such that (26) dn+1 ≈γdα n with 1< α ≤2 and γ >0, (26) so as to avoid the slow linear convergence found in the steepest descent approach. In order to achieve fast convergence and avoid zigzag circumstances, a better option might be to restrict in the vicinity of beta point methods that aim at reconstructing the Hessian of the Lagrangian. 7. Critical Discussion It was asserted in [2] that in SQP methods the approximating matrix for the Hessian is unreliable and tends to become indefinite. This is because beta point Hessian can be indefinite there, as it is not necessarily positively definite. In regular circumstances, assuming that the Hessian is definitely positive, semi-definite only in the tangential space at the MPP is valid. y p , y g p Reduced Hessian methods, which aim at reconstructing only the Hessian projected onto the tangential space, seem to be more reliable [88]. Given that the constrained minimum is regular, this projected matrix is positively deﬁnite. According to [17], it is difﬁcult to recommend which approach to use, as these variations are problem-speciﬁc. Methods such as reduced Hessian, quasi-Newton, or conjugate gradients may have better convergence characteristics close to the beta point but are less robust compared to the HL– RF algorithm for problems with LSF that is not sufﬁciently smooth. Here, all approaches discussed are variants of or associated with the SQP method. Alternative methods are the augmented Lagrangian approaches but they seem to be inefﬁcient for the beta point search. Due to the volume of research, developments in the area of SRA are tending toward the use of RS and SM/MM. The PF of most realistic engineering structures today are highly non-linear and implicit (i.e., cannot be solved without the use of FEA) and, thus, analytical methods are incapable of solving them efﬁciently. Also, the MCS cannot efﬁciently provide accurate solution of LS with very low failure probabilities. Despite the fact that numerical techniques, such as IS, LHS, and SS, have been developed to reduce computational costs, this drawback, among others, still exists. There have been several alternative methods that involve a combination of two or more reliability methods to overcome these drawbacks. The RSM provides an efﬁcient vehicle to combine the high-ﬁdelity FEA modeling with conventional SRA methods. SM SRA methods are ideal for cases requiring implicit PF to be evaluated point-wise using FEM. The PCE (Polynomial Chaos Expansion) has been used as a regression technique to model accurately the global behavior of computational Reduced Hessian methods, which aim at reconstructing only the Hessian projected onto the tangential space, seem to be more reliable [88]. Given that the constrained minimum is regular, this projected matrix is positively deﬁnite. 7. Critical Discussion The extrema of f (x) under g(x) are found under some regularity conditions under the stationary Lagrangian function (L(x, λ) = f (x) + λg(x)) points with: ∇xL(x, λ) = 0 g(x) = 0 (23) (23) in such a way that determining these points involves altering the minimization algorithms. In the SQP (sequential quadratic programming) method, the original problem is replaced in each iteration step by a quadratic function in which its minimum under the linearized constraint must be calculated, i.e., in such a way that determining these points involves altering the minimization algorithms. In the SQP (sequential quadratic programming) method, the original problem is replaced in each iteration step by a quadratic function in which its minimum under the linearized constraint must be calculated, i.e., minxTAkx + bT k x + c under g(xk) + ∇g(xk)T(x −xk) = 0, (24) (24) provided the original constraint condition g(x) = 0 is not violated by much and this method predicts a direction whereby the target function f (x) is decreasing accurately. In order to achieve this, a merit function M(x, C) in the direction given by p is minimized via adjusting the step length. Depending on a parameter C, the merit function is that in which its constrained minimum approximately coincides at least with the unconstrained minimum of the target function for parameter C of sufﬁciently large magnitude. g p y g g Alternatively, the augmented Lagrangian method, which involves replacing the con- strained optimization problem by a sequence of unconstrained optimization problems, can be utilized. The Lagrangian L (x, λ) function is replaced by the minimized augmented La- grangian L(x, λ, µ) = f (x) + λg(x) + µg(x)2 and then by adjusting the parameters λ and µ. The unconstrained minima of L (x, λ, µ), given a large enough µ, are also the minima of the constrained problem [88]. In order to determine the minimum of the augmented Lagrangian, any approach for the unconstrained optimization can be applied. In order to overcome the deﬁciencies of non-convergence encountered in running the original HL–RF algorithm, several modiﬁcations were developed. These are critically discussed below [17]. 7. Critical Discussion According to [17], it is difﬁcult to recommend which approach to use, as these variations are problem-speciﬁc. Methods such as reduced Hessian, quasi-Newton, or conjugate gradients may have better convergence characteristics close to the beta point but are less robust compared to the HL– RF algorithm for problems with LSF that is not sufﬁciently smooth. Here, all approaches discussed are variants of or associated with the SQP method. Alternative methods are the augmented Lagrangian approaches but they seem to be inefﬁcient for the beta point search. g g g pp y p Due to the volume of research, developments in the area of SRA are tending toward the use of RS and SM/MM. The PF of most realistic engineering structures today are highly non-linear and implicit (i.e., cannot be solved without the use of FEA) and, thus, analytical methods are incapable of solving them efﬁciently. Also, the MCS cannot efﬁciently provide accurate solution of LS with very low failure probabilities. Despite the fact that numerical techniques, such as IS, LHS, and SS, have been developed to reduce computational costs, this drawback, among others, still exists. There have been several alternative methods that involve a combination of two or more reliability methods to overcome these drawbacks. The RSM provides an efﬁcient vehicle to combine the high-ﬁdelity FEA modeling with conventional SRA methods. SM SRA methods are ideal for cases requiring implicit PF to be evaluated point-wise using FEM. The PCE (Polynomial Chaos Expansion) has been used as a regression technique to model accurately the global behavior of computational Metals 2021, 11, 50 21 of 37 21 of 37 models, while the KIM with high ﬁdelity models local variations, owing to its interpolation capabilities. Global methods, such as the Active Learning Reliability Method combining Kriging and MCS (AK-MCS) and the EGRA [56], have been found to be very efﬁcient, as they give accurate results for implicit, time-consuming, unaffordable PFs, as demonstrated in [14]. For the treatment of problems, such as complex geotechnical designs and those involving high-frequency vibrations and of elastic stability, SFEM has been found to be ideal as it accounts for spatially varying stochastic structural heterogeneity [80,83]. Numerous studies were carried out on reliability analysis in other related areas such as in geotechnical engineering, ﬂuid mechanics, rock engineering, etc. 7. Critical Discussion The scope of the present study is limited to only reliability assessment with respect to metal offshore jacket structures associated with structural engineering. Applications of the T-D SRA for the design of modern metal structures such as those deployed offshore (i.e., in environments characterized by highly stochastic loads and resistance properties, thus necessitating the need for SRA to account for such uncertainties systematically) include [2,3,44,47–49,89–94]. In [95], the fatigue reliability of ﬁxed offshore platforms was investigated by analyzing different failure scenarios. The analysis was divided into a ﬁnite number of sub-scenarios in order to evaluate the occurrence probability of a special scenario. The Palmgren–Miner’s rule and S-N curve were employed to estimate the accumulated fatigue damages in the LSF. Extensive research has been carried out in the area of probabilistic fracture mechanics approach to FRA, some of which include: According to [96], to accurately evaluate the effect of an inspection and repair strategy of structures subjected to degradation resulting from crack growth, application of FM models are required to describe crack propagation. The reliability methods applied to account for inherent uncertainties with respect to selecting an optimal tool for making the proper decision, which enables a balance between design criteria, inspection, and repair plans, include SN and FM formulations. These include a crack-growth formulation based on bilinear crack-growth law, assuming both crack-growth law segments to be correlated and non-correlated in the POF calculation. The FORM and SORM, as well as MCS, were used to illustrate the effect of inspection in the updated reliabilities. It was reported that crack initiation time, initial crack size, and the crack aspect ratio play signiﬁcant roles in the calibration of FM methods. Dong et al. [91] investigated the fatigue reliability of welded multi-planar tubular joints of the support structure of a ﬁxed-jacket, offshore wind turbine in a water depth of 70 m. The long-term statistical distribution of hot-spot stress ranges was ﬁtted using a two-parameter Weibull function by combining time-domain simulation for the inherent environmental conditions (wind/sea states) in operational condition. The SN-Miner– Palmgren approach-based fatigue design criteria were satisﬁed by normalizing the load histories and, thus, the estimated safety levels refer to fatigue design of tubular joints that meet design criteria. The FM analysis of crack growth was applied for the reliability analysis. 7. Critical Discussion According to the results revealed from the study, the structural reliability prediction produced following the SN-curve method gave conservative results compared to those produced using the FM approach at the start of the service life of the structure, but toward the end of the design life, the results may be overly optimistic. Hence, the authors inferred from the foregoing that it is ideal for applying the S-N curve approach during the design stage, while the FM approach should be used as the structure approaches failure. Furthermore, other fatigue reliability assessment exercises performed included estimating the maximum inspection time for the support structure determined via updating the developed reliability framework with structural health monitoring/condition monitoring (SHM/CM) data. Hence, the updated reliability assessment provides valuable information for making decisions concerning the inspection, maintenance, and repair (IMR) of OWT jacket support structures. A subcategory under this, which is an area open to further investigations due to the sparse amount of research, is the damage tolerance approach for probabilistic pitting- corrosion fatigue life prediction performed by [2], wherein comprehensive mechanistic- based probabilistic models for pitting-corrosion fatigue life prediction by including all stages were presented, and the FORM was implemented with the proposed models. In [2], a generic RA framework that combines parametric FEA modeling, RSM, and RA speciﬁcally for complex OWT jacket-type support structures in the presence of highly stochastic variables and taking into consideration, speciﬁcally, T-D phenomena such as fatigue as well as degradation mechanisms such as corrosion was developed. Two ANNs were trained to relate various stochastic variables for predicting the PF. The two ANNs were employed in order to have an intermediate predictor for a stochastic variable, which is more advantageous because of the added interpretability of results. An advantage of the proposed methodology is that the ﬁrst ANN architecture enabled a signiﬁcant reduction in the computational cost, which would have been required to simulate global behavior of h h ll i h l b l b i d [4] p q g the support structure, thus allowing other global parameters to be incorporated [4]. pp g g p p Despite the high volume of research in this area, it has proven troublesome to establish a model that adequately describes the growth of short cracks, particularly for corrosive environments. Different uncertainties were reported in the two-stage crack-growth law with the (lower) near-threshold segment having the largest variability. 7. Critical Discussion The hot-spot stress range was increased, assumed to be due to changes of the nominal stress and stress concentration factors produced by thickness thinning (wastage) effects of selected components with a general uniform corrosion model in order to account for the corrosion-induced crack-growth rate. Also investigated were factors such as the effects of geometry function and corrosion-induced material degradation on the reliability evaluation. The inﬂuence of inspection and repair with and without considering corrosion were investigated based on the quality of inspection in terms of probability of crack detection curves. FM models were applied to describe crack propagation models, which were calibrated based on SN-data since the initiation of cracks and their initial stages at growth are subject to uncertainties that are hard to quantify. The FORM was applied in the calculation of the RI at the welded joint, representing the failure-critical hot-spot location where the most cumulative fatigue damage occurs. According to [97], the FM approach is more complicated than the SN-curve design, which increases the risk of gross errors. In [4,90], it was assumed that the errors due to the complexity of the FM approach may be reduced as a result of using the ANSYS SMART Metals 2021, 11, 50 22 of 37 22 of 37 Fracture© FEA facility. Hence, in their study, owing to the high ﬁdelity of the FEA model, it was assumed that this accounted for model uncertainties that may exist due to errors if analytical calculations were to be used. The non-intrusive formulation used therein enables an enhanced analysis, leading to more accurate results as it utilizes the 3D simulation method, as established in [2,4]. Fracture© FEA facility. Hence, in their study, owing to the high ﬁdelity of the FEA model, it was assumed that this accounted for model uncertainties that may exist due to errors if analytical calculations were to be used. The non-intrusive formulation used therein enables an enhanced analysis, leading to more accurate results as it utilizes the 3D simulation method, as established in [2,4]. In [4,90], SN-curve and FM approaches to fatigue reliability assessment were compared. A non-intrusive stochastic framework that includes 3D parametric FEA model of OWT jacket support structures was developed, taking account of soil–structure interactions. 7. Critical Discussion • MVFOSM method is a straightforward procedure, while the HL method needs several iterations to converge, especially for non-linear problems Hasofer and Lind-Rackwitz Fiessler Method Widely used approximate analytical method since it provides a good balance between efﬁciency and accuracy in realistic engineering RA. • It may yield unstable results, such as severe oscillations and chaotic solutions for highly non-linear problems (gives non-convergence problems). • Cannot be used for implicit LS and also limited to only one dominant failure mechanism Second-Order Reliability Method Ideal for cases where the LSS has large or irregular curvatures (high non-linearity), the POF estimated by FORM, using the RI β, can produce inaccurate and unreliable results. By introducing second-order Taylor series expansions (or other polynomials), this drawback may be overcome. • It is basically a more time-consuming and complex process • For highly non-linear PF with wide input data sets the estimated FORM/SORM RI results may not be sufﬁciently precise as a result of non-normal to normal transformations and multiple MPPs as well as the application of only ﬁrst/second-order terms to calculate the original PFs. • As a result of the parabolic approximation, there is an additional error in such SORMs besides the quadratic approximation error. The applicability range of this method is diminished as a result of the following reasons: The applicability range of this method is diminished as a result of the following reasons: • Non-linearity or large variations may not be handled by this method efﬁciently because linearization of the LSF about the mean values may result in inaccurate results. y • The MVFOSM algorithm is dependent on different (mathematically equivalent) formulations of the same problem; for both non-linear and linear LSF expressions. • The MVFOSM algorithm is dependent on different (mathematically equivalent) formulations of the same problem; for both non-linear and linear LSF expressions. First Order Reliability Method (FORM) • For cases where the LSS is characterized by irregular/large curvatures (high non-linearity), the POF estimated by FORM, using the RI β, may yield inaccurate and unreliable results. • For cases where the LSS is characterized by irregular/large curvatures (high non-linearity), the POF estimated by FORM, using the RI β, may yield inaccurate and unreliable results. FORM approximation gives adequate outcome when the function is nearly linear close to the MPP, and the LSS has only one minimal distance point. 7. Critical Discussion • MVFOSM method is a straightforward procedure, while the HL method needs several iterations to converge, especially for non-linear problems Hasofer and Lind-Rackwitz Fiessler Method Widely used approximate analytical method since it provides a good balance between efﬁciency and accuracy in realistic engineering RA. • It may yield unstable results, such as severe oscillations and chaotic solutions for highly non-linear problems (gives non-convergence problems). • Cannot be used for implicit LS and also limited to only one dominant failure mechanism Second-Order Reliability Method Ideal for cases where the LSS has large or irregular curvatures (high non-linearity), the POF estimated by FORM, using the RI β, can produce inaccurate and unreliable results. By introducing second-order Taylor series expansions (or other polynomials), this drawback may be overcome. • It is basically a more time-consuming and complex process • For highly non-linear PF with wide input data sets the estimated FORM/SORM RI results may not be sufﬁciently precise as a result of non-normal to normal transformations and multiple MPPs as well as the application of only ﬁrst/second-order terms to calculate the original PFs. • As a result of the parabolic approximation, there is an additional error in such SORMs besides the quadratic approximation error. Table 1. Capabilities and limitations of the most common level III approximation structural reliability methods. Method Capabilities Limitations First Order Reliability Method (FORM) Mean Value First Order Second Moment Reliability Method • Simplest and least expensive reliability method • A simplistic technique used in calculating reliability indices, using a minimum representation of basic variables The applicability range of this method is diminished as a result of the following reasons: • Non-linearity or large variations may not be handled by this method efﬁciently because linearization of the LSF about the mean values may result in inaccurate results. • The MVFOSM algorithm is dependent on different (mathematically equivalent) formulations of the same problem; for both non-linear and linear LSF expressions. Hasofer and Lind Method FORM approximation gives adequate outcome when the function is nearly linear close to the MPP, and the LSS has only one minimal distance point. • For cases where the LSS is characterized by irregular/large curvatures (high non-linearity), the POF estimated by FORM, using the RI β, may yield inaccurate and unreliable results. 7. Critical Discussion The larger variability may be because the knuckle region is very close to the short-crack regime and, thus, the inherent uncertainty of the ∆K threshold, whereby the material experiences no crack growth below this. It is not clear how the material behaves around this region. Coupled with the difﬁculty of carrying out measurements of such fatigue, growth rates in tests entails that further research is required on this. It was suggested in [2,4] that one way to overcome this drawback is by introducing effective initial ﬂaw size (EIFS) concepts, which are alternative solutions to the K-T (Kitagawa-Takahashi) diagram. Several studies have been performed on applying GA in SRA and on sensitivity analysis and design optimization, referred to as the stochastic/structural reliability sen- sitivity analysis (SSA) and RBDO (Reliability-based Design Optimization), respectively. The scope of this study was limited to only SRA methods, due to the myriads of research performed in these areas (i.e., GA, SSA, and RBDO). It should be noted that this review did not cover all the papers published on this subject, but attempted to present the main techniques, with the intent of producing an important document, which serves as a guide to designers/researchers on this subject. 23 of 37 Metals 2021, 11, 50 Table 1. Capabilities and limitations of the most common level III approximation structural reliability methods. Method Capabilities Limitations Order Reliability ethod (FORM) Mean Value First Order Second Moment Reliability Method • Simplest and least expensive reliability method • A simplistic technique used in calculating reliability indices, using a minimum representation of basic variables The applicability range of this method is diminished as a result of the following reasons: • Non-linearity or large variations may not be handled by this method efﬁciently because linearization of the LSF about the mean values may result in inaccurate results. • The MVFOSM algorithm is dependent on different (mathematically equivalent) formulations of the same problem; for both non-linear and linear LSF expressions. Hasofer and Lind Method FORM approximation gives adequate outcome when the function is nearly linear close to the MPP, and the LSS has only one minimal distance point. • For cases where the LSS is characterized by irregular/large curvatures (high non-linearity), the POF estimated by FORM, using the RI β, may yield inaccurate and unreliable results. 7. Critical Discussion β y y • MVFOSM method is a straightforward procedure, while the HL method needs several iterations to converge, especially for non-linear problems • It may yield unstable results, such as severe oscillations and chaotic solutions for highly non-linear problems (gives non-convergence problems). Widely used approximate analytical method since it provides a good balance between efﬁciency and accuracy in realistic engineering RA. Widely used approximate analytical method since it provides a good balance between efﬁciency and accuracy in realistic engineering RA. Widely used approximate analytical method since it provides a good balance between efﬁciency and accuracy in realistic engineering RA. Hasofer and Lind-Rackwitz Fiessler Method p ) • Cannot be used for implicit LS and also limited to only one dominant failure mechanism Ideal for cases where the LSS has large or irregular curvatures (high non-linearity), the POF estimated by FORM, using the RI β, can produce inaccurate and unreliable results. By introducing second-order Taylor series expansions (or other polynomials), this drawback may be overcome. It is basically a more time consuming and complex process • For highly non-linear PF with wide input data sets the estimated FORM/SORM RI results may not be sufﬁciently precise as a result of non-normal to normal transformations and multiple MPPs as well as the application of only ﬁrst/second-order terms to calculate the original PFs. Second-Order Reliability Method g • As a result of the parabolic approximation, there is an additional error in such SORMs besides the quadratic approximation error. 24 of 37 Metals 2021, 11, 50 Table 2. Capabilities and limitations of the most common level III direct structural reliability methods. Method Capabilities Limitations Analytical Integration Ideal for simple failure surface • Only possible for some special cases of limited practical interest • Due to rapidly increasing computational demands as the number of dimension increases (the so-called ‘curse of dimensionality’) it has not found great favor in reliability computations. Numerical Integration Standard routines are found in most computer systems Not always feasible, owing to the growth-off errors and excessive computational times de Monte Carlo Simulation Technique (MCS) Most versatile, clear, and well understood exact method available Requires no partial derivative of LSF; therefore, the method can be used for implicit LSF. 7. Critical Discussion • Computationally unaffordable/exorbitant for real engineering problems (for very small POFs) • Simplest MC approach for reliability problems but not the most efﬁcient • It becomes inefﬁcient when the LS needs a continuous calculation procedure such as FEA, in which case it will be impractical. Table 2. Capabilities and limitations of the most common level III direct structural reliability methods. Capabilities • Increases the accuracy for the same number of runs • IS distribution hv(·) may not be chosen well, such as being too ﬂat or being skewed • Extremely concave LSFs may result in inefﬁcient sampling • A maximum likelihood unique point x∗may not be identiﬁable: • If fx(·) has a ‘ﬂat’ contour region • If the PF G(·) = 0 coincides with a contour of fx(·) in a region of interest, or • If the PF G(·) = 0 is non-smooth (e.g., Ripple-like) and has a series of candidate locations for x∗ • There can be more than one point of local maximum likelihood such as when fx(·) is not uni-modal • The application of IS is sometimes referred to as an art that must be applied with caution. Adaptive Sampling • Involves a prior analysis to locate the design point for problems of high dimensionality in particular; a search algorithm with the constraint that the point must lie along the LS would ﬁx the desirable location of hv(·). It is also possible to modify hv(·), depending on the information being obtained during the search process. • Can be applied for multiple failure mechanisms and small probabilities of failure. Application of the directional sampling and adaptive sampling is limited to a moderate number of random variables Table 2. Cont. od Capabilities Limitations h Importance Sampling Technique • Reduce computational time compared to the crude MCS • Has been demonstrated to be very robust and effective for a range of possible LSF shapes. • Increases the accuracy for the same number of runs • IS distribution hv(·) may not be chosen well, such as being too ﬂat or being skewed • Extremely concave LSFs may result in inefﬁcient sampling • A maximum likelihood unique point x∗may not be identiﬁable: • If fx(·) has a ‘ﬂat’ contour region • If the PF G(·) = 0 coincides with a contour of fx(·) in a region of interest, or • If the PF G(·) = 0 is non-smooth (e.g., Ripple-like) and has a series of candidate locations for x∗ • There can be more than one point of local maximum likelihood such as when fx(·) is not uni-modal • The application of IS is sometimes referred to as an art that must be applied with caution. Method Adaptive Sampling • Involves a prior analysis to locate the design point for problems of high dimensionality in particular; a search algorithm with the constraint that the point must lie along the LS would ﬁx the desirable location of hv(·). It is also possible to modify hv(·), depending on the information being obtained during the search process. • Can be applied for multiple failure mechanisms and small probabilities of failure. Capabilities Limitations • Due to rapidly increasing computational demands as the number of dimension increases (the so-called ‘curse of dimensionality’) it has not found great favor in reliability computations. Ideal for simple failure surface Not always feasible, owing to the growth-off errors and excessive computational times Numerical Integration Standard routines are found in most computer systems • Computationally unaffordable/exorbitant for real engineering problems (for very small POFs) • Simplest MC approach for reliability problems but not the most efﬁcient Crude Monte Carlo Simulation Technique (MCS) 25 of 37 Metals 2021, 11, 50 Table 2. Cont. Method Capabilities Limitations nce tion ques MCS with Importance Sampling Technique • Reduce computational time compared to the crude MCS • Has been demonstrated to be very robust and effective for a range of possible LSF shapes. • Increases the accuracy for the same number of runs • IS distribution hv(·) may not be chosen well, such as being too ﬂat or being skewed • Extremely concave LSFs may result in inefﬁcient sampling • A maximum likelihood unique point x∗may not be identiﬁable: • If fx(·) has a ‘ﬂat’ contour region • If the PF G(·) = 0 coincides with a contour of fx(·) in a region of interest, or • If the PF G(·) = 0 is non-smooth (e.g., Ripple-like) and has a series of candidate locations for x∗ • There can be more than one point of local maximum likelihood such as when fx(·) is not uni-modal • The application of IS is sometimes referred to as an art that must be applied with caution. Adaptive Sampling • Involves a prior analysis to locate the design point for problems of high dimensionality in particular; a search algorithm with the constraint that the point must lie along the LS would ﬁx the desirable location of hv(·). It is also possible to modify hv(·), depending on the information being obtained during the search process. • Can be applied for multiple failure mechanisms and small probabilities of failure. Application of the directional sampling and adaptive sampling is limited to a moderate number of random variables Table 2. Cont. Method Capabilities Limitations nce tion ques MCS with Importance Sampling Technique • Reduce computational time compared to the crude MCS • Has been demonstrated to be very robust and effective for a range of possible LSF shapes. Capabilities Limitations Variance Reduction Techniques Adaptive Sampling • Involves a prior analysis to locate the design point for problems of high dimensionality in particular; a search algorithm with the constraint that the point must lie along the LS would ﬁx the desirable location of hv(·). It is also possible to modify hv(·), depending on the information being obtained during the search process. • Can be applied for multiple failure mechanisms and small probabilities of failure. Application of the directional sampling and adaptive sampling is limited to a moderate number of random variables Application of the directional sampling and adaptive sampling is limited to a moderate number of random variables Metals 2021, 11, 50 26 of 37 Table 2. Cont. Method Capabilities Limitations Variance Reduction Techniques Conditional Expectation Techniques Directional Simulation • Has the advantage of simulation in polar coordinate standard normal (y) space. • Is recommended for convex safe sets • Is applicable speciﬁcally to LSS which are nearly spherical (in standard normal space y) • The amount of sampling required usually is reduced considerably compared to IS in Cartesian coordinates • The special case of hyperspherical LS surface in standard normal space, only one directional sample is required (to ﬁx the radius) and would give the exact result immediately. The technique is not very efﬁcient for one or a few planar LS. Axis Orthogonal Simulation Technique Is recommended for convex failure sets They typically require a large number of response function evaluations, which makes them impractical if the response function is expensive to evaluate. Design Point Simulation Makes use of the FORM design point which makes it less cumbersome in the search for the POF Subset Simulation • Efﬁciently deals with small failure probabilities • Converts the simulation of a rare event into sequence simulation of more frequent events. • Additional complexity as it involves the use of Markov Chain MCS Method. • Inefﬁcient for expensive PF. Table 2. Cont. Table 2. Cont. Method Limitations Limitations Capabilities They typically require a large number of response function evaluations, which makes them impractical if the response function is expensive to evaluate. 27 of 37 Metals 2021, 11, 50 Table 3. Capabilities and limitations of the most common advanced approximation structural reliability methods. Method Capabilities Limitations Parallel System A parallel system fails when all the links (potential failure modes) fail. The most consistent function of the parallel system is for modelling the sequential failure of components in a single failure path leading to structural failure Redundant members are introduced which introduces a computationally intensive procedure Series System Ideal for pipelines Failure of one component leads to failure of the system tic Finite Method Perturbation Method The perturbation techniques are desirable owing to their efﬁciency in terms of computation times and accuracy Too mathematically intensive Neumann Expansion Solution • Can be referred to as a computational scheme to lessen the amount of calculation time and to increase the efﬁciency of standard techniques as FORM/SORM and MC • Is adopted to prevent repeated inversion of the random system-stiffness-matrix while undergoing the MC simulation approach. • The matrix ko has to be decomposed only once for all samples in conjunction with the MCS. Due to this single matrix decomposition, computing time can be signiﬁcantly reduced. Determining the covariance matrix among all elements of the ﬂuctuation part of the stiffness matrix involves prohibitively high computational effort. Response Surface Method • It is desirable when the simulation is used to determine reliability results • Ideal for implicit LSFs • It is not restrained by the number of random variables • It is simple to perform with high accuracy • May sometimes be inefﬁcient for highly non-linear failure function • Restricted to small dimensions of uncertainty space • Inefﬁcient for large complex structures • If the initial point of choice in the RSM is unreasonable, and the objective function is highly non-linear, then the rate of convergence of the RSM is slow. • Therefore, the RSM for SR estimation in a wide range of applications has been limited. Branch and Bound Method Are useful for the elastic-plastic analysis of frame structures where effects of plasticity like the formation of plastic hinges give sharp changes in the stiffness behavior Its application is limited s and limitations of the most common advanced approximation structural reliability methods. Its implementation involves high computational efforts, and sufﬁcient model sparsity cannot be guaranteed. Limitations Determining the covariance matrix among all elements of the ﬂuctuation part of the stiffness matrix involves prohibitively high computational effort. Metals 2021, 11, 50 28 of 37 Table 3. Cont. Method Capabilities Limitations Surrogate Mod- els/Response Surface Model/Meta- Models Polynomial Regression Models The most widely used due to their simple formulations and implementation • Higher-order polynomials tend to show severe oscillations • They sometimes face non-convergence issues for cases with an increasing number of support points centered at the mean value of the basic random variables Approaches Based on: Radial Basis Function • SM has the advantage of been more affordable compare with other exact methods • Has the merit of deﬁning failure conditions of structural systems such that RA can be conducted with high ﬁdelity. • Have been validated to be the best interpolation methods compared to others by using examples of different kinds of scattered data. Computationally efﬁcient but at the expense of accuracy Local Interpolation Model (Polyhedra) • These are highly ﬂexible to local approximations and should converge in the long run to the exact LSF. • It is possible to construct approximation whose reliability estimates converge fast to the true value, as a result of their availability which can be augmented incrementally by providing additional support points. It is an approximate method Artiﬁcial Neural Network • Possess the capability of describing failure conditions of structural systems such that RA can be carried out with high ﬁdelity. • They are ﬂexible in nature and have the ability to capture complex non-linear relationships between input and output through appropriate learning. • Have practical advantages over classical RSM due to their superior mapping capabilities and the ﬂexibility in the functional form • Overﬁtting could occur, whereby the number of hidden nodes is too large for the number of training samples • Have the main challenge of suitably choosing the learning parameters that help restrain under or overﬁtting, as both are equally disastrous. Support vector Machine In comparison to ANNs, SVM employs the theory of minimizing the structure risk to avoid the problems of excessive study, calamity data, local minimum value etc. Its implementation involves high computational efforts, and sufﬁcient model sparsity cannot be guaranteed. Table 3. Cont. Table 3. Cont. Method Capabilities Limitations Limitations • Higher-order polynomials tend to show severe oscillations • They sometimes face non-convergence issues for cases with an increasing number of support points centered at the mean value of the basic random variables The most widely used due to their simple formulations and implementation Computationally efﬁcient but at the expense of accuracy Computationally efﬁcient but at the expense of accuracy • Overﬁtting could occur, whereby the number of hidden nodes is too large for the number of training samples • Have the main challenge of suitably choosing the learning parameters that help restrain under or overﬁtting, as both are equally disastrous. Artiﬁcial Neural Network • Have practical advantages over classical RSM due to their superior mapping capabilities and the ﬂexibility in the functional form In comparison to ANNs, SVM employs the theory of minimizing the structure risk to avoid the problems of excessive study, calamity data, local minimum value etc. Its implementation involves high computational efforts, and sufﬁcient model sparsity cannot be guaranteed. 29 of 37 Metals 2021, 11, 50 Table 3. Cont. Capabilities Limitations Table 4. Capabilities and limitations of the other common structural reliability methods. Limitations • In a lot of cases, reliability aspects are T-V. For instance, some kind of degrading mechanism on the resistance side or ﬂuctuating loads may be the reason. • Fatigue and corrosion phenomenon are T-V • Complicated and demands access to simulation procedure • Linear and non-linear SFEM is not yet practical if the original random ﬁeld is discretized into a high dimensional random vector Limitations Limitations p Surrogate Mod- els/Response Surface Model/Meta- Models Approaches Based on: Moving Least Squares • The optimization problem of the overdetermined system of equations can be solved by using the least-squares approach. • As a result of the approximate character of the classical MLS method, the exact representation of support point values cannot be achieved, which is a positive aspect. • If the input data is noisy, it is owing to the fact that due to the approximative character of the classical MLS method, the support point values cannot be represented exactly. • The application of this in the framework of RA this property result in signiﬁcant errors in the calculation of the POF Kriging Models • Have more general approximation capability, i.e., predicts highly accurate POF values compared to the polynomial regression models • Have important features that have been explored in the context of SRA such as the interpolation capability, the ﬂexibility to estimate arbitrary functions with high precision as well as being capable of providing a measure of local uncertainty for the model predictions. • Provides prediction uncertainty measure which has been used in the development of effective adaptive SMs for SRA with active reﬁnement algorithms for the DoE. Signiﬁcantly more complex compared to polynomial regression models • If the input data is noisy, it is owing to the fact that due to the approximative character of the classical MLS method, the support point values cannot be represented exactly. • The application of this in the framework of RA this property result in signiﬁcant errors in the calculation of the POF Signiﬁcantly more complex compared to polynomial regression models 30 of 37 Metals 2021, 11, 50 Table 4. Capabilities and limitations of the other common structural reliability methods. 8. Conclusions This work presents the state of the art in methods used for structural reliability analysis (SRA) based on a systematic review focusing mainly on literature from 2005 to 2020. The paper focused on presentations of methods and their variations, aiming to qualify their advantages and limitations with applicability to design of metal offshore jacket structures. Improvements to the fundamental analytical methods facing issues with highly non-linear performance function (PF) structures were modiﬁed with conjugate search direction approach, Saddle Point Approximation (SPA), the introduction of the merit function, subset simulation (SS), and evidence theory, among others. To reduce the computational burden of MCS, the approximation methods can be applied to estimate implicitly or explicitly the limit state functions (LSFs) involving Finite Element Analysis (FEA). Combinations of advanced approximation methods and RA methods are also found in literature, as they can be suitable for complex, highly non-linear problems. • The FORM was improved by the development of the conjugate search direction, ﬁnite-based Armijo search direction method, Hybrid Relaxed HL–RF, stability trans- formation method (STM) with chaos feedback control, STM with chaos feedback control, and STM with chaotic conjugate search direction, among others. The combi- nation of Maximum Entropy Fitting Method and the FORM was applied to problems of implicit LSFs. The SORM is an improvement on the FORM, to provide solutions to highly non-linear LSFs. A new SORM for RA was developed using the SAP in order to overcome some of the issues inherent in the traditional SORM. • The MCS method was improved by the development of interval MC method, which combines simulation process with interval analysis, new MC-based methods involving the use of brute force MCS methods for complicated structural systems, IMC-IFEM, merging IS with directional simulation, etc. Improvements in variance reduction techniques were achieved, such as the development of interval importance sampling (IS) method, which applies the IS technique and imprecise probability, and the LHS- based quasi-random polar sampling technique. q p p g q • The advanced approximation modeling methods include the well-established Re- sponse Surface Models/Method (RSM) and the Surrogate Models (SM) as well as the Stochastic RSM (SRSM). The SRSM is a model for the RA of complex systems with low Probability of Failure (POF) for which approximate methods are inaccurate and for which Monte Carlo Simulation (MCS) is too computationally intensive. Capabilities • It is an extension of the conventional RSM, the PF of structures are replaced with polynomials while the traditional one employs polynomial sequences, and the SRSM uses Polynomial Chaos. y • Describes uncertainties in a non-explicit way, assuming the analysis code to be a “black box” such that access to the analysis code will not be required. It is widely used in chemical Engineering. Its application in Structural Engineering is still burgeoning. y q • The mean square error is minimized by selecting the collocation points and from high probability regions, therefore, resulting in fewer function evaluations for enhanced precision Intrusive Spectral Stochastic FEM Gives more reliable results • Complicated and deman procedure • Linear and non-linear SF original random ﬁeld is dimensional random vec Time variant reliability Methods • In a lot of cases, reliability aspects are T-V. For instance, some kind of degrading mechanism on the resistance side or ﬂuctuating loads may be the reason. • Fatigue and corrosion phenomenon are T-V Practical application of T-V re appears rather limited, partia computer codes are available Intrusive Spectral Stochastic FEM Gives more reliable results Practical application of T-V reliability methodology appears rather limited, partially because only very few computer codes are available Time variant reliability Methods Metals 2021, 11, 50 31 of 37 31 of 37 Funding: This research received no external funding. 8. Conclusions The efﬁ- ciency of the RSMs developed for implicit LSFs studied herein include the Collocation Based SRSM, novel SRSM combining FEA, MPR, and FORM/SORM, incorporating the SRSM with Saddle point approximation (SPA), among others. Examples of SM include the Kriging, Adaptive Kriging, EGRA, Support vector machines, ANN, RBF, etc. These can be combined with conventional reliability methods for problems of implicit LSFs. Kriging and Adaptive Kriging interpolation models were combined with the FORM, Line sampling, IS, SS, MCS, etc. p g • This study focused speciﬁcally on the probabilistic fatigue and fracture mechanics approaches because the fatigue limit state in most cases is the design-driving crite- rion for structural components of offshore jacket structures. Consequently, the SRA of structures considering pitting-corrosion fatigue phenomenon was identiﬁed as particularly of note and is recommended as an area open to further investigation. p g • This study focused speciﬁcally on the probabilistic fatigue and fracture mechanics approaches because the fatigue limit state in most cases is the design-driving crite- rion for structural components of offshore jacket structures. Consequently, the SRA of structures considering pitting-corrosion fatigue phenomenon was identiﬁed as particularly of note and is recommended as an area open to further investigation. Author Contributions: Conceptualization, methodology, software, formal analysis, investigation, re- sources, data curation, writing—original draft preparation, and writing—review and editing, A.A.S.; conceptualization, methodology, formal analysis, investigation, resources, writing—review and editing, supervision, and project administration, A.K.; methodology, formal analysis, investigation, resources, writing—review and editing, supervision, project administration, and funding acquisition, A.M. All authors have read and agreed to the published version of the manuscript. Metals 2021, 11, 50 32 of 37 Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available because it also forms part of an ongoing study. Acknowledgments: Author Abdulhakim A. Shittu would like to acknowledge the Petroleum Technology Development Fund (PTDF), Nigeria, for doctoral study scholarship, award number: PTDF/ED/PHD/SAA/1142/17. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. The data are not publicly available because it also forms part of an ongoing study. Institutional Review Board Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. 8. Conclusions The data are not publicly available because it also forms part of an ongoing study. Acknowledgments: Author Abdulhakim A. Shittu would like to acknowledge the Petroleum Technology Development Fund (PTDF), Nigeria, for doctoral study scholarship, award number: PTDF/ED/PHD/SAA/1142/17. Conﬂicts of Interest: The authors declare no conﬂict of interest. 8. Conclusions Abbreviations AI Analytical integration AK-IS active learning kriging with importance sampling AK-MCS Active learning kriging with Monte Carlo simulation ALM Active learning methods ANN Artiﬁcial neural networks ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASVM Adaptive support vector machine BM Bending moment CDF Cumulative density function CGF Cumulant generating function CM Computational models CSRSM Collocation-based stochastic response surface method DNV Det Norske Veritas DoE Design of experiment EGRA Efﬁcient global reliability analysis FAL Finite-based Armijo line search direction FCG Fatigue crack growth FEA Finite element analysis FEM Finite element method FLS Fatigue limit state FM Fracture mechanics FORM First-order reliability method FR Fletcher and Reeves method FRA Fatigue reliability analysis GA Genetic algorithm HL Hasofer and Lind method HL–RF Hasofer Lind–Rackwitz Fiessler method HRHL–RF Hybrid relaxed Hasofer Lind–Rackwitz Fiessler method HSAC Hybrid self-adaptive conjugate ICE Institution of Civil Engineers IFEM Interval ﬁnite element method IMC Interval Monte Carlo simulation IS Importance sampling ISKRA Improved sequential kriging reliability analysis ISO International Organisation for Standardisation KIM Kriging interpolation model KL Karhunen–Leove expansion LCoE Levelized cost of energy LEFM Linear-elastic fracture mechanics LHS Latin hypercube sampling LIF Least improvement function LS Limit state(s) LSF Limit state function LSS Limit state surface MC Monte Carlo Metals 2021, 11, 50 33 of 37 MCMC Markov Chain Monte Carlo MCS Monte-Carlo simulation MEM Maximum entropy ﬁtting method MFEM Multi-scale ﬁnite element method MLS Moving least square MM Meta-model(s) MPP Most probable failure point MPR Multivariate (quadratic) polynomial regression MVFOSM Mean value ﬁrst-order second moment method NF Number of ﬁtting points NI Numerical integration OWT Offshore wind turbine PCE Polynomial chaos expansion PC-Kriging Polynomial chaos-kriging PDF Probability density function PF Performance function PLS Partial least squares PMA Performance measure approach POF Probability of failure PSMFEM Perturbation-based stochastic multi-scale ﬁnite element method RA Reliability analysis/assessment RBF Radial basis function RBO Reliability-based optimization RHL–RF Relaxed Hasofer Lind–Rackwitz Fiessler method RI Reliability index RS Response surface RSM Response surface method/model SAC Self-adaptive conjugate SFEM Stochastic ﬁnite element method SGFEM Stochastic Galerkin FEM SHM Structural health monitoring SKRA Sequential kriging reliability Analysis SLS Serviceability limit state SM Surrogate modelling SORM Second-Order reliability method SPA Saddle point approximation SR Structural reliability SRA Structural reliability assessment SRBDO System reliability-based design optimization SRE Structural reliability evaluation SRSM Stochastic response surface method SS Subset simulation SSFEM Spectral stochastic ﬁnite element Method STM Stability Transformation Method SVM Support vector machines SVR Support vector regression T-D Time-dependent T-I Time-independent/time-invariant T-V Time-variant ULS Ultimate limit state Nomenclature hX(x) Density function Pf POF Xi Random variable a0, a1, . 8. Conclusions . . , a16 Regression coefﬁcients for a quadratic regression fX(x) The joint PDF of the random variables X gX(X), gu(x) Limit-state function xi Realization of the random variable Xi αi Direction cosine µx Mean of random variables 34 of 37 Metals 2021, 11, 50 34 of 37 σx Standard deviation of random variables Φ(·) Cumulative density function of the standard normal distribution φ(·) Density function of standard normal distribution I(·) Failure domain identiﬁer N Number of samples Pr(·) Probability function U Standard normal form of variables V Vector V Pf Target coefﬁcient of variation of failure probability X n-dimensional random vector β Reliability/safety index References Time-dependent reliability of aging structures in the presence of non-stationary loads and degradation. Struct. Saf. 2015, 52, 132–141. [CrossRef] g pp 10. Ditlevsen, O.; Madsen, H.O. Structural Reliability Methods; Wiley: New York, NY, USA, 2007; ISBN 0 , ; , y ; y , , , ; 11. Li, Q.; Wang, C.; Ellingwood, B.R. Time-dependent reliability of aging structures in the presence of non-stationary loads and degradation. Struct. Saf. 2015, 52, 132–141. [CrossRef] 11. Li, Q.; Wang, C.; Ellingwood, B.R. Time-dependent reliabili degradation. Struct. Saf. 2015, 52, 132–141. [CrossRef] g f 12. Li, G.; Lu, Z.; Li, L.; Ren, B. Aleatory and epistemic uncertainties analysis based on non-probabili solution. Appl. Math. Model. 2016, 40, 5703–5716. [CrossRef] 13. Wei, P.; Song, J.; Bi, S.; Broggi, M.; Beer, M.; Lu, Z.; Yue, Z. Non-intrusive stochastic analysis with parameterized imprecise probability models: II. Reliability and rare events analysis. Mech. Syst. Signal Process. 2019, 126, 227–247. [CrossRef] p y y y y g , , [ ] 14. Huang, C.; El Hami, A.; Radi, B. Overview of Structural Reliability Analysis Methods—Part III: Global Reliability Methods. Incert. Fiabilité des Systèmes Multiphysiques 2016, 17, 1–10. [CrossRef] 14. Huang, C.; El Hami, A.; Radi, B. Overview of Structural Reliability Analysis Methods—Part III: Glob Fiabilité des Systèmes Multiphysiques 2016, 17, 1–10. [CrossRef] y p y q 15. Huang, C.; El Hami, A.; Radi, B. Overview of Structural Reliability Analysis Methods—Part I: Local Reliability Methods. Incert. Fiabilité des Systèmes Multiphysiques 2017, 17, 1–10. [CrossRef] 16. Nowak, A.S.; Collins, K.R. Reliability of Structures, 2nd ed.; CRC Press: Boca Raton, FL, USA, 2012. 17. Breitung, K. 40 years FORM: Some new aspects? Probabilistic Eng. Mech. 2015, 42, 71–77. [CrossRef] 17. Breitung, K. 40 years FORM: Some new aspects? Probabilistic Eng. Mech. 2015, 42, 71–77. [CrossRef] 18 Shi P; Mahadevan S Damage tolerance approach for probabilistic pitting corrosion fatigue life prediction Eng Fract Mech 17. Breitung, K. 40 years FORM: Some new aspects? Probabilistic Eng. Mech. 2015, 42, 71 77. [CrossRef] 18. Shi, P.; Mahadevan, S. Damage tolerance approach for probabilistic pitting corrosion fatigue life pre 2001, 68, 1493–1507. [CrossRef] 18. Shi, P.; Mahadevan, S. Damage tolerance approach for probabilistic pitting corrosion fatigue life prediction. Eng. Fract. Mech. 2001, 68, 1493–1507. [CrossRef] 19. Ayyub, B.M.; McCuen, R.H. Probability, Statistics, and Reliability for Engineers and Scientists, 3rd ed Boca Raton, FL, USA, 2011; ISBN 9781439895337. 20. Low, B.K.; Tang, W.H. References 1. Leimeister, M.; Kolios, A. A review of reliability-based methods for risk analysis and their application in the offshore wind industry. Renew. Sustain. Energy Rev. 2018, 91, 1065–1076. [CrossRef] y gy 2. Shittu, A.A.; Mehmanparast, A.; Shaﬁee, M.; Kolios, A.; Hart, P.; Pilario, K.E. Structural Reliability Assessment of Offshore Wind Turbine Support Structures Subjected to Pitting Corrosion-Fatigue: A Damage Tolerance Modelling Approach. Wind Energy 2020, 23, 2004–2026. [CrossRef] 2. Shittu, A.A.; Mehmanparast, A.; Shaﬁee, M.; Kolios, A.; Hart, P.; Pilario, K.E. Structural Reliability Assessment of Offshore Wind Turbine Support Structures Subjected to Pitting Corrosion-Fatigue: A Damage Tolerance Modelling Approach. Wind Energy 2020, 23, 2004–2026. [CrossRef] [ ] 3. Shittu, A.A.; Mehmanparast, A.; Wang, L.; Salonitis, K.; Kolios, A. Comparative Study of Structural Reliability Assessment Methods for Offshore Wind Turbine Jacket Support Structures. Appl. Sci. 2020, 10, 860. [CrossRef] 3. Shittu, A.A.; Mehmanparast, A.; Wang, L.; Salonitis, K.; Kolios, A. Comparative Study of Structural Reliability Assessment Methods for Offshore Wind Turbine Jacket Support Structures. Appl. Sci. 2020, 10, 860. [CrossRef] 4. Shittu, A.A. Structural Reliability Assessment of Complex Offshore Structures Based on Non-Intrusive Stochastic Methods; Cranﬁeld University: Bedford, UK, 2020. elchers, R.; Andre, T. Structural Reliability Analysis and Prediction, 3rd ed.; Wiley: Hoboken, NJ, USA, 2018. 6. Choi, S.-K.; Grandhi, R.; Canﬁeld, R. Reliability-Based Structural Design; Springer: New York, NY, USA, 6. Choi, S.-K.; Grandhi, R.; Canﬁeld, R. Reliability-Based Structural Design; Springer: New York, NY, USA, 2006. 7. Kolios, A. A multi-Conﬁguration Approach to Reliability Based Structural Integrity Assessment for Ulti University: Cranﬁeld, UK, 2010. Conﬁguration Approach to Reliability Based Structural Integrity Assessment for Ultimate Strength; Cranﬁeld d, UK, 2010. 8. Vrouwenvelder, T.; Karadeniz, H. Overview of structural reliability methods. In Safety and Reliability of Industrial Products, Systems and Structures; Soares, G.C., Ed.; Taylor & Francis Group: San Francisco, CA, USA, 2010; pp. 181–189. ISBN 9780415663922. 9. Karadeniz, H. Review of Structural Reliability Methods with Applicability in Practice. In Proceedings of the the Sixteenth (2006) International Offshore and Polar Engineering Conference, San Francisco, CA, USA, 28 May–2 June 2006; The International Society of Offshore and Polar Engineers: San Francisco, CA, USA, 2006; pp. 393–402. g pp 10. Ditlevsen, O.; Madsen, H.O. Structural Reliability Methods; Wiley: New York, NY, USA, 2007; ISBN 0471960861. g pp 10. Ditlevsen, O.; Madsen, H.O. Structural Reliability Methods; Wiley: New York, NY, USA, 2007; ISBN 0471960861. 11. Li, Q.; Wang, C.; Ellingwood, B.R. References 2014, 136, 1–9. [CrossRef] 34. Gaspar, B.; Naess, A.; Leira, B.J.; Soares, C.G. System reliability analysis by monte carlo based method an models. J. Offshore Mech. Arct. Eng. 2014, 136, 1–9. [CrossRef] 35. Jahani, E.; Muhanna, R.L.; Shayanfar, M.A.; Barkhordari, M.A. Reliability assessment with fuzzy random variables using interval monte carlo simulation. Comput. Civ. Infrastruct. Eng. 2014, 29, 208–220. [CrossRef] 36. Dai, H.; Zhang, H.; Wang, W. A support vector density-based importance sampling for reliability assessment. Reliab. Eng. Syst. Saf. 2012, 106, 86–93. [CrossRef] f 37. Shayanfar, M.A.; Barkhordari, M.A.; Barkhori, M.; Barkhori, M. An adaptive directional importance sampling method for structural reliability analysis. Struct. Saf. 2018, 70, 14–20. [CrossRef] 38. Zhang, H. Interval importance sampling method for ﬁnite element-based structural reliability assessment under parameter uncertainties. Struct. Saf. 2012, 38, 1–10. [CrossRef] f Mukherjee, S.; Khor, E.H. Probabilistic ﬁnite element analysis using ANSYS. Struct. Saf. 2006, 28, 17–43 39. Reh, S.; Beley, J.D.; Mukherjee, S.; Khor, E.H. Probabilistic ﬁnite element analysis using ANSYS [CrossRef] 39. Reh, S.; Beley, J.D.; Mukherjee, S.; Khor, E.H. Probabilistic ﬁnite element analysis using ANSYS. Struct. Saf. 2006, 28, 17–43. [CrossRef] l S b C C l h f S b S l b bili i h [CrossRef] 40. Papaioannou, I.; Betz, W.; Zwirglmaier, K.; Straub, D. MCMC algorithms for Subset Simulation. Probabilistic Eng. Mech. 2015, 41, [ ] 40. Papaioannou, I.; Betz, W.; Zwirglmaier, K.; Straub, D. MCMC algorithms for Subset Simulation. Probabilistic Eng. Mech. 2015, 41, 89–103. [CrossRef] 41. Schneider, R.; Thöns, S.; Straub, D. Reliability analysis and updating of deteriorating systems with subset simulation. Struct. Saf. 2017, 64, 20–36. [CrossRef] 42. Bucher, C.; Most, T. A comparison of approximate response functions in structural reliability analysis. Probabilistic Eng. Mech. 2008, 23, 154–163. [CrossRef] 43. Salonitis, K.; Kolios, A. Reliability assessment of cutting tool life based on surrogate approximation methods. Int. J. Adv. Manuf. Technol. 2014, 71, 1197–1208. [CrossRef] 44. Kolios, A.; Di Maio, L.F.; Wang, L.; Cui, L.; Sheng, Q. Reliability assessment of point-absorber wave energy converters. Ocean Eng. 2018, 163, 40–50. [CrossRef] , A. Force-based reliability estimation of remaining cutting tool life in titanium milling. Int. J. Adv. Manuf 321–3333. [CrossRef] 45. Salonitis, K.; Kolios, A. Force-based reliability estimation of remaining cutting tool life in titanium m Technol. 2020, 106, 3321–3333. [CrossRef] 46. Kolios, A.J.; Brennan, F. Reliability Deterioration of Offshore/Marine Steel Structures due to Effects of Corrosion. References Efﬁcient Spreadsheet Algorithm for First-Order Reliability Method. J. Eng. Mech. 2007, 133, 1378–1387. [CrossRef] 21. Keshtegar, B.; Chakraborty, S. A hybrid self-adaptive conjugate ﬁrst order reliability method for robust structural reliability analysis. Appl. Math. Model. 2018, 53, 319–332. [CrossRef] 22. Keshtegar, B.; Meng, Z. A hybrid relaxed ﬁrst-order reliability method for efﬁcient structural reliability analysis. Struct. Saf. 2017, 66, 84–93. [CrossRef] 23. Keshtegar, B.; Chakraborty, S. An efﬁcient-robust structural reliability method by adaptive ﬁnite-step length based on Armijo line search. Reliab. Eng. Syst. Saf. 2018, 172, 195–206. [CrossRef] 24. Yang, D. Chaos control for numerical instability of ﬁrst order reliability method. Commun. Nonlinear Sci. Numer. Simul. 2010, 15, 3131–3141. [CrossRef] 35 of 37 Metals 2021, 11, 50 25. Keshtegar, B. Chaotic conjugate stability transformation method for structural reliability analysis. Comput. Methods Appl. Mech. Eng. 2016, 310, 866–885. [CrossRef] g , , [ ] 26. Pedroso, D.M. FORM reliability analysis using a parallel evolutionary algorithm. Struct. Saf. 2017, g , , [ ] 26. Pedroso, D.M. FORM reliability analysis using a parallel evolutionary algorithm. Struct. Saf. 2017, 65, 84–99. [CrossRef] 27. Shi, X.; Teixeira, A.P.; Zhang, J.; Soares, C.G. Structural reliability analysis based on probabilistic respo Maximum Entropy Method. Eng. Struct. 2014, 70, 106–116. [CrossRef] py g 28. Rackwitz, R. Reliability analysis—A review and some perspectives. Struct. Saf. 2001, 23, 365–395. [Cro 28. Rackwitz, R. Reliability analysis—A review and some perspectives. Struct. Saf. 2001, 23, 365–395. [CrossRef] 29. Huang, X.; Li, Y.; Zhang, Y.; Zhang, X. A new direct second-order reliability analysis method. Appl. Math. Model. 2018, 55, 68–80. [CrossRef] 29. Huang, X.; Li, Y.; Zhang, Y.; Zhang, X. A new direct second-order reliability analysis method. Appl. Math. Model. 2018, 55, 68–80. [CrossRef] ] z, H. Stochastic Analysis of Offshore Steel Structures; Springer: London, UK, 2013; ISBN 9781849961899. , y f ff ; p g , , ; 31. Cardoso, J.B.; de Almeida, J.R.; Dias, J.M.; Coelho, P.G. Structural reliability analysis using Monte Carlo simulation and neural networks. Adv. Eng. Softw. 2008, 39, 505–513. [CrossRef] 32. Zhang, H.; Mullen, R.L.; Muhanna, R.L. Interval Monte Carlo methods for structural reliability. Struct. Saf. 2010, 32, 183–190. [CrossRef] 33. Naess, A.; Leira, B.J.; Batsevych, O. Reliability analysis of large structural systems. Probabilistic Eng. Mech. 2012, 28, 164–168. [CrossRef] ] .; Naess, A.; Leira, B.J.; Soares, C.G. System reliability analysis by monte carlo based method and ﬁnite elemen . Offshore Mech. Arct. Eng. References Wen, Z.; Pei, H.; Liu, H.; Yue, Z. A Sequential Kriging reliability analysis method with characteristics of adaptive sampling regions and parallelizability. Reliab. Eng. Syst. Saf. 2016, 153, 170–179. [CrossRef] 55. Gaspar, B.; Teixeira, A.P.; Guedes Soares, C. Adaptive surrogate model with active reﬁnement combining Kriging and a trust region method. Reliab. Eng. Syst. Saf. 2017, 165, 277–291. [CrossRef] fﬁ d l f l b l l d g g y f 56. Bichon, B.J. Efﬁcient Surrogate Modeling for Reliability Analysis and Design; Vanderbilt University: Vande , J fﬁ g g f y y g ; y , , , 57. Richmond, M.; Sobey, A.; Pandit, R.; Kolios, A. Stochastic assessment of aerodynamics within offshore wind farms based on machine-learning Renew Energy 2020, 161, 650–661 [CrossRef] 57. Richmond, M.; Sobey, A.; Pandit, R.; Kolios, A. Stochastic assessment of aerodynamics within offsho machine-learning. Renew. Energy 2020, 161, 650–661. [CrossRef] 58. Zhao, H.; Yue, Z.; Liu, Y.; Gao, Z.; Zhang, Y. An efﬁcient reliability method combining adaptive importance sampling and Kriging metamodel. Appl. Math. Model. 2015, 39, 1853–1866. [CrossRef] pp 59. Cadini, F.; Santos, F.; Zio, E. An improved adaptive kriging-based importance technique for sampling multiple failure regions of low probability. Reliab. Eng. Syst. Saf. 2014, 131, 109–117. [CrossRef] 60. Huang, X.; Chen, J.; Zhu, H. Assessing small failure probabilities by AK–SS: An active learning method combining Kriging and Subset Simulation. Struct. Saf. 2016, 59, 86–95. [CrossRef] 61. Bichon, B.J.; Eldred, M.S.; Swiler, L.P.; Mahadevan, S.; McFarland, J.M. Efﬁcient Global Reliability Analysis for Nonlinear Implicit Performance Functions. AIAA J. 2008, 46, 2459–2468. [CrossRef] 62. Pan, Q.; Dias, D. An efﬁcient reliability method combining adaptive Support Vector Machine and Monte Saf. 2017, 67, 85–95. [CrossRef] efﬁcient reliability method combining adaptive Support Vector Machine and Monte Carlo Simulation. Struct [CrossRef] 63. Dai, H.; Zhang, B.; Wang, W. A multiwavelet support vector regression method for efﬁcient reliability assessment. Reliab. Eng. Syst. Saf. 2015, 136, 132–139. [CrossRef] y f [ ] 64. Chojaczyk, A.A.; Teixeira, A.P.; Neves, L.C.; Cardoso, J.B.; Guedes Soares, C. Review and application of Artiﬁcial Neural Networks models in reliability analysis of steel structures. Struct. Saf. 2015, 52, 78–89. [CrossRef] y y f 65. Gomes, H.M.; Awruch, A.M. Comparison of response surface and neural network with other methods for structural reliability analysis. Struct. Saf. 2004, 26, 49–67. [CrossRef] 66. Nie, J.; Ellingwood, B.R. Finite element-based structural reliability assessment using efﬁcient directional simulation. J. References In Proceedings of the Asranet 2012, London, UK, 2–4 July 2012; p. 10. J y p 47. Ivanhoe, R.O.; Wang, L.; Kolios, A. Generic framework for reliability assessment of offshore wind turbine jacket support structures under stochastic and time dependent variables. Ocean Eng. 2020, 216, 107691. [CrossRef] 48. Wang, L.; Kolios, A. A generic framework for reliability assessment of offshore wind turbine monopiles considering soil-solid interaction and harsh marine environments. In Proceedings of the Progress in the Analysis and Design of Marine Structures- Proceedings of the 6th International Conference on Marine Structures, MARSTRUCT 2017, Lisbon, Portugal, 8–10 May 2017; pp. 931–938. 49. Kolios, A.; Wang, L. Advanced reliability assessment of offshore wind turbine monopiles by combining reliability analysis method and SHM/CM technology. In Proceedings of the 28th International Ocean and Polar Engineering Conference, Sapporo, Japan, 1–15 June 2018; pp. 1412–1419. 49. Kolios, A.; Wang, L. Advanced reliability assessment of offshore wind turbine monopiles by combining reliability analysis method and SHM/CM technology. In Proceedings of the 28th International Ocean and Polar Engineering Conference, Sapporo, Japan, 1–15 June 2018; pp. 1412–1419. J p J pp 50. Huang, X.; Liu, Y.; Zhang, Y.; Zhang, X. Reliability analysis of structures using stochastic response surface method and saddlepoint approximation. Struct. Multidiscip. Optim. 2017, 55, 2003–2012. [CrossRef] p pp 50. Huang, X.; Liu, Y.; Zhang, Y.; Zhang, X. Reliability analysis of structures using stochastic response surface method and saddlepoint approximation. Struct. Multidiscip. Optim. 2017, 55, 2003–2012. [CrossRef] 51. Gavin, H.P.; Yau, S.C. High-order limit state functions in the response surface method for structural reliability analysis. Struct. Saf. 2008, 30, 162–179. [CrossRef] 51. Gavin, H.P.; Yau, S.C. High-order limit state functions in the response surface method for structural reliability analysis. Struct. Saf. 2008, 30, 162–179. [CrossRef] f 52. Kim, D.H.; Lee, S.G. Reliability analysis of offshore wind turbine support structures under extreme Renew. Energy 2015, 79, 161–166. [CrossRef] 53. Rücker, W.; Faber, M.H. Support Structure Reliability of Offshore Wind Turbines Utilizing an Adaptive Response Surface Method. In Proceedings of the ASME 2010 29th International Conference on Ocean Offshore Arctic Engineering, Shanghai, China, 1–6 June 2010. [CrossRef] 53. Rücker, W.; Faber, M.H. Support Structure Reliability of Offshore Wind Turbines Utilizing an Adaptive Response Surface Method. In Proceedings of the ASME 2010 29th International Conference on Ocean Offshore Arctic Engineering, Shanghai, China, 1–6 June 2010. [CrossRef] Metals 2021, 11, 50 36 of 37 54. 81. Li, Y.-J.; Huang, B.; Li, C.-Q. Hybrid perturbation-Galerkin methods for structural reliability analysi 2017, 48, 59–67. [CrossRef] References Eng. Mech. 2005, 131, 259–267. [CrossRef] , M.D. Surrogate-enhanced stochastic search algorithms to identify implicitly deﬁned functions for reliability 2016, 62, 1–11. [CrossRef] 67. Sundar, V.S.; Shields, M.D. Surrogate-enhanced stochastic search algorithms to identify implicitly deﬁne analysis. Struct. Saf. 2016, 62, 1–11. [CrossRef] 67. Sundar, V.S.; Shields, M.D. Surrogate-enhanced stochastic search algorithms to identify implicitly deﬁned functions for reliability analysis. Struct. Saf. 2016, 62, 1–11. [CrossRef] analysis. Struct. Saf. 2016, 62, 1 11. [CrossRef] 68. Zhang, H.; Ellingwood, B.R.; Rasmussen, K.J.R. System reliabilities in steel structural frame design by inelastic analysis. Eng. Struct. 2014, 81, 341–348. [CrossRef] ood, B.R.; Rasmussen, K.J.R. System reliabilities in steel structural frame design by inelastic analysis. Eng –348. [CrossRef] y f 68. Zhang, H.; Ellingwood, B.R.; Rasmussen, K.J.R. System reliabilities in steel structural frame design b Struct. 2014, 81, 341–348. [CrossRef] 69. DNV-OS-C101 Design of Offshore Steel Structures, General (LRFD Method); National Association of Corrosion Engineers: Huston, TX, USA, 2015. 70. Mehmanparast, A.; Taylor, J.; Brennan, F.; Tavares, I. Experimental investigation of mechanical and fracture properties of offshore wind monopile weldments: SLIC interlaboratory test results. Fatigue Fract. Eng. Mater. Struct. 2018, 41, 2485–2501. [CrossRef] 70. Mehmanparast, A.; Taylor, J.; Brennan, F.; Tavares, I. Experimental investigation of mechanical and fracture properties of offshore wind monopile weldments: SLIC interlaboratory test results. Fatigue Fract. Eng. Mater. Struct. 2018, 41, 2485–2501. [CrossRef] 71. Mehmanparast, A.; Brennan, F.; Tavares, I. Fatigue crack growth rates for offshore wind monopile weldments in air and seawater: SLIC inter-laboratory test results. Mater. Des. 2017, 114, 494–504. [CrossRef] 71. Mehmanparast, A.; Brennan, F.; Tavares, I. Fatigue crack growth rates for offshore wind monopile weldments in air and seawater: SLIC inter-laboratory test results. Mater. Des. 2017, 114, 494–504. [CrossRef] y 72. Igwemezie, V.; Mehmanparast, A. Waveform and frequency effects on corrosion-fatigue crack growth behaviour in modern marine steels. Int. J. Fatigue 2020, 134, 105484. [CrossRef] g 73. Igwemezie, V.; Dirisu, P.; Mehmanparast, A. Critical assessment of the fatigue crack growth rate sensitiv ture in ferrite-pearlite steels in air and marine environment. Mater. Sci. Eng. A 2019, 754, 750–765. [Cro ie, V.; Dirisu, P.; Mehmanparast, A. Critical assessment of the fatigue crack growth rate sensitivity to material p 74. Igwemezie, V.; Mehmanparast, A.; Kolios, A. Current trend in offshore wind energy sector and material requirements for fatigue resistance improvement in large wind turbine support structures–A review. Renew. Sustain. Energy Rev. 2019, 101, 181–196. References Improvement of equivalent component approach for reliability analyses of series systems. Struct. Saf. 2017, 68, 65–72. [CrossRef] 85. Zhang, H.; Shayan, S.; Rasmussen, K.J.R.; Ellingwood, B.R. System-based design of planar steel frames, I: Reliability framework. J. Constr. Steel Res. 2016, 123, 135–143. [CrossRef] Frangopol, D.M. An efﬁcient time-dependent reliability method. Struct. Saf. 2019, 81, 101864. [CrossRef] 86. Gong, C.; Frangopol, D.M. An efﬁcient time-dependent reliability method. Struct. Saf. 2019, 81, 101 87. Yang, D.Y.; Teng, J.G.; Frangopol, D.M. Cross-entropy-based adaptive importance sampling for time-dependent reliability analysis of deteriorating structures. Struct. Saf. 2017, 66, 38–50. [CrossRef] 88. Nocedal, J.; Wright, S.J. Numerical Optimization; Springer: New York, NY, USA, 1999. 88. Nocedal, J.; Wright, S.J. Numerical Optimization; Springer: New York, NY, USA, 1999. 89. Jiang, Z.; Hu, W.; Dong, W.; Gao, Z.; Ren, Z. Structural reliability analysis of wind turbines: A rev [CrossRef] 90. Shittu, A.A.; Mehmanparast, A.; Pilario, K.E.; Hart, P.; Kolios, A.J. Comparative study between SN and fracture mechanics approach on reliability assessment of offshore wind turbine jacket foundations. 2020. under review. 91. Dong, W.; Moan, T.; Gao, Z. Fatigue reliability analysis of the jacket support structure for offshore wind turbine considering the effect of corrosion and inspection. Reliab. Eng. Syst. Saf. 2012, 106, 11–27. [CrossRef] 92. Yeter, B.; Garbatov, Y.; Soares, C.G. Fatigue reliability of an offshore wind turbine supporting structure accounting for inspection and repair. In Analysis and Design of Marine Structures; Soares, G., Shenoi, E., Eds.; Taylor & Francis Group: London, UK, 2015; pp. 737–747. 93. Yeter, B.; Garbatov, Y.; Soares, C.G. Reliability of Offshore Wind Turbine Support Structures Subjected to Extreme Wave-Induced Loads and Defects. In Proceedings of the ASME 2016 35th International Conference on Ocean, Offshore and Arctic Engineering, Busan, Korea, 14–18 June 2016. 94. Yeter, B.; Garbatov, Y.; Guedes Soares, C. System reliability of a jacket offshore wind turbine subjected to fatigue. In Proceedings of the 5th International Conference on Marine Structures, MARSTRUCT, Lisbon, Portugal, 25–27 March 2017; pp. 939–950. g 95. Gholizad, A.; Golafshani, A.A.; Akrami, V. Structural reliability of offshore platforms considering fatigue damage and different failure scenarios. Ocean Eng. 2012, 46, 1–8. [CrossRef] g 96. Ayala-uraga, E.; Moan, T. Fatigue reliability-based assessment of welded joints applying consistent fracture mechanics formula- tions. Int. J. Fatigue 2007, 29, 444–456. [CrossRef] 97. Ziegler, L.; Muskulus, M. References [CrossRef] manparast, A.; Kolios, A. Materials selection for XL wind turbine support structures: A corrosion-fatigue ruct. 2018, 61, 381–397. [CrossRef] 75. Igwemezie, V.; Mehmanparast, A.; Kolios, A. Materials selection for XL wind turbine support structu perspective. Mar. Struct. 2018, 61, 381–397. [CrossRef] 76. Jacob, A.; Mehmanparast, A.; D’Urzo, R.; Kelleher, J. Experimental and numerical investigation of fatigue crack growth behaviour of S355 steel weldments. Int. J. Fatigue 2019, 128, 105196. [CrossRef] 77. Jacob, A.; Oliveira, J.; Mehmanparast, A.; Hosseinzadeh, F.; Kelleher, J.; Berto, F. Residual stress measurements in offshore wind monopile weldments using neutron diffraction technique and contour method. Theor. Appl. Fract. Mech. 2018, 96, 418–427. [CrossRef] 78. Bocher, M.; Mehmanparast, A.; Braithwaite, J.; Shaﬁee, M. New shape function solutions for fractu offshore wind turbine monopile foundations. Ocean Eng. 2018, 160, 264–275. [CrossRef] 79. Khajeian, A.; Mahmoudi, A.H.; Mehmanparast, A. Shot peening effects on residual stresses redistribution of offshore wind monopile multi-pass weldments. Mar. Struct. 2019, 66, 106–120. [CrossRef] p p 80. Aldosary, M.; Wang, J.; Li, C. Structural reliability and stochastic ﬁnite element methods: State-of-the-art review and evidence- based comparison. Eng. Comput. 2018, 35, 2165–2214. [CrossRef] 80. Aldosary, M.; Wang, J.; Li, C. Structural reliability and stochastic ﬁ based comparison. Eng. Comput. 2018, 35, 2165–2214. [CrossRef] 80. Aldosary, M.; Wang, J.; Li, C. Structural reliability and stochastic ﬁnite element methods: State-of-the-art review and evidence- based comparison. Eng. Comput. 2018, 35, 2165–2214. [CrossRef] p g p 81. Li, Y.-J.; Huang, B.; Li, C.-Q. Hybrid perturbation-Galerkin methods for structural reliability analysis. Probabilistic Eng. Mech. 2017, 48, 59–67. [CrossRef] 81. Li, Y.-J.; Huang, B.; Li, C.-Q. Hybrid perturbation-Galerkin methods for structural reliability analysis. Probabilistic Eng. Mech. 2017, 48, 59–67. [CrossRef] 82. Feng, J.; Wu, D.; Gao, W.; Li, G. Hybrid uncertain natural frequency analysis for structures with random and interval ﬁelds. Comput. Methods Appl. Mech. Eng. 2018, 328, 365–389. [CrossRef] 82. Feng, J.; Wu, D.; Gao, W.; Li, G. Hybrid uncertain natural frequency analysis for structures with random and interval ﬁelds. Comput. Methods Appl. Mech. Eng. 2018, 328, 365–389. [CrossRef] 37 of 37 Metals 2021, 11, 50 37 of 37 83. Papadopoulos, V.; Papadrakakis, M.; Deodatis, G. Analysis of mean and mean square response of general linear stochastic ﬁnite element systems. Comput. Methods Appl. Mech. Eng. 2006, 195, 5454–5471. [CrossRef] y p pp g , , [ ] 84. Gong, C.; Zhou, W. References Comparing a fracture mechanics model to the SN-curve approach for jacket-supported offshore wind turbines: Challenges and opportunities for lifetime prediction. In Proceedings of the ASME 2016 35th International Conference on Ocean, Offshore and Arctic Engineering, OMAE2016, Busan, Korea, 19–24 June 2016; American Society of Mechanical Engineers (ASME): Busan, Korea, 2016; pp. 1–10.
https://openalex.org/W2120285536	https://www.investigacionesgeograficas.com/article/view/2009-n48-turismo-rural-y-desarrollo-territorial-en-espacios-indi/pdf, https://www.redalyc.org/pdf/176/17620925007.pdf, https://rua.ua.es/dspace/bitstream/10045/15328/1/IG_48_07.pdf	es		Turismo rural y desarrollo territorial en espacios indígenas de México	Investigaciones geográficas/Investigaciones geográficas	2,009	cc-by	8,621	Investigaciones geográficas, nº 48 pp. 189 - 208 ISSN: 0213-4691 Instituto de Geografía Universidad de Alicante TURISMO RURAL Y DESARROLLO TERRITORIAL EN ESPACIOS INDÍGENAS DE MÉXICO José Pedro Juárez Sánchez y Benito Ramírez Valverde Colegio de Postgraduados, Campus Puebla, México María Guadalupe Galindo Vega Directora de Cultura de San Pedro Cholula, Puebla, México RESUMEN México es uno de los principales productores de café a escala mundial, lo que genera una gran cantidad de divisas al país. Sin embargo, es producido en su mayoría por indígenas en pequeñas plantaciones y viven en condiciones de extrema pobreza. Esta situación se ha acentuado a raíz de las crisis recurrentes del cultivo. Este estudio se realizó en cuatro municipios con plantaciones de café, ubicados en la sierra Nororiente de Puebla, México. La investigación se centró en la posibilidad de impulsar el agroturismo en espacios marginados, con población pobre e indígena. Los resultados muestran que el 80 por ciento de los entrevistados desea participar en actividades agroturísticas. Se concluye que mediante el apoyo a este tipo de proyectos por parte de las autoridades gubernamentales, el agroturismo puede representar una actividad complementaria a la producción de café que proporcione recursos económicos y que permita contribuir al incremento y calidad de la producción y, consecuentemente, a mejorar las condiciones de vida de la familia campesina. Palabras clave: Café, agroturismo, indígena, pobreza, marginación, Puebla ABSTRACT Mexico is one of the world main coffee producers which generates an important income for the country. However, it is mainly produced by indigenous groups in small land areas and in extremely poor conditions. This situation has become Fecha de recepción: 30 de septiembre de 2009 N.º 48 B.indd 189 Fecha de aceptación: 29 de enero de 2010 23/9/10 20:16:36 190 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega even worse due to the frequent crisis in the crop price. This study was carried out in four municipalities where coffee is grown, located in the Sierra Nororiental of Puebla, Mexico. The research focused on the possibility of impelling agrotourism in marginalized spaces with poor indigenous population. Results show that 80% of the interviewed people are willing to participate in agrotourism activities. It is concluded that the support given by the governmental authorities to this kind of projects can make of agrotourism a supplementary activity to the production of coffee. This is expected to increase the economic resources and the quality of the production thus improves life conditions of the peasants. Key words: coffee agrotourism, poverty, marginalization, Puebla. 1. Introducción A principios de la década de los 80´s del siglo pasado, organismos multinacionales como el Banco Mundial (BM), Banco Interamericano de Desarrollo (BID) y el Fondo Monetario Internacional (FMI) impulsaron políticas de corte neoliberal mediante programas de ajuste estructural, principalmente en los países subdesarrollados. Esta estrategia trajo como consecuencia el abandono del modelo de Industrialización Sustitutiva de Importaciones (ISI), para implantar una política proexportadora basada en productos no tradicionales (Ramos, 1993). Con el nuevo modelo económico se fortalece a las ciudades, con el fin de comercializar y promocionar una imagen global urbana competitiva frente a otras ciudades. Éstas se transforman en productos de una city marketing atrayendo servicios, actividades comerciales, inversiones, además de la promoción creciente del turismo urbano (Sassano, 2001: 101). Además, este modelo repercutió directamente en la calidad de vida de gran parte de la población; al respecto, diversos analistas coinciden en determinar como causa generadora en las últimas décadas de la pobreza rural, al conjunto de procesos asociados con el modelo de desarrollo que han adoptado los países latinoamericanos (IICA y BID, 2002: 12). Una de sus repercusiones ha sido el incremento de la población pobre, por lo menos en el caso de México. En este sentido, Echeverría (1998: 2) menciona que la pobreza rural aumentó de 46 por ciento en 1992 a 47 por ciento en 1994, y Lustig et al. (1997: 12) indican que entre 1989 y 1994 la pobreza, tanto extrema como moderada, registró un aumento entre los trabajadores rurales, en el sector primario y en las regiones Sur y Sureste del país. En 2004, el 28 por ciento de los habitantes en zonas rurales se encontraba en niveles de pobreza extrema y el 57 por ciento en situación de pobreza moderada; aunque sólo una cuarta parte de la población mexicana vive en zonas rurales, en estas regiones reside el 60,7 por ciento de la población en pobreza extrema y el 46,1 por ciento de los moderadamente pobres (BIRF y BM, 2005: 69). Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 190 23/9/10 20:16:36 Turismo rural y desarrollo territorial en espacios indígenas de México 191 Los habitantes de las comunidades indígenas han sido los más afectados, no sólo por el impacto de la política agrícola seguida por el Estado mexicano, sino también por el descenso del precio internacional del café, ya que el 65 por ciento de sus productores son indígenas minifundistas y producen un tercio de la producción en México. Son los pobres de los pobres. Ante este nuevo orden económico, los pobladores del medio rural marginados de los beneficios de la globalización, impulsan nuevas formas de generación de ingresos para sobrevivir en una agricultura subsumida en la crisis. Entre las estrategias que han desarrollado destaca la diversificación de las actividades agrícolas, así como el surgimiento de actividades no agrícolas y los movimientos migratorios a las grandes ciudades y al extranjero. Se puede decir que la agricultura minifundista está en crisis y no es capaz de generar ingresos suficientes para sostener a los miembros de una familia; al respecto, Pérez (2001: 22) menciona que existe una pérdida relativa de la significación económica y social de los sectores primario y secundario, y una evidente terciarización en los espacios rurales. Es importante destacar que la terciarización del medio rural va en aumento, pero, sobre todo, como una meta del nuevo modelo de desarrollo impulsado por los organismos multinacionales. Pérez concluye que lo rural ya no es equivalente a lo agrícola y, al mismo tiempo, que la llamada tercera revolución agrícola implica que lo agrícola no sea exclusivamente la producción primaria. Todo esto conduce a la desagrarización de la actividad productiva. En ese sentido, se fomenta el financiamiento a proyectos turísticos en espacios rurales que presentan un gran valor natural o cultural, pero hasta el momento han sido insuficientes. Mediante estos proyectos se pretende fomentar y diversificar los atractivos turísticos del país, para incrementar la afluencia turística, generar empleos y mejorar los servicios básicos (boletín de prensa 2006). Un ejemplo de lo anterior es el municipio de estudio –Cuetzalan-, declarado como Pueblo Mágico ubicado en la sierra Nororiente de Puebla. En donde las actividades turísticas en estos espacios son bien vistas y consideradas como un instrumento para mejorar las condiciones de vida de la población que las habita. 2. Turismo rural y desarrollo territorial en México En los últimos años, la forma de promover el desarrollo ha cambiado en Latinoamérica y especialmente en México; esta nueva visión obedece fundamentalmente al cambio de modelo de desarrollo de Industrialización Sustitutiva de Importaciones (ISI) por uno de corte neoliberal. La implementación de este modelo trajo consigo profundos cambios estructurales en la economía, como es el fomento del libre mercado. Para ello fue necesario debilitar a los estados nacionales; ahora el mercado se ha convertido no sólo en el principal, sino en el único Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 191 23/9/10 20:16:36 192 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega dispositivo de regulación económica (Linck, s/f: 1-2). El neoliberalismo cree en la apertura de las economías nacionales a los mercados globales sin intervención del Estado. Ello ha traído consigo el sacrificio de sectores no competitivos como el industrial, en aras de la competencia internacional (Kay, 1998: 10). El sector agrícola no es la excepción. Al sector agropecuario en el modelo ISI se le asignó el papel de producir alimentos baratos para la población urbana y generar divisas mediante su exportación. Ahora la agricultura no sólo es explotada, sino excluida. Esta situación agravó la crisis del sector agrícola, y ante esto, en los espacios rurales se impulsa el surgimiento del desarrollo del territorial rural, que se define como un proceso de transformación productiva e institucional de un espacio rural, con el fin de disminuir la pobreza rural (Schejtman y Berdagué, 2004: 30). El termino de desarrollo territorial rural se basa en la revaloración del espacio rural y su geografía, como unidad de gestión que permita integrar una economía multisectorial, dimensiones políticas, sociales, culturales y ambientales, que han venido construyendo una institucionalidad dinámica, aunque compleja, y que ofrece la posibilidades de una respuesta a los errores que ha mostrado el desarrollo rural en las ultimas décadas (Echeverri y Ribero, 2003: 23). En este modelo de desarrollo se dejan de operar políticas sectoriales por políticas sustentadas en el territorio, ya que la agricultura de tipo minifundista no es capaz de generar los ingresos suficientes para satisfacer las necesidades de quienes la practican, debido al escaso apoyo que reciben del Estado y a los bajos precios de sus productos agropecuarios. En el caso mexicano, en su implementación fue necesario que se abandonara la forma de conducir la economía por parte del Estado, es decir, se dejó la planificación tradicional que se caracterizaba por ser normativa, indicativa y sectorizada, por una planificación del desarrollo que, según sus promotores, busca mejorar la calidad de vida de su población, mediante la distribución equitativa del ingreso, y que además piensa en la conservación y regeneración de los recursos naturales y promueve la participación de los pobladores en el proceso de toma de decisiones. Es decir, que se apuesta por una planificación de abajo hacia arriba, y para su implementación, el Estado descentralizó funciones y atribuciones a los gobiernos de los estados y municipios, con el objetivo de impulsar el desarrollo en los territorios, pero México está lejos de alcanzar una distribución equitativa del ingreso; si acaso se tiene una diversificación de las actividades económicas en algunas regiones rurales del país. Por otro lado, Silva (2002: 106-107) menciona que la generación de excedentes productivos en el mundo desarrollado, la creciente apertura del comercio internacional de productos agrarios, la constatación del papel desempeñado por Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 192 23/9/10 20:16:36 Turismo rural y desarrollo territorial en espacios indígenas de México 193 los cultivos en la conformación de los paisajes rurales, la creciente demanda de espacios libres, está conduciendo a cambiar el papel tradicionalmente asignado a la agricultura, y a que se superponga a otros más relacionados con la provisión de bienes ambientales y culturales. A pesar de ello, la producción agropecuaria y forestal siguen siendo la principal actividad económica en los espacios rurales de México y, de acuerdo al desarrollo territorial rural, es posible promover su dinamización mediante el impulso de la agroindustrialización de los productos agropecuarios, de la producción de artesanías y de la participación de la población local en el turismo rural, es decir, se tiene que promover la diversificación productiva en las unidades de producción e impulsar el Empleo Rural No Agrícola (ERNA). Esta nueva forma de enfrentar la crisis en la agricultura hace que tomen un lugar primordial otras actividades no agrícolas, como el turismo, en el desarrollo territorial rural, que actúa no sólo como generador de ingresos, sino también en la creación de empleos, mejoramiento de la calida de vida de la población local y en la conservación de los recursos naturales. En este sentido, Font (2006: 7) menciona que el turismo juega un rol importante en la economía de muchos países de ingresos bajos, donde la llegada de turistas internacionales superaron los 175 millones en 2005. En términos de ingresos, corresponden a un valor para estos países de 85 mil millones de dólares. De hecho, el turismo es uno de los mayores sectores generadores de divisas en un gran número de países en desarrollo y la primera fuente de divisas en 46 de los 50 países menos adelantados. La OMT (2003) argumenta que se ha convertido en un sector importante para la mayoría de los países en desarrollo y, en algunos casos, ha desbancado a la agricultura comercial y a otras industrias primarias como principal fuente de rentas nacionales, empleo e ingreso por exportaciones. En Kenya, el turismo ha desplazado a la producción de té, café y productos hortícolas y se ha convertido en el principal generador de divisas, al igual que ha ocurrido en Costa Rica. El desarrollo del turismo ha contribuido al desarrollo de Botswana, que dejó de figurar entre los países menos desarrollados. Así mismo, el turismo ha sido vital para la economía cubana a partir de que se redujera el apoyo económico de la antigua Unión de Repúblicas Socialistas Soviéticas. Los ingresos que genera el turismo han ayudado a las economías del Caribe a afrontar el descenso de los precios de sus principales productos de exportación como el plátano y el azúcar. La incorporación del turismo en China como medio de desarrollo económico le está permitiendo alcanzar un alto índice de crecimiento en algunas zonas. Es por eso que la creciente multifuncionalidad de los espacios rurales se está viendo acompañada de la asunción de nuevas funciones, en donde el turismo puede ser una de ellas. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 193 23/9/10 20:16:36 194 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega Es importante mencionar que en las economías rurales deprimidas la agricultura sigue siendo la actividad dominante, es por ello, que la política territorial tiene que ir dirigida a ésta, pero a su vez tiene que buscar la diversificación de las actividades no agrícolas, sin caer en su total dependencia. En este enfoque y con relación al cambio de las funciones del medio rural, Vera et al. (1997: 134) comentan que el turismo rural hace una contribución valiosa a las economías rurales que se observa en términos financieros, de empleo, en la contribución de la financiación a la conservación del medio ambiente, de estímulo y motivación a la adopción de nuevas formas de trabajo y contribuye a dar una nueva vitalidad a las economías. Se debe tener cuidado en no caer en ideas falsas que se tienen del turismo rural, al cual le atribuyen la solución a los problemas del desarrollo rural; es por ello importante tener en cuenta las observaciones de José Luis Andrés (2000: 47), al decir que el turismo rural es complementario más que una alternativa. Juárez y Ramírez (2007) indican que el turismo es sólo una fuente de ingresos más que se tiene que promover en los territorios rurales, pero que no se tienen que descuidar las actividades productivas que realizan los agricultores. Éstas tienen que fortalecerse y, a su vez, crear nuevas oportunidades de trabajo que permitan la generación de ingresos no agrícolas. En este mismo sentido se pronuncia Ascanio (2004: 157), al decir que el turismo rural no debería desplazar las actividades agrícolas y culturales de estos espacios geográficos, sino, por el contrario, se deben estimular y fortalecer. Pero el turismo sólo traerá beneficios económicos si va dirigido a los habitantes de la comunidad, ya que muchas veces, y específicamente en el ecoturismo, los beneficiados son los operadores de empresas turísticas. Lo que se busca es fortalecer en el medio rural las actividades agrícolas e impulsar la diversificación de actividades no agrícolas, entre ellas el turismo. La promoción del turismo rural en México se da en dos vertientes. La primera es a través de la inversión del gobierno federal en programas de desarrollo turístico de carácter general, en el cual involucra de manera marginal al turismo rural (Juárez y Ramírez, 2006) y en donde el turismo de sol y playa es prioritario en la economía y en la política turística mexicana. En esta vertiente destaca el programa de Pueblos Mágicos al que pertenece el municipio de Cuetzalan y que forma parte de este estudio. En la segunda se encuentra el turismo rural, que se basa fundamentalmente en los inmuebles de las antiguas haciendas del México porfiriano; éstas han iniciado su operación con grandes presupuestos para su remodelación y rescate: es el caso de las haciendas henequeneras en Yucatán y del programa Haciendas y Casas Rurales de Jalisco (Amaya, 2005: 54), así como las haciendas cocoteras en el estado de Tabasco. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 194 23/9/10 20:16:36 Turismo rural y desarrollo territorial en espacios indígenas de México 195 En el país, en el 2000, existían 15 empresas comunitarias apoyadas por el Fondo Nacional de Empresas Sociales (FONAES) de un total de 442 empresas operadoras de ecoturismo y turismo de aventura. Esta actividad se concentra principalmente en siete estados: Quintana Roo (17,2%), Distrito Federal (14.2%), Baja California Sur (8,9%), Oaxaca (8,7%), Chiapas (8,5%), Jalisco (8,4%) y Veracruz (7,6%) (SECTUR, 2001). Actualmente existen un total de 1.239 empresas y proyectos orientados a atender el ecoturismo y turismo rural, de las cuales 872 están en operación y 367 están en etapa de desarrollo. De dicha oferta, 325 son empresas privadas y 914 comunitarias. Entre los estados con mayor oferta destacan Chiapas, Distrito Federal, Oaxaca, Michoacán, Puebla, Veracruz y Quintana Roo (boletín de prensa, 2006). Este tipo de empresas están ubicadas espacialmente en la denominada Mesoamérica, que se caracteriza por tener un importante patrimonio natural y cultural capaz de atraer a los turistas. 3. El turismo en el estado de Puebla Bajo el modelo neoliberal, el Estado reestructuró la forma de intervenir en el fomento del desarrollo económico y social del país; en el sector turístico promueve fundamentalmente el turismo de playa en los Centros Integralmente Planeados (CIP) como Cabo San Lucas y Loreto en el estado de Baja California Sur; Cancún en Quintana Roo y Bahías de Huatulco en Oaxaca. Este tipo de turismo constituye un elemento básico para la economía de México. En menor intensidad promueve el turismo cultural o de ciudad en los principales centros poblacionales del país; entre ellos destaca la capital de la república, Guadalajara, Monterrey, Guanajuato y Puebla, quedando excluidos los territorios rurales poseedores de gran potencial turístico. En estas ciudades se abandona, paulatinamente las actividades del sector secundario y se impulsa la prestación de servicios; entre ellos destaca el turismo, ya sea cultural, de negocios, salud, congresos y compras. Esta actividad es apoyada económicamente por el Estado; quieren hacer de las ciudades parques temáticos para atraer al turismo nacional e internacional, relegando a un segundo término al turismo rural. Esta política debe orientarse a aprovecharse el espacio rural para diversificar la oferta turística, con el objetivo de impulsar el desarrollo territorial rural; es el caso del territorio poblano. En las últimas décadas el estado de Puebla ha buscado fortalecer su economía a través de las actividades turísticas y se han observado dos variantes en su evolución: la ciudad capital ha avanzado en la vertiente de turismo de negocios y congresos; para ello se construyó el centro de convenciones de Puebla, donde se realizan importantes eventos: ferias, convenciones, congresos, etc., configurándose como el más importante y moderno segmento turístico de este tipo del centro Sur de la república mexicana. También se está impulsando el turismo de Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 195 23/9/10 20:16:36 196 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega «ciudades y de compras», para ello se han realizado importantes obras; entre ellas destaca la construcción del centro comercial Ángelopolis y el mejoramiento de la imagen urbana del centro histórico de la capital del estado. En la segunda vertiente se menciona al turismo arqueológico, etnográfico, ecológico y de aventura, que no ha sido aprovechado a plenitud por el bajo desarrollo empresarial, de infraestructura y de subvenciones del estado dentro de una política de inversión. Otro importante punto es la construcción de carreteras de cuota, que han permitido disminuir el tiempo de traslado de la capital a las principales ciudades donde existen centros industriales de importancia en el estado. Con las inversiones en infraestructura se espera que se beneficie y se incremente la afluencia de turistas para visitar el interior del estado. El turismo en Puebla, específicamente el de ciudad, ha tenido momentos importantes, ha logrado colocarse entre los primeros lugares en ocupación hotelera en la categoría de «ciudades del interior», con más de 12 puntos arriba de la media nacional y por encima de destinos como Oaxaca, Monterrey y Guadalajara (plan estatal de desarrollo, Puebla, 2005: 79). En el 2006 la afluencia de turistas al estado fue de 5.751.141, el 70,3% se concentró en la ciudad de Puebla. Se puede observar que al porcentaje restante se le puede catalogar como turismo rural, teniendo en cuenta que en las estadísticas se considera a los municipios de San Andrés y San Pedro Cholula como parte de la metrópoli Angelopolitana. El tipo de turistas que visitan Puebla son mayoritariamente nacionales; sólo el 18,3% es de procedencia extranjera. La derrama económica en este sector fue del orden de los 437,56 millones de dólares, en el interior del estado representó el 33% del gasto total (SECTUR, 2006). El estado de Puebla tiene una infraestructura hotelera compuesta por 426 hoteles; el 34,5% se concentra en la capital y aproximadamente representan el 3,2% de la oferta nacional. Se tienen 13.203 habitaciones registradas en el Sistema de Información Turística Estatal (SITE), que equivalen al 2,5% de la oferta nacional; los hoteles en el interior del estado tienen como máximo 3 estrellas, salvo los municipios de Izúcar de Matamoros, Atlixco, Cuetzalan, Huauchinango, Tehuacán, Xicotepec y Zacatlán, que cuentan con hoteles de 4 y 5 estrellas. La actividad turística en el estado genera 24.142 empleos de manera directa y 72.425 indirectos y se concentran fundamentalmente en la capital. El proceso de turistificación que se promueve en la capital del estado se ha visto reforzado con el crecimiento y la diversificación de la oferta restaurantera; de 732 se incrementó a 976 establecimientos en el periodo de 1999 a 2003. Se puede decir que el estado de Puebla tiene como polo turístico a su ciudad capital –municipio de Puebla–, ello se manifiesta en la centralización de la oferta hotelera, servicios, derrama económica de los turistas y en la generación de empleos. El turismo rural debido a la política turística se puede decir que es embrionario. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 196 23/9/10 20:16:36 Turismo rural y desarrollo territorial en espacios indígenas de México 197 Al respecto se tiene que en el estado de Puebla se promovieron diversos destinos turísticos en el ámbito rural, en el 2006, mediante el Convenio de Reasignación de Recursos entre el gobierno estatal con la Federación; éste asignó 1.143.773,11 dólares para proyectos turísticos. En Zacatlán, específicamente en el Valle de Piedras Encimadas, se invirtieron 265.876,86 dólares para el mantenimiento de edificios; en Cuetzalan, que pertenece al programa de Pueblos Mágicos y es catalogado como uno de los más exitosos, se invirtieron 326.075,41 dólares para el mejoramiento de la imagen urbana y señalización. En Xicotepec de Juárez se llevó a cabo el proyecto de aprovechamiento turístico del centro ceremonial Xochipila con un importe de 343.388,12 dólares, y en Jonotla se construyó el mirador «El Peñón» con una inversión de 208.441,17 dólares (Bretón, J. 2007: 5). En el Peñón se promueve el turismo religioso en él se venera la Aparición de la Virgen de Guadalupe, y uno de los grupos de visitantes más destacados son los originarios de esta comunidad que migraron al área metropolitana de la capital de la república. Como se puede observar, se tienen grandes inversiones, pero son insuficientes, ya que sólo apoyaron a 5 municipios localizados en el Norte del estado. En suma, se puede decir que el turismo rural en Puebla es marginal, pero tiene grandes posibilidades de desarrollarse y aprovechar las ventajas comparativas de sus territorios, como sería la belleza natural y cultural, en beneficio de la población local. 4. Objetivos y Metodología En esta investigación se plantea que el impulso del turismo rural, específicamente el agroturismo, representa una actividad complementaria que permita mejorar los ingresos de la población local, sin descuidar su principal actividad agropecuaria, que en este caso es la producción de café bajo minifundio. Es decir, que en los municipios de estudio se tiene que fomentar el desarrollo agrícola e implantar una política de desarrollo territorial rural que diversifique las actividades que se desarrollan en el ámbito rural. En este trabajo se busca relacionar la producción de café por indígenas y el agroturismo como complemento a la actividad agrícola. La región de estudio contempla cuatro municipios que fueron seleccionados por tener plantaciones de cafetales, recursos naturales de gran valor para incursionar en actividades turísticas y población indígena considerada de muy alta marginación; ellos son: 1) Cuetzalan; 2) Huehuetla; 3) Huitzilan; y 4) Ixtepec. Se encuentran enclavados en la sierra Nororiente y se ubican entre los 20º de latitud Norte y los 97º de latitud Oeste. Tienen una altitud sobre el nivel del mar que oscila de los 540 a los 1000 m.s.n.m.; en la figura 1 se aprecia la ubicación de los municipios en el contexto nacional y estatal. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 197 23/9/10 20:16:36 198 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega Figura 1. Localización geográfica de los municipios de estudio Fuente: Síntesis Geográfica del Estado de Puebla, 2000. Elaboración propia Los principales núcleos de población de los municipios de estudio se concentran en las cabeceras municipales, los otros poblados se caracterizan por ser pequeños y dispersos, propios de los municipios rurales. El municipio de Cuetzalan tiene 45.010 habitantes, Huehuetla 16.130, Huitzilan 6.589 e Ixtepec 11.670 pobladores. Los centros poblacionales con mayor importancia poblacional o económica tienen mercado municipal, electricidad, agua potable, clínicas de salud, teléfono, Internet y biblioteca pública. En el aspecto educativo tienen escuelas de primaria, secundaria, preparatoria y una universidad intercultural en Huehuetla. La actividad económica que predomina en los municipios es la producción de café; la población económicamente activa se emplea principalmente como jornalero, trabajador por su cuenta y trabajador familiar no remunerado. En el caso de Cuetzalan el turismo constituye la segunda actividad económica; tiene 27 hoteles: 11 son de una estrella, 3 son clasificados como de dos, 6 tienen tres, 1 posee cuatro y otro tiene cinco estrellas. El municipio de Huehuetla tiene un hotel de dos estrellas. Cuetzalan cuenta con restaurantes con excelente servicio, en Huehuetla sólo se tienen fondas; Cuetzalan cuenta con servicios de gasolineras, venta de refacciones para automóviles y farmacias. Los otros municipios carecen de estos servicios. Existen carreteras seguras, que han reducido hasta en un 40% el tiempo de traslado de la ciudad de Puebla a la zona de estudio. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 198 23/9/10 20:16:37 Turismo rural y desarrollo territorial en espacios indígenas de México 199 Estos resultados señalan que los espacios turísticos son incipientes en la región y los servicios turísticos están en manos de las personas que tienen mayor poder económico y social. Los recursos turísticos de los municipios son principalmente: naturaleza, destacando la presencia de selvas; arqueología, representada por los vestigios arqueológicos de la antigua ciudad del Totonacapan; espeleología; culturas vivas, donde se acentúan los grupos indígenas totonacos y nahuas; práctica de la caficultura, su agroindustrialización y producción de maíz en ladera; y la ciudad de Cuetzalan, que es considerada Pueblo Mágico por sus características históricas, culturales y arquitectónicas. Para calcular el tamaño de muestra de este estudio se utilizó un muestreo estratificado aleatorio, con distribución proporcional al tamaño de los municipios. El universo de muestreo estuvo constituido por los productores de café de los cuatro municipios en estudio, tomando como base el censo de caficultores realizado en 2001. La ecuación para estimar el tamaño de muestra es presentada por Gómez (1979) y se especifica de la siguiente forma: Ecuación 1. Expresión matemática para calcular el tamaño de muestra en un estratificado aleatorio con distribución proporcional donde: d = Precisión = Confiabilidad N = Tamaño de la población; Ni = Tamaño de la población del estrato i 2 = Varianza del estrato i s i con Con una precisión del 15% de la media y una confiabilidad del 95%, se obtuvo el tamaño de la muestra de 216 productores, y mediante un procedimiento aleatorio se seleccionó a los productores de café incluidos en la muestra. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 199 23/9/10 20:16:39 200 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega En este estudio se revisan variables de tipo social, para conocer si los agricultores indígenas están dispuestos a diversificar sus actividades, y entre ellas destaca el turismo rural; ello permitirá conocer si el agroturismo representa para los caficultores un componente complementario al desarrollo rural. Para analizar los datos se utilizó estadística descriptiva, se realizaron pruebas de t y chi – cuadrada (χ2), y en los resultados se hicieron comparaciones entre indígenas productores de café interesados y no interesados en realizar actividades turísticas. 5. La producción de café y el agroturismo El trabajo de investigación se centró en la posibilidad de impulsar el turismo rural, específicamente el agroturismo, en espacios marginados y con población indígena productora de café. Se indagó sobre el interés en participar en actividades turísticas por parte de los caficultores de cuatro municipios. El trabajo considera central la comparación entre productores con interés y aquéllos que no lo tienen, en participar en actividades de agroturismo, con la finalidad de entender las motivaciones de los productores y así hacer propuestas turísticas que apoyen el desarrollo y mejora de las condiciones de vida de las familias indígenas. El idioma representa un atractivo importante en el turismo cultural; al respecto, los habitantes de los municipios de Cuetzalan y Huitzilan en su totalidad hablan náhuatl, en Huehuetla el 35% de los productores dijo que su lengua materna es el Totonaco y el 30% en Ixtepec. En el 2000, en México, de acuerdo al Instituto Nacional para la Evaluación de la Educación (INEE) existían un total de 10.185.060 indígenas y representaban el 10,45% del total nacional (INEE, 2006: 63). El estado de Puebla ocupa el cuarto lugar en población de habla indígena, concentra el 9.4% del total nacional, habitan el territorio poco más de 550 mil indígenas y representan el 13% de la población total. Las principales etnias indígenas son las siguientes: náhuatl (73,7%), totonaco (17,8%), popoloca (2,6%), mazateco (2,1%), otomí (1,5%), mixteco (1,5%) y tepehua (0,5%). Esto indica que se tiene un excelente potencial turístico en los municipios de estudio para atraer turistas en este segmento del mercado. Las plantaciones de café de los campesinos que quieren participar en actividades turísticas de acuerdo al trabajo de campo indican que tienen en promedio 1,54 y 1 hectárea los que no tienen interés en esta actividad. No existe diferencia estadística entre los grupos (t = 1,769; p = ,078) con respecto al número de hectáreas en posesión. En la sierra Nororiente el promedio de la tierra dedicada al café es de 0,85 hectáreas (Romero, 2006: 3), la cual es muy similar a la que tienen los que no desean participar en proyectos turísticos. En este sentido, destaca Cánoves, et al. (2005: 69) que es importante tener tierras para practicar el agroturismo y que en el caso español, en la modalidad de agroturismo, se exige tener una determinada superficie dedicada a la agricultura. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 200 23/9/10 20:16:39 Turismo rural y desarrollo territorial en espacios indígenas de México 201 Con respecto a la propiedad de la tierra, el 92.1% dijo que sus tierras pertenecen al régimen de la propiedad privada; este resultado posibilita la inversión de capital en los municipios, ya que algunos autores mencionan que el tipo de propiedad, y en especial la pública, frena la inversión, y no es la excepción en el sector turístico. Al respecto, Echeverría y Bello (2002: 3) mencionan que la importancia de la tierra en el medio rural abarca aspectos económicos, ambientales, sociales y culturales. Además, comentan que es un medio para acumular bienestar, para atraer inversiones y permite generar ingresos en actividades agrícolas y no agrícolas. El cuadro 1 muestra que el 80% de los caficultores están dispuestos a incursionar en actividades turísticas; además, pone de manifiesto que en los municipios con mayor infraestructura hotelera y experiencia en servicios turísticos, como Cuetzalan y Huehuetla, los entrevistados están menos interesados en esta actividad, debido a que piensan que tienen que invertir mucho dinero en la infraestructura de alojamiento, restaurantes y espacios recreativos y que no están a su alcance. Otra situación es que los productores consideran que es una actividad que está siendo explotada por los pobladores que tienen mayores recursos económicos en la comunidad y, por lo tanto, está fuera de sus posibilidades. Es claro que para este tipo de turismo se pueden realizar inversiones de bajo costo, en donde participe la comunidad activamente en la construcción de su desarrollo. Tabla 1. Productores a los que les gustaría participar en actividades turísticas Productores a los que les gustaría participar en actividades S Municipio Total N F % F % F % Cuetzala 7 7 2 2 10 10 Huehueti 2 6 1 3 3 10 Ixtepe 2 9 2 7 3 10 Huitzila 4 9 2 4 4 10 16 8 4 2 21 10 Total Fuente: Elaboración propia, encuesta a agricultores en 2005 Es importante destacar que las personas que desean participar en actividades turísticas perciben que son más pobres que las del otro grupo, debido a que el 40% mencionó que son los más pobres de la comunidad, contra el 17% de los no participantes. Es por ello que el 92% de los participantes consideraron que pueden mejorar su ingreso y el 63% de los que no desean participar opinó lo mismo. Berdegue, et al. (2001: 185) encuentran que la ausencia de fuentes no agrícolas de ingreso, en los hogares rurales pobres sería varias veces mayor la Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 201 23/9/10 20:16:39 202 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega magnitud de su pobreza que los que tienen alguna fuente de ingreso no agrícola. Las mismas investigaciones confirman que los hogares rurales pobres recurren al empleo no agrícola no sólo para elevar su ingreso total, sino para amortiguar durante el año las fluctuaciones del ingreso, que es una de las características de la pobreza rural. Ello muestra que el agroturismo puede ser una actividad complementaria a la actividad agrícola que realizan cotidianamente los caficultores y, especialmente, los que consideran que tienen mayores niveles de pobreza. El turismo rural puede revitalizar la economía de este tipo de agricultores. En el turismo rural existen diferentes modalidades; una de ellas es el agroturismo, el cual consideramos es el más apropiado para que participen de manera más inmediata en la prestación de servicios turísticos los caficultores; al respecto, Sayadi y Calatraba (2001: 132) mencionan que en términos generales es un tipo de turismo rural en donde el componente importante de la oferta turística es la acogida, alojamiento, gastronomía, ocio y la participación en la explotación agraria. Ernesto Barrera (2006: 74) es más específico y menciona que se caracteriza porque el visitante participa activamente de las actividades productivas. Esta conceptualización es la que se apega más a nuestro estudio; es por ello importante conocer la tecnificación de las actividades más relevantes del proceso productivo que se realiza en las plantaciones de café y que se considera que les gustaría conocer y practicar a los turistas. El conocimiento del proceso productivo del café representa una actividad pedagógica para los turistas, ya que en el mundo rural el paisaje y la tranquilidad no son suficientes para retener a gran parte de la población urbana. Este tipo de turismo es actualmente muy apreciado en los países desarrollados. En contraparte, en los países con menor desarrollo existen zonas rurales deprimidas económicamente, pero con potencial turístico. Un ejemplo es el caso de las fincas cafetaleras en Colombia, donde los propietarios reciben y alojan a visitantes y les hacen participar en el proceso productivo del café (Gurría, 2000). Las personas que desean participar en turismo tienen 15,4 años trabajando sus plantaciones de café y el otro grupo tiene 18,9 años. Se encontró que no existe diferencia estadística entre los grupos (t = -1,937; p = ,054), ello muestra que los caficultores tienen experiencia en el manejo de las plantaciones y que puede convertirlas en producto turístico si se maneja adecuadamente. Un trabajo importante en el proceso productivo del café es el control de las malezas, y que interesaría realizar a los turistas; el 82% de los que quieren participar en actividades turísticas lo realiza; de este porcentaje el 83,5% lo hace de manera manual y el 6,5% utiliza productos biológicos. El 88% de los no interesados controlan las yerbas, sólo el 2,7% de éstos realizan esta actividad empleando productos químicos y el 94,6% manualmente. Con respecto a la aplicación de Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 202 23/9/10 20:16:39 Turismo rural y desarrollo territorial en espacios indígenas de México 203 fertilizante químico, el 27,2% de los que desean participar en turismo lo aplican y el 9,5% de los no interesados. El fertilizante orgánico es más empleado, ya que el 45% de los interesados y el 21,4% de no interesados lo manejan. La poda en los cafetales no es muy común, el 45 y el 21% de los que desean no participar mencionaron que la realizan. El 77 y el 83% de los interesados y no interesados en actividades turísticas dijo tener plagas o enfermedades en sus plantaciones; la principal plaga que tienen los cafetales es la broca; en el grupo de los interesados lo mencionó el 62,4% y en el otro grupo el 65,7%. La controlan mediante prácticas culturales, productos químicos y de manera biológica. Estos resultados ponen de manifiesto que la tecnificación de las plantaciones de café no es muy adecuada, es por ello que se tiene que capacitar a los caficultores en el control de malezas, aplicación de fertilizante, en la poda y en el control de plagas, que permita tener la plantación en condiciones óptimas para los posibles visitantes. Para hacer atractivas las prácticas productivas deben de proporcionarle una orientación didáctica, ya que se plantea un turismo activo en el que a priori se requiere una implicación participativa del turista (Vera, et al. 1997, 55). Es por ello que una de las vertientes a seguir en las plantaciones es la producción de café orgánico. Al respecto, se encuentra que 79,1% de los que desean participar han escuchado hablar de la producción de café orgánico y el 85,3% de los no participantes. Estas personas consideran que representa un buen negocio económico producirlo; además, desearían iniciar su producción y también mencionaron que saben cómo producirlo. Éste es otro producto interesante para que se promueva el consumo y comercialización del café orgánico entre los turistas, beneficiando aún más a las comunidades rurales. A su vez, la capacitación también repercutirá en el incremento de la producción y calidad del café a cosechar, lo que representa una gran importancia, debido a que, por un lado, se benefician los productores con la presencia y consumo de turistas, y por otro lado, se mejora la tecnología, la productividad, la calidad del producto y los ingresos obtenidos de la venta del café. Este tipo de prácticas más la recolección del café, y específicamente el orgánico, son un excelente producto que interesa a los turistas y que la población local puede brindar. Pero no sólo con capacitar se logrará que la tecnología sea adoptada por los agricultores, es necesario diseñar una estrategia de desarrollo agrícola que contemple la asesoría técnica, el crédito y la provisión de insumos. Otro producto que se puede brindar en las fincas cafetaleras, es mostrar cómo se procesa el café para el consumo. Al respecto, las personas que desean participar en la prestación de servicios turísticos, el 38,5% mencionó que lo vende en la planta, el 36,7% en cereza y el 21% en la planta y en cereza. Los que no Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 203 23/9/10 20:16:39 204 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega quieren participar, el 52% lo vende en la planta, el 38% en cereza y el 9,5% en ambos. Las principales causas por las que no proporcionan valor agregado, mencionan que no tienen equipo (38%) y porque no tienen dinero para comprar éste (36%). En el otro grupo mencionan que no tienen el equipo (55,6%) y que no disponen de capital (29,6%). El grupo que desea participar, el 45,6% comentó que le agradaría proporcionar valor agregado a su producción de café y el 35,7% de los no interesados realizó comentarios similares. La gran mayoría comenta que le agradaría vender su producto proporcionándole valor agregado, como es la venta del producto en forma de café pergamino, y muy pocos, café molido. Como podrá observarse, es importante que se impulse entre los cafetaleros el proporcionar valor agregado a su producción, específicamente en el beneficiado que consiste en: despulpar y lavar el café; secado en zaranda; trillado (en mortero de madera); selección de grano; tostado y molido, y empacado. Estas prácticas propias de la agroindustria son demasiado interesantes y didácticas; en ellas se busca que los turistas se involucren realizando cada una de ellas y, a su vez, conozcan como se produce el café que degustan. Con ello se busca relacionar las plantaciones de café con el turismo para proporcionar valor añadido a estos espacios. Para su impulso se requiere apoyo económico a los agricultores para la compra de equipo, desde el beneficiado hasta el empaque del café; esta actividad agroturística traerá beneficios económicos directos a los cafeticultores, al mejorar su ingreso por la venta de café y por la atención de turistas en su hogar. Un ejemplo es el caso colombiano, que ante el deterioro de sus condiciones de vida, los habitantes del eje cafetero buscaron alternativas de diversificación del ingreso. El caficultor, ante la necesidad de transformar sus explotaciones agropecuarias, se reconvirtió y recurrió a estrategias adaptativas diversas que implicaron una nueva organización productiva y espacial. Es así como los caficultores decidieron que sus plantaciones de café no sólo servían para exportar «the richest coffee in the world», sino también, para atraer turistas y generar ingresos complementarios (Ramírez, 2002: 2). Es un ejemplo que se puede aprovechar con los pequeños caficultores de la localidad. También es importante que se aprovechen los determinantes exógenos sobre la prestación de servicios que están demandando –casas de fin de semana, restaurantes tradicionales, espacios de paseo, pesca y diversión- los habitantes de las grandes ciudades como la de México y Puebla. El turismo rural, y en especial el agroturismo, es un instrumento que puede impulsar la economía de los caficultores y con ello mejorar sus condiciones de vida al incorporar a los turistas en las prácticas agrícolas y agroindustialización que realizan los caficultores. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 204 23/9/10 20:16:39 Turismo rural y desarrollo territorial en espacios indígenas de México 205 6. Conclusiones Se encuentra que existe un porcentaje importante de caficultores que están dispuestos a incursionar en actividades turísticas; para ello es necesario diseñar programas de financiamiento para participar en el servicio de hospedaje y en la infraestructura complementaria para los turistas. Además, se tienen que promover proyectos que contemplen su capacitación para incursionar en el turismo, con temas de administración de casa rurales, contabilidad, recursos humanos, entre otros no menos importantes. Es necesario desarrollar políticas de turismo rural, para aprovechar que en los municipios de estudio los caficultores participen en este tipo actividades para complementar el ingreso que obtienen de su producción. Es importante combinar la actividad agrícola con el agroturismo, ya que en conjunto, coadyuvarán a mejorar la calidad de vida de los habitantes de la localidad. Para ello se deben emprender proyectos agroturísticos que apoyen con recursos económicos y capacitación a los caficultores. En primer lugar, en la producción de café convencional y orgánico; en éste tienen que participar la Secretaría de Agricultura, Ganadería, Pesca y Alimentación (SAGARPA) y el Consejo Poblano del Café mediante la capacitación de los caficultores en el proceso productivo y a enseñarles a proporcionarle valor agregado. En segundo lugar, se tiene que diseñar una estrategia de financiamiento para la compra de insumos para aplicar lo aprendido en la capacitación; también se debe financiar la compra de maquinaria y equipo para dar valor agregado al café. Se puede decir que se tienen que reconvertir las unidades de producción y casas rurales para uso turístico para lograr que se incrementen los ingresos de los caficultores. Si Cuetzalan es un ejemplo exitoso, es necesario extender la actividad turística a los municipios circunvecinos de manera ordenada y planificada para no degradar los recursos naturales, que de una u otra manera, en el mediano plazo, se realizarán esfuerzos aislados y desordenados por explotar la activad turística. Es necesario planificar de manera integral el espacio turístico rural, donde los habitantes de las comunidades participen activamente en su propio desarrollo. Pero también es importante promover entre los turistas que visitan la ciudad de Puebla el turismo de interior y ofrecerles a Cuetzalan –Pueblo Mágico–, debido a su experiencia en este ramo, y hacer de ella una región turística polinuclear. Para ello se propone que se desarrolle una ruta turística –del café–, que tenga como polo de desarrollo a Cuetzalan y que en su hinterland los caficultores ofrezcan en sus casas productos agroturísticos, basados fundamentalmente en el atractivo de incorporar a los turistas a realizar trabajos en el proceso productivo del café y en su participación en proporcionarle valor agregado. Se esperaría una fuerte aceptación de esta ruta. Con este tipo de productos se diversifica la oferta turística y se contribuye a romper la estacionalidad de la actividad turística. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 205 23/9/10 20:16:39 206 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega 8. Bibliografía Amaya, C. M. (2005): «Desafíos y oportunidades del turismo rural en México». En Desarrollo rural y turismo, Editores: Dachari, A. C., Orozco, J. y Arnaiz, S. México, Universidad de Guadalajara, Centro Universitario de la Costa y Universidad de Buenos Aires, Facultad de Agronomía, pp. 51–57. André Sarasa, J. L. (2000): «Aportaciones básicas del turismo al desarrollo rural», en Cuadernos de Turismo, n° 6, pp. 45-59. Ascanio Guevara, A. (2004): «Turismo y desarrollo de la comunidad, un primer paso para rescatar la identidad cultural», en Pasos. Revista de Turismo y Patrimonio Cultural, vol. 2 n° 2, pp. 155–161. Berdegué, J. A.; Reardon, T. y Escobar, G. (2002): La creciente importancia del empleo y el ingreso rurales no agrícolas. En Echeverrí, R. G. (editor) Desarrollo de las economías rurales en América Latina y el Caribe. Banco Interamericano de Desarrollo, 275 pp. Boletín de Prensa, octubre (2006): Nuevo León, México. Bretón Ávalos, J. J. (2007): Avances y resultados 2006. Secretaría de Turismo, México, 12 pp. Cánoves Valiente, G.; Herrera Jiménez, L. y Villarino Pérez, M. (2005): «Turismo rural en España. Paisajes y usuarios, nuevos usos y nueves visiones», en Cuadernos de Turismo, nº 15, pp. 63–76. Damiani, O. (2002): Pequeños productores rurales y agricultura orgánica: lecciones aprendidas en América Latina. FIDA, 75 pp. Echeverría, R. G. y Bello, N. (2002): «Opciones para fomentar el acceso a tierras». En Banco Interamericano de Desarrollo (comp.) El acceso a la tierra en la agenda de desarrollo rural. Serie de Informes Técnicos del Departamento de Desarrollo Sostenible, BID, pp. 1-23. Echeverri, R. y Ribero, M. P. (2002): Nueva ruralidad. Visión del territorio en América Latina y el caribe. Instituto Interamericano de Cooperación para la Agricultura, 207 pp. BIRF y BM (2005): Generación de ingreso y protección social para los pobres. Banco Internacional de Reconstrucción y Fomento / Banco Mundial, 218 pp. Font, X. (2006): «Turismo Sostenible: factor fundamental para la erradicación de la pobreza en Colombia». Informe del IV Seminario Internacional Organización Mundial del Turismo, 52 pp. Gobierno del estado de Puebla (2005): Plan estatal de desarrollo 2005-2011. Gobierno del estado de Puebla, 272 pp. Gómez, R. (1979): «Introducción al muestreo». Chapingo, México, Tesis de Maestría en Ciencias en Estadística. Centro Estadística y Cálculo. Colegio Postgraduados. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 206 23/9/10 20:16:39 Turismo rural y desarrollo territorial en espacios indígenas de México 207 Gurría Di-Bella, M. (2000): «El turismo rural sostenible como una oportunidad de desarrollo de las pequeñas comunidades de los países en desarrollo», en Cuarta Feria Ecoturística y de Producción, 15 – 23 Julio, Buena Noche de Hato Nuevo, Manoguayabo, Santo Domingo, República Dominicana, disponible en web: http://www.kiskeya-alternative.org/publica/diversos/rural-tur-gurria. html, pp. 10. Kay, C. (1998): «Estructuralismo y teoría de la dependencia en el periodo neoliberal. Una perspectiva latinoamericana», en Nueva Sociedad, n° 158, pp. 100-119. Linck, T. s/f, «La economía y la política en la apropiación de los territorios». INRA – Systèmes de Décision pour le Développement, disponible en web: www.gis-syal.agropolis.fr/PDF/TLinck.pdf IICA y BID (2003): Experiencia de atención a la pobreza y municipalización en el desarrollo rural sostenible. IICA y BID, 222 pp. INEE (2006): Panorama educativo de México 2005. Indicadores del Sistema Educativo Nacional. Instituto Nacional para la Evaluación de la Educación, 549 pp. Juárez Sánchez, J. P. y Ramírez Valverde, B. (2007): «El turismo rural como complemento al desarrollo rural en zonas indígenas de México», en Scripta Nova, abril. Lustig, N. y Székely, M. (1997): México: evolución económica, pobreza y desigualdad. Washington, D.C., PENUD, BID y la CEPAL, 47 pp. OMT (2003): Turismo y atenuación de la pobreza. Organización Mundial del Turismo, 119 pp. Pérez, E. (2005): «Hacia una nueva visión de lo rural», en GIARRACCA, N. (comp.) ¿Una nueva ruralidad en América Latina? CLACSO, 384 pp. Ramírez Vallejo, J. (2002): «Agroturismo como alternativa de diversificación del ingreso en el eje cafetalero colombiano», en Harvard Review of Latin America, Tourism in the Américas. Development, Cultura and Identity, págs. 5. Ramos, J. (1993): «Crecimiento, crisis y viraje estratégico», en Revista de la CEPAL, nº. 50, agosto, pp. 34–49. Romero, J. L. (2006): Modelos productivos regionales durante la crisis de bajos precios del café, en el estado de Puebla. Consejo Poblano del Café, Gobierno el estado de Puebla, 44 pp. Sassano, S. (2001): «Transformaciones de un espacio urbano: El caso del mercado de abasto de Buenos Aires», en Anales de Geografía de la Universidad Complutense, n° 21, pp. 99-118. Sayadi, S. y Calatraba, J. (2001): «El agroturismo y desarrollo rural: situación actual y potencial y estrategias en zonas de montaña del Sureste español» en Cuadernos de turismo, n° 7, pp. 131 – 157. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 207 23/9/10 20:16:40 208 José Pedro Juárez Sánchez, Benito Ramírez Valverde y M.ª Guadalupe Galindo Vega SECTUR (2001): Resumen Ejecutivo. Estudio estratégico de viabilidad del segmento ecoturístico en México. SECTUR e ITAM, 14 pp. SECTUR (2006): El turismo 2001 – 2006. Secretaría de Turismo del Estado de Puebla. Gobierno del estado de Puebla, 33 pp. Schejtman, A. y Berdegué, J. (2004): Desarrollo territorial rural. Chile, Centro Latinoamericano para el Desarrollo Rural, Debates y Temas Rurales, nº 1, 53 pp. Silva Pérez, R. (2002): «Estrategias de inserción de las áreas rurales en la economía mundial. Una aproximación desde Andalucía», en Boletín de la Asociación de Geógrafos Españoles, nº 33, pp. 103-131. Vera, F.; López Palomeque, F.; Marchena, M. y Antón, S. (1997): Análisis territorial del turismo. España, Ariel, 443 pp. Investigaciones geográficas, nº 48, pp. 189 - 208 N.º 48 B.indd 208 23/9/10 20:16:40
https://openalex.org/W4362429133	https://figshare.com/articles/journal_contribution/Supplementary_Figure_3_from_Biological_and_Clinical_Relevance_of_miRNA_Expression_Signatures_in_Primary_Plasma_Cell_Leukemia/22445967/1/files/39897015.pdf	English	null	Supplementary Figure 5 from Biological and Clinical Relevance of miRNA Expression Signatures in Primary Plasma Cell Leukemia	null	2,023	cc-by	208	Supplementary Figure 3. MiRNA/target anticorrelation in MM. Significant negative correlation of ACVR1 mRNA with (A) miR-301a and (B) miR-301b levels. The spots represent mature miRNA expression (scale on left-sided vertical axis), and the grey bars represent target gene expression (scale on right-sided vertical axis) in the dataset of 36 primary myeloma and 18 pPCL patients. Horizontal axis: patient samples ordered according to increasing miRNA expression. (C) Luciferase assay showing decreased luciferase activity in cells transfected with pre-miR-301a and pre-miR-301b compared to cells transfected with pre-miR negative control. All experiments were performed in triplicate. Results were plotted as relative luciferase activity mean + standard deviation. Supplementary Figure 3. MiRNA/target anticorrelation in MM. Significant negative correlation of ACVR1 mRNA with (A) miR-301a and (B) miR-301b levels. The spots represent mature miRNA expression (scale on left-sided vertical axis), and the grey bars represent target gene expression (scale on right-sided vertical axis) in the dataset of 36 primary myeloma and 18 pPCL patients. Horizontal axis: patient samples ordered according to increasing miRNA expression. (C) Luciferase assay showing decreased luciferase activity in cells transfected with pre-miR-301a and pre-miR-301b compared to cells transfected with pre-miR negative control. All experiments were performed in triplicate. Results were plotted as relative luciferase activity mean + standard deviation.
https://openalex.org/W1509308160	https://revistas.esan.edu.pe/index.php/jefas/article/download/332/209	Spanish; Castilian	null	CIA. minera Milpo S.A.A.: valorización de una empresa minera	Cuadernos de difusión	2,007	cc-by	15,705	CÍA. MINERA MILPO S. A. A. VALORIZACIÓN DE UNA EMPRESA MINERA* Alfredo Mendiola Universidad ESAN amendio@esan.edu.pe Resumen Bailey Investments Inc. (BII) estaba reestructurando su portafolio de inversión y examinando la posi- bilidad de ampliar su exposición hacia el sector minero peruano. Especíﬁ camente, estaba interesada en invertir en la Cía. Minera Milpo, de la cual tenía muy buenas referencias. Por tal motivo, solicitó a MC Perú, de la cual es una de sus principales clientas extranjeras, que le elaborara un informe sobre la conveniencia de realizar esta inversión. Esperaba una comparación entre el precio de mercado y el valor de la acción común de Milpo. El analista Ernesto Valdez, a quien se le encargó este trabajo, sabía que disponía de un plazo muy breve (tres días), de modo que no habría tiempo para conversar con funcionarios de Milpo, y su análisis tendría que centrarse en la información pública disponible en Internet. Palabras clave: valor de la empresa, evaluación de inversiones, mercado de capitales, minería, precios de metales. Abstract Bailey Investments Inc. (BII) was restructuring their investment portfolio and examining the possibility of extending their exposure towards the Peruvian mining industry. Speciﬁ cally, they were interested in investing in the Milpo Mining Company, about which they had very good references. For this reason, they asked MC Peru, who is one of their primary foreign customers, to create a report about the convenience for realizing this investment. They expected a comparison between the market price and the value of ordinary Milpo stock. Analyst Ernesto Valdez, who was in charge of this work, knew that he would only have a short time (three days) to do so, meaning that there would not be any time to converse with Milpo representatives, and his analysis would have to focus on public information available through the Internet. Key words: value of the company, investment evaluation, market capital, mining, prices of metals. * El autor desea agradecer a Lourdes Valdez y Liz Villanes por sus comentarios a la versión inicial de este caso, y a Pedro P. Campaña por la recolección de la información. Cuad. Difus. 12 (22), jun. 2007 1. MC Perú es una empresa consultora peruana que se especializa en temas ﬁ nancieros. Tiene gran experiencia y goza de prestigio en el medio. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 148 1. La solicitud los precios de las acciones mineras, que a lo largo del año 2006 alcanzaron, en algu- nos casos, más de 300%. Esta situación se reﬂ ejó en el aumento de la capitalización bursátil de las acciones comunes de las empresas mineras, que pasaron de 16 715 a 32 039 millones de dólares entre diciembre de 2005 y diciembre de 2006, incremento equivalente a algo más de 90% en un año. Era un caluroso día de marzo de 2007 y Ernesto Valdez, analista senior de MC Perú1, tomaba muy preocupado una taza de café. Hacía algunos minutos el gerente de Estudios Económicos le había enco- mendado un trabajo urgente y conﬁ den- cial solicitado por una de las principales clientas extranjeras de MC Perú: Bailey Investments Inc. (BII). Paralelamente, en el caso de las accio- nes de inversión2, la capitalización bursátil pasó de 1 177 a 2 491 millones de dólares en el mismo periodo3, lo que signiﬁ có un crecimiento de más de 110%. BII se encontraba en proceso de rees- tructurar su portafolio de inversión y estaba examinando la posibilidad de ampliar su exposición hacia el sector minero peruano. Tenía por política efectuar inversiones con un horizonte de mediano plazo (dos a tres años), y en años anteriores habían tenido problemas en el Este asiático al focalizar sus esfuerzos en inversiones especulativas de corto plazo. Su política era muy clara: buscaba invertir en acciones de empresas en crecimiento, que contaran con un ﬂ ujo de caja operativo sostenido y costos con- trolados, y estuvieran localizadas en países emergentes estables. Durante el 2006, la acción común de Milpo obtuvo un índice de lucratividad de 317,86%4 y su precio pasó de 3,5 a 10,8 soles entre enero y diciembre. En este mismo periodo, el precio de la acción de inversión pasó de 1,98 a 7,40 soles (ver cuadro 1 y gráﬁ co 1). Es necesario men- cionar que el índice de lucratividad incluye el retorno por las ganancias de capital y dividendos. De manera más especíﬁ ca, este cliente tenía muy buenas referencias de Cía. Mi- nera Milpo y le había solicitado a MC Perú que le elaborara un informe en el que se examinara la conveniencia de invertir en esta empresa. BII había sido muy preciso en su pedido: esperaba una comparación entre el precio de mercado y el valor de la acción común de Milpo. 2. 2. Las acciones comunes corresponden a la parte alí- cuota del capital social de una sociedad anónima que, incorporada a un título representativo, otorga a su propietario la calidad de socio y es transmi- sible o negociable. Las acciones de inversión son valores mobiliarios emitidos por empresas indus- triales, mineras y pesqueras que conﬁ eren a sus titulares el derecho a participar en los dividendos por distribuir, de acuerdo con su valor nominal. Constituyen la cuenta Participación Patrimonial del Trabajo. Su antecedente fueron las acciones laborales. 1. La solicitud Las acciones comunes corresponden a la parte alí- cuota del capital social de una sociedad anónima que, incorporada a un título representativo, otorga a su propietario la calidad de socio y es transmi- sible o negociable. Las acciones de inversión son valores mobiliarios emitidos por empresas indus- triales, mineras y pesqueras que conﬁ eren a sus titulares el derecho a participar en los dividendos por distribuir, de acuerdo con su valor nominal. Constituyen la cuenta Participación Patrimonial del Trabajo. Su antecedente fueron las acciones laborales. El interés de BII en ampliar su nivel de inversiones en la Bolsa de Valores de Lima se relacionaba con el incremento de 3. Fuente: < http://www.bvl.com.pe/pubdif/infmen/ 200612a4.htm>. 4. Fuente: < http://www.bvl.com.pe/pubdif/infmen/ 200612b3.htm>. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 149 Fuente: Bolsa de Valores de Lima. Cuadro 1. Cotizaciones de las acciones de Minera Milpo, 2006 Empresa Tipo de Precio a Precio a ﬁ nes Volumen acción inicios de enero de diciembre negociado en el año (soles) (soles) (millones de soles) Milpo Común 3,50 10,80 356,7 Milpo Inversión 1,98 7,40 12,5 Cuadro 1. Cotizaciones de las acciones de Minera Milpo, 2006 Gráﬁ co 1. Cotización de las acciones de Minera Milpo 10 8 6 4 2 0 Precios (soles) Dic. Feb. Mar. May. Jul. Ago. Oct. Nov. Ene. 2005 2006 2006 2006 2006 2006 2006 2006 2007 Meses Acciones de inversión 14 12 10 8 6 4 2 0 Precios (soles) Dic. Mar. Jul. Oct. Ene. 2005 2006 2006 2006 2007 Meses Acciones comunes Precios (soles) Gráﬁ co 1. Cotización de las acciones de Minera Milpo Valdez sabía que la empresa se encontraba en una excelente situación económico- ﬁ nanciera, pero diversos temas le daban vueltas en la cabeza: ¿cuál sería la evolu- ción de las cotizaciones de los minerales a partir del año 2007?, ¿qué tendencias eco- nómicas, políticas y sociales se perﬁ larían en el país luego de que el APRA asumiera el BII deseaba que el informe se emitiera en un plazo máximo de tres días útiles, lo cual limitaba la búsqueda de información que podía hacer Ernesto Valdez; no habría tiempo para conversar con funcionarios de Milpo, y su análisis tendría que centrarse en la información pública disponible en Internet. Al margen de este aspecto, el señor Cuad. Difus. 12 (22), jun. 5. Notas a los estados ﬁ nancieros por los años ter- minados el 31 de diciembre de 2005 y de 2004. <www.conasev.gob.pe>. 1. La solicitud 2007 Alfredo Mendiola Alfredo Mendiola 150 poder, por segunda vez, en julio de 2006?, ¿darían resultado las exploraciones que Milpo estaba llevando a cabo?, ¿cuál era el horizonte de actividad de las minas que ahora se encontraban en operación?, ¿serían exitosos los proyectos de exploración?, ¿cuáles eran los costos de operación que tendría que considerar?, ¿cuál sería la tasa de descuento apropiada en el caso de valuar un ﬂ ujo de caja descontado? Al cierre del año 2006, el grupo operaba tres unidades mineras: Mina El Porvenir, de concentrados de zinc, plomo y cobre, Mina y Reﬁ nería Iván, de cátodos de cobre, y Mina Chapi, de sulfato de cobre. Además, su mina Cerro Lindo entraría en operación a partir de junio de 2007 y produciría con- centrados de zinc, plomo y cobre. En el sector minero [peruano] Milpo ocupó en el año 2006 la cuarta posición como pro- ductor de zinc y como productor de plomo, participando con el 7% de la producción nacional de dichos metales […]. Con el ini- cio de las actividades de nuestra subsidiaria Minera Pampa de Cobre S. A., titular de la Unidad Minera Chapi, hemos ingresado al ranking nacional de productores de cobre (Cía. Minera Milpo, Memoria 2006). Todas estas preguntas le diﬁ cultarían llegar a la conclusión de su informe: ¿cuál era el valor de la acción común de Milpo?, ¿estaba sobrevaluado el precio de la acción común? 6. Sigla en inglés: Earnings Before Interests, Tax, Depreciation and Amortization. 3. Las unidades mineras El proceso productivo se inicia con la extracción del mineral, medido en toneladas métricas secas (TMS). El término TMS indi- ca que el peso del mineral extraído es neto, libre de cualquier contenido de agua. 2. La empresa La Cía. Minera Milpo fue constituida en 1949 y es: Durante el 2006, los resultados econó- micos de la empresa fueron interesantes. Tal como se indica en la memoria de la empresa, la utilidad neta fue de 96,3 millones de dólares (400% más que el año anterior); y las EBITDA6, de 190,3 millones de dólares (200% más que el año anterior). Este crecimiento de los resultados fue con- secuencia del incremento de las cotizacio- nes internacionales de los metales. ... titular de la unidad minera El Porvenir, ubicada en el departamento de Pasco, y a través de la participación en la Subsidiaria Milpo Finance and Investments Inc. posee el 100% de la participación en Minera Rayrock Ltda., una entidad dedicada a la actividad minera que desarrolla actividades en Antofagasta, Chile5. El grupo económico tiene varios pro- yectos de exploración en diversas fases de avance. Entre ellos, Hilarión, que está en la segunda fase de perforación, luego de hallarse evidencias de zinc, plomo, plata y oro, y Pukaqaqa, que está en tercera fase de perforación, en busca de cobre y oro. Este grupo económico opera diferentes minas de tamaño mediano y está en proceso de ampliación de operaciones sobre la base de nuevos proyectos mineros. Sus minas tienden a ser polimetálicas, pues incluyen metales como zinc, cobre, plomo, plata y oro. Antes del proceso de explotación, es necesario desarrollar actividades de exploración que permitirán establecer con Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 151 precisión las reservas y los recursos7 de minerales económicamente explotables. El nivel de reservas determina las expectativas de «vida» de la operación minera. Se debe señalar que el proceso de exploración es continuo y que una empresa minera siempre debe buscar «reponer» como reservas cada tonelada de mineral extraída. concentrado, que contiene una alta pro- porción de metales y es espesado y ﬁ ltrado para ser despachado. El otro producto es el relave, que por ser contaminante debe ser tratado apropiadamente; puede ser alma- cenado en canchas de relave o rebombeado a la mina para mejorar el sostenimiento. 3.1. Mina El Porvenir La Mina El Porvenir es la unidad minera emblemática de la empresa. Localizada en Cerro de Pasco, a más de 4 100 m. s. n. m., opera desde hace más de treinta años. Pro- duce concentrados de zinc, plomo y cobre, con contenidos de oro y plata, y mantiene operaciones subterráneas. Luego, el mineral extraído es chancado en dos etapas: chancado primario y chanca- do secundario, después de lo cual es molido hasta el nivel de gralunometría deseado. A continuación, el mineral molido es llevado a las celdas de ﬂ otación, donde el contenido metálico se extrae con la ayuda de reactivos químicos. En el año 2006, la planta concentradora trató 1,39 millones de toneladas métricas secas de mineral, semejante al nivel de producción alcanzado el año anterior (ver cuadro 2). El proceso de ﬂ otación deja dos pro- ductos. El más importante de ellos es el Cuadro 2. Mina El Porvenir: Extracción y beneﬁ cio Fuente: Cía. Minera Milpo, Memoria 2006. Cuadro 2. Mina El Porvenir: Extracción y beneﬁ cio Cuadro 2. Mina El Porvenir: Extracción y beneﬁ cio Año TMS Zn Pb Cu Ag Oz/Tc Costo de producción (dólares por TMS) 2001 1 067 744 7,7% 2,3% 0,3% 3,40 22,24 2002 1 217 291 8,0% 2,6% 0,3% 3,70 18,59 2003 1 313 346 7,7% 2,5% 0,2% 3,50 18,13 2004 1 342 451 7,6% 1,4% 0,2% 2,20 19,93 2005 1 395 991 6,9% 1,6% 0,3% 2,50 19,61 2006 1 390 940 6,1% 1,7% 0,3% 2,50 20,46 Fuente: Cía. Minera Milpo, Memoria 2006. partir de información menos completa que en el caso anterior. Los recursos corresponden a aquellas reservas en las que se tiene evidencia de la concentración del mineral. Cuando por rasgos geológicos del yacimiento se inﬁ ere la existencia de depósitos mineralizados se dice que los recur- sos son inferidos. 7. Las reservas de minerales probadas son aquellas de las cuales se tiene certeza de continuidad, incluyendo ley y volumen. Son estimadas a partir de los resultados obtenidos en los trabajos de muestreo, labores y sondajes. Las reservas de minerales probables son aquellas en las que existe riesgo de discontinuidad en cuanto a geometría, volumen y ley, pues han sido determinadas a Cuad. Difus. 12 (22), jun. 2007 152 Alfredo Mendiola Alfredo Mendiola El mineral de cabeza es procesado y convertido en concentrados de zinc (Zn), plomo (Pb) y cobre (Cu) según los volúme- nes indicados en los cuadros 3, 4 y 5. El mineral de cabeza es procesado y convertido en concentrados de zinc (Zn), plomo (Pb) y cobre (Cu) según los volúme- nes indicados en los cuadros 3, 4 y 5. Las cifras de extracción del cuadro 2 corresponden a toneladas de mineral de cabeza8 secas extraídas y procesadas en la planta concentradora. El costo de produc- ción que se presenta en este mismo cuadro se reﬁ ere al costo promedio de extracción y procesamiento del mineral de cabeza. Cuadro 3. Mina El Porvenir concentrado de zinc Año TMS Ley Zn Ley Ag Oz/Tc 2000 127 798 54,7% 3,20 2001 137 261 54,9% 2,60 2002 163 602 54,7% 2,60 2003 168 348 55,3% 2,70 2004 173 560 54,5% 1,90 2005 161 980 54,3% 2,00 2006 143 982 54,3% 2,20 Cía. Minera Milpo, Memoria 2006. Cuadro 3. Mina El Porvenir concentrado de zinc Cuadro 4. 8. Corresponde al material extraído de la mina y transportado a la planta concentradora para su procesamiento. Este mineral tiene una ley Cía. Minera Milpo S. A. A.: valoración de una empresa minera Cía. Minera Milpo S. A. A.: valoración de una empresa minera 153 En cuanto a las reservas y los recursos de la unidad, estos ascienden a más de 20 millones de toneladas métricas secas según el detalle del cuadro 6. con reservas probadas y probables que le permitirán operar los siguientes siete años. Este horizonte no toma en cuenta posibles descubrimientos por labores de explora- ción. En octubre del 2006, Golder Associa- tes S. A. realizó una auditoría de reservas y en su informe señala que las reservas y los recursos (medidos e indicados) de Milpo alcanzan 10,3 millones de toneladas mé- tricas secas de mineral de cabeza. A lo largo del año 2006, las inversio- nes alcanzaron un total de 5,4 millones de dólares: Las principales fueron: el recrecimiento de la presa de relaves, la adquisición de un sistema de monitoreo sísmico, la adquisición de bombas centrífugas, la compra de celdas de ﬂ otación, la adquisición de un analizador de leyes en línea, la compra de diferentes equipos como jumbos, scissor bolter, scaler (lo que ha permitido la completa mecaniza- ción de las actividades del ciclo de minado como son la perforación, el desate y el sos- tenimiento), entre otras (Cía. Minera Milpo, Memoria 2006: 24). Cuadro 2. Mina El Porvenir: Extracción y beneﬁ cio Mina El Porvenir: concentrado de plomo Año TMS Au Oz/Tc Pb Ag Oz/Tc 2000 28 429 0,37 68,5% 87,10 2001 31 737 0,24 68,8% 79,00 2002 40 444 0,19 67,6% 73,60 2003 41 992 0,18 67,1% 72,50 2004 24 293 0,16 65,7% 73,30 2005 30 093 0,13 66,5% 72,70 2006 31 115 0,14 65,7% 72,20 Cía. Minera Milpo, Memoria 2006. Cuadro 4. Mina El Porvenir: concentrado de plomo Cuadro 5. Mina El Porvenir: concentrado de cobre Cuadro 5. Mina El Porvenir: concentrado de cobre Año TMS Cu Au Oz/Tc Ag Oz/Tc 2000 n. d. n. d. n. d. n. d. 2001 1 343 26,1% 0,86 69,0 2002 3 953 26,0% 1,07 70,9 2003 3 869 25,5% 1,02 76,0 2004 3 911 25,9% 0,42 50,3 2005 4 768 26,2% 0,40 40,7 2006 4 491 25,4% 0,44 40,1 Cía. Minera Milpo, Memoria 2006. Cuad. Difus. 12 (22), jun. 2007 8. Corresponde al material extraído de la mina y transportado a la planta concentradora para su procesamiento. Este mineral tiene una ley (contenido) de metal que es recuperado en la planta concentradora a través de diferentes pro- cesos químicos. (contenido) de metal que es recuperado en la planta concentradora a través de diferentes pro- cesos químicos. Cuad. Difus. 12 (22), jun. 2007 Cuad. Difus. 12 (22), jun. 2007 Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo, Memoria 2006. 3.2. Mina Iván (Chile) Esta unidad minera se encuentra ubica- da en Antofagasta, Chile. El 100% de su capital es de propiedad de Minera Rayrock Ltda., empresa subsidiaria del grupo Milpo. La mina Iván produce cáto- dos de cobre de alta pureza en su propia reﬁ nería, cuya capacidad es de 40 tonela- das diarias. El material procesado en la reﬁ nería corresponde tanto a material ex- traído en operaciones propias como a ma- terial comprado a terceros. De acuerdo con diferentes estimacio- nes, la unidad minera El Porvenir cuenta Cuadro 6. Mina El Porvenir: reservas y recursos Cuadro 6. Mina El Porvenir: reservas y recursos Cuadro 6. Mina El Porvenir: reservas y recursos RESERVAS Y RECURSOS AL AÑO 2006 Categoría TM Zn Pb Cu Ag Oz/Tc Reservas 6 672 063 7,78% 1,57% 0,32% 2,74 Recursos 6 189 445 6,36% 1,32% 0,35% 2,34 Total 12 861 508 7,10% 1,45% 0,33% 2,55 Inferidos 13 109 281 5,61% 0,27% 0,57% 0,97 RESERVAS DE MINERAL Año TMS Zn Pb Cu Ag Oz/Tc 2000 6 457 382 6,90% 2,10% 3,60 2001 6 243 018 7,00% 2,10% 3,60 2002 6 655 957 7,40% 2,00% 3,70 2003 8 218 014 9,00% 1,30% 0,29% 2,60 2004 7 577 527 8,60% 1,50% 0,30% 2,90 2005 7 569 598 8,30% 1,60% 0,31% 2,70 2006 6 672 063 7,78% 1,57% 0,32% 2,74 Cía. Minera Milpo, Memoria 2006. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 154 Cuadro 7. Mina Iván (Chile): producción Año Producción de Mineral procesado cátodos (toneladas) (toneladas) 2000 12 822 – 2001 13 858 – 2002 9 295 1 162 063 2003 8 568 1 073 853 2004 10 052 1 312 981 2005 9 166 1 100 855 2006 10 538 1 020 000 Cía. Minera Milpo, Memoria 2006. Cuadro 7. Mina Iván (Chile): producción De acuerdo con Class & Asociados, las reservas y los recursos de mina Iván le permitirán un horizonte de operaciones de diez años, sin considerar los recursos inferidos ni la incorporación de nuevas reservas como consecuencia de actividades de exploración. 3.3. Mina Chapi La mina Chapi es propiedad de la empresa Minera Pampa de Cobre S. A., una subsi- diaria (al 100%) de Milpo. Las operacio- nes se localizan en el departamento de Moquegua y la producción se inició en enero de 2006. De acuerdo con la infor- Las reservas y los recursos de esta uni- dad ascienden a más de 7 millones de tone- ladas métricas secas de mineral de cabeza, según se muestra en el cuadro 8. Cuadro 8. Mina Iván (Chile): reservas y recursos Cuadro 8. Mina Iván (Chile): reservas y recursos RESERVAS Óxidos Sulfuros Año Toneladas CuT CuS Toneladas CuT CuS 2001 4 748 1,15% – 10 098 1,68% – 2002 1 201 500 1,43% – 5 801 000 1,75% – 2003 860 135 1,37% 1,04% 943 623 2,72% 0,33% 2004 802 863 1,26% 0,96% 498 506 2,09% 0,16% 2005 1 224 942 0,53% 0,35% 471 742 2,09% 0,14% 2006 1 082 857 0,64% 0,44% 471 742 2,09% 0,14% RECURSOS (MEDIDOS + INDICADOS) Óxidos Sulfuros Año TM CuT CuS TM CuT CuS 2001 7 244 1,52% 6 446 1,91% 2002* 2 492 000 1,29% 8 790 000 1,70% 2003 908 563 1,56% 1,23% 3 490 217 2,26% 0,16% 2004 684 441 1,48% 1,18% 3 613 965 2,30% 0,15% 2005 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% 2006 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% * Los recursos incluyen las reservas. RECURSOS INFERIDOS Óxidos Sulfuros Año TM CuT CuS TM CuT CuS 2005 4 124 915 0,63% 0,39% 1 818 458 1,84% 0,16 Cuadro 8. Mina Iván (Chile): reservas y recursos RESERVAS 2004 802 863 1,26% 0,96% 498 506 2,09% 0,16% 2005 1 224 942 0,53% 0,35% 471 742 2,09% 0,14% 2006 1 082 857 0,64% 0,44% 471 742 2,09% 0,14% RECURSOS (MEDIDOS + INDICADOS) Óxidos Sulfuros Año TM CuT CuS TM CuT CuS 2001 7 244 1,52% 6 446 1,91% 2002* 2 492 000 1,29% 8 790 000 1,70% 2003 908 563 1,56% 1,23% 3 490 217 2,26% 0,16% 2004 684 441 1,48% 1,18% 3 613 965 2,30% 0,15% 2005 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% 2006 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% * Los recursos incluyen las reservas. 4. Los proyectos mación disponible, la vida útil de la mina es de 14 años. Tal como se indica en la Memoria 2006 de Milpo, esta unidad: La Cía. Minera Milpo tiene un proyecto en fase de implementación: Cerro Lindo, y dos proyectos en fase de exploración: Hilarión y Pukaqaqa. ... produce sulfato de cobre en su propia planta de cristalización, la misma que tiene una capacidad instalada de 80 toneladas de cobre al día. Adicionalmente como parte de su proceso productivo, se incluye la explo- tación de sulfuros, en mina subterránea y de óxidos, en tajo abierto. Ambos minerales son chancados, aglomerados y transportados para ser dispuestos en pilas de lixiviación. Luego, el material lixiviado es alimentado a la Planta de Extracción por Solventes y de Cristalización (Cía. Minera Milpo, Memoria 2006). 3.3. Mina Chapi RECURSOS (MEDIDOS + INDICADOS) Óxidos Sulfuros Año TM CuT CuS TM CuT CuS 2001 7 244 1,52% 6 446 1,91% 2002* 2 492 000 1,29% 8 790 000 1,70% 2003 908 563 1,56% 1,23% 3 490 217 2,26% 0,16% 2004 684 441 1,48% 1,18% 3 613 965 2,30% 0,15% 2005 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% 2006 1 484 695 1,23% 0,96% 5 233 498 2,07% 0,21% * Los recursos incluyen las reservas. RECURSOS INFERIDOS Óxidos Sulfuros Año TM CuT CuS TM CuT CuS 2005 4 124 915 0,63% 0,39% 1 818 458 1,84% 0,16 2006 4 860 914 0,66% 0,44% 2 920 707 1,58% 0,16 Fuente: Cía. Minera Milpo, Memoria 2006. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 155 Cerro Lindo Este proyecto, 100% de propiedad de Milpo, se encuentra localizado al sur de la ciudad de Lima, en la provincia de Chincha, y entrará en operación a media- dos del año 2007. Tal como se indica en la memoria de la empresa, se estima una pro- ducción anual promedio de 146 mil tone- ladas métricas secas de concentrado de zinc, 14 800 toneladas métricas secas de concentrados de plomo y 39 500 toneladas métricas secas de concentrados de cobre. Con estas cantidades de concentrados, se espera que el grupo Milpo duplique su fac- turación anual. Las reservas y los recursos de la mina Chapi, a diciembre de 2006, se muestran en el cuadro 9. De acuerdo con la información disponible, con este nivel de reservas la uni- dad minera podrá operar durante 14 años. Tal como se indica en la memoria de Milpo, en las actividades de minado sub- terráneo se extrajeron 225 mil toneladas de sulfuros, mientras en el tajo abierto se extrajeron 650 mil toneladas de óxidos. La producción del año 2006 fue de 22 760 toneladas de sulfato de cobre, lo que equi- vale a alrededor de 5 mil toneladas de cobre ﬁ no. Esta unidad facturó 26,2 millones de dólares en el año 2006. El reporte de Class & Asociados indica que la inversión total en este proyecto se estimó en 90 millones de dólares, monto que ya ha sido invertido casi en su tota- lidad. La estructura del ﬁ nanciamiento del Pro- yecto Cerro Lindo se encuentra todavía en evaluación por parte de la compañía, ha- biendo solicitado a Conasev, dentro de las Cuad. Difus. 12 (22), jun. 2007 Cuadro 9. Mina Chapi: reservas Óxidos Sulfuros Categoría TM CuT TM CuT Reservas 199 189 0,84% 11 068 933 1,30% Recursos (medidos + indicados) 76 325 1,29% 3 158 078 1,40% Recursos inferidos 573 047 1,17% 26 076 673 1,24% Fuente: Cía. Minera Milpo, Memoria 2006. Cuadro 9. Mina Chapi: reservas Óxidos Sulfuros Categoría TM CuT TM CuT Reservas 199 189 0,84% 11 068 933 1,30% Recursos (medidos + indicados) 76 325 1,29% 3 158 078 1,40% Recursos inferidos 573 047 1,17% 26 076 673 1,24% Fuente: Cía. Minera Milpo, Memoria 2006. Cuadro 9. Mina Chapi: reservas Cuadro 9. Mina Chapi: reservas Cuad. Difus. 12 (22), jun. Proyecto Hilarión Los recursos de este proyecto se mues- tran en el cuadro 10. Este proyecto es también 100% de propie- dad de Milpo. Se encuentra ubicado a 50 kilómetros al sur del yacimiento Antami- na, en el departamento de Áncash. El total de recursos (medidos, indicados, inferi- dos) es de 12,2 millones de toneladas y contiene 5,25% de zinc, 1,44% de plomo y 2,11 onzas de plata. Como se indica en la Memoria 2006 de Milpo, el objetivo del actual programa de exploración es encon- trar entre 7 y 10 millones de toneladas de recursos medidos e indicados. Cerro Lindo 2007 Alfredo Mendiola 156 alternativas, la inscripción del Primer Pro- grama de Bonos Corporativos, que fuera aprobado por la Junta Obligatoria Anual de Accionistas en marzo 2005, hasta por US$70 millones. Se cuenta con una estructura ﬁ nan- ciera de mediano plazo con el Citibank N.A. por US$ 40 millones9. cuprífero. A ﬁ nes de 2004, Milpo ﬁ rmó un contrato de joint venture con la empresa canadiense Tiomin Resources Inc. para culminar la exploración avanzada de este proyecto. Milpo mantendrá el 51% de la participación y se encargará de las activi- dades de exploración y, eventualmente, de explotación del yacimiento. 5. Estados ﬁ nancieros De acuerdo con el reporte de Class & Aso- ciados, a lo largo del año 2006 el creci- miento sostenido del precio de los metales determinó que las ventas de Milpo aumen- taran más del doble. Esta situación ha per- mitido que Milpo: • Continúe ﬁ nanciando el proyecto Ce- rro Lindo con recursos propios. 9. Class & Asociados S. A. Fundamento de clasiﬁ ca- ción de riesgo de Compañía Minera Milpo S. A. A. Sesión de Comité n.° 02/2007 del 2 de febrero de 2007. 11. De acuerdo con esta especiﬁ cación: “Grade A cathode, LME spot price, CIF European ports, US$ per metric tone”. Proyecto Pukaqaqa • Haya prepagado en setiembre de 2006 el íntegro del saldo de la deuda por 30 millones de dólares que mantenía con Este proyecto, ubicado en el departamen- to de Huancavelica, es primordialmente Cuad. Difus. 12 (22), jun. 2007 Class & Asociados S. A. Fundamento de clasiﬁ ca- ción de riesgo de Compañía Minera Milpo S. A. A. Cuadro 10. Proyecto Pukaqaqa: reservas de mineral Cut-off Miles de Cobre Oro Cobre Oro Categoría (% Cu) toneladas % g/t (miles de libras) (miles de onzas) Medidos 0,3 191 0,58 0,08 2 441 0,48 0,4 170 0,61 0,08 2 287 0,44 Indicados 0,3 16 933 0,54 0,07 201 590 38,65 0,4 13 796 0,58 0,07 176 408 32,82 Med. + Ind. 0,3 17 124 0,54 0,07 204 031 39,14 0,4 13 966 0,58 0,07 178 696 33,27 Inferidos 0,3 114 654 0,60 0,10 1 516 601 353,88 0,4 93 559 0,66 0,10 1 361 319 312,83 Fuente: Cía. Minera Milpo, Memoria 2006. Sesión de Comité n.° 02/2007 del 2 de febrero de 2007. Cuadro 10. Proyecto Pukaqaqa: reservas de mineral Cut-off Miles de Cobre Oro Cobre Oro Categoría (% Cu) toneladas % g/t (miles de libras) (miles de onzas) Medidos 0,3 191 0,58 0,08 2 441 0,48 0,4 170 0,61 0,08 2 287 0,44 Indicados 0,3 16 933 0,54 0,07 201 590 38,65 0,4 13 796 0,58 0,07 176 408 32,82 Med. + Ind. 0,3 17 124 0,54 0,07 204 031 39,14 0,4 13 966 0,58 0,07 178 696 33,27 Inferidos 0,3 114 654 0,60 0,10 1 516 601 353,88 0,4 93 559 0,66 0,10 1 361 319 312,83 Fuente: Cía. Minera Milpo, Memoria 2006. Cuadro 10. Proyecto Pukaqaqa: reservas de mineral Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 157 enero de 2004. Por ejemplo, la cotización del cobre11 pasó de 2 421,48 a 6 680,97 dólares por tonelada entre enero de 2004 y diciembre de 2006, lo que equivale a un incremento de 135%. Es necesario resal- tar que, en julio de 2006, el precio de este metal llegó 7 726,74 dólares por tonelada. Durante el mismo periodo, la cotización del zinc aumentó en 331% y la del plomo en 122%. Citibank del Perú, mientras que su sub- sidiaria Minera Rayrock Ltda. Proyecto Pukaqaqa hizo lo propio con la obligación por 10 millo- nes de dólares mantenida con WestLB Ag, con lo que el nivel consolidado de deuda al cierre del tercer trimestre es de 2,5 millones de dólares. • Fortalezca su patrimonio, luego de ca- pitalizar los resultados correspondien- tes al ejercicio 2005 por 52,9 millones de soles. Según el Financial Times, al 30 de marzo de 2007 los precios de corto plazo de los metales en el mercado de Londres (London Metal Exchange) eran estables (ver cuadro 13). Por ejemplo, el precio por tonelada de cobre a tres meses mos- traba una ligera tendencia a la baja de 1%, mientras que el zinc presentaba un ligero incremento de 0,20%. Estos movimientos no son signiﬁ cativos. • Registre un nivel de caja superior al observado al cierre del año 2005, con las consecuentes mejoras sobre la com- pañía. En el anexo 1 se presentan los estados ﬁ nancieros consolidados de Milpo y sus subsidiarias para los años comprendidos entre 2003 y 2006. Estos estados ﬁ nan- cieros están en dólares estadounidenses. Es necesario hacer notar que los estados financieros correspondientes a los años 2003 y 2004 fueron convertidos a dólares a partir de los estados ﬁ nancieros en soles. El tipo de cambio empleado fue el promedio de compra y venta de cierre del año. Sin embargo, de acuerdo con el London Metal Exchange, en el mediano plazo, los precios de los metales no preciosos mues- tran una tendencia hacia la baja, aﬁ rmación basada en el hecho de que el mercado de futuros es un buen pronosticador de los precios spot de los metales. En el cuadro 14 se presentan los precios de los futuros de los metales a 15 y 27 meses para el cobre, el zinc y el plomo. Por ejemplo, en este cuadro se indica que el precio del cobre disminuirá 7% a 15 meses y 16% a 27 meses; de igual manera, el precio del plomo disminuirá 18% a 15 meses, mientras que el zinc disminuirá 6,5% a 15 meses y 15% a 27 meses; similar información se presenta en el cuadro 15. 10. Información tomada de <www.imf.org>. 12. ScotiaBank. Reporte Empresarial, Sociedad Mi- nera Cerro Verde S. A. A, 22 de enero de 2007. 6. Cotizaciones de los metales A partir del año 2004, el crecimiento sos- tenido de los precios de los metales deter- minó que los resultados de las empresas mineras peruanas, en general, fueran espec- taculares. En el cuadro 1110 se presenta la evolución de las cotizaciones internacio- nales de los metales más importantes. Para mayor detalle, en el cuadro 12 se muestra el incremento porcentual de las cotizacio- nes de los metales tomando como base Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 158 por sus siglas en inglés) que ha desincenti- vado las compras de parte de las fundidoras locales. El ScotiaBank12, en el Reporte Empre- sarial de la Cía. Minera Cerro Verde, indica una posición similar a la antes descrita con respecto al precio de los metales; más especíﬁ camente, el cobre. En el caso de los metales preciosos (oro y plata), las perspectivas son diferentes. De acuerdo con el New York Mercantile Exchange (NYMEX), los precios de los futuros del oro y la plata en el año 2007 tendrían una tendencia creciente; aunque no se habían formado precios para fechas posteriores. En todo caso, según la infor- mación presentada en el cuadro 16, para ﬁ nes de 2007 se espera un incremento en el precio spot de 4% para el caso del oro, y de 6% para la plata. De acuerdo con nuestros estimados, en los próximos tres años la cotización interna- cional del cobre registraría una tendencia a la baja, pero manteniéndose por encima de sus niveles de largo plazo, producto de un mercado con mayor disponibilidad del mineral (superávit físico) y menores compras especulativas, tal como se viene observando en los últimos meses […]. En el corto plazo consideramos que se podrían registrar alzas en la cotización del mineral producto de una mayor presencia de China en el mercado internacional, tras la caída de sus reservas locales, y una menor contrac- ción de sus importaciones de concentrados tras los bajos costos de tratamiento (TC/RC En el anexo 2 se presentan diversos artículos relacionados con la evolución de los precios de los metales y un artículo de Moody’s sobre el Perú. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 159 Cuadro 11. 6. Cotizaciones de los metales Cotizaciones históricas de los principales metales Aluminio Cobre Hierro Plomo Níquel Estaño Zinc 2004 Enero 1 608,89 2 421,48 37,90 753,56 15 089,33 6 432,49 1 015,89 Febrero 1 685,23 2 751,72 37,90 885,15 15 099,35 6 662,70 1 085,79 Marzo 1 657,35 3 000,28 37,90 878,14 13 786,54 7 602,21 1 101,80 Abril 1 731,68 2 926,98 37,90 747,03 12 725,90 8 909,00 1 028,91 Mayo 1 625,27 2 728,46 37,90 804,10 11 228,61 9 391,63 1 030,98 Junio 1 682,04 2 689,05 37,90 863,73 13 599,36 9 158,64 1 018,86 Julio 1 707,87 2 816,80 37,90 931,34 15 020,07 9 021,91 988,10 Agosto 1 692,10 2 844,20 37,90 916,62 13 639,62 9 021,43 976,80 Septiembre 1 731,02 2 903,17 37,90 932,16 13 430,36 9 015,68 980,03 Octubre 1 830,37 3 009,40 37,90 932,86 14 378,48 9 043,21 1 066,95 Noviembre 1 817,35 3 130,31 37,90 966,31 14 089,51 9 038,65 1 100,23 Diciembre 1 852,92 3 139,79 37,90 972,34 13 764,98 8 473,71 1 182,14 2005 Enero 1 836,21 3 168,10 65,00 954,23 14 563,75 7 705,65 1 245,55 Febrero 1 882,80 3 247,10 65,00 972,98 15 415,60 8 106,33 1 323,11 Marzo 1 987,52 3 378,90 65,00 1 001,69 16 239,90 8 442,24 1 373,96 Abril 1 892,01 3 389,81 65,00 980,48 16 138,33 8 133,81 1 297,81 Mayo 1 741,45 3 241,90 65,00 985,29 17 002,25 8 099,25 1 245,54 Junio 1 731,94 3 529,73 65,00 982,65 16 113,18 7 604,36 1 273,12 Julio 1 783,26 3 608,48 65,00 857,52 14 587,60 7 180,93 1 196,86 Agosto 1 871,27 3 791,91 65,00 887,42 14 962,00 7 228,14 1 300,75 Septiembre 1 837,69 3 850,66 65,00 932,77 14 154,55 6 770,58 1 396,66 Octubre 1 934,14 4 056,17 65,00 999,40 12 431,12 6 415,48 1 483,22 Noviembre 2 056,97 4 278,16 65,00 1 017,84 12 235,05 6 173,72 1 610,65 Diciembre 2 250,90 4 577,03 65,00 1 120,15 13 490,45 6 762,50 1 819,36 2006 Enero 2 383,30 4 743,86 77,35 1 258,14 14 660,81 7 067,36 2 091,77 Febrero 2 453,38 4 974,98 77,35 1 267,44 14 974,50 7 788,87 2 219,75 Marzo 2 432,48 5 123,67 77,35 1 193,89 14 925,48 7 949,02 2 427,66 Abril 2 623,86 6 404,44 77,35 1 170,56 18 028,89 8 859,66 3 068,34 Mayo 2 852,07 8 059,19 77,35 1 167,50 21 131,33 8 793,17 3 544,64 Junio 2 490,95 7 222,77 77,35 963,61 20 585,91 7 858,77 3 197,59 Julio 2 511,83 7 726,74 77,35 1 053,26 26 185,71 8 356,44 3 320,74 Agosto 2 461,55 7 690,25 77,35 1 179,32 30 468,86 8 436,68 3 339,97 Septiembre 2 484,38 7 622,64 77,35 1 346,53 29 702,62 8 975,02 3 394,06 Octubre 2 657,15 7 497,41 77,35 1 525,66 32 551,14 9 809,50 3 829,60 Noviembre 2 702,14 7 029,30 77,35 1 626,02 31 891,59 10 038,41 4 378,61 Diciembre 2 823,67 6 680,97 77,35 1 709,16 34 440,53 11 125,95 4 381,45 2007 Enero 2 799,06 5 689,34 77,35 1 664,34 36 821,59 11 331,45 3 784,86 Cuadro 11. Fuente: Fondo Monetario Internacional. Fuente: Fondo Monetario Internacional. 6. Cotizaciones de los metales 1254.8 +7.2 1257.0 1 245.7 2.14 10.9 Total 3.61 11.5 PALLADIUM NYMEX (100 troy oz; $/troy oz) Jun. 357.25 +1.25 357.85 354.95 0.56 14.4 Sep. 362.75 +1.25 – – 0.00 1.09 Total 0.56 15.6 SILVER COMEX (5 000 troy oz; $/troy oz) May. 1345.0 +11.0 1347.5 1332.5 17.7 58.5 Jul. 1357.9 +11.2 1362.0 1345.5 1.50 17.7 Total 19.6 111.6 Sources: London Bullion Market Association, Reuters. Gold Fix at 10:30 & 15:00. Silver ﬁ x at 12:00. Closing ﬁ gures at 17:00 BASE METALS Cuadro 13. Precios de los metales según el London Metal Exchange al 30 de marzo de 2007 LONDON METAL EXCHANGE Cash 3 Mth Kerb PM Day’s High Open Tumover S/tonne Ofﬁ cial Ofﬁ cial 3 Mth close Low (3 Mth) interest (Lots) (Lots) Aluminium 2791-92 2803-4 2778-80 2815/2 776 515,710 87,300 Alum alloy 2230-40 2260-70 2240-50 2260/2 240 6,874 6,578 Amer alloy 2150-55 2180-90 2190-200 2195/2 190 24,687 2,514 Copper 6939-40 6859,5-60,5 6850-60 6935/6 785 238,791 51,265 Lead 1935-36 1925-26 1910-12 1935/1 910 63,237 9,246 Nickel 45400-500 43795-800 44800-5 000 44800/43 514 46,419 1,274 Tin 13650-75 13650-700 13400-450 13700/13 300 15,475 8,570 Zinc 3279,5-80,5 3285-90 3250-55 3303/3 250 142,674 37,484 Spot: 1,9582 3 mths: 1,9678 0 mths: 1,9654 9 mths 1,9533. LME AM Ofﬁ cial £/$ rate 1,9585. LME Closing £/$ rates: 1,9648. Karb close at 17:00. Source: Amalgamated Metal Trading www amt co uk For further trading information see: www Jme co uk BASE METALS Cuadro 13. Precios de los metales según el London Metal Exchange al 30 de marzo de 2007 LONDON METAL EXCHANGE LONDON BULLION MARKET Gold (troy oz) $ price £ equiv € equiv Close 663.60-664.10 Opening 662.40-662.90 Morning ﬁ x 663.50 339.04 498.24 Afternoon ﬁ x 661.75 338.28 497.71 Day’s High 665.30 Day’s Low 660.90 Previous close 659.90-660.40 Loco Ldn Mean Gold Lending Rates (US$) 1 mth …... 5.32 6 mths …... 5.33 3 mths …... 5.35 12 mths …... 5.22 Gold Leasing Rates (US$) 1 mth ….. –0.06 6 mths ….. 0.00 3 mths…... –0.02 12 mths ….. 0.03 Silver Fix p/troy oz US cts equiv. Spot 681.92 1335.00 Silver Lending Rates 1 mth ….. 5.30 6 mths ….. 4.90 3 mths …. 5.20 12 mths …. 4.30 Sources: London Bullion Market Association, Reuters. Gold Fix at 10:30 & 15:00. Silver ﬁ x at 12:00. Closing ﬁ gures at 17:00 LONDON BULLION MARKET 6. Cotizaciones de los metales Cotizaciones históricas de los principales metales Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 160 Cuadro 12. Incremento porcentual en las cotizaciones de los principales metales Aluminio Cobre Hierro Plomo Níquel Estaño Zinc 2004 Enero 1,00 1,00 1,00 1,00 1,00 1,00 1,00 Febrero 1,05 1,14 1,00 1,17 1,00 1,04 1,07 Marzo 1,03 1,24 1,00 1,17 0,91 1,18 1,08 Abril 1,08 1,21 1,00 0,99 0,84 1,39 1,01 Mayo 1,01 1,13 1,00 1,07 0,74 1,46 1,01 Junio 1,05 1,11 1,00 1,15 0,90 1,42 1,00 Julio 1,06 1,16 1,00 1,24 1,00 1,40 0,97 Agosto 1,05 1,17 1,00 1,22 0,90 1,40 0,96 Septiembre 1,08 1,20 1,00 1,24 0,89 1,40 0,96 Octubre 1,14 1,24 1,00 1,24 0,95 1,41 1,05 Noviembre 1,13 1,29 1,00 1,28 0,93 1,41 1,08 Diciembre 1,15 1,30 1,00 1,29 0,91 1,32 1,16 2005 Enero 1,14 1,31 1,72 1,27 0,97 1,20 1,23 Febrero 1,17 1,34 1,72 1,29 1,02 1,26 1,30 Marzo 1,24 1,40 1,72 1,33 1,08 1,31 1,35 Abril 1,18 1,40 1,72 1,30 1,07 1,26 1,28 Mayo 1,08 1,34 1,72 1,31 1,13 1,26 1,23 Junio 1,08 1,46 1,72 1,30 1,07 1,18 1,25 Julio 1,11 1,49 1,72 1,14 0,97 1,12 1,18 Agosto 1,16 1,57 1,72 1,18 0,99 1,12 1,28 Septiembre 1,14 1,59 1,72 1,24 0,94 1,05 1,37 Octubre 1,20 1,68 1,72 1,33 0,82 1,00 1,46 Noviembre 1,28 1,77 1,72 1,35 0,81 0,96 1,59 Diciembre 1,40 1,89 1,72 1,49 0,89 1,05 1,79 2006 Enero 1,48 1,96 2,04 1,67 0,97 1,10 2,06 Febrero 1,52 2,05 2,04 1,68 0,99 1,21 2,19 Marzo 1,51 2,12 2,04 1,58 0,99 1,24 2,39 Abil 1,63 2,64 2,04 1,55 1,19 1,38 3,02 Mayo 1,77 3,33 2,04 1,55 1,40 1,37 3,49 Junio 1,55 2,98 2,04 1,28 1,36 1,22 3,15 Julio 1,56 3,19 2,04 1,40 1,74 1,30 3,27 Agosto 1,53 3,18 2,04 1,56 2,02 1,31 3,29 Septiembre 1,54 3,15 2,04 1,79 1,97 1,40 3,34 Octubre 1,65 3,10 2,04 2,02 2,16 1,52 3,77 Noviembre 1,68 2,90 2,04 2,16 2,11 1,56 4,31 Diciembre 1,76 2,76 2,04 2,27 2,28 1,73 4,31 2007 Enero 1,74 2,35 2,04 2,21 2,44 1,76 3,73 Fuente: Fondo Monetario Internacional. dro 12. Incremento porcentual en las cotizaciones de los principales metales Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. 6. Cotizaciones de los metales A.: valoración de una empresa minera 161 LME WAREHOUSE STOCKS (tonnes) Alluminum –950 to 811,325 Alluminum alloy –180 to 96,040 Copper –1,800 to 178,075 Lead +125 to 33,250 Nickel +126 to 5,418 Zinc –300 to 106,275 Tin –10 to 9,635 LONDON BULLION MARKET Gold (troy oz) $ price £ equiv € equiv Close 663.60-664.10 Opening 662.40-662.90 Morning ﬁ x 663.50 339.04 498.24 Afternoon ﬁ x 661.75 338.28 497.71 Day’s High 665.30 Day’s Low 660.90 Previous close 659.90-660.40 Loco Ldn Mean Gold Lending Rates (US$) 1 mth …... 5.32 6 mths …... 5.33 3 mths …... 5.35 12 mths …... 5.22 Gold Leasing Rates (US$) 1 mth ….. –0.06 6 mths ….. 0.00 3 mths…... –0.02 12 mths ….. 0.03 Silver Fix p/troy oz US cts equiv. Spot 681.92 1335.00 Silver Lending Rates 1 mth ….. 5.30 6 mths ….. 4.90 3 mths …. 5.20 12 mths …. 4.30 Cuadro 13. Precios de los metales según el London Metal Exchange al 30 de marzo de 2007 LONDON METAL EXCHANGE Cash 3 Mth Kerb PM Day’s High Open Tumover S/tonne Ofﬁ cial Ofﬁ cial 3 Mth close Low (3 Mth) interest (Lots) (Lots) Aluminium 2791-92 2803-4 2778-80 2815/2 776 515,710 87,300 Alum alloy 2230-40 2260-70 2240-50 2260/2 240 6,874 6,578 Amer alloy 2150-55 2180-90 2190-200 2195/2 190 24,687 2,514 Copper 6939-40 6859,5-60,5 6850-60 6935/6 785 238,791 51,265 Lead 1935-36 1925-26 1910-12 1935/1 910 63,237 9,246 Nickel 45400-500 43795-800 44800-5 000 44800/43 514 46,419 1,274 Tin 13650-75 13650-700 13400-450 13700/13 300 15,475 8,570 Zinc 3279,5-80,5 3285-90 3250-55 3303/3 250 142,674 37,484 Spot: 1,9582 3 mths: 1,9678 0 mths: 1,9654 9 mths 1,9533. LME AM Ofﬁ cial £/$ rate 1,9585. LME Closing £/$ rates: 1,9648. Karb close at 17:00. Source: Amalgamated Metal Trading www.amt.co.uk For further trading information see: www.Jme.co.uk HIGH GRADE COPPER (Comex) 26/07/06 Sett Day’s High Low Vol Open price change int Apr. 314.35 +6.70 315.25 308.35 0.61 2.34 May. 314.60 +5.95 316.00 308.80 6.76 44.8 Jun. 314.40 +5.80 315.00 312.00 0.17 1.94 Jul. 313.90 +5.65 315.10 308.25 1.04 16.6 Total 8.81 75.4 PRECIOUS METALS GOLD COMEX (100 troy oz; $/troy oz) Sett Day’s High Low Vol 0 int price change 000s 000s Apr. 663.0 +1.5 667.7 659.1 37.4 9.71 Jun. 669.0 +1.4 673.8 666.5 86.9 200.7 Total 131.2 345.9 PLATINUM NYMEX (50 troy oz; $/troy oz) Apr. 1248.3 +5.7 1248.0 1 240.4 1.40 0.65 Jul. PRECIOUS METALS GOLD COMEX (100 troy oz; $/troy oz) GOLD COMEX (100 troy oz; $/troy oz) Sett Day’s High Low Vol 0 int price change 000s 000s Apr. 663.0 +1.5 667.7 659.1 37.4 9.71 Jun. 669.0 +1.4 673.8 666.5 86.9 200.7 Total 131.2 345.9 PLATINUM NYMEX (50 troy oz; $/troy oz) Apr. 1248.3 +5.7 1248.0 1 240.4 1.40 0.65 Jul. 1254.8 +7.2 1257.0 1 245.7 2.14 10.9 Total 3.61 11.5 PALLADIUM NYMEX (100 troy oz; $/troy oz) Jun. 357.25 +1.25 357.85 354.95 0.56 14.4 Sep. 362.75 +1.25 – – 0.00 1.09 Total 0.56 15.6 SILVER COMEX (5 000 troy oz; $/troy oz) May. 1345.0 +11.0 1347.5 1332.5 17.7 58.5 Jul. 1357.9 +11.2 1362.0 1345.5 1.50 17.7 Total 19.6 111.6 Cuad. Difus. 12 (22), jun. 2007 162 Alfredo Mendiola Cuadro 14. Precio de los metales a mediano plazo (marzo de 2007, dólares por tonelada) Cobre Plomo Zinc Spot 6 450 1 913 3 270 A 15 meses 6 015 1 575 3 058 A 27 meses 5 409 – 2 764 Fuente: London Metal Exchange. Cuadro 14. Precio de los metales a mediano plazo (marzo de 2007, dólares por tonelada) Cuadro 15. Precios promedio y de cierre de los metales (marzo de 2007, dólares por tonelada) Primario- Aleación Cobre Plomo Níquel Estaño Zinc Special Aluminio Aluminio High Grade Cash Buyer 2 760,20 2 187,82 6 449,16 1 912,18 46 282,05 13 878,86 3 270,34 Cash Seller & 2 761,73 2 193,05 6 452,48 1 914,05 46 324,77 13 892,95 3 271,30 Settlement Cash Mean 2 760,97 2 190,43 6 450,82 1 913,11 46 303,41 13 885,91 3 270,82 3-months Buyer 2 756,91 2 219,32 6 424,48 1 882,98 43 594,77 13 798,86 3 273,32 3-months Seller 2 758,20 2 228,64 6 427,30 1 885,98 43 631,59 13 820,91 3 276,50 3-months Mean 2 757,56 2 223,98 6 425,89 1 884,48 43 613,18 13 809,89 3 274,91 15-months Buyer 2 609,86 2 310,45 6 010,45 1 573,18 35 440,91 12 578,18 3 055,91 15-months Seller 2 614,86 2 320,45 6 020,45 1 578,18 35 540,91 12 628,18 3 060,91 15-months Mean 2 612,36 2 315,45 6 015,45 1 575,68 35 490,91 12 603,18 3 058,41 27-months Buyer 2 430,14 2 338,86 5 404,32 30 888,64 2 761,68 27-months Seller 2 435,14 2 348,86 5 414,32 30 988,64 2 766,68 27-months Mean 2 432,64 2 343,86 5 409,32 30 938,64 2 764,18 Cuadro 15. Fuente: London Metal Exchange. PRECIOUS METALS Precios promedio y de cierre de los metales (marzo de 2007, dólares por tonelada) Las siguientes equivalencias para libras esterlinas han sido calculadas en función de conversiones diarias. siguientes equivalencias para libras esterlinas han sido calculadas en función de iones diarias. Tipos de cambio Copper Cash Seller & Settlement £ 3 313,41 Conversión de monedas Copper 3-months Seller £ 3 301,66 Stg/$ 1,9468 Lead Cash Seller & Settlement £ 983,13 $/JY 117,2532 Lead 3-months Seller £ 969,01 Euro 1,3243 Fuente: London Metal Exchange. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 163 163 Cuadro 16. Precios promedio de cierre al 2 de abril de 2004 (dólares por onza) Precios del oro Last Open high Open low High Low Most recent 2007 Abril 663,50 665,50 665,50 665,50 656,10 665,7s Mayo 667,00 667,80 667,80 669,00 660,90 668,4s Junio 669,30 669,00 668,50 672,30 661,70 671,5s Agosto 677,10 676,70 676,70 678,40 668,50 677,7s Octubre 684,50 683,00 683,00 684,50 677,90 683,9s Diciembre 690,00 689,60 689,60 690,00 680,00 690,0s Precios de la plata Last Open high Open low High Low Most recent 2007 Abril 13,02 n. a. n. a. 13,10 13,02 13,293s Mayo 13,32 13,36 13,34 13,50 13,02 13,350s Junio n. a. n. a. n. a. n. a. n. a. 13,408s Agosto 13,48 13,52 13,46 13,60 13,18 13,477s Octubre 13,36 13,74 13,74 13,74 13,36 13,595s Diciembre 13,76 13,83 13,83 13,84 13,42 13,748s Fuente: New York Mercantile Exchange. Cuadro 16. Precios promedio de cierre al 2 de abril de 2004 (dólares por onza) 7. Estructura de ﬁ nanciamiento, accionistas y gerencia sus proyectos y para usos generales cor- porativos. A lo largo del año 2006, Milpo y Rayrock prepagaron los préstamos bancarios otor- gados por Citibank del Perú S. A. (por 30 millones de dólares) y WestLB AG (por 10 millones de dólares). Como consecuencia del prepago, la empresa tuvo que efectuar un pago adicional por 345 mil dólares que se registró como gasto ﬁ nanciero (año 2006). Las deudas de largo plazo se des- componen a su vez en: Desde el 2004, Milpo ha mantenido una corriente sostenida de pago de dividendos a sus accionistas. 2006 2005 2004 2003 Dividendos (miles de dólares 6 204 2 490 1 896 – estadounidenses) Al 31 de diciembre de 2006, el capital social de Milpo estaba compuesto por 484 032 930 acciones comunes autori- zadas, emitidas y pagadas, cuyo valor nominal era de 1 sol cada una (equivalente a 0,3046 de dólar). En el cuadro 17 se detalla el movimiento de acciones comunes para los ejercicios 2005 y 2006, de acuerdo con el dictamen de los auditores. 2006 2005 Porción corriente 1 894 3 362 Porción no corriente 1 428 39 448 A la fecha, Milpo se encuentra evaluan- do la conveniencia de obtener un préstamo por 40 millones de dólares para ﬁ nanciar Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 164 Cuadro 17. Movimiento de acciones comunes, 2005 y 2006 2006 2005 En circulación al inicio del año 431 893 560 30 898 966 Emitidas por: Incremento por cambio de valor nominal – 342 360 543 Capitalización de la reexpresión del ejercicio 2004 – 17 118 069 Incremento del capital por fusión 16 618 – Capitalización de resultados acumulados 52 122 752 41 515 982 Subtotal 484 032 930 431 893 560 Acciones en tesorería (96 234 909) (85 871 564) Acciones en circulación al ﬁ nal del año 387 798 021 346 021 996 Cuadro 17. Movimiento de acciones comunes, 2005 y 2006 Acciones en circulación al ﬁ nal del año inversión en caso de aumento de capital por nuevos aportes, incrementar la cuenta de acciones de inversión por capitalización de cuentas patrimoniales, redimir sus acciones en cualquiera de los casos de liquidación de la Compañía. Las acciones de inversión no conﬁ eren acceso al Directorio ni a las Juntas Generales de Accionistas.14 El principal accionista de la Cía. Mine- ra Milpo es Votorantim Metais-Cajamarqui- lla S. 14. Ernst & Young. Compañía Minera Milpo y subsidiarias. Notas a los estados ﬁ nancieros con- solidados al 31 de diciembre de 2006 y 2005. 13. Votorantim Metais mantiene una posición líder en la producción de zinc y níquel en América Latina, ya que es la tercera mayor productora de aceros largos del Brasil y la quinta productora de zinc a nivel mundial. Esta empresa cuenta con ocho unidades industriales y con minas propias en el Brasil. Fuente: Class & Asociados. Informe de Fundamento de Clasiﬁ cación de Riesgo de Compañía Minera Milpo. Sesión de Comité n.° 02/2007 del 2 de febrero de 2007. 7. Estructura de ﬁ nanciamiento, accionistas y gerencia A., que concentra el 19,93% del capital. Esta empresa se constituyó como conse- cuencia de la compra de la reﬁ nería de zinc de Cajamarquilla por Votorantim Metais de Brasil13. Los otros accionistas importantes son Milpo Finance and Investments Inc., que tiene el 10,50%, Cuyuma S. A., que posee el 9,33%, e IN-Cartadm, cuya parti- cipación es de 8,21%. El resto de los accio- nistas mantiene una participación menor de 8% del capital social. Según Class & Asociados: Cuadro 18. Movimiento de acciones de inversión, 2005 y 2006 se ha reducido, aunque todavía no se logra el grado de inversión necesario para atraer capitales extranjeros. En todo caso, la eco- nomía peruana es más fuerte y está mejor preparada para soportar las condiciones económicas y ﬁ nancieras internacionales adversas que se esperan. La misma fuente indica que el riesgo de no pago del país se ha reducido, las reservas internacionales netas han crecido y las exportaciones se han mantenido a un ritmo creciente en los últimos años. Auditoría, de Compensación, de Seguridad y Salud Ocupacional; de Gobierno Corpo- rativo; y de Medio Ambiente y Responsa- bilidad Social. Al 30 de setiembre de 2006, el número de trabajadores de la compañía ascendió a 961 personas, entre obreros, empleados, ejecutivos y profesionales (806 al cierre de 2005). 8. Perspectivas económicas del país Las perspectivas económicas del Perú en el mediano plazo (dos o tres años) son muy prometedoras. De acuerdo con el área de estudios económicos del Banco Bilbao Vizcaya, sucursal del Perú (ver cuadro 19), se espera que el crecimiento de la economía se mantenga en alrededor de 6% durante los siguientes dos años, con niveles de inﬂ ación menores de 3% al año y un proceso de devaluación mínimo. Tal como indica el FMI, el crecimiento eco- nómico se ha reﬂ ejado en la reducción de las vulnerabilidades económicas del país, que ahora presenta superávit en cuenta corriente, mercados cambiarios más ﬂ exi- bles, mayores reservas internacionales y reducción del déﬁ cit ﬁ scal. Según Class & Asociados: ... la empresa mantiene contacto con sus accionistas a través del Directorio, cuyos miembros para el período 2006-2008, fue- ron ratiﬁ cados en la Junta Obligatoria de Accionistas celebrada en marzo de 2008 […]. Respecto a los cuadros gerenciales, se observó cierto nivel de rotación de los ejecutivos durante los primeros tres tri- mestres del ejercicio 2005, hasta que se aprobó en setiembre una nueva estructura organizacional como parte de la estrategia de crecimiento de la compañía. En junio de 2006, se realizó el cambio más reciente en la plana gerencial con el nombramiento del Gerente de Seguridad y Salud Ocupacional […]. Desde el año 2005, el Directorio de la empresa es apoyado por los Comités de En el cuadro 18 se precisa el movimien- to de acciones de inversión. Es necesario hacer notar que: ... de acuerdo a la ley, las acciones de in- versión atribuyen a sus titulares derecho a participar en la distribución de dividendos, efectuar aportes a ﬁ n de mantener su pro- porción existente en la cuenta acciones de Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 165 Cuadro 18. Movimiento de acciones de inversión, 2005 y 2006 2006 2005 En circulación al inicio del año 6 419 730 4 623 488 Emitidas por: Incremento por cambio de valor nominal – 924 698 Capitalización de la reexpresión del ejercicio 2004 – 254 447 Capitalización de resultados acumulados 774 760 617 097 Subtotal 7 194 490 6 419 730 Acciones en tesorería (488 464) (450 652) Acciones en circulación al ﬁ nal del año 6 706 026 5 969 078 Cuadro 18. Movimiento de acciones de inversión, 2005 y 2006 En este sentido, Fitch es muy preciso: En este sentido, Fitch es muy preciso: La suave transición a la administración García, las condiciones económicas favora- bles, así como un acuerdo stand-by preven- tivo con el FMI han reducido los riesgos políticos relacionados con el mantenimien- to de la política económica de Perú. Sin embargo, la naturaleza concentrada de la base exportadora del país y las potenciales implicaciones negativas que esto podría tener sobre la posición externa y ﬁ scal de Perú en caso de un shock de los precios de los commodities, continúan como una debi- lidad crediticia clave. No obstante, Fitch considera que Perú está mejor preparado para enfrentar precios de los minerales más bajos, debido a que los aumentos recientes y esperados de los volúmenes de exportación de bienes tradicionales y no tradicionales De igual manera, Moody’s ha estable- cido que el riesgo de inversión en el Perú Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 166 Sin embargo, las grandes debilidades del Perú se centran alrededor de los proble- mas sociales. El crecimiento poblacional es sostenido, pero se observa concentración en las ciudades de la costa, lo que evidencia no solo el centralismo, sino el atraso en el cual viven las otras regiones del país. La diferen- cia de ingresos entre ricos y pobres es muy grande, la más pronunciada de la región. Otro problema social muy importante es la ofensiva del narcotráﬁ co, al punto que dos líderes cocaleros han sido elegidos parlamentarios. Finalmente, es necesario notar que las normas legales establecen con precisión los cuidados ambientales que debe tener en cuenta la minería en el Perú. En todo caso, diversas unidades mineras han tenido problemas con las comunidades nativas establecidas en los alrededores de sus operaciones. podrían ser suﬁ cientes para superar una reducción gradual en los términos comer- ciales de Perú15. En el ámbito político, en el año 2006 el Partido Aprista16 ganó las elecciones presi- denciales, y después de más de medio año en el poder no está claro cuál será el rumbo que seguirá el país. Aunque el partido de gobierno no logró mayoría en el Parla- mento, lo que le permitiría cierto respaldo para el cambio del marco legal vigente, el nivel de aceptación del presidente García se encuentra cercano al 60%. Estimado. * Proyectado. Fuente: Banco Bilbao Vizcaya-Banco Continental. En este sentido, Fitch es muy preciso: En todo caso, hay algunos temas en la agenda política que son cruciales para la estabilidad de largo plazo del país: a) lucha contra la pobreza, b) descentralización, c) reforma del Estado y d) generación de empleo. Las autoridades gubernamentales están tratando de atacar estos problemas, pero los resultados de estas políticas se verán en el largo plazo. Cuadro 19. Perú: previsiones anuales 2004 2005 2006 2007* 2008*** PIB (%) 4,8 6,4 7,4 6,3 6,0 Inﬂ ación IPC (% ﬁ n de año) 3,5 1,5 1,1 1,9 2,5 Balanza comercial (m.M.$) 2,8 5,3 8,5 7,0 6,0 Cuenta corriente (m.M.$) 0,0 1,1 1,7 0,7 –0,7 % PIB 0,0 1,4 1,9 0,7 –0,7 Reservas (m.M.$, ﬁ n de año) 12,6 14,1 17,3 18,0 18,5 Tipo de cambio (ﬁ n de año vs. US$) 3,28 3,42 3,20 3,22 3,30 Saldo sector público (% PIB) –1,1 –0,3 1,0 –0,5 –0,8 Tasa de interés* (ﬁ n de año) 3,0 3,3 4,5 5,0 5,0 Tipo de cambio efectivo real (ﬁ n de año, dic. 97=100) BBVA-MAP (ﬁ n de año, jun. 95=100) Tasa interbancaria en moneda nacional Fuente: Banco Bilbao Vizcaya-Banco Continental. * Estimado. *** Proyectado. Cuadro 19. Perú: previsiones anuales Cuad. Difus. 12 (22), jun. 2007 Cuad. Difus. 12 (22), jun. 2007 167 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 167 Anexo 1 Compañía Minera Milpo S. A. A. En este sentido, Fitch es muy preciso: y subsidiarias Balance General (miles de dólares) 2006 2005 2004 2003 Activo Activo corriente Efectivos y equivalentes de efectivo 51 231 33 799 31 954 14 349 Cuentas por cobrar comerciales (neto) 23 439 8 201 7 739 3 922 Otras cuentas por cobrar (neto) 11 433 8 439 1 220 1 563 Existencias (neto) 18 768 13 921 11 406 9 363 Gastos diferidos 692 1 230 4 698 5 650 Total activo corriente 105 563 65 590 57 017 34 847 Activo no corriente Inversiones ﬁ nancieras 1 855 1 497 1 045 13 097 Otras cuentas por cobrar – 250 250 262 Inmuebles, maquinaria y equipo (neto) 100 947 86 076 78 331 74 666 Activos intangibles (neto) 98 372 64 331 46 651 42 556 Activo por impuesto a la renta y 2 751 519 2 156 453 participaciones diferidos Otros activos – – 1 380 – Total activo no corriente 203 925 152 673 129 813 131 033 Total activo 309 488 218 263 186 830 165 880 Pasivo y Patrimonio Pasivo corriente Obligaciones ﬁ nancieras 1 894 3 362 2 855 8 720 Cuentas por pagar comerciales 17 046 10 584 10 274 12 736 Impuesto a la renta y participaciones corrientes 23 413 9 621 – – Otras cuentas por pagar 19 403 7 045 8 085 8 308 Pasivos por instrumentos ﬁ nancieros derivados 2 886 14 428 14 660 13 306 Total pasivo corriente 64 642 45 040 35 875 43 070 Pasivo no corriente Obligaciones ﬁ nancieras 1 428 39 448 41 384 51 775 Instrumentos derivados – – 7 280 6 141 Pasivo por impuesto a la renta y 5 134 3 887 3 682 2 542 participaciones diferidos Otras cuentas por pagar 8 267 3 340 1 961 1 236 Total pasivo no corriente 14 829 46 675 54 307 61 694 Total pasivo 79 471 91 715 90 182 104 765 Patrimonio neto Capital 124 553 109 953 101 382 70 013 Acciones de inversión 2 184 1 955 1 768 1 429 Reservas legales 18 322 8 695 6 774 4 868 Otras reservas 30 6 963 (407) (10 607) Utilidades no distribuidas 84 928 20 999 3 735 3 449 Pérdida no realizada por instrumentos derivados – (22 017) (16 605) (8 038) Total patrimonio neto atribuible a la matriz 230 017 125 548 96 648 61 116 Total Pasivo y Patrimonio 309 488 218 263 186 830 165 880 Cuad. En este sentido, Fitch es muy preciso: Difus. 12 (22), jun. 2007 Alfredo Mendiola 168 Compañía Minera Milpo S. A. A. y subsidiarias Estado de Ganancias y Pérdidas (miles de dólares) 2006 2005 2004 2003 Ingresos operacionales Ventas netas (ingresos operacionales) 279 961 122 060 97 892 77 876 Otros ingresos operacionales – – 1 636 2 878 Total de ingresos brutos 279 961 122 060 99 528 80 754 Costos operativos (76 265) (47 989) (49 584) (41 728) Depreciación (10 149) (6 638) (7 879) (7 121) Total costos operacionales (86 414) (54 627) (57 463) (48 850) Utilidad bruta 193 547 67 433 42 064 31 904 Gastos de ventas (6 501) (5 887) (4 733) (4 626) Gastos de administración (6 867) (5 099) (4 551) (5 317) Exploración y amortización de proyectos (7 173) (4 264) (3 366) (1 890) Total gastos operacionales (20 541) (15 250) (12 650) (11 833) Utilidad operativa 173 006 52 183 29 415 20 071 Ganancia en la venta de inversiones – – 8 315 529 Ingresos ﬁ nancieros 1 199 772 257 204 Ganancia (pérdida) por instrumentos (28 611) (17 135) (14 982) (2 709) ﬁ nancieros derivados Gastos ﬁ nancieros (4 046) (3 447) (3 425) (3 768) Participación en los resultados de partes – – 2 383 2 049 relacionadas por el método de participación Aporte económico extraordinario temporal (3 333) – – – Otros (neto) (5 915) (346) (2 907) (421) Ganancia (pérdida) por translación 479 (1 095) (383) (447) Amortización del crédito mercantil – – – (277) Provisión por desvalorización de inversiones – – – (121) (40 227) (21 251) (10 741) (4 961) Resultado antes de participaciones y del 132 779 30 932 18 673 15 110 impuesto a la renta Participación de los trabajadores (11 302) (3 533) (689) (1 213) Impuesto a la renta (25 207) (8 161) (1 612) (683) Utilidad neta 96 270 19 238 16 373 13 215 Compañía Minera Milpo S. A. A. y subsidiarias Estado de Ganancias y Pérdidas (miles de dólares) Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 169 Compañía Minera Milpo S. A. A. En este sentido, Fitch es muy preciso: y subsidiarias Estado de Flujo de Efectivo (miles de dólares) 2006 2005 2004 2003 Actividades de operación Venta de bienes o servicios 264 723 121 594 95 927 88 801 (ingresos operacionales) Otros cobros de efectivo relativos a la actividad – – 1 060 – Proveedores de bienes y servicios (56 629) (35 308) (50 676) (36 565) Remuneraciones y beneﬁ cios sociales (24 472) (17 921) (12 553) (14 783) Tributos (29 053) (4 022) (2 311) (1 800) Intereses y rendimientos (no incluidos en ﬁ nanciación) 4 046 (3 447) – – Regalías (271) – (3 425) (4 461) Otros pagos de efectivo relativos a la actividad (30 420) (27 321) (24 068) 2 654 Efectivo y equivalente de efectivo provenientes 119 832 33 575 3 955 33 845 de actividades de operación Actividades de inversión Venta de valores e inversiones permanentes – – 48 380 3 120 Venta de inmuebles, maquinaria y equipo – – 329 736 Compra de inmuebles, maquinaria y equipo (28 520) (15 073) (9 220) (2 229) Desembolsos por actividades de exploración y desarrollo (27 945) (13 549) (2 628) (15 998) Otras adquisiciones (358) (453) (1 533) – Efectivo y equivalente de efectivo provenientes (56 823) (29 075) 35 328 (14 371) de actividades de inversión Actividades de ﬁ nanciación Aumento de préstamos bancarios 1 204 1 767 1 259 – Pago de deudas de largo plazo (40 577) (1 917) (21 833) (9 389) Dividendos pagados a accionistas de la matriz (6 204) (2 490) (1 896) – Efectivo y equivalente de efectivo provenientes de (45 577) (2 640) (22 470) (9 389) actividades de ﬁ nanciación Aumento (disminución) neto de efectivo y 17 432 1 860 16 813 10 085 equivalente de efectivo Saldo efectivo y equivalente de efectivo al 33 799 31 939 15 140 5 056 inicio del ejercicio Saldo efectivo y equivalente de efectivo al 51 231 33 799 31 954 15 140 ﬁ nalizar el ejercicio Compañía Minera Milpo S. A. A. y subsidiarias Estado de Flujo de Efectivo (miles de dólares) Actividades de inversión Venta de valores e inversiones permanentes – – 48 380 3 120 Venta de inmuebles, maquinaria y equipo – – 329 736 Compra de inmuebles, maquinaria y equipo (28 520) (15 073) (9 220) (2 229) Desembolsos por actividades de exploración y desarrollo (27 945) (13 549) (2 628) (15 998) Otras adquisiciones (358) (453) (1 533) – Efectivo y equivalente de efectivo provenientes (56 823) (29 075) 35 328 (14 371) de actividades de inversión Actividades de ﬁ nanciación Aumento de préstamos bancarios 1 204 1 767 1 259 – Pago de deudas de largo plazo (40 577) (1 917) (21 833) (9 389) Dividendos pagados a accionistas de la matriz (6 204) (2 490) (1 896) – Efectivo y equivalente de efectivo provenientes de (45 577) (2 640) (22 470) (9 389) actividades de ﬁ nanciación Aumento (disminución) neto de efectivo y 17 432 1 860 16 813 10 085 equivalente de efectivo Saldo efectivo y equivalente de efectivo al 33 799 31 939 15 140 5 056 inicio del ejercicio Saldo efectivo y equivalente de efectivo al 51 231 33 799 31 954 15 140 ﬁ nalizar el ejercicio Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 170 A continuación, se presenta el detalle de las ventas de concentrados y de los costos de operación para el periodo 2003-2006. Todas las cifras están en dólares estadounidenses, excepto los volúmenes vendidos, que se encuentran en toneladas métricas secas. A continuación, se presenta el detalle de las ventas de concentrados y de los costos de operación para el periodo 2003-2006. Todas las cifras están en dólares estadounidenses, excepto los volúmenes vendidos, que se encuentran en toneladas métricas secas. Actividades de inversión 2006 2005 2004 2003 Ventas netas Concentrado de zinc 142 109 55 914 – – Concentrado de plomo 41 696 25 425 – – Concentrado de cobre 10 124 5 741 – – Cátodos y sulfato de cobre 85 335 33 387 – – Otros 697 1 593 – – 279 961 122 060 – – Volúmenes vendidos (TMS) Concentrado de plomo 30 878 31 594 25 471 42 287 Concentrado de zinc 142 418 165 925 174 658 168 103 Concentrado de cobre 4 590 5 115 3 978 3 782 Cátodos de cobre y sulfato de cobre 28 324 9 160 10 092 8 521 Costos operativos Compra de concentrados y mineral a terceros 21 351 5 699 7 461 5 878 Suministros 11 981 10 959 10 477 9 061 Gastos de personal 14 413 13 193 11 973 9 905 Compra de cátodos de cobre a terceros 166 767 – – Gastos de producción Contratistas 17 563 10 449 11 471 9 333 Consumo de electricidad 3 856 3 192 3 166 3 613 Mantenimiento y reparación 2 100 1 888 1 520 1 490 Alquiler de maquinaria 1 941 1 439 1 399 1 194 Otros costos operativos 2 957 403 2 118 1 254 76 328 47 989 49 584 41 728 Gastos generales y administrativos Gastos de personal 3 384 2 328 2 401 3 021 Servicios prestados por terceros 1 361 1 039 797 669 Remuneración del directorio y la gerencia 1 591 1 311 631 645 Programa de reestructuración – – 121 105 Tributos 25 27 200 63 Cargas diversas de gestión 506 394 403 814 6 867 5 099 4 551 5 317 Gastos de ventas Fletes 4 619 4 297 3 341 2 750 Gastos de embarque 954 995 891 1 293 Otros 928 595 501 583 6 501 5 887 4 733 4 626 Costos operativos Cuad. Difus. 12 (22), jun. 2007 171 Cía. Minera Milpo S. A. A.: valoración de una empresa minera BASE METAL PRICES ARE HEADED DOWNWARDS March 13, 2007 THE world slowdown has been delayed - not cancelled. That is the message from the metals analysts at BNP Paribas in London and they say - with the exception of tin - base metals are headed for a fall. Headed by ex-Commonwealth Bank commodity strategist David Thurtell, the BNP team believes that the conditions remain in place for a slowdown: real interest rates are rising in the major economies and the US housing sector is in recession. Moreover, ﬁ scal policy has been tightened in two of the larger European economies and a number of central banks have continued to remove their stimulatory monetary stance, most notably the People’s Bank of China, the Bank of Japan, the European Central Bank (ECB) and the Bank of England. “We suspect that activity in Germany will weaken signiﬁ cantly in 2007,” the BNP analysts said in the latest quarterly base metals report, The Base Case. The reasoning behind this is the rise in the euro and the ECB raising interest rates. Adding to world woes is the fall in housing prices in the US, with mortgage defaults rising and the practice of mortgage equity withdrawl (otherwise known as using your house as an ATM) has all but dried up. This is the background to BNP Paribas’ outlook for the metals: This is the background to BNP Paribas’ outlook for the metals: * Aluminium will see a surplus this year with the level of Chinese exports being the key to prices. After being roughly in balance in 2006, global supply is set to exceed demand by between 300,000 tonnes and 400,000 tonnes in 2007, as a result of which the London Metal Exchange price should register modest falls, the bank said. The extent to which the price declines in 2007 depended on the pace of US and European economic growth over the coming six months. But with the decline of net Chinese exports, the LME price was expected to to be underpinned at $US2150/tonne. Firmer world growth and lower Chinese exports could see the aluminium market move back into balance in 2008 and the price rebound to an average $US2250/tonne. * Copper prices have already been hit by reduced Chinese and US use of the red metal and concerns about future supplies have eased considerably. BNP Paribas believes the market will ease further through 2007 and 2008. BASE METAL PRICES ARE HEADED DOWNWARDS March 13, 2007 The bank noted that copper’s fall in the past ﬁ ve months had been as spectacular as its rise. This was due, in part, to the signiﬁ cant positions taken in the market by speculators. The ﬁ rst stage of the sharp decline in copper prices was demand-led; increasing supplies of the metal would drive the next down-leg in the copper price cycle, surpluses being around 200,000 tonnes in 2007 and 2008. Spot copper should average $US5600/tonne in 2007 and $US4650/tonne in 2008 - compared to $US6732 in 2006. * Copper prices have already been hit by reduced Chinese and US use of the red metal and concerns about future supplies have eased considerably. BNP Paribas believes the market will ease further through 2007 and 2008. The bank noted that copper’s fall in the past ﬁ ve months had been as spectacular as its rise. This was due, in part, to the signiﬁ cant positions taken in the market by speculators. The ﬁ rst stage of the sharp decline in copper prices was demand-led; increasing supplies of the metal would drive the next down-leg in the copper price cycle, surpluses being around 200,000 tonnes in 2007 and 2008. Spot copper should average $US5600/tonne in 2007 and $US4650/tonne in 2008 - compared to $US6732 in 2006. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 172 * Nickel supply problems have pushed prices higher but BNP expects prices to weaken. Produc- tion growth looked likely to strengthen in 2008. CVRD’s Onca Puma nickel mine in Brazil’s Amazon may open as much as two months earlier than the targeted November 2008 and it will add 55,000 tonnes a year to world nickel supply. PT Aneka Tambang, Indonesia’s second largest nickel miner, aims to lift production by 22,000 tonnes in 2008. Nickel prices should average a still very high $US33,000/tonne in 2007 but fall to $US20,500/tonne in 2008. * The outlook for zinc will be affected by China turning exporter and the bank believes the metal’s market will ease signiﬁ cantly in the second half of this year, allowing the supply of zinc to progressively catch up with demand as 2007 progresses. Prices should average $US3325/tonne this year, $US2700/tonne in 2008. * While lead supplies have seen the market tighten and demand remaining solid, especially in China, the bank sees a signiﬁ cant easing in this metal’s price in 2007. BASE METAL PRICES ARE HEADED DOWNWARDS March 13, 2007 At around $US1820/tonne last week, the price was double that recorded in November 2005. (However, LME prices rose overnight, lead being up 8 per cent to $US1915/tonne on news that Toronto-based Ivernia announced that shipments of lead concentrate from its Magellan mine in Western Australia had been halted, pending the outcome of an investigation into the recent death of several birds in Western Australia, possibly from lead poisoning. Magellan produced 63,200 tonnes of lead metal in concentrate last year.) BNP Paribas predicts that the lead market will ease signiﬁ cantly in the second half of 2007 as a wave of new lead-zinc mines provides more concentrate feed. The spot price should average $US1600/tonne in 2007 and $US1310/tonne in 2008. * But tin is the odd man out, with the bank expecting prices to push still higher. The price has gone from $US10,150/tonne in November to $US13,700/tonne overnight due to Indonesia clamping down on illegal production and concerns about supply out of Bolivia. The bank believes that the clamp-down in Indonesia will underpin tin prices at relatively high levels over the next two years, and growth in Chinese output this year will be insufﬁ cient to make up the gap. With stocks falling, the LME spot should push to $US15,000 by mid-2007, but ease next year. ANNOUNCED TODAY * Elkedra Diamonds (EDN) said it will spin-off its uranium and base metals prospects into a new company. The so far unnamed new ﬁ rm will hold 8000sq km of exploration licences, mostly in the Northern Territory. * Gold and base metals explorer Montezuma Mining Co (MZM) has decided it does not want to be left out of the uranium hayride. The company has taken a 70 per cent interest in the Robinson Range uranium project in a deal with Greater Paciﬁ c Gold (GPN). The project is located 125km north of Meekatharra in Western Australia. * West Australian Metals (WME) has begun drilling at its Marenica uranium project in Namibia. Meanwhile, the company has been an additional exploration licence at the Scaddan uranium project in Western Australia, bringing the area 75km northeast of Esperance to 300sq km. * West Australian Metals (WME) has begun drilling at its Marenica uranium project in Namibia. Meanwhile, the company has been an additional exploration licence at the Scaddan uranium project in Western Australia, bringing the area 75km northeast of Esperance to 300sq km. NSW GAS DEPENDENCE END IN SIGHT COAL seam gas, rather than the coal itself, is now being investigated as a future source of power generation feed in NSW. But the move by Eastern Star Gas (ESG) and government-owned utility Macquarie Generation has much wider ramiﬁ cations than just feedstock for one power station. ESG claims that the days of NSW being totally dependent on gas from other parts of Australia - namely the Cooper Basin and Bass Strait - may soon be over. The junior and its partner, Houston-based Gastar, have signed a memorandum of understanding with Macquarie to look at piping gas to the Bayswater power station in the Hunter Valley. The two gas explorers are developing a coal seam gas ﬁ eld in the Gunnedah Basin, lying east of the towns of Gunnedah and Narrabri. The utility, with generating capacity of 2640 megawatts, is now coal-fueled but the plan is for Macquarie to use coal seam gas to reheat the steam from the existing generators and then pump the recycled energy through a new set of generators to expand power output. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 173 ESH managing director Dennis Morton said the company plans to build a pipeline the 250km to Bayswater and possibly the 370km to Newcastle. He believes there is enough gas in the coal seams at Gunnedah to not only supply the power utility but also the Sydney-Newcastle-Wollongong market. http://www.thebulliondesk.com Monday, April 2, 2007 9:14 GMT Daily Report By James Moore The Bullion Desk http://www.asiamoney.com/default.asp?Page=7&PUBID=185&ISS=23606&SID=679650 http://www.asiamoney.com/default.asp?Page=7&PUBID=185&ISS=23606&SID=679650 Research report Source: asiamoney Date: 14th March 2007 Research report Source: asiamoney Date: 14th March 2007 GOLD SUPPORTED BY OIL/IRAN BUT VULNERABLE TO CORRECTION Despite volatility in other markets gold remained conﬁ ned to a $7 range Friday as month end posi- tioning kept the market capped while the heightened tensions in the Middle east kept the market underpinned. In the currencies the greenback ﬁ nished lower across the board as news of import tariffs negated the impact of strong PCE and manufacturing data. EUR/USD settled at 1.3355 having touched a high of 1.3401, while the Yen closed at 117.78. Economic data today will show the ISM Index for March, forecast at 51.5%, while the line-up for the rest of the week includes the ISM Service Index Wednesday and Non-farm payrolls Friday. Tensions between Iran and west continued to boost oil prices with NYMEX crude trading to $66.78 before settling at $65.87/barrel. Gold’s dip Thursday led to some light proﬁ t taking as Friday’s session got underway in Asia howe- ver the yellow metal remained underpinned as the ongoing discussions between London and Tehran, and rising energy prices prompted safe-haven buying interest. Gold posted the days low at $660.70 in Asia and tracked largely sideways across Asia, brieﬂ y touching $665 on the US opening before ﬁ rm US data led gold back towards the days low. Bargain hunter and physical players again provided support with gold spiking a little later as the dollar fell sharply as the US announced import duties on goods from China. Gold jumped to a high of $667.70 but quickly ran into proﬁ t taking, trading out the day between $662-65. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 174 Gold closed the ﬁ rst quarter at $663, posting a gain of $6.20 on the week and $27.30 on the quarter, but while the metal has lacked upside momentum over the past few weeks we are now moving into what is traditionally the strongest period for physical demand, with the start of the monsoon/wedding season later in the month. Further resistance is expected between $670-72 however successful clearance should enable gold to challenge $675-78 before targeting February’s high around $689. CFTC data Friday showed a modest increase in the COMEX/CBOT non-commercial net long, rising 12.862- tonnes to 318.995-t. GOLD SUPPORTED BY OIL/IRAN BUT VULNERABLE TO CORRECTION Gold closed the ﬁ rst quarter at $663, posting a gain of $6.20 on the week and $27.30 on the quarter, but while the metal has lacked upside momentum over the past few weeks we are now moving into what is traditionally the strongest period for physical demand, with the start of the monsoon/wedding season later in the month. Further resistance is expected between $670-72 however successful clearance should enable gold to challenge $675-78 before targeting February’s high around $689. CFTC data Friday showed a modest increase in the COMEX/CBOT non-commercial net long, rising 12.862- tonnes to 318.995-t. The bullish mood in copper and modest gains in gold led silver to ﬁ nish Friday up 10-cents at $13.37, posting a gain of 24-cent on the week and 53-cents on the quarter. Silver is still lacking direction of its own for the moment as traders look to gold and the base metals. $13.45 chart resistance remains a sticking point and with technical indicators pointing lower there is the risk silver may test back to $12.80 before $14 is re-tested. The non-commercial net long increased 2.57M/oz in the week of March 27th to 160.39M/ozs. Platinum rallied back to challenge the $1245 chart resistance Friday, setting the days high of $1246 on the AM ﬁ x before settling at $1242. Platinum continues to post solid gains as the metal remains buoyed by its tight supply/demand balance. The white metal posted a 9.5% gains across Q1 and may look to challenge February’s $1259 high once $1245 is cleared. Friday’s Commitment of Traders report reﬂ ected the modest fund proﬁ t taking seen across the past two-weeks; however the metal remains comfortable straddling $350, while chart indicators suggest a move back to $360. The non-commercial net long fell 13,700/ozs to 665k/ozs. Ghee Peh, Regional base metals and mining analyst, UBS securities hee Peh, Regional base metals and mining analyst, UBS securities With the growth of commodities as an asset class and the increase in investment ﬂ ows, one should expect more volatility. The base metal and gold price moves for the ﬁ rst two months of this year have demonstrated this. We believe that the much misunderstood and low-proﬁ le coal sector is the chief investment opportunity for 2007. A tenet of commodities is that if supply does not match demand then prices will go up. The role of speculators on the demand and supply sides and of substitute commodities can go some way to mitigate any upside price moves. In our view, there is an absence of speculation and economic substitutes for coal, so the tenet applies. Cuad. Difus. 12 (22), jun. 2007 Cuad. Difus. 12 (22), jun. 2007 Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 175 We have had a positive outlook both on the regional and China thermal coal price, and recent events have conﬁ rmed this. The 2007 Australia-Japan coal benchmark has been settled at US$55.5/tonne, up US$3 year-on-year, while in January 2007 China turned net importer of coal for the ﬁ rst time with exports of 3.3 metric tonnes, (down 20% year-on-year), but imports of 4.7 metric tonnes (up 81% year-on-year). As recently as 2003, China was the world’s second largest coal exporter. What happened? Applying the above tenet, China coal supply is not keeping up with demand. The Chinese coal supply faces four challenges: mining, with the majority of mines deep underground; geological, with the 55% lower energy value of coal; infrastructure, with coal located in the northwest and southwest; and policy, with authorities looking to impose new safety and environment related charges on producers. The next step is to ask if there are any factors that could explain the 10% rise in China spot coal price since November. Coal is not a commodity that facilitates the use of derivatives for speculation. On the issue of substitutes, because coal is much cheaper than diesel and natural gas in generating electricity there is some room for a coal price rise before other energy alternatives become viable. We expect that China will continue importing coal and reducing exports. This should have a beneﬁ cial impact on the regional coal price. We note that Tata Power has bid for Indonesian coal assets. Coal may not keep its low proﬁ le for long. Michael Lewis, MD and global head of commodities research, Deutsche Bank For many commodities, the start of the current bull run dates back to November 2001. As a result, this bull cycle has been one of the most powerful and durable on record. However, like the popular English nursery rhyme, successive bottles to this rally appear to be falling one by one. This is increasing market concern that nominal and real commodity price appre- ciation has now been exhausted. We are maintaining our bullish assessment towards the grains and precious metals complex. We also expect crude oil as well as industrial metal prices to recover into the second quarter. • Oil: After the weather induced sell-off in crude oil, we expect oil prices to recover on the back of strong world gross domestic product growth and further downgrades to non-OPEC oil production growth. • Grains: Tightening inventories, rising use of corn for ethanol production, China's increasing appetite for cattle feed and the possibility of La Niña sustaining droughts across the Americas poses further upside price risks this year. • Industrial metals: This was the best performing commodity sector last year. We continue to ascribe to a structural shift in long-term prices and consequently expect the back-ends of these markets to remain underpinned. • Precious metals: The risk of a rebound in the US dollar is the main event risk this year. Even so, we still expect producer de-hedging and rising fabrication demand to push gold prices and implied volatility higher into 2008. • US natural gas: This was the worst performing commodity in terms of spot price performance last year. However, we expect that disappointing US gas production growth will come to the rescue and push prices higher. • Commodity indices: We expect another year of divergent commodity index returns in 2007. We are maintaining our long Deutsche Bank Commodities Index (DBCI)-Mean Reversion vs. short Goldman Sachs Commodity Index as well as our long DBLCI-Optimum Yield vs. short DBLCI trading recommendations. Tobias Merath, commodity analyst, Credit Suisse With the correction in commodity prices last year and the current recovery, volatility in the markets has further intensiﬁ ed. Commodity price volatility has been rising since 2002, and the Dow Jones AIG Commodity Index now exhibits volatility around 20%. So the pattern on the commodity mar- kets contrasts with other asset classes, such as equities or bonds, where volatility has followed a downward trend. This is quite pronounced in the stock markets. The volatility of the MSCI World Index of 7% exceeds that of the JPMorgan World Bond Index by just 200 basis points. The drop in volatility is attributable to numerous factors. Academic literature points to the reduced variability of interest and inﬂ ation rates. Robust growth and expanding investment in infrastructure have led to a tight supply-side situation on many markets. This has triggered movements in prices. At the same time, the reduced variability of interest rates underpins economic growth and should also lead to elevated volatility in commodity markets in coming months. But large price movements spark a greater need for hedging on the producer, as well as the con- sumer, side. Investors are thus in a position to fulﬁ l this growing need and, in this way, are also able to earn a risk premium. There are still investment opportunities in the commodity universe. We expect the price of gold to increase above US$700. An investment in selected base metals should also become attractive this year, when global economic growth picks up. Low inventories and demand from China should lead to higher prices for copper and zinc. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 176 http://www.moodys.com/moodys/cust/displaySummary.asp?busLineId=12&original= 1&document_id=1506300000004913 http://www.moodys.com/moodys/cust/displaySummary.asp?busLineId=12&original= 1&document_id=1506300000004913 Francisco Blanch, Head of Global Commodity Research, Merrill Lynch The current economic expansion has been one of the broadest, longest, strongest and least volatile in history, with annual global GDP growth rates ranging from 4.1% to 5.3% in the last four years. This has lent signiﬁ cant support to base metals prices. Looking forward, we believe the strong economic momentum will likely continue over the next few quarters. In the near-term, ﬁ rm Chinese domestic consumption, a stabilisation in the US housing market and an end to inventory adjustment in the US auto and electronics sector will further support base metal prices. Moreover, copper, zinc and nickel inventories are still very low, so small changes to the underlying supply and demand conditions could result in signiﬁ cant price swings, suggesting that prices will remain extremely volatile. Cuad. Difus. 12 (22), jun. 2007 Cía. Minera Milpo S. A. A.: valoración de una empresa minera 177 We expect aluminium prices to remain strong in the near-term, supported by low global inventories in terms of forward demand, falling Chinese exports and strengthening global demand. However, the combination of lower alumina and energy prices together with a build-up in smelter capacity around the world suggests that sharp price rises from current levels are unlikely. While the copper market should be in surplus this year, copper prices are likely to be driven higher over coming months by the strong growth in Chinese consumption and the bottoming out of the US housing market downturn. Fundamentals for zinc are strong and we expect prices to continue to rise in 2007, while nickel has weak short-term, but strong long-term fundamentals. In the medium-term, our view is that base metals could resume super-cyclical price adjustments beyond 2007. During a super cycle, prices spike in order to choke off soaring demand in the face of extremely low inventories and a limited capacity cushion, typically the result of years of underinvest- ment in supply infrastructure. The combination of a less volatile global macro economy, strong commodity-intensive emerging markets growth, and a rapidly expanding global urban population, could add upward pressure to base metal prices in 2008. Should global economic growth remain at elevated rates of 4% to 5%, demand will likely catch up again quickly with supply across many of the industrial metals markets. PERU UPGRADE REVIEW AT BA3 BONDS, BA2 CEILING PERU UPGRADE REVIEW AT BA3 BONDS, BA2 CEILING PRESS RELEASE New York, March 08, 2007 – Moody’s Investors Service has placed on review for possible upgrade Peru’s foreign currency ratings. The review will examine the extent to which reduced external credit vulnerabilities and the presence of lower external debt indicators have led to a sustained improve- ment in Peru’s relative credit standing, particularly under medium-term scenarios that incorporate less favorable international economic and ﬁ nancial conditions. The review includes Peru’s Ba2 foreign currency country ceiling for bonds and the government’s Ba3 foreign-currency bond rating. The country bond ceiling is based on the government bond rating and Moody’s assessment of a moderate risk of a payments moratorium in the event of a government bond default. The country’s B1 ceiling for foreign currency bank deposits also has been placed on review for an upgrade. The Baa1 local currency deposit ceiling and the A3 local currency bond ceiling - the highest possible rating that could be assigned to obligors and obligations denominated in local currency within the country - are not on review. Cuad. Difus. 12 (22), jun. 2007 Alfredo Mendiola 178 “Peru’s external credit indicators have improved signiﬁ cantly, as evidenced by a continuous decline in external debt ratios, a strengthened international reserve position, and robust export growth,” said Moody’s Vice President Mauro Leos. “Peru’s external credit indicators are improving and converging towards the mean values for Ba-rated countries,” noted Leos. Leos said that as part of the review process, Moody’s will evaluate the government’s ability to manage adverse external shocks to public ﬁ nances and the external accounts, as well as those resulting from socio-political constraints present in Peru. Risks from a less benign international environment could stem from (i) a sustained reduction in commodity prices, metals prices in particular; (ii) a deceleration in world economic growth; and (iii) increased ﬁ nancial volatility. Because of credit risks related to Peru’s high share of foreign currency-denominated government debt, the review will also include an assessment of efforts by the authorities to address this situation. The review will evaluate the anticipated changes in the currency composition of government debt and the likely implications for the government’s credit risk proﬁ le. PRESS RELEASE At the same time, the review will evaluate the anticipated impact on Peru’s external and ﬁ scal accounts of upcoming projects in the mining and energy sectors that, once operational, are expected to have a positive inﬂ uence on medium-term growth and export prospects. Lastly, given the presence of a banking system characterized by a relatively high, albeit declining, degree of ﬁ nancial dollarization, the review will evaluate the authorities’ ability to provide support if a ﬁ nancial stress scenario were to materialize. Cuad. Difus. 12 (22), jun. 2007
https://openalex.org/W3038105848	https://www.revhosp.org/hospitalidade/article/download/877/pdf	Portuguese	null	Visitar/Acolher: Arquitetura, Turismo e Encontros	Revista Hospitalidade	2,020	cc-by	6,967	Visitar/Acolher: Arquitetura, Turismo e Encontros Visiting/Sheltering: Architecture, Tourism and Encounters Visitar/Acoger: Arquitectura, Turismo y Encuentros Felipe Loureiro1 Roberto Bartholo 2 Fernanda Barcelos 3 Flávia Mattos 4 Resumo: O artigo desenvolve uma abordagem teórica que compreende a produção arquitetônica como uma produção de presença (GUMBRECHT, 2010), com implicações tanto para a visitação quanto para a acolhida. A possibilidade de diálogos e encontros em meio a práticas turísticas é pensada a partir da perspectiva apresentada por Buber no clássico “Eu e Tu” (1977). A formatação típica do que Pedro Abreu chama de “turismo cultural híper- moderno” (ABREU & MALHEIRO, 2011) busca poupar do turista o “trabalho” de interagir com as presenças que configuram e habitam o sítio visitado, restringindo sua experiência à dimensão de um entretenimento espetacularizado. Para que a experiência do turista no contexto do turismo cultural híper-moderno possa ser transformadora, é preciso afirmá-la em sua inteireza como uma experiência estética e sinestésica que acontece como um encontro face a face entre um “Eu” e um “Tu”. Palavras-Chave: Turismo, Arquitetura, Presença, Entretenimento, Arte. Abstract: The paper develops a theoretical approach based on the notion that the production of architecture is a production of presence (GUMBRECHT, 2010), concluding with possible consequences of this approach to the study of touristic experience. The possibilities for dialogues and encounters amid and around the touristic experience is discussed according to the perspective proposed by Buber in “I and Thou” (1977). The typical structure of what Pedro Abreu calls “hyper-modern cultural tourism” (ABREU & MALHEIRO, 2011) tries to save tourists from the “trouble” of interacting with the presences that configure the visited site, limiting the experience to the dimension of entertainment. The paper aims at proposing that, even in the context of hyper-modern cultural tourism, touristic experiences can still be as transformative as the experience of a work of art – it is only necessary to preserve the possibility of encounters between an “I” and a “Thou”. Key words: Tourism, Architecture, Presence, Entertainment, Art. Resumen: El artículo desarrolla un enfoque teórico que entiende la producción arquitectónica como una producción de presencia, concluyendo con las posibles consecuencias de este enfoque para la experiencia turística. Las posibilidades de diálogos y encuentros en la experiencia turística son pensadas a partir de la perspectiva propuesta por Buber en “Yo y Tú” (1977). g 4 Universidade Federal do Rio de Janeiro. ORCID: 0000-0002-4903-7889. E-mail: flaviamattosbr@gmail g 3 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-2845-8219. E-mail: ftbarcelos@gmail.com 1 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-4515-0304. E-mail: loureiro.fgsf@gmail.com. 2 95 1 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-4515-0304. E-mail: loureiro.fgsf@gmail.com. 2 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-2258-2198. E-mail: bartholo.roberto@gmail.com. 3 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-2845-8219. E-mail: ftbarcelos@gmail.com. 4 Universidade Federal do Rio de Janeiro. ORCID: 0000-0002-4903-7889. E-mail: flaviamattosbr@gmail.com. g al do Rio de Janeiro. ORCID: 0000-0002-4903-7889. E-mail: flaviamattosbr@gmail.com. Volume 17, n. 02, (mai-ago) de 2020 ISSN 1807-975X 1 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-4515-0304. E-mail: loureiro.fgsf@gmail.com. 2 Universidade Federal do Rio de Janeiro. ORCID: 0000-0003-2258-2198. E-mail: bartholo.roberto@gmail.com. 3 Universidade Federal do Rio de Janeiro ORCID: 0000 0003 2845 8219 E mail: ftbarcelos@gmail com LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 ISSN 1807-975X Visitar/Acolher: Arquitetura, Turismo e Encontros Visiting/Sheltering: Architecture, Tourism and Encounters Visitar/Acoger: Arquitectura, Turismo y Encuentros El formato típico de lo que Pedro Abreu llama "turismo cultural hipermoderno" (ABREU & MALHEIRO, 2011) busca ahorrar a los turistas el "trabajo" de interactuar con las presencias que dan forma y habitan el lugar visitado, limitando la experiencia a la dimensión del entretenimiento. La propuesta que concluye el artículo parte de la idea de que, incluso en el contexto del turismo cultural hipermoderno, la experiencia del turismo puede ser tan transformadora como la experiencia de una obra de arte: es suficiente para preservar la posibilidad de encuentros. entre un "yo" y un "tú". Palabras clave: Turismo, Arquitectura, Presencia, Entretenimiento, Arte. 95 ISSN 1807-975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 detalhes são programados antecipadamente, configura um novo tipo de turismo – o “turismo cultural híper-moderno” (ABREU & MALHEIRO, 2011). O turista segue roteiros pré- determinados, sendo poupado do “trabalho” de interagir com as presenças que configuram e habitam o lugar visitado. Neste tipo de visitação, é pouco provável que ocorram encontros e diálogos buberianos. Assim, a experiência turística restringe-se ao entretenimento consumista, numa suspensão temporária da rotina cotidiana, através da distração. Ainda assim, sendo o diálogo buberiano um acontecimento e não uma produção - ou seja, sendo impossível capturá-lo em causalidades de cunho determinista -, pode ser que existam, mesmo no contexto adverso do turismo cultural híper-moderno, tempo e espaço para a imprevisibilidade da ocorrência surpreendente deste tipo de interação. Mas, para isso, é preciso ultrapassar os limites da programação das experiências. 1 O que virá Em um artigo publicado em 2001 com o título “Ir ao Mesmo Lugar”, Umberto Eco (2017) afirma que: (...) o turismo representa para muitos um modo de se reapropriar do mundo. Só que antes a experiência da viagem era decisiva, voltávamos diferentes do que éramos ao partir, enquanto agora só se encontra gente que volta sem ter sido tocada nem minimamente pela fascinação do Outro Lugar. Retornam e só pensam nas próximas férias, não falam de nenhuma iluminação transformadora. Esta observação, feita num artigo de jornal “não acadêmico”, não serve como diagnóstico científico das práticas turísticas contemporâneas, e tampouco oferece o autor respaldo bibliográfico ou estatístico para tal proposição. De qualquer forma, artistas são antenas sensíveis para as transformações culturais do seu tempo, e, assim, Eco nos provoca a colocar questões muito mais do que nos oferece respostas. Neste pequeno texto, o autor sugere alguns fatores que podem ser determinantes para as transformações pelas quais tem passado as práticas turísticas contemporâneas - nas quais a possibilidade de experiências interpessoais transformadoras estaria sendo reduzida em prol de uma maior aproximação a distrações consumistas espetacularizadas. Este artigo busca reagir à provocação colocada por Eco, e esta reação parte da perspectiva introduzida por Martin Buber no seu livro clássico "Eu e Tu" (original de 1923, citado neste artigo a partir da edição brasileira de 1977). Lugares visitados "distraidamente", sem que a visitação aconteça como um encontro que provoca relação vinculante, são tratados como um "Isso", e não como um "Tu". Isso implica a inexistência de relações dialogais e o predomínio de interações funcionais no fenômeno da visitação. A objetificação do lugar – seja um edifício, um monumento ou uma cidade – bloqueia a abertura do visitante para a possibilidade de experiências transformadoras associadas a interferências dialogais recíprocas. A arquitetura é um ato potencialmente criador de presenças, tanto a do edifício construído como a de número indeterminado de presenças que encontrem ali abrigo (habitantes, visitantes etc.). As possibilidades dialogais associadas ao ato arquitetônico são decorrentes da abertura relacional aos encontros com estas presenças. Assim, o acontecimento de uma “iluminação transformadora” tal como referida por Eco implica diálogo, abertura à alteridade, vulnerabilidade à interferência recíproca de presenças. Para Pedro Abreu, a dinâmica das visitas curtas e aceleradas, nas quais mesmo os menore 96 O turismo tem por antecedentes recentes a Grand Tour - viagem transeuropeia (mas com especial ênfase em Roma) com que os jovens iluministas endinheirados, dos séculos XVIII, e XIX completavam a sua educação. A partir do fim da Segunda Guerra Mundial o turismo teve um crescimento exponencial, sobretudo devido à facilitação das condições laborais, com a generalização, na Europa Ocidental, do período de férias pago. Este era, contudo, um turismo generalista com predominância estival e balnear, sem efeitos notórios na arquitectura. Desde finais dos anos 70 e sobretudo nos anos 80 e 90 - nos tempos da civilização pós- moderna - dá-se uma alteração, significativa para as cidades: o turismo cultural emerge como um turismo de massas; aos centros históricos, aos grandes monumentos do património mundial, acorre já não apenas uma elite culta, mas população de todos os âmbitos socioculturais. A segunda fase do turismo cultural, sobretudo enquanto turismo de cidades e/ou de monumentos, suscitará múltiplas complicações. A mais significativa, porventura, será a transformação a que o turismo de massas irá forçar o próprio turismo cultural: dando início a uma terceira fase - que, talvez, recorrendo a Lipovetsky, pudéssemos denominar de “híper-moderna”. Desta fase, emergente a partir de finais dos anos 90 e que nos é LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 ISSN 1807-975X A observação de Eco citada na Introdução claramente se refere ao “turismo de cidades e/ou de monumentos”, ou seja, ao turismo cultural. É deste tipo de turismo (e não de experiências turísticas mais próximas do turismo “estival e balnear”, por exemplo) que ele esperava ouvir relatos de “iluminações transformadoras”. Eco lamenta o fato de que as cidades e monumentos parecem estar perdendo seu principal apelo – e, para o autor, um dos fatores que parecem contribuir para a aparente indiferença dos turistas contemporâneos em relação a estas “atrações” é justamente a padronização provocada pela massificação do turismo cultural: Penso em alguns lugares mágicos de Paris, como Saint-Germain, onde pouco a pouco estão desaparecendo os velhos restaurantes, as livrarias à meia-luz, as lojinhas dos velhos artesãos, substituídos por lojas de estilistas internacionais, que por sua vez são as mesmas que podemos encontrar na Fifth Avenue em Nova York, em Londres, em Milão. (...) até os grandes monumentos correm o risco de ficar cada vez mais parecidos (pelo menos aos olhos dos turistas), pois estão se transformando em mero suporte para jogos de luz de estilo internacional (ECO, 2017). Penso em alguns lugares mágicos de Paris, como Saint-Germain, onde pouco a pouco estão desaparecendo os velhos restaurantes, as livrarias à meia-luz, as lojinhas dos velhos artesãos, substituídos por lojas de estilistas internacionais, que por sua vez são as mesmas que podemos encontrar na Fifth Avenue em Nova York, em Londres, em Milão. (...) até os grandes monumentos correm o risco de ficar cada vez mais parecidos (pelo menos aos olhos dos turistas), pois estão se transformando em mero suporte para jogos de luz de estilo internacional (ECO, 2017). Esta padronização se dá pela inserção destas cidades e monumentos em uma lógica que pertence mais à indústria do entretenimento que ao turismo cultural – ao menos à segunda fase identificada por Abreu & Malheiro. Esta padronização já fora prevista por Guy Debord, que usou o termo “sociedade do espetáculo” para identificar este contexto sociocultural. 2 De onde viemos Buscamos estudar a transformação de práticas turísticas traçando a emergência de um turismo cultural híper-moderno. Para isso, nos apoiamos na filosofia relacional de Martin Buber, ousando propor uma transposição dos conceitos buberianos para o campo da arquitetura. A pesquisa acerca do turismo cultural foi norteada pelo artigo de Abreu e Malheiro (2011), segundo o qual a evolução do turismo cultural pode ser dividida nas seguintes fases: O turismo tem por antecedentes recentes a Grand Tour - viagem transeuropeia (mas com especial ênfase em Roma) com que os jovens iluministas endinheirados, dos séculos XVIII, e XIX completavam a sua educação. A segunda fase do turismo cultural, sobretudo enquanto turismo de cidades e/ou de monumentos, suscitará múltiplas complicações. A mais significativa, porventura, será a transformação a que o turismo de massas irá forçar o próprio turismo cultural: dando início a uma terceira fase - que, talvez, recorrendo a Lipovetsky, pudéssemos denominar de “híper-moderna”. Desta fase, emergente a partir de finais dos anos 90 e que nos é hoje asfixiadoramente contemporânea, não se conseguiu ainda bem desvendar as consequências. 97 Para Debord (2005), a padronização das cidades – e, consequentemente, da experiência de visitar estas cidades - é uma consequência inevitável da incorporação, no turismo, de uma lógica de consumo: Subproduto da circulação das mercadorias, a circulação humana considerada como um consumo, o turismo, reduz-se fundamentalmente à distracção de ir ver o que se tornou banal. A ordenação económica da frequentação de lugares diferentes é já por si mesma a garantia da sua equivalência. A mesma modernização que retirou da viagem o tempo, retirou-lhe também a realidade do espaço (DEBORD, 2005, p. 121). 98 A transformação da circulação em consumo se espalha, portanto, para o turismo cultural – até então, um tipo de “circulação” que “manifestava uma sadia curiosidade pelos lugares.” (ABREU & MALHEIRO, 2011). Esta curiosidade que contaminava os viajantes fazia parte do ambiente cultural que os cercava – “O apelo do desconhecido, a fome de saber, forneciam suporte à dinâmica deste turismo cultural e nele encontravam resposta, porquanto ele se correlacionava com várias exigências antropológicas.” (ABREU & MALHEIRO, 2011). No contexto da 98 O mundo é duplo para o homem, segundo a dualidade de sua atitude. A atitude do homem é dupla de acordo com a dualidade das palavras-princípio que ele pode proferir. As palavras-princípio não são vocábulos isolados, mas pares de vocábulos. Uma palavra-princípio é o par EU-TU. A outra é o par EU-ISSO no qual, sem que seja alterada a palavra-princípio, pode-se substituir ISSO por ELE OU ELA. Deste modo, o EU do homem é também duplo. Pois, o EU da palavra-princípio EU-TU é diferente daquele da palavra-princípio, EU- ISSO (BUBER, 1977, p. 29). ISSN 1807-975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 de algo, mesmo que esse algo seja o outro, pertence ao âmbito da relação Eu-Isso.” (BARTHOLO, 2009). Ou seja, em uma verdadeira relação do tipo Eu-Tu, nos relacionamos com o Outro real, presente, e não com uma idéia que temos sobre ele – “O Eu não constrói uma representação do Tu, e sim o encontra. O intervalo onde se dá a relação face a face é o entre (Zwischen), e não é concebível como um espaço vazio, independente.” (BARTHOLO, 2009). de algo, mesmo que esse algo seja o outro, pertence ao âmbito da relação Eu-Isso.” (BARTHOLO, 2009). Ou seja, em uma verdadeira relação do tipo Eu-Tu, nos relacionamos com o Outro real, presente, e não com uma idéia que temos sobre ele – “O Eu não constrói uma representação do Tu, e sim o encontra. O intervalo onde se dá a relação face a face é o entre (Zwischen), e não é concebível como um espaço vazio, independente.” (BARTHOLO, 2009). O surgimento deste intervalo, deste espaço interpessoal, é a própria hospitalidade, e "requer moradia e memória: um lar aberto ao outro" (BARTHOLO, 2015, p. 184). As referências espaciais presentes nestas afirmações levaram, naturalmente, a uma pesquisa acerca da transposição desta abordagem conceitual para o domínio da arquitetura – tanto no nível da concepção arquitetônica como no nível da percepção do espaço em geral, incluindo o ambiente construído e as paisagens naturais. As principais referências encontradas nesta pesquisa foram a tese de doutorado do Prof. Pedro Marques de Abreu (2007) - na qual o autor apresenta uma teoria sobre a percepção da arquitetura e das obras de arte como um todo, usando como ferramenta de análise um processo de “leitura” de monumentos -, e o livro “A History of Architecture: Settings and Rituals”, de Spiro Kostof. A contribuição destes textos para este artigo será apresentada na seção seguinte, já que o método que orientou o desenvolvimento deste trabalho consistiu em uma reflexão de caráter filosófico sobre o cruzamento entre os conceitos apresentados acerca do turismo cultural e da filosofia de Buber no contexto da arquitetura. 100 5 No original: “the I- You relation enters essentially into every aspect of the moral life” (Scruton, 2017, p. 51). ISSN 1807-975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 sociedade de consumo – ou sociedade do espetáculo, como aponta Debord -, a curiosidade do Grand Tour é substituída pela “obrigação” de se ocupar o tempo livre – o tempo de consumo, que complementa e alimenta o tempo de produção – com visitas a lugares que hão de se tornar cada vez mais similares. É claro que, neste contexto, a experiência do lugar será quase que necessariamente superficial – o lugar será muito provavelmente tratado como um “Isso”. Assim, a conclusão aponta para os conceitos apresentados por Martin Buber: De acordo com perspectiva buberiana, podemos dizer que experimentamos o “Isso”, enquanto encontramos o “Tu”; e que o “Eu” que experimenta não é o mesmo “Eu” que encontra. O mundo do Isso é obviamente necessário, e tem seu próprio valor: “O mundo do Isso abrange todo o espaço de experiência humana com objetos de conhecimento objetivo, manejo operativo prático e apropriação utilitária” (BARTHOLO, 2009). Porém, o tipo de contato com um Isso é necessariamente instrumental. Não se trata de uma relação entre iguais, mas sim um relação necessariamente hierarquizada, na qual alguém usa e alguém é usado: “Ao ente issificado se imputa o papel de servir como um anônimo artigo de troca, que se pode experienciar, analisar e instrumentalizar, mas com quem não se estabelece uma verdadeira relação vinculante” (BARTHOLO, 2009). Por outro lado, a relação do tipo Eu-Tu pode ser vinculante justamente por ser, essencialmente, uma relação entre iguais. E, além disso, estes iguais estão igualmente abertos à influência do Outro - “a relação Eu-Tu pressupõe a confrontação imediata, face a face, com um ente exterior que é radicalmente um outro, e em tanto que tal percebido na relação.” (BARTHOLO, 2009). Nesta relação de abertura mútua, é necessário suspender preconcepções que poderiam influenciar ou mesmo impedir a abertura para o outro, uma vez que “Ter uma idéia 99 ISSN 1807 975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 direitos e deveres e aos padrões de reciprocidade nas relações (SCRUTON, 2017). Resumidamente, podemos dizer que, se tenho direito a algo, alguém tem o dever de oferecer este algo a mim. Logo, a própria noção de uma sociedade baseada em direitos e deveres pressupõe a centralidade de relações do tipo Eu-Tu – o que Darwall chama de “ponto de vista da segunda pessoa” (DARWALL, 2006)6. Mas a dinâmica da relação Eu-Tu não está limitada ao âmbito estrito das relações inter- humanas, tal como reconhece, originalmente, o próprio Buber (1977). Esta dinâmica pode dizer respeito a todo o âmbito da gestão e política do patrimônio histórico e cultural - construído ou natural, material ou imaterial. Afinal, pode-se experimentar uma obra – seja ela um edifício, uma pintura, uma sinfonia etc. – como um Isso ou como um Tu (ABREU, 2007). Esta possibilidade em aberto tem profundas implicações para a concepção e percepção de obras de arte, e também para as práticas e experiências turísticas – principalmente, mas não exclusivamente, no âmbito do turismo cultural. O ato arquitetônico é criador de novas presenças: em primeiro lugar, é criada a presença da própria edificação em si - seja ela uma casa, um estádio, uma praça pública ou um mero poste -, mas o surgimento desta presença também traz consigo a possibilidade de fazer presentes outras (inclusive imprevistas) presenças. Isto fica evidente se considerarmos uma casa ou um edifício de escritórios: em ambos, temos cômodos ou espaços a serem ocupados por pessoas que vão dormir, comer, trabalhar, se relacionar com outras pessoas, etc.; ou seja, os espaços criados pela presença da obra de arquitetura permitem que pessoas e “coisas” se tornem presentes. Porém, o mesmo fenômeno ocorre com construções bem mais simples. Em alguns contextos, um poste pode se tornar um ponto de referência “no qual” – ou ao redor do qual – pessoas se encontram ou se concentram, e pode inclusive adquirir uma relevância simbólica e cultural que o aproxima de construções mais complexas. Tanto na escala da casa como no caso do poste, há um espaço criado pela construção. 6 No original: “second-person standpoint” (Darwall, 2006). 3 Percurso O exercício que norteou este texto foi pensar novos domínios a partir da dualidade buberiana Eu-Tu / Eu-Isso. Não pretendemos aqui ineditismo, uma vez que esta iniciativa já foi empreendida por outros antecedentemente, tais como os filósofos Stephen Darwall e Roger Scruton, para quem “a relação Eu-Tu entra essencialmente em todos os aspectos da vida moral” (SCRUTON, 2017), p. 51)5. A extensão da perspectiva buberiana para o domínio ético-moral, onde ganham realce questões como responsabilidade e liberdade, foi empreendida por Darwall (2006), dando destaque aos processos de trocas e negociações interpessoais associados aos 100 ) 100 ISSN 1807-975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 Cruzando as noções introduzidas por Buber e Abreu, e tomando uma expressão usada por Gumbrecht (2010), concluímos que a essência da arquitetura é a “produção de presença”. A arquitetura é, portanto, condição de possibilidade para o surgimento – ou acontecimento – de encontros, de relações do tipo Eu-Tu. Cruzando as noções introduzidas por Buber e Abreu, e tomando uma expressão usada por Gumbrecht (2010), concluímos que a essência da arquitetura é a “produção de presença”. A arquitetura é, portanto, condição de possibilidade para o surgimento – ou acontecimento – de encontros, de relações do tipo Eu-Tu. Reconhecemos que, à primeira vista, esta proposição pode parecer um tanto exagerada. Podemos imaginar, por exemplo, um encontro entre um Eu e um Tu ocorrendo em um ambiente natural, sem nenhuma intervenção humana aparente. Mas também podemos, a partir da abordagem construída até aqui, considerar que, se há a criação de um espaço relacional no qual um Eu encontra um Tu, já há, aí, arquitetura. Embora esta ideia possa parecer um tanto radical, Spiro Kostof – historiador da arquitetura e professor em Berkeley durante mais de 25 anos – considerava que a primeira arquitetura não foi a “cabana primitiva” sobre a qual tanto se especulou nos últimos séculos, mas sim algo aparentemente mais simples: a fogueira. Compreendendo a arquitetura como “o dom de criar lugares para algum propósito humano” (KOSTOF, 1995, p. 17)7, Kostof já considera a fogueira da caverna de Escale, na França - que acredita-se ter cerca de 500.000 anos -, como um exemplo de arquitetura (KOSTOF, 1995, p. 21). Na experiência típica do turismo cultural híper-moderno, há uma tendência à coisificação do ambiente construído, à transformação de edifícios, monumentos e pessoas em “Issos”. Ou seja, predominam arquiteturas despojadas desse espaço relacional fundamental. No caso extremo, podem mesmo deixar de ser arquitetura – ainda que não se lhes removam ou destruam uma pedra ou um tijolo. As novas tecnologias comunicativas redesenham possibilidades relacionais humanas. A cultura digital na sociedade contemporânea em rede possibilita novos modos de presença. Assim, busca-se experimentar fragmentos de presença através de fotografias, vídeos, visitas virtuais e relatos de outros visitantes, transformando a visitação do turista numa “visita de confirmação”. 102 7 No original: “"Architecture, in the end, is nothing less than the gift of making places for some human purpose” (Kostof, 1995, p. 17). ISSN 1807 975X É este o “entre”, “O intervalo onde se dá a relação face a face”, que por sua vez não é um espaço no sentido abstrato - um “vazio” - mas sim um espaço fértil para o surgimento de encontros. 101 10 8 No original: “feed the hunger and shelter the destitute from the nowhere of a website” (Wyschogrod, 2003, p. 41). ISSN 1807-975X No entanto, não se pode "alimentar os famintos e abrigar os desabrigados no lugar nenhum de 102 ISSN 180 9 X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 um website" (WYSCHOGROD, 2003, p. 41)8. A experiência presencial não pode, portanto, ser anulada, mas ela pode sim ser disciplinada - ou seja, busca-se anular seus riscos, retirando-lhe todas as facetas de imprevisibilidade. Desse modo, o encontro turístico se transforma em experiência pré-formatada. O diálogo e seus riscos se dissolvem numa nuvem de discursos que eclipsa a abertura para a alteridade. Mesmo um monumento famoso como a Torre Eiffel ou o Coliseu se transforma em (mais um) mero objeto, algo a ser apenas visto e fotografado. Neste tipo de relação, é difícil que haja abertura para a interferência recíproca causada pela relação face a face com a presença de uma alteridade que se apresenta como um Tu. A obra arquitetônica, transformada em objeto de consumo espetacularizado, serve apenas de suporte para uma visitação turística de confirmação: a confirmação de que ela é Isso mesmo. um website" (WYSCHOGROD, 2003, p. 41)8. A experiência presencial não pode, portanto, ser anulada, mas ela pode sim ser disciplinada - ou seja, busca-se anular seus riscos, retirando-lhe todas as facetas de imprevisibilidade. Desse modo, o encontro turístico se transforma em experiência pré-formatada. O diálogo e seus riscos se dissolvem numa nuvem de discursos que eclipsa a abertura para a alteridade. Mesmo um monumento famoso como a Torre Eiffel ou o Coliseu se transforma em (mais um) mero objeto, algo a ser apenas visto e fotografado. Neste tipo de relação, é difícil que haja abertura para a interferência recíproca causada pela relação face a face com a presença de uma alteridade que se apresenta como um Tu. A obra arquitetônica, transformada em objeto de consumo espetacularizado, serve apenas de suporte para uma visitação turística de confirmação: a confirmação de que ela é Isso mesmo. De produto singular, o património cultural e arquitectónico passou a ser considerado pela indústria do turismo como um produto de mercado, criando-se, por conseguinte um sistema de produção institucionalizado, sujeito aos mais variados interesses. Efectivamente, esta indústria transferiu o enfoque para o sujeito que o usufrui momentaneamente e não para o usufruidor tradicional (ABREU & MALHEIRO, 2011). ISSN 180 9 X De produto singular, o património cultural e arquitectónico passou a ser considerado pela indústria do turismo como um produto de mercado, criando-se, por conseguinte um sistema de produção institucionalizado, sujeito aos mais variados interesses. Efectivamente, esta indústria transferiu o enfoque para o sujeito que o usufrui momentaneamente e não para o usufruidor tradicional (ABREU & MALHEIRO, 2011). A transformação do monumento em produto turístico busca, obviamente, satisfazer às expectativas do turista, e não dos habitantes locais – que, imagina-se, foram justamente os primeiros a reconhecer o caráter monumental da obra. Não é difícil encontrar exemplos de igrejas nas quais já não se pode rezar, parques nos quais se tornou difícil encontrar uma sombra livre no verão, restaurantes “de bairro” que passaram a exigir reservas com meses de antecedência. Em casos como estes, destroem-se possibilidades de se redescobrir o Tu, ou de se recordar do encanto da sedução originária que fez do lugar uma “atração”. Este tipo de coisificação é um fenômeno contemporâneo que não se limita à arquitetura ou à experiência do turista. A transição de uma cultura de base literária para uma nova cultura de base imagética (FLUSSER, 2011) tende a substituir o “ler” pelo “assistir”, e “assistir apassiva” (SENNETT, 2003, p. 16). Na leitura de um romance ou poema, a imaginação está ativa e aberta – é literalmente informada (moldada) pelo conteúdo do texto, que a provoca e enriquece. Mesmo 103 ISSN 1807 975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 um trecho descritivo dará origem a diferentes imagens, tão numerosas quanto numerosos forem seus leitores. Quando assistimos um vídeo, filme ou série, a imaginação não precisa trabalhar tanto – muito já nos é dado, somos poupados de muito trabalho (FLUSSER, 2011). Embora uma cena possa tocar, provocar e comover cada espectador de uma forma diferente, a imagem da cena já está muito mais acabada que o a do trecho literário, e demanda muito menos esforço. Segundo Vilém Flusser, os aparelhos que nos oferecem estas imagens – câmeras fotográficas, televisores, computadores etc. – têm justamente esta função: O escritor informa objetos durante seu jogo: coloca letras sobre páginas brancas. Tais letras são símbolos decifráveis. Aparelhos fazem o mesmo. Há aparelhos, porém, que o fazem “melhor” que escritores, pois podem informar objetos com símbolos que não significam fenômenos, como no caso das letras, mas que significam movimentos dos próprios objetos. Tais objetos assim informados vão decifrando os símbolos e passam a movimentar-se. Por exemplo: podem executar os movimentos de trabalho. Podem, portanto, substituir o trabalho humano. Emancipam o homem do trabalho, liberando-o para o jogo (FLUSSER, 2011, p. 39). O escritor informa objetos durante seu jogo: coloca letras sobre páginas brancas. Tais letras são símbolos decifráveis. Aparelhos fazem o mesmo. Há aparelhos, porém, que o fazem “melhor” que escritores, pois podem informar objetos com símbolos que não significam fenômenos, como no caso das letras, mas que significam movimentos dos próprios objetos. Tais objetos assim informados vão decifrando os símbolos e passam a movimentar-se. Por exemplo: podem executar os movimentos de trabalho. Podem, portanto, substituir o trabalho humano. Emancipam o homem do trabalho, liberando-o para o jogo (FLUSSER, 2011, p. 39). O escritor informa objetos durante seu jogo: coloca letras sobre páginas brancas. Tais letras são símbolos decifráveis. Aparelhos fazem o mesmo. Há aparelhos, porém, que o fazem “melhor” que escritores, pois podem informar objetos com símbolos que não significam fenômenos, como no caso das letras, mas que significam movimentos dos próprios objetos. Tais objetos assim informados vão decifrando os símbolos e passam a movimentar-se. Por exemplo: podem executar os movimentos de trabalho. Podem, portanto, substituir o trabalho humano. Emancipam o homem do trabalho, liberando-o para o jogo (FLUSSER, 2011, p. 39). ISSN 1807 975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 predisponham de graus de atenção diferentes em relação a uma obra de arquitetura, pode-se oferecer a ambos esta abertura, esta possibilidade para se ver “além da aparência”. É justamente esta experiência mais profunda, este encontro, que é dificultada pela transformação da arquitetura em Isso, em produto, em entretenimento. predisponham de graus de atenção diferentes em relação a uma obra de arquitetura, pode-se oferecer a ambos esta abertura, esta possibilidade para se ver “além da aparência”. É justamente esta experiência mais profunda, este encontro, que é dificultada pela transformação da arquitetura em Isso, em produto, em entretenimento. O entertainment - entretenimento - é uma indústria que tem por finalidade o preenchimento do tempo do ciclo consumpção-produção remanescente ao trabalho e ao descanso. Não se o pode considerar como ócio (na significação clássica do termo) enquanto este tinha uma razão de ser externa a este ciclo - introduzir humanidade na vida activa, mundana, da espécie humana. Pelo contrário o entretenimento existe dentro daquele ciclo e, tal como a alimentação e o descanso, tem por finalidade a reposição das condições que permitem ao indivíduo voltar a trabalhar, pertence, portanto, ao processo biológico da vida. (...) Na medida em que parece cumprir a mesma função, a arte e a arquitectura começam no nosso tempo a serem avaliadas pelos mesmos critérios que o entretenimento é (ABREU & MALHEIRO, 2011). O entertainment - entretenimento - é uma indústria que tem por finalidade o preenchimento do tempo do ciclo consumpção-produção remanescente ao trabalho e ao descanso. Não se o pode considerar como ócio (na significação clássica do termo) enquanto este tinha uma razão de ser externa a este ciclo - introduzir humanidade na vida activa, mundana, da espécie humana. Pelo contrário o entretenimento existe dentro daquele ciclo e, tal como a alimentação e o descanso, tem por finalidade a reposição das condições que permitem ao indivíduo voltar a trabalhar, pertence, portanto, ao processo biológico da vida. (...) Na medida em que parece cumprir a mesma função, a arte e a arquitectura começam no nosso tempo a serem avaliadas pelos mesmos critérios que o entretenimento é (ABREU & MALHEIRO, 2011). Os critérios de avaliação do entretenimento são, obviamente, distintos dos critérios que usamos, até aqui, para definir a arquitetura. ISSN 1807 975X Um edifício pode “entreter” adequadamente sem oferecer o espaço relacional que é necessário para encontros do tipo Eu-Tu. No contexto atual, este “entreter” pode se limitar, por exemplo, ao oferecimento de um cenário interessante para fotografias. Assim, um monumento pode ser reduzido a um mero pano de fundo para fotografias que serão publicadas e compartilhadas nas redes sociais. Nesse caso, a lógica do “assistir” é levada ao extremo – embora ainda se experimente ao menos a aparência externa do edifício, o objetivo principal é transformar este edifício não apenas em um Isso, mas em uma imagem. ISSN 1807 975X O escritor informa objetos durante seu jogo: coloca letras sobre páginas brancas. Tais letras são símbolos decifráveis. Aparelhos fazem o mesmo. Há aparelhos, porém, que o fazem “melhor” que escritores, pois podem informar objetos com símbolos que não significam fenômenos, como no caso das letras, mas que significam movimentos dos próprios objetos. Tais objetos assim informados vão decifrando os símbolos e passam a movimentar-se. Por exemplo: podem executar os movimentos de trabalho. Podem, portanto, substituir o trabalho humano. Emancipam o homem do trabalho, liberando-o para o jogo (FLUSSER, 2011, p. 39). O mesmo pode ocorrer no encontro com um edifício, um monumento ou uma cidade como um todo. Podemos ler, ou apenas assistir. O ritmo do turismo cultural híper-moderno privilegia o assistir – além do “capturar” e compartilhar -, e, portanto, exige menos do turista. Porém, obviamente também lhe oferece menos. Em contrapartida, Abreu fala sobre o processo de “leitura” da arquitetura como “algo mais do que um complexo de sensações desencadeada pela obra (...) ou do que a interpretação positivista desta. O conceito de ‘leitura’ pressupõe uma articulação densa entre o pessoal e o objectivo, realizada de uma forma que seja intersubjectivamente acessível.” (ABREU, 2007, p. 24). Este processo de leitura proposto por Abreu visa atender a dois públicos distintos: o indivíduo comum que aborda a obra de arquitectura procurando tão-somente que ela manifeste a correspondência que promete e que a torna útil à vida real; e o profissional do restauro, que requer um instrumento que lhe permita determinar quais os objectos e quais os aspectos de um objecto que devem ser conservados (ABREU, 2007, p. 29). o indivíduo comum que aborda a obra de arquitectura procurando tão-somente que ela manifeste a correspondência que promete e que a torna útil à vida real; e o profissional do restauro, que requer um instrumento que lhe permita determinar quais os objectos e quais os aspectos de um objecto que devem ser conservados (ABREU, 2007, p. 29). No contexto do turismo cultural, podemos claramente pensar também nestes dois públicos: o turista típico, “comum”, e o visitante “especializado”. Em ambos os casos, “O método de leitura serve para nos adentrarmos para além da aparência da obra e das primeiras impressões por ela suscitadas” (ABREU, 2007, p. 39) – ou seja, embora o estes dois tipos de visitante 104 ISSN 1807-975X LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 ‘terciária’, já que não dominava. Atualmente, ocupa o centro da cena.” (FLUSSER, 2011, p. 35- 36). Ou seja, mesmo o turista que organiza sua própria viagem acaba recorrendo a aparelhos – sites e aplicativos de viagens, por exemplo -, o que o introduz na lógica dos programas. É claro que a questão central não é necessariamente o uso (ou não) de aparelhos - afinal, um site pode ser utilizado da mesma forma que um guia de viagens impresso. Da mesma forma, não faria sentido sugerir que os turistas deveriam viajar sem nenhum tipo de programação. O que realmente importa neste caso é perceber que a experiência do turismo foi em grande parte absorvida por duas lógicas dominantes: a lógica da indústria do entretenimento, que oferece distrações para preencher as horas vagas dos funcionários contemporâneos; e a lógica programática dos aparelhos, que está presente tanto nos aparatos eletrônicos que usamos para uma infinidade de fins como nos sistemas de administração, controle e distribuição de informações. (Re)moldada por estas duas forças, a experiência do turismo perde seu caráter transformador, reduzindo a relevância não só da atividade em si, mas também de tudo o que se experimenta através dela – edifícios, monumentos, cidades, culturas etc. Incapazes de nos abrirmos para o Tu que habita em cada lugar, podemos criar a ilusão de conhecer aquilo que apenas vimos - mas que não nos tocou. Pior que não conhecer o Outro é achar que o conhece – e assim, também deixamos de conhecer mais sobre nós mesmos. Uma experiência turística que não seja excessivamente programada pode deixar pedaços de tempo nos quais haja espaço – literalmente – para acontecimentos imprevistos, como os encontros de que falava Buber. Uma visitação na qual se busca ler – e não apenas assistir – aquilo que se vê, permitindo-se modificar pelo Tu que se revela nas pessoas e nas coisas, pertence certamente ao domínio da arte – o habitat natural da iluminação transformadora citada por Eco. Recuperar, na experiência turística, a força transformadora do encontro, é uma abertura para relações mais livres, plurais e diversas com o mundo. 4 Chegada Para que o turismo possa ser transformado em um produto da indústria do entretenimento, a experiência do turista precisa ser programada. Isto fica claro no caso de pacotes turísticos que incluem uma programação completa – voos, traslados, passeios, refeições etc. -, a ser seguida pelos turistas, que na maioria das vezes são reunidos em grupos. Porém, a possibilidade das “visitas de confirmação” evidencia que a lógica programática se espalhou para outros tipos de visitação, incluindo aqueles que não parecem, a princípio, se tratar de “produtos”. Segundo Flusser, “A maioria da sociedade está empenhada nos aparelhos dominadores, programadores e controladores. Outrora, antes que aparelhos fossem inventados, a atividade deste tipo se chamava 105 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 ISSN 1807-975X ISSN 1807-975X Isto é fecundo e enriquecedor, mas, em contrapartida, é também arriscado e surpreendente. A mera confirmação do já sabido não permite tal fecundidade. Ela sufoca o diálogo em discursos, ao submeter a vida à programação. A arte de visitar é a arte de se colocar disponível a encontros como um co-ator, e não como espectador. 106 SCRUTON, R. On Human Nature. Princeton : Princeton University Press, 2017. Referências ABREU, P. M. Palácios da Memória II: A Revelação da Arquitectura. Volume I – Secção Teórica: O Processo de Leitura do Monumento. 2007. Tese (Doutorado em Arquitetura) - Faculdade de Arquitectura, Universidade Técnica de Lisboa, Lisboa, 2007. ABREU, P. M.; MALHEIRO, J. B. (An)Estética do Turismo: ou a mediatização do Património. In: GAZZANEO, L. M. (org.). Espaços Culturais e Turísticos em Países Lusófonos – cultura e turismo. Rio de Janeiro: Universidade Federal do Rio de Janeiro – Faculdade de Arquitectura e Urbanismo, 2011. p. 125-147. BARTHOLO, R. Sobre o sentido da proximidade: implicações para um turismo situado de base comunitária. In: BARTHOLO, R.; SANSOLO, D.G.; BURSZTYN, I. (Org.). Turismo de Base Comunitária: Diversidade de Olhares e Experiências Brasileiras. 1ed. Rio de Janeiro: Letra e Imagem, 2009. p. 45-54. _______. Sobre Tendas e Casas: Hospitalidade, Turismo e Emigração em Perspectiva Filosófica. In: BRAYNER, A. (org.). Vilém Flusser: Filosofia do Desenraizamento. Porto Alegre: Clarinete, 2015. _______. Turismo, teoria e liberdade de ir e vir. In: IRVING, M.; CALABRE, L.; BARTHOLO, R.; LIMA, M.; MORAES, E.; EGREJAS, M.; LIMA, D. (org.). Turismo, natureza e cultura. Interdisciplinaridade e política pública. 1ed. Rio de Janeiro: Fundação Casa de Rui Barbosa, 2017. p. 28-30. BUBER, M. Eu e Tu. São Paulo: Editora Moraes, 1977. DARWALL, S. The Second-Person Standpoint: Morality, Respect and Accountability. Londres: Harvard University Press, 2006. DEBORD, G. A Sociedade do Espetáculo. Lisboa: Edições Antipáticas, 2005. ECO, U. Pape Satàn Aleppe: Crônicas de Uma Sociedade Líquida. Rio de Janeiro: Record, 2017. FLUSSER, V. Filosofia da Caixa Preta: Ensaios para uma Futura Filosofia da Fotografia. São Paulo: Annablume, 2011. GUMBRECHT, H. U. Produção de Presença: O Que O Sentido Não Consegue Transmitir. Rio de Janeiro: Contraponto, 2010. KOSTOF, S. A History of Architecture: Settings and Rituals. Nova York: Oxford University Press, 1995. SCRUTON, R. On Human Nature. Princeton : Princeton University Press, 2017. 107 LOUREIRO, F.; BARTHOLO, R.; BARCELOS, F.; MATTOS, F. Visitar/Acolher: Arquitetura, Turismo e Encontros. Revista Hospitalidade. São Paulo, volume 17, n.02, p. 95-108, 2020. Doi: https://doi.org/10.21714/2179-9164.2020.v17n2.006 SENNETT, R. Carne e Pedra: O corpo e a cidade na civilização ocidental. Rio de Janeiro: Record, 2003. WYSCHOGOROD, E. Autochthony and Welcome: Discourses of Exile in Levinas and Derrida. Journal of Philosophy and Scripture, vol. 1, issue 1, 2003. SENNETT, R. Carne e Pedra: O corpo e a cidade na civilização ocidental. Rio de Janeiro: Record, 2003. SENNETT, R. Carne e Pedra: O corpo e a cidade na civilização ocidental. Rio de Janeiro Record, 2003. WYSCHOGOROD, E. Autochthony and Welcome: Discourses of Exile in Levinas and Derrida. Journal of Philosophy and Scripture, vol. 1, issue 1, 2003. WYSCHOGOROD, E. Autochthony and Welcome: Discourses of Exile in Levinas and Derrida. Journal of Philosophy and Scripture, vol. 1, issue 1, 2003. x’ Artigo recebido em: 21/01/2020 Avaliado em: 30/01/2020 Aprovado em: 22/06/2020 108
https://openalex.org/W2911273410	https://www.frontiersin.org/articles/10.3389/fmars.2018.00530/pdf	English	null	Future Directions in Eubalaena spp.: Comparative Research to Inform Conservation	Frontiers in marine science	2,019	cc-by	26,713	Future Directions in Eubalaena spp.: Comparative Research to Inform Conservation Rob Harcourt1, Julie van der Hoop2, Scott Kraus3 and Emma L. Carroll4,5 1 Department of Biological Sciences, Macquarie University, Sydney, NSW, Australia, 2 Zoophysiology, Department of Bioscience, Aarhus University, Aarhus, Denmark, 3 Anderson Cabot Center for Ocean Life, New England Aquarium, Boston, MA, United States, 4 Sea Mammal Research Unit, Scottish Oceans Institute, University of St Andrews, St Andrews, United Kingdom, 5 School of Biological Sciences, The University of Auckland, Auckland, New Zealand All three extant right whales [Eubalaena australis (Southern; SRW), glacialis (North Atlantic; NARW), and japonica (North Paciﬁc; NPRW)] were heavily exploited, and the status of the two northern hemisphere species remains precarious. Recently, limited gains made by the NARW have been reversed and urgent changes to management approaches are needed if extinction is to be averted. By contrast, some SRW populations are recovering. Given their close phylogenetic relationship, morphological, demographic, and ecological similarities, the contrasting recovery rates between populations and species provide an opportunity to apply a comparative approach to inform the differences in recovery as follows. (1) Recovery: All right whale species were internationally protected in 1931, but NARW, eastern NPRW and some SRW populations have barely recovered from whaling, while others are doing so at maximal rates. Are these differences a legacy of extreme depletion (e.g., loss of genetic diversity and cultural knowledge) or primarily due to anthropogenic factors (e.g., high mortality from ship strike and ﬁsheries entanglement)? If modern anthropogenic threats are not affecting remote SRW populations, can these serve as baseline populations for comparison with NARW and NPRW? (2) Linking individuals to population-level responses: In wild mammals, strong links exist between reproductive indices and environmental conditions within the context of life-history strategies. Individual identiﬁcation of whales provides the ability to track survival, reproduction and other demographic parameters, and their population-level consequences, providing the tools with which to uncover these links. Robust life-history analyses are now available for NARW and several SRW populations, linking demography with environmental conditions, providing the potential for teasing out important inﬂuencing factors. (3) Adapting to shifting resources: Recent reproductive declines in NARW appear linked to changing food resources. While we know some large-scale movement patterns for NARW and a few SRW populations, we know little of mesoscale movements. For NPRW and some SRW populations, even broad-scale movements are poorly understood. In the face of climate change, can methodological advances help identify Eubalaena distributional and migratory responses? Future Directions in Eubalaena spp.: Comparative Research to Inform Conservation (4) Emergent diseases and the vulnerability of populations under stress: Marine mammals are vulnerable to infectious diseases, Edited by: Jeremy Kiszka, Florida International University, United States Reviewed by: Robert L. Brownell, Southwest Fisheries Science Center (NOAA), United States Daniel M. Palacios, Oregon State University, United States Correspondence: Rob Harcourt robert.harcourt@mq.edu.au Specialty section: This article was submitted to Marine Megafauna, a section of the journal Frontiers in Marine Science Received: 04 August 2018 Accepted: 31 December 2018 Published: 30 January 2019 Citation: Harcourt R, van der Hoop J, Kraus S and Carroll EL (2019) Future Directions in Eubalaena spp.: Comparative Research to Inform Conservation. Front. Mar. Sci. 5:530. doi: 10.3389/fmars.2018.00530 Specialty section: This article was submitted to Marine Megafauna, a section of the journal Frontiers in Marine Science Specialty section: This article was submitted to Marine Megafauna, a section of the journal Frontiers in Marine Science Received: 04 August 2018 Accepted: 31 December 2018 Published: 30 January 2019 Received: 04 August 2018 Accepted: 31 December 2018 Published: 30 January 2019 REVIEW published: 30 January 2019 doi: 10.3389/fmars.2018.00530 published: 30 January 2019 doi: 10.3389/fmars.2018.00530 Introduction – Exploitation/Devastation Introduction Exploitation/Devastation The Eubalaena genus comprises three species of right whale that have the dubious honor of being named by whalers for their desirability: North Atlantic (Eubalaena glacialis), North Paciﬁc (E. japonica), and southern (E. australis) right whales (Braham and Rice, 1984; Rosenbaum et al., 2000). The ‘right’ whale to kill, these species are large baleen whales that tend to ﬂoat when dead and yielded long baleen plates (known as whalebone historically), and much valuable oil that was used in industries from cosmetics to commercial lubricant (Allen, 1916; International Whaling Commission [IWC], 2001). The family (Balaenidae) also includes a fourth member, the bowhead whale (Balaena mysticetus), which was also heavily exploited but the Paciﬁc population has had a major recovery and numbers more than the combined populations of the other three species (Givens et al., 2016). Exploitation of right whales began in earnest in the eastern North Atlantic with Basque shore and Basque pelagic whaling as early as 1000 AD (Reeves and Smith, 2006). The Basque shore-whaling techniques were exported to the South Atlantic by 1603. In the North and South Paciﬁc, following the decline of Basque whaling, American shore and pelagic whaling methods predominated, appearing in the southern Indo-Paciﬁc by 1805. Within 20 years the French whaling vessel Gange became the ﬁrst pelagic whaling vessel to venture north of 50◦N where it harpooned the ﬁrst NPRW in 1835 (Reeves and Smith, 2006). NARW have a much longer and poorly documented catch history, and the ability to estimate historical abundance for this species has proved challenging. Reeves et al. (2007) used catch records, whale oil, and baleen import records to estimate at least 5500 (and possibly twice that number) NARW were killed in the western North Atlantic between 1634 and 1950 and suggested a population of at least a few thousand whales in the mid-1600’s. Early studies indicating that Basque whalers took tens of thousands of NARW in the 1500’s from eastern Canada have been corrected by genetic analyses of bones from the Labrador whaling stations. These show that Basque whalers took primarily bowheads, and rarely caught right whales, leading to suggestions that small population size may be a long-term characteristic of this species (Rastogi et al., 2004; McLeod et al., 2008). Nevertheless, Reeves et al. Citation: January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 1 Perspectives on Right Whale Research and Conservation Harcourt et al. particularly when subjected to stressors such as ﬁshing gear entanglements, acoustic disturbance, and prey shortages. New tools to assess large whale health include body condition imaging, viromes, microbiomes, as well as metabolic and stress hormones. Comparative analysis of the three Eubalaena spp. could identify causes of varying recovery. (5) Comparative synthesis and cumulative effects: The lack of a good analytical approach for cumulative effects is an urgent bio-statistical problem in conservation biology. Without such a framework every stressor is managed in isolation, limiting efﬁcacy. We propose a comparative synthesis to inform future cumulative effect analyses and outline future research priorities to achieve these goals. Keywords: right whale, conservation, cumulative effects analysis, conservation technology, threats, recovery Frontiers in Marine Science \| www.frontiersin.org RECOVERY The decline of the SRW is the best understood of all the right whales. It is estimated that prior to whaling there were approximately 120,000 SRW found in 12 wintering grounds (Figure 1, International Whaling Commission [IWC], 2001, 2012; Jackson et al., 2008). However, between 1790 and 1971 up to 150,000 SRW were killed, leading to a hemispheric decline to as few as 400 whales around 1920 (International Whaling Commission [IWC], 2001, 2012; Jackson et al., 2008). Full protection from commercial whaling was provided for all right whale species under the Convention for Regulation of Whaling (CRW) in 1931 under the League of Nations. Japan did not accede to the CRW, but the CRW provided the framework for the future regulation of whaling that continues until today as the 1946 International Convention for the Regulation of Whaling (ICRW). Both the Soviet Union and Japan acceded to the ICRW. However, like the CRW, the ICRW was not perfect and even as the various species recovered from historical whaling, they were targeted illegally by the Soviet whaling ﬂeet in the 1960s and 1970s. For the SRW, this killed half the extant population at the time (Tormosov et al., 1998; Jackson et al., 2008). Today, the SRW is estimated to have recovered to 12,000–15,000 individuals across its circumpolar distribution (Figure 1; International Whaling Commission [IWC], 2012). Introduction – Exploitation/Devastation (1992) suggested that fewer than 100 right whales were alive at the end of the 1700’s, and Katona and Kraus (1999) hypothesized that only a few dozen survived by the early 1900’s (Figure 2). As of 2015, there were an estimated p Shore whaling targeted right whales on their wintering grounds, as these are generally in sheltered coastal waters. Shore stations set up in these regions routinely killed calves as a way to secure the larger, more valuable, female, as this description from 1844 shows; “the object is to harpoon the calf ...the maternal aﬀection of the whale causes it to follow the calf till it gets ashore” (New Zealand Blue Book, 1841). Females show ﬁdelity to nursery grounds, frequently returning to the same bay to calve, so shore stations were successful for several years until they had wiped out those whales philopatric to a particular region. Oﬀshore whaling, although termed Basque or American (“Yankee”) whaling due to its origins, was in fact conducted by many nations, including British, French, Norwegian and Dutch, and was indiscriminate in its killing of right whales (Reeves et al., 1999; Richards, 2009; Smith et al., 2012). January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 2 Perspectives on Right Whale Research and Conservation Harcourt et al. FIGURE 1 \| Change in distribution and abundance of southern right whales. (A) Shows historical and contemporary wintering distributions (Figure 1 from Carroll et al., 2018), and (B) shows decline in abundance and subsequent recovery (solid line is the mean, dashed line shows upper and lower 95% CI). Modiﬁed Figure 1 from Jackson et al. (2008). Contemporary sightings are divided into regions where large aggregations are seen during winter: Argentina (ARG), Brazil (BZL), South Africa (SAF), southwest Australia (SWA), south central Australia (SCA), and New Zealand sub-Antarctic (NZSA) and regions where sightings are typically of small numbers of individuals per year. The large aggregations are IWC management units and correspond to historical whaling grounds, although another 5 whaling grounds show little sign of recovery. Summer feeding areas are poorly described and so not shown. 58 individuals in the western NARW population (Pace et al whole totaled 100 200 whales (Wada 1973; Sekiguchi et al 2014; FIGURE 1 \| Change in distribution and abundance of southern right whales. Frontiers in Marine Science \| www.frontiersin.org Introduction – Exploitation/Devastation Whales were occasionally seen and killed north of New England up to Greenland. In the northeast Atlantic, the species was found off Greenland, Iceland and the United Kingdom in the summer months, and the Bay of Biscay and Cintra Bay, Western Sahara, during the winter. There are now no regular aggregations in the northeast Atlantic, although there are occasional sightings. In the northwest Atlantic, there is a well-described winter calving ground off the coast of Florida and Georgia, United States (SEUS), and summer feeding grounds around the Bay of Fundy and Nova Scotia, off the coast of North America (NAF). The North Paciﬁc and North Atlantic right whales are, by IWC convention, divided into east and west stocks. Recent genetic data supports differentiation of eastern and western stocks of North Paciﬁc right whales (Pastene et al., 2018). (B) Modeled population trajectory of North Atlantic right whales off the eastern United States, with Maximum Sustainable Yield rates of 0.03 (solid line) and 0.01 and 0.05 (dashed lines; modiﬁed from Reeves et al., 1992). Black points are from modern observations of the minimum number alive from the North Atlantic right whale catalog (NARWC, 2018). stock from photographic (18 identiﬁed individuals) and genotype (21 identiﬁed individuals) data through 2008 were 31 (95% CI: 23–54) and 28 whales (95% CI: 24–42), respectively (Wade et al., 2011b) and the photo-ID and genetic catalogs have 20 and 23 unique individuals, respectively (1997–2011; LeDuc et al., 2012; Muto et al., 2017). Conﬁrmed sightings since 1996 (Goddard and Rugh, 1998), and a recent increase in eﬀort have led to more frequent detections in the region, including calves (Wade et al., 2006, 2011a; Zerbini et al., 2015; Ford et al., 2016). logbooks have been useful in describing the distribution of the species prior to and during their exploitation (Dawbin, 1986; Du Pasquier, 1986; Reeves et al., 1992; Clapham et al., 2004; Reeves et al., 2007; Smith et al., 2012; Carroll et al., 2014; Rocha et al., 2014; Thomas et al., 2016) as well as identifying potential habitats that members of the modern populations may be re- exploiting. While this short summary demonstrates the dramatic impact of whaling, it does not capture the spatially variable patterns of decline and, in some cases, recovery. Introduction – Exploitation/Devastation (A) Shows historical and contemporary wintering distributions (Figure 1 from Carroll et al., 2018), and (B) shows decline in abundance and subsequent recovery (solid line is the mean, dashed line shows upper and lower 95% CI). Modiﬁed Figure 1 from Jackson et al. (2008). Contemporary sightings are divided into regions where large aggregations are seen during winter: Argentina (ARG), Brazil (BZL), South Africa (SAF), southwest Australia (SWA), south central Australia (SCA), and New Zealand sub-Antarctic (NZSA) and regions where sightings are typically of small numbers of individuals per year. The large aggregations are IWC management units and correspond to historical whaling grounds, although another 5 whaling grounds show little sign of recovery. Summer feeding areas are poorly described and so not shown. 458 individuals in the western NARW population (Pace et al., 2017). whole totaled 100–200 whales (Wada, 1973; Sekiguchi et al., 2014; Ovsyanikova et al., 2015). For the western stock, surveys by Miyashita and Kato (1998) in summers of 1989, 1990, and 1992 in the Okhotsk Sea estimated abundance at 922 (95% CI 404– 2108). More recently, surveys by Hakamada and Matsuoka (2016) suggest an abundance estimate of 1,147 individuals on the feeding ground oﬀthe southern end of the Kamchatka Peninsula in 2011 and 2012. In contrast, the eastern stock is on the order of 20–30 times smaller. Mark-recapture estimates for the eastern NPRW Two centuries of aboriginal, commercial, and then illegal Soviet whaling signiﬁcantly reduced the NPRW (Scarﬀ, 1991, 2001; Ivashchenko and Clapham, 2012; Ivashchenko et al., 2013). There are two distinct populations of NPRW, on the eastern and western sides of the Paciﬁc (Gregr and Coyle, 2009), which are genetically distinct (Pastene et al., 2018). Early estimates based on sightings data suggested the species as a January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 3 Perspectives on Right Whale Research and Conservation Harcourt et al. FIGURE 2 \| (A) Change in distribution of North Paciﬁc and North Atlantic right whales. The North Paciﬁc right whale was hunted across the North Paciﬁc, although the whaling grounds were primarily summering grounds and no wintering areas were identiﬁed. Introduction – Exploitation/Devastation This mystery continues today, as the contemporary distribution of these two stocks are characterized by small aggregations seen in summer feeding grounds in the Gulf of Alaska (GOA) and southeast Bering Sea (BS) in the east, and the Sea of Okhotsk (SOO) and the Kamchatka Peninsula (KP) in the west. The North Atlantic right whale was hunted around the New England coast during spring months in the northwest Atlantic, and around the southeast United States coast during winter. Whales were occasionally seen and killed north of New England up to Greenland. In the northeast Atlantic, the species was found off Greenland, Iceland and the United Kingdom in the summer months, and the Bay of Biscay and Cintra Bay, Western Sahara, during the winter. There are now no regular aggregations in the northeast Atlantic, although there are occasional sightings. In the northwest Atlantic, there is a well-described winter calving ground off the coast of Florida and Georgia, United States (SEUS), and summer feeding grounds around the Bay of Fundy and Nova Scotia, off the coast of North America (NAF). The North Paciﬁc and North Atlantic right whales are, by IWC convention, divided into east and west stocks. Recent genetic data supports differentiation of eastern and western stocks of North Paciﬁc right whales (Pastene et al., 2018). (B) Modeled population trajectory of North Atlantic right whales off the eastern United States, with Maximum Sustainable Yield rates of 0.03 (solid line) and 0.01 and 0.05 (dashed lines; modiﬁed from Reeves et al., 1992). Black points are from modern observations of the minimum number alive from the North Atlantic right whale catalog (NARWC, 2018). FIGURE 2 \| (A) Change in distribution of North Paciﬁc and North Atlantic right whales. The North Paciﬁc right whale was hunted across the North Paciﬁc, although the whaling grounds were primarily summering grounds and no wintering areas were identiﬁed. This mystery continues today, as the contemporary distribution of these two stocks are characterized by small aggregations seen in summer feeding grounds in the Gulf of Alaska (GOA) and southeast Bering Sea (BS) in the east, and the Sea of Okhotsk (SOO) and the Kamchatka Peninsula (KP) in the west. The North Atlantic right whale was hunted around the New England coast during spring months in the northwest Atlantic, and around the southeast United States coast during winter. Frontiers in Marine Science \| www.frontiersin.org Migratory Culture, Extirpation and Recovery Satellite tagging in Argentinean wintering grounds showed that some whales will visit both the Patagonian Shelf and South Georgia in a single season (Zerbini et al., 2016). This is consistent with ﬁndings reported by Rowntree et al. (2008), who used isotopic signals in baleen plates to suggest a mixture of foraging strategies. Some whales appeared to specialize in either low- or high-latitude foraging grounds but not both, while others exhibited a mixture of latitudes over the 6–8 years represented in the baleen record (Rowntree et al., 2008). It remains to be seen whether right whales are able to learn from each other, and horizontally transmit foraging-ground preferences. Horizontal transmission of behavior or information could facilitate spread of innovation; in this case, potentially new foraging grounds (Dautenhahn and Nehaniv, 2002; Keith and Bull, 2017); and could buﬀer right whales from changes in prey distribution and abundance driven by climate. While vertical transmission (cow to calf) of migratory destinations provides valuable information on suitable foraging and nursery areas in a vast ocean (Whitehead, 2007, 2010), decisions conserved across generations can become problematic in the face of environmental change (Keith and Bull, 2017). This is because of the potential for ecological or evolutionary traps; a behavior that originally increased ﬁtness (e.g., ﬁdelity to rich feeding ground) becomes a hindrance in the face of rapid environmental change (e.g., changes in distribution of food resources due to climate variation and/or anthropogenic activities: Schlaepfer et al., 2002; Keith and Bull, 2017). Migratory Culture, Extirpation and Recovery y SRW have a circumpolar distribution and rates of recovery have varied from near maximum growth rates (rm) in some populations (southwest Australia, South Africa, East South America, and subantarctic New Zealand), to poor or not measurable in others. We now know that SRW show a form of migratory culture, with females transmitting preferences for both winter calving/breeding areas and summer foraging areas to their calves during the 1st year of life (Valenzuela et al., 2009; Carroll et al., 2015, 2016). This has led to genetic structuring across the species’ migratory network (Patenaude et al., 2007; Carroll et al., 2015; Carroll et al., 2018), and has been invoked to explain the species’ patchy recovery (Clapham et al., 2008; Carroll et al., 2011). Essentially, when whales inhabiting a region were extirpated, the memory of that area as a good migratory destination was also lost. This loss of ‘cultural memory’, exacerbated by low density and loss of adjacent populations, mean it is unlikely that once-inhabited areas will be recolonized on a timeframe relevant to management (i.e., decades; Clapham et al., 2008). This may help explain the lack of recovery around mainland New Zealand and east Australia wintering grounds compared with the strongly recovering populations in the New Zealand sub-Antarctic Islands and southwest Australia (Carroll et al., 2011, 2015). Comparable ﬁndings from the NARW show female philopatry to summering grounds (Malik et al., 1999), and population structuring that is probably due to migratory patterns of behavior (Schaeﬀ et al., 1993). NPRW also show signiﬁcant diﬀerentiation in maternally inherited DNA markers between Eastern and Western stocks that are likely partially attributable to female philopatry and stock identity (Pastene et al., 2018). However, historical bottlenecks may have inﬂuenced population structuring and migratory behavior prior to major whaling impacts (Waldick et al., 2002). However, some feeding grounds of the SRW appear to have been completely abandoned, and this may be due to a loss of cultural memory rather than resource quality. For example, New Zealand right whales historically used foraging grounds around the Chatham Rise and Louisville and Kermadec ridges (Smith et al., 2012; Jackson et al., 2016), but there are no contemporary records of the whales from this region (Carroll et al., 2014). Fortunately, like NARW, there is evidence of mixed foraging strategies within populations of SRW with whales visiting multiple foraging grounds, both within and between years. Introduction – Exploitation/Devastation There are several aspects of right whale ecology, population dynamics, as well as anthropogenic impacts that need to be considered when All contemporary right whale populations reﬂect remnants of their pre-whaling populations. As such, they appear to occupy a portion of their historical ranges (Figures 1, 2). Whaling January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 4 Perspectives on Right Whale Research and Conservation Harcourt et al. Seyboth et al., 2016). The Patagonian Shelf, and South Georgia, south of the Polar Front, are an important feeding ground for the Argentina SRW population based on stable isotope, photo- ID and satellite track data. This high productivity means that some of these foraging grounds are shared with whales that winter in South Africa (Table 1; Best et al., 1993; Moore et al., 1999; Rowntree et al., 2001; Valenzuela et al., 2009; Mate et al., 2011; Zerbini et al., 2016). Both of these populations are thriving; although the Argentinean population is now increasing at a slower rate (Crespo et al., 2018) and it has been subject to a recent die oﬀ(Rowntree et al., 2013). In the NARW, there are strong links between prey availability, condition, and fecundity: poor prey years lead to low population health and body condition, and longer inter-birth intervals (Hlista et al., 2009; Miller et al., 2011, 2012; Fortune et al., 2013; Rolland et al., 2016; Meyer-Gutbrod and Greene, 2018). NARW move between habitats, especially foraging grounds, many times within a season (Mate et al., 2011; Baumgartner et al., 2017). Connectivity to alternative feeding grounds could buﬀer a population against poor conditions in a single feeding ground, ensuring that some of the population had suﬃcient resources to reproduce. contemplating why there has not been uniform recovery across populations or stocks. Each of these are explored with the intent of providing an overarching discussion of the roles of intrinsic and extrinsic drivers of recovery in the diﬀerent species of right whales. Frontiers in Marine Science \| www.frontiersin.org Inverse Density Dependence: Genetic Diversity and Stochastic Demographic Inﬂuences on Recovery North Paciﬁc right whale Best and Schell, 1996; Rowntree et al., 2001, 2008; Valenzuela et al., 2009, 2010, 2018; Carroll et al., 2015 Baker et al., 1999; Patenaude et al., 2007; Valenzuela et al., 2009; Carroll et al., 2011, 2014, 2015, 2018 Pastene et al., 2018 Payne, 1986; Barendse and Best, 2014 Argüelles et al., 2016 ratio and a loss of genetic diversity that appeared to be midway between that observed for right whales in the North Atlantic and the Southern Hemisphere. The analysis also suggested a degree of separation between eastern and western populations, a male:female ratio of 2:1, and a low eﬀective population size for the Eastern North Paciﬁc stock, which LeDuc et al. (2012) considered to be at “extreme risk” of extirpation. In contrast, the NARW has long been recognized as having low levels of genetic diversity, although genetic erosion does not seem to have coincided with 19th century whaling (Waldick et al., 2002; McLeod et al., 2010). It is possible that the NARW historically had a lower eﬀective population size and hence harbored lower genetic diversity than its more numerous southern counterpart. Today, the high variability in male reproductive success seen in NARW, less prominent in SRW, could contribute to further lowering eﬀective population size by limiting genetic input from all apart from those few successful males (Frasier et al., 2007; Carroll et al., 2012). There is some evidence from the whaling era to support density- dependent changes in growth curves and fecundity in sperm whales (Kasuya, 1991) and other cetacean species (Fowler, 1984), but these data may be erroneous (NOAA, 2017) and cannot be relied upon. However, the lack and/or slow recovery of many populations of SRW, has led to speculation about the role of inverse density dependence at low densities or the Allee eﬀect (Baker and Clapham, 2004). There are generally three primary categories of Allee eﬀects; inbreeding, or a loss of heterozygosity; demographic stochasticity; and the reduction in cooperative interactions when there are fewer individuals (Allee, 1931; Courchamp et al., 1999). p Loss of genetic diversity due to demographic bottlenecks depends on the severity of the event, in terms of eﬀective population size, and its duration, in terms of generations (Nei, 1975; Allendorf, 1986). Genetic erosion can lead to inbreeding, maladaptation, and loss of adaptive potential, increasing the risk of population extinction (Leroy et al., 2017). Best and Schell, 1996; Rowntree et al., 2001, 2008; Valenzuela et al., 2009, 2010, 2018; Carroll et al., 2015 Baker et al., 1999; Patenaude et al., 2007; Valenzuela et al., 2009; Carroll et al., 2011, 2014, 2015, 2018 Pastene et al., 2018 Payne, 1986; Barendse and Best, 2014 Argüelles et al., 2016 Frontiers in Marine Science \| www.frontiersin.org Inverse Density Dependence: Genetic Diversity and Stochastic Demographic Inﬂuences on Recovery Standard density-dependent population theory, exempliﬁed by logistic growth models (Tsoularis and Wallace, 2002), holds that populations reduced to low densities should in the absence of migration increase at close to the maximum biological rate. This information is critical to understanding the variable rates of recovery of all right whales; foraging ground quality seems to be a strong determinant of the reproductive rate of SRW populations, and hence recovery rates (Leaper et al., 2006; January 2019 \| Volume 5 \| Article 530 5 Perspectives on Right Whale Research and Conservation Harcourt et al. TABLE 1 \| Tools for monitoring right whale occurrence and movements. Method Scale North Atlantic right whale Southern right whale North Paciﬁc right whale Satellite tags 10s–1000s km Mate et al., 1997; Baumgartner et al., 2003; Baumgartner and Mate, 2003, 2005; Schick et al., 2009 Childerhouse et al., 2010; Mate et al., 2011; Zerbini et al., 2016 Zerbini et al., 2015 Passive acoustic monitoring 10s–1000s km Davis et al., 2017 and studies cited therein Webster et al., 2016; Jacobs et al., 2018; Rayment et al., 2018 Waite et al., 2003; Mellinger et al., 2004; Rone et al., 2012; Širovi ´c et al., 2015; Wright et al., 2018 Autonomous vehicles with on-board acoustics 10s–100s km Baumgartner, 2014; Davis et al., 2016 Stable isotopes 100s–1000s km Lysiak, 2009; Lysiak et al., 2018 Best and Schell, 1996; Rowntree et al., 2001, 2008; Valenzuela et al., 2009, 2010, 2018; Carroll et al., 2015 Genetics 100s–1000s km Schaeff et al., 1993; Malik et al., 1999; Frasier et al., 2007 Baker et al., 1999; Patenaude et al., 2007; Valenzuela et al., 2009; Carroll et al., 2011, 2014, 2015, 2018 Pastene et al., 2018 Shore-based tracking 10s–1000s km Hain et al., 2013 Payne, 1986; Barendse and Best, 2014 Time-depth recorders 1–1000s km Winn et al., 1995; Goodyear, 1996; Nowacek et al., 2001, 2004; Baumgartner and Mate, 2003; Parks et al., 2012; Nousek-McGregor et al., 2013; van der Hoop et al., 2013b, 2017b; Baumgartner et al., 2017 Argüelles et al., 2016 High-resolution orientation, kinematics 1–10s km Nowacek et al., 2001, 2004; Parks et al., 2012; Nousek-McGregor et al., 2013; van der Hoop et al., 2013b, 2017b TABLE 1 \| Tools for monitoring right whale occurrence and movements. Anthropogenic Impacts on Recovery The level of human-induced mortality to NPRW is currently unknown, due to their rare occurrence and poorly known distribution (Muto et al., 2017); however, threats are assumed to be similar to other right whale populations and bowhead whales (Reeves et al., 2012) and their observed distribution does overlap heavy vessel traﬃc and ﬁshing areas. The threats from vessels will only increase as Arctic shipping increase in the near future and the shipping lanes will be through the summer feeding habitat in the southeastern Bering Sea, as it will for their cousin the bowhead whale (Reeves et al., 2012; Pirotta V. et al., 2018). Despite the low likelihood of observing anthropogenic mortality or serious injury to NPRW in the eastern population, some live individuals show scars as evidence of ﬁsheries interactions (Ford et al., 2016). For the much larger western NPRW population, 25 whales have been reported entrapped in ﬁshing gear or stranded from Japan, South Korea, and Russia since 1996 and at least 10 of these are conﬁrmed ﬁshing mortalities (Brownell et al., 2001; Burdin et al., 2004: Brownell and Mallette, 2018). The role of anthropogenic impacts on right whale recovery is best understood in the NARW, due to comprehensive population monitoring over the last four decades (Pettis, 2009; Rolland et al., 2016) and consistent necropsy eﬀort (Moore et al., 2004, 2013; van der Hoop et al., 2013a). The toll of human-induced mortality is, and continues to be, high; from 1970 to 2009, 44% (38/87) of detected and diagnosed NARW mortalities were due to ship strikes and 35% (31/87) were due to entanglements (van der Hoop et al., 2013a), despite widespread regulations attempting to reduce these threats. Entanglement has now surpassed vessel strike as the leading cause of death to the population (van der Hoop et al., 2014): from 2010 to 2015, 15% of diagnosed mortalities were due to ship strikes and 85% were due to entanglements (Pettis and Hamilton, 2015; Hayes et al., 2017). Demographic models emphasize the importance of individual reproductive females in the population (e.g., Fujiwara and Caswell, 2001) and suggest that a loss of 4–6 females per year will lead the population to extinction (Meyer-Gutbrod and Greene, 2018). In 2017, at least 17 NARW mortalities were detected, including ﬁve diagnosed vessel-strike mortalities, and four entanglement mortalities, leading to the declaration of an Unusual Mortality Event (NOAA, 2017). Synthesis Whaling nearly extirpated all right whale species worldwide. Numerous factors have inﬂuenced the recolonization and recovery of species and populations. The lack of obvious diﬀerences in anthropogenic impacts and to date, climate change, between SRW wintering grounds has led to intrinsic factors, such as migratory culture, extirpation and Allee eﬀects, being inferred as strong drivers and/or inhibitors of recovery. In what is potentially the world’s smallest whale population, the eastern NPRW, demographic stochasticity could play a deciding role in population persistence if the proportion of females in the population does not increase. Finally, the NARW has been reduced to a single, small population that is highly susceptible to extrinsic factors inﬂuencing recovery. The combination of anthropogenic impacts, both lethal and sublethal, and environmental ﬂuctuations, appear to shape health and reproductive success in the NARW (Corkeron et al., 2018). This has only been elucidated by long-term, scientiﬁc monitoring projects using a range of approaches combined with intensive necropsy eﬀorts; next, we discuss how this has worked and how it may be applied elsewhere. The situation is comparatively positive for SRW populations. At its best, <1% of SRW in New Zealand show anthropogenic scarring (W. Rayment, pers. comm.) compared with the 83% for the NARW. Nevertheless, particularly for the smaller, more vulnerable populations there may still be critical impacts. While there have been ten ship-strikes with at least four mortalities and 28 entanglements with two deaths in Australian waters, the fatalities include mother and calf pairs from the small, remnant southeast population (Kemper et al., 2008; Carroll et al., 2015; Lanyon and Janetzki, 2016; Peel et al., 2018; Tulloch et al., unpublished). Vessel strikes to the Brazilian population still occur at low levels (<0.5 per year) but between 1999 and 2014, 38 entanglement cases were reported (Figueiredo et al., 2017; Groch, 2018). In Argentina, 6% of whales identiﬁed in 2007, 2008, 2009 seasons in Golfo Nuevo showed evidence of vessel interactions via scarring (International Whaling Commission [IWC], 2011). The potential for increased negative interactions between recovering right whale populations and increased human development has Inverse Density Dependence: Genetic Diversity and Stochastic Demographic Inﬂuences on Recovery Simulation studies suggest that the centuries-long demographic bottleneck due to whaling did reduce mitochondrial genetic diversity in SRW (Jackson et al., 2008), with Indo-Paciﬁc populations showing signiﬁcantly lower mtDNA diversity than their South Atlantic counterparts (Carroll et al., 2018). However, it is not yet resolved whether this has impacted nuclear DNA diversity, as microsatellite studies show high levels of heterozygosity in the extant SRW wintering grounds (Carroll et al., 2018). In the North Paciﬁc, LeDuc et al. (2012) analyzed 49 biopsy samples from 24 right whales, of which all but one were from the eastern North Paciﬁc. The analysis revealed a male-biased sex Demographic stochasticity in inverse density dependent eﬀects are most likely to enact through sex ratio ﬂuctuations in very small populations and this may be impeding the recovery of the eastern NPRW. Wade et al. (2011b) estimated the eastern NPRW population contained only eight females (95% CI: 7–18) and 20 males (95% CI: 17–37). Fujiwara and Caswell (2001) suggested that demographic stochasticity was less likely to impact the NARW population, but Pace et al. (2017) have shown a strong trend to a male biased sex ratio that is worrisome. The current estimate of breeding females is 105, in a population of 458. Equivalent data for SRW is lacking for the January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 6 Perspectives on Right Whale Research and Conservation Harcourt et al. smallest sub-populations where demographic stochasticity might be enacted. been identiﬁed as an issue of concern in Brazil (Figueiredo et al., 2017), New Zealand (Suisted and Neale, 2004), and eastern Australia (Carroll et al., 2015). Anthropogenic Impacts on Recovery As of 2009, 83% of NARW showed evidence of entanglement; 26% showed new entanglement scars every year, and 59% had been entangled more than once (Knowlton et al., 2012a). Rates of serious injury from entanglement (i.e., those that will likely lead to death, NOAA, 2008) have increased from on average 0.7/year (2000- 2009) to >3/year (2010–2017) (van der Hoop et al., 2014; Hayes et al., 2017). The sublethal eﬀects of entanglements have been quantiﬁed with respect to individual energetic cost (van der Hoop et al., 2016a; van der Hoop et al., 2017a), stress (Hunt et al., 2016; Pettis et al., 2017; Rolland et al., 2017; Lysiak et al., 2018) and reproductive output (Knowlton et al., 2012b), as is their role in combination with ﬂuctuations in prey availability on population level health (Rolland et al., 2016). Frontiers in Marine Science \| www.frontiersin.org Estimating Demographic Parameters Right whales typically have a 3-year reproductive cycle; 1 year gestation, 1 year calving and 1 year resting (Knowlton et al., 1994). Right whale calves grow extremely fast compared with other cetacean species, with SRW calves estimated to grow between 2.2 and 3.5 cm per day (Whitehead and Payne, 1978; Best and Rüther, 1992; Christiansen et al., 2018). During their typically fasting wintering period, lactating females are estimated to lose up to 3.4 cm of blubber thickness (Miller et al., 2011) or 25% of their body volume (Christiansen et al., 2018). NARW have signiﬁcantly thinner blubber than SRW (Miller et al., 2011). Unsurprisingly these factors all demonstrate that female health and body condition are strong determinants of calving rates. Estimating Demographic Parameters The ability to identify and track individuals across time is fundamental to assessments of population abundance and growth using open capture-recapture models for eastern NPRW and New Zealand SRW (Carroll et al., 2011, 2013; Wade et al., 2011b). These models are appropriate when heterogeneity in capture probability can be modeled by sex (i.e., genotype mark- recapture) or where there is no segregation by demographic class (e.g., feeding ground studies). Heterogeneity in female capture probability on wintering grounds is associated with reproductive cycles, i.e., females are more likely to be sighted in calving years, means that standard capture-recapture models cannot be used for estimating abundance or demographic trends in long-term photo-identiﬁcation studies. Instead, stage-structured demographic models that estimate population parameters based on sightings of females in calving years are used to estimate abundance and trend in Argentina and South Africa (Cooke et al., 2015; Brandão et al., 2018). In southwest Australia, the total number of calving females seen in 3 years of aerial surveys has been assumed to represent the total number of reproductive females (see section “Linking Reproductive Success and Health to Environmental and Anthropogenic Factors”), which is then subject to a correction factor to estimate the total population size (Bannister, 2017). Population trends, estimated from the Since 1984, nearly all of the new calves in the NARW population have been observed due to aerial and ship surveys on calving grounds (Browning et al., 2010). Using this comprehensive dataset for years 1988–2008, Rolland et al. (2016) showed that females that transition from resting to pregnant had a signiﬁcantly higher annual mean health score compared with those that did not calve (Rolland et al., 2016). Catalogs The NARW is one of the best-studied whale populations in the world, with concentrated survey eﬀorts in the species’ known habitats since 1979 producing a near-complete census of the entire population (Hamilton et al., 2007). Abundance estimates have typically been produced by calculating the total cataloged population minus the cumulative number of individuals detected or presumed dead (Kraus et al., 2001). However, habitat use patterns have changed in recent years, meaning that capture probabilities have declined. Therefore, Pace et al. (2017) conducted a mark-recapture estimate of abundance of NARW, and showed the population increased at ∼2.8% from 270 whales in 1990 to 482 whales in 2010. Since then, however, there are strong indications that the population is declining, with abundance estimated to be 458 whales in 2015 (95% credible interval: 444-471, Pace et al., 2017). The foundation of long-term monitoring projects is the identiﬁcation of individuals, typically using natural markings, but also DNA proﬁles. In right whales, individuals are identiﬁed using callosity patterns, lip ridges, unusual color or patches, or scarring (Kraus et al., 1986), from both aerial and boat-based photographs (Patenaude et al., 1998). Photo-identiﬁcation studies started on NARW in the United States in the 1960s (Kraus et al., 1986), on SRW in South Africa, southwest Australia, and Argentina in the 1970s (Payne et al., 1981; Best, 1990), in Brazil in the 1980s (Groch et al., 2005), in New Zealand in the 1990s (Patenaude et al., 1998), and in Chile in the 2000s (Vernazzani et al., 2014). NOAA/NMFS began compiling the NPRW catalog in 2008, and this now includes photos from 1979 to present (Kennedy et al., 2012). DNA proﬁles, constructed using a combination of hyper- variable microsatellite markers and sex-speciﬁc loci, have also been used to identify individuals primarily in NARW, NPRW and SRW in the Indo-Paciﬁc region (Frasier et al., 2007; Carroll et al., 2011; Wade et al., 2011b; Carroll et al., 2012, 2015). Introduction The predictable aggregations of right whales that once attracted whalers are now the foundation of some of the longest- running, continuous whale studies in the world. These long- term datasets began with pioneering work that used natural markings to identify individuals and follow them through time (Kraus et al., 1986; Payne, 1986), and have been extended into multi-disciplinary studies (e.g., genetics, health) to gain January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 7 Perspectives on Right Whale Research and Conservation Harcourt et al. a comprehensive understanding of population and individual status. mark-recapture or stage-structured models, or from the linear regression analyses of aerial count data, show that most of the recovering SRW populations have been growing at ∼7% per annum (Table 1). Linking Reproductive Success and Health to Environmental and Anthropogenic Factors The strength of these catalogs is derived from the ability to use individual recognition to record all of the signiﬁcant events that occur in an individual’s life. This is best exempliﬁed in the NARW where sightings locations, entanglement history, body condition, calving, tagging history and many other factors such as health and paternity can and are recorded. Having this detailed information provides us with immensely greater conﬁdence in estimating demographic parameters and linking factors such as entanglement, environmental parameters, foraging success with reproductive output, health status and likelihood of survivorship. Cataloging females through multiple calving events in SRW, and the ability to monitor almost all individual NARW, have provided powerful datasets from which to assess environmental and anthropogenic drivers of reproductive success and health. This, in turn, provides the ability to determine drivers of population growth and persistence. In particular, the low or negative growth rate of NARW stands in contrast to the higher growth rates of SRW (Table 1), and has been explicitly linked to changes in calving intervals between the two species (Corkeron et al., 2018). ADAPTING TO SHIFTING RESOURCES et al., 2011, 2012; Rolland et al., 2016). Recovery of condition post calving may, however, be impeded by external factors and chronic entanglement in ﬁshing gear can result in the consumption of endogenous lipid energy reserves on the same order of magnitude as is consumed during lactation (van der Hoop et al., 2016a). et al., 2011, 2012; Rolland et al., 2016). Recovery of condition post calving may, however, be impeded by external factors and chronic entanglement in ﬁshing gear can result in the consumption of endogenous lipid energy reserves on the same order of magnitude as is consumed during lactation (van der Hoop et al., 2016a). The distribution of right whale species reﬂects evolutionary, demographic, ecological, habitat and anthropogenic processes that come together over various scales of space and time (e.g., Forcada et al., 2012). Within each species, individuals move through diﬀerent habitats that provide appropriate conditions for speciﬁc ecological and physiological needs, often governed by life-history stages or characteristics (e.g., Corkeron and Connor, 1999). All three right whale species show some migratory tendency, though the identiﬁcation of calving, breeding and foraging grounds varies for each species. Moving between these grounds can require multi-month migrations over thousands of kilometers (Figure 1). For the most part these migratory pathways between grounds remain only partially deﬁned and may indeed be very diﬀuse. Within grounds, individuals will move to fulﬁll their needs of foraging, socializing, courtship (Best et al., 1993; Kraus and Hatch, 2001) and calving (Patenaude and Baker, 2001; Elwen and Best, 2004; Torres et al., 2013; Rayment et al., 2015). As reported above, within a season, some (though emphatically not all) individuals may move between diﬀerent habitats that support the same needs (e.g., movement between diﬀerent foraging grounds), depending on proximity and suitability (Vanderlaan, 2010; Schick et al., 2013), and perhaps, experience or cultural knowledge. As foraging success helps determine body condition, work in SRW has investigated the link between foraging ground conditions and reproductive success. Leaper et al. (2006) showed a relationship between breeding success at the Argentinean wintering ground and sea surface temperature (SST) anomalies at their feeding ground in South Georgia. The SST anomalies in late summer to autumn prior to winter conception seemed to have the largest impact, and the authors hypothesized it was due to relationship between SST and krill biomass (Trathan et al., 2003). Estimating Demographic Parameters This and other work shows that the greatest ﬂuctuations in NARW body condition and health scores are linked to the cost of lactation in the calving cycle, with the lowest estimated health scores in resting and lactating females, and blubber thicknesses increasing with the years since the last calving event (Pettis et al., 2004; Miller January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 8 Perspectives on Right Whale Research and Conservation Harcourt et al. Factors Driving Distribution Factors Driving Distribution The single most important habitat feature for the NARW is the occurrence of concentrated patches of copepod prey (Murison and Gaskin, 1989; Wishner et al., 1995; Mayo et al., 2001; Baumgartner and Mate, 2003). Similarly, the probability of detecting NPRW in the Southeastern Bering Sea is strongly associated with the abundance of their primary prey, Calanus marshallae, with patch densities around right whales as high as 105 copepods m−3 (Baumgartner et al., 2013b). Copepod patches result from a combination of physical (supply, removal, and resuspension) and biological (diapause and density) mechanisms occurring across scales. Together, these factors can lead to predictable and persistent, high-energy food sources, or can cause high inter-annual variability in the abundance, concentration, and energy density in the same foraging habitat (Zakardjian, 2003; Jiang et al., 2007; Davies et al., 2013; Davies et al., 2014). The same features that tend to concentrate NARW prey (Calanus spp.) have not necessarily been shown to concentrate NPRW or SRW prey, nor have NPRW or SRW been associated with similar foraging-type environments (Baumgartner et al., 2013b). The latter do not appear to be obligative frontal foragers (Reid et al., 2000), although Mate et al. (2011) reported that some SRW tagged in South Africa were associated with the Subtropical Convergence. There is limited information on foraging behavior or prey in NPRW and SRW (Hamner et al., 1988), though current eﬀorts are underway. For example, D’Agostino et al. (2018) describe seasonal phytoplankton and mesozooplankton dynamics along with environmental factors in the northern Patagonian gulfs in addition to mesozooplankton composition and abundance measured close to foraging SRW. Frontiers in Marine Science \| www.frontiersin.org ADAPTING TO SHIFTING RESOURCES Variation in krill abundance and density, linked with the phase of the Antarctic Oscillation and El Nino events, were also found to be signiﬁcantly correlated with the number of calves observed on the Brazilian wintering ground (Seyboth et al., 2016). g g ( y ) A similar relationship has been found in the NARW, where calving success is strongly correlated with the abundance of its primary prey Calanus ﬁnmarchicus, which in turn is linked to the North Atlantic Oscillation (Greene and Pershing, 2004; Hlista et al., 2009; Meyer-Gutbrod and Greene, 2018). Recent analyses have sought to understand the relative contribution of anthropogenic impacts and food availability on NARW reproduction and population growth (e.g., Meyer-Gutbrod and Greene, 2018). This work oﬀers an important example to understanding right whale population persistence and recovery more generally. It has become particularly relevant on a global scale as the strong population growth in some SRW’s appears to have slowed in recent years, with lower counts since 2015 (Charlton, 2017; Brandão et al., 2018). In Australia this appears to be due to a redistribution of calving females to historical calving areas such as Fowlers Bay (Charlton, 2017). But in South Africa this has been attributed to an increase in the calving interval, with models suggesting females are resting longer than in the past (Brandão et al., 2018). Furthermore, major die oﬀs of calves at the Argentinean wintering ground may reduce population growth in the future, and the drivers of these events remain as yet unknown (Rowntree et al., 2013; Marón et al., 2015). NARW In the western North Atlantic, individual right whales routinely range from Florida to the Gulf of St. Lawrence (Kraus and Rolland, 2007); however, they can be considered “condition- dependent partial migrators” (Krzystan et al., 2018). Recent sightings of “western” NARW have been reported from Iceland, Norway, Portugal and the Azores. The calving ground is in the coastal waters of the southeastern United States during the winter months, although there are at least two records of calves being born around Cape Cod in the spring. Non-calving females, and adult and juvenile males are also observed on the calving grounds (Schick et al., 2013). Winter aggregations also have been observed in the middle of the Gulf of Maine (Cole et al., 2013) and south of Cape Cod (Leiter et al., 2017), but the location of most of the non-calving whales is unknown. Historical whaling records suggest a winter-spring presence along the United States Mid- Atlantic coast (NY – DE; Reeves et al., 1999). In the spring, aggregations of right whales occur in Massachusetts Bay, and sometimes in the great South Channel east of Cape Cod. In the summer and fall, right whales are observed in the Bay of Fundy (between Maine and Nova Scotia), in Roseway Basin (50 km south of Nova Scotia), and recently in the southern Gulf of St. Lawrence. Photo-identiﬁcation and some limited satellite tagging data show that seasonal movements into, out of, between, and around the so-called critical habitats are frequent and extensive (Baumgartner and Mate, 2005; Brillant et al., 2015). There is signiﬁcant substructuring within the NARW population. Schaeﬀ et al. (1993) inferred from genetic and photo-identiﬁcation data that about one-third of the cows in the population do not bring their calves to the Bay of Fundy nursery area. This suggests that another summer and fall nursery area may exist, although its location is unknown. Studies of social and cultural drivers remain in their infancy, currently restricted to SRW (Valenzuela et al., 2009; Carroll et al., 2015). Application to NARW may help characterize site ﬁdelity and transmission of migratory patterns, e.g., Fundy vs. non- Fundy females in NARW (Schaeﬀet al., 1993; Malik et al., 1999), and the segregation of mature males to, e.g., Browns Bank and Roseway Basin prior to 1993 (Clapham et al., 1999). Contemporary Distribution SRW The SRW are circumpolar in the Southern Hemisphere between about 12◦S and 65◦S, albeit with a large historical and contemporary discontinuation between New Zealand and Chile. The austral winter distribution of SRW (Figure 1) includes the New Zealand sub-Antarctic Auckland and Campbell Islands (Carroll et al., 2013; Torres et al., 2017), with far fewer whales reported around mainland New Zealand (Carroll et al., 2014). In Australian coastal waters, SRW occur along the southern coastline including Tasmania, generally as far north as Sydney on the east coast and Perth on the west coast (Carroll et al., 2011). Similarly, SRW migrate up the east coast of South America to Brazil to breed in coastal shallows (De Oliveira et al., 2009), and along the west coast through Chile into Peru as far as 12◦S (and possibly 4◦S) (Aguayo-Lobo et al., 2008; Van Waerebeek et al., 2009), and on both the east and west coasts of southern Africa (Best et al., 1993; Mate et al., 2011, Figure 1). When in coastal habitat, SRW are usually within 2 km of shore and distinctly clumped in aggregation areas (Figure 1). SRWs tend to group into diﬀerent demographic class along the coast, with cow- calf pairs and single adults or socializing groups using diﬀerent sections of the coast or region, e.g., cow-calf pairs concentrate in Port Ross, Auckland Islands, while Campbell Island seems to contain more single whales (Torres et al., 2017). Synthesis Increases in the number of cow-calf pairs displaces social groups and solitary individuals to other habitat in Argentina (Rowntree, 2001; Arias et al., 2018) New Zealand (Carroll et al., 2014; Torres et al., 2017), and Australia (Charlton, 2017). Frontiers in Marine Science \| www.frontiersin.org Synthesis Long-term monitoring studies have linked environmental factors such as prey availability and its proxies, as well as anthropogenic impacts, with right whale health, body condition and reproductive success. The analyses conducted in NARW using over three decades of detailed sightings histories of individually identiﬁed whales showcases the way forward for explaining the impact of environmental variation and human impacts on reproduction and population persistence. As some SRW populations begin to experience a decline in reproductive success with population-level implications, we suggest adopting a similar strategy and in particular targeting eﬀorts toward monitoring those that appear most vulnerable. Given that SRW travel between feeding grounds shared by more than one population, we also recommend developing a uniﬁed catalog strategy between regions and research groups. January 2019 \| Volume 5 \| Article 530 9 Perspectives on Right Whale Research and Conservation Harcourt et al. and feeding grounds is known from photo-identiﬁcation with movement from southwest Australia to south of 60◦(Bannister et al., 1999), and between Península Valdés and South Georgia (Best et al., 1993). Illegal Soviet whaling in the 1960s and 1970s supports these inferences, as SRWs were killed during the austral summer around South Georgia and other high latitude areas (Tormosov et al., 1998). Stomach content data from these whales showed that SRW were feeding on copepods north of 40◦S and euphausiids south of 50◦S and a mixture in between (Tormosov et al., 1998). The summer distribution of contemporary SRW is one of the least well understood aspects of the species’ biology. Calving areas share many features between SRW and NARW. Mother-calf pairs select shallow areas closer to shore (Winn et al., 1986; Elwen and Best, 2004; Keller et al., 2012; Rayment et al., 2012; Rayment et al., 2015), with calm waters (Winn et al., 1986; Rowntree et al., 2001), protection from ocean swell and seasonal winds (Elwen and Best, 2004), and soft substrates, when compared to unaccompanied or non-calving individuals. Historical and contemporary calving grounds for either NPRW stock remain a mystery. y y Changes in distribution of diﬀerent demographic classes have been linked with changes in density in multiple SRW populations. Increases in the number of cow-calf pairs displaces social groups and solitary individuals to other habitat in Argentina (Rowntree, 2001; Arias et al., 2018) New Zealand (Carroll et al., 2014; Torres et al., 2017), and Australia (Charlton, 2017). NPRW The seasonal distribution of NPRW is largely understood through analysis of historical catch records and sightings data (Miyashita and Kato, 1998; Clapham et al., 2004), along with more recent visual and acoustic survey eﬀort since the early 2000s. Critical Habitat for the eastern NPRW has been established in the southeast Bering Sea and in a small portion of the Gulf of Alaska based on consistent sightings in these areas (NOAA, 2008). Aerial and shipboard surveys, satellite tracking and acoustic monitoring eﬀorts are largely focused in the southeast Bering Sea and conﬁrm it is an important summer foraging area (Wade et al., 2006, 2011a; Zerbini et al., 2015). There is considerably less survey eﬀort in In summer SRW probably forage between about 40◦S and 65◦S, including Magellan Strait and along the Antarctic Peninsula (Aguayo-Lobo et al., 2008; Table 1). Foraging and summer distribution have been determined from multiple sightings (Table 1); stable isotope studies (e.g., Rowntree et al., 2001; Valenzuela et al., 2009) and a small number of satellite tracking studies (Mate et al., 2011, see section “Migratory Culture, Extirpation and Recovery”). Direct movement between wintering January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 10 Perspectives on Right Whale Research and Conservation Harcourt et al. the northern Gulf of Alaska compared to the southeast Bering Sea, but individual sightings and acoustic detections (Waite et al., 2003; Mellinger et al., 2004; Allen and Angliss, 2012; Širovi´c et al., 2015) including further south (Ford et al., 2016) suggest that alternative foraging grounds may exist. Eastern NPRW arrive in the southeast Bering Sea as early as May, move into mid- shelf waters throughout the summer, and are detected through December (Munger et al., 2008). Visual sightings and year- round acoustic detections of NPRW in Unimak Pass suggest this region may be used as a corridor between summer foraging and unknown winter grounds (Wright, 2015, 2016; Wright et al., 2017). rates, and sound production varies between habitats (Matthews et al., 2001; Parks et al., 2011). NPRW Few historical sightings suggest that in winter, eastern NPRW ranged from between the Hawaiian Islands to the west coast of North America (Klumov, 1962) and for western NPRW, from oﬀshore Kamchatka south to the Yellow Sea (Omura et al., 1969; Scarﬀ, 1986; Brownell et al., 2001). Contemporary sightings remain within these ranges, and acoustic detections indicate winter residence in Alaskan waters (Wright, 2015, 2016). Mitochondrial DNA comparisons suggest that the populations in the eastern and western North Paciﬁc are genetically distinct (Pastene et al., 2018). Calving areas remain unknown, and have never been reported by whalers; modeling studies have identiﬁed coastal areas in California, Hawaii, southern China and northern Vietnam as potential calving regions of based on habitat features (Good and Johnston, 2010). There is a single photo-identiﬁcation match between Hawaii and the Bering Sea (Kennedy et al., 2012). NPRW Photo identiﬁcation of individuals has been used to demonstrate longer-range movements of all three right whale species (e.g., Knowlton et al., 1992; Best et al., 1993; Pirzl et al., 2009; Kennedy et al., 2012), as have genetic data in SRW (Carroll et al., 2014), but these paired observations essentially only provide information on the presence of the individual at both places at both times, usually lacking information on residence time, and how the animal moved from A to B (e.g., Whitehead, 2001). The diﬀerent right whale catalogs contain individual sightings at scales from <1 day to 17 years (Hamilton et al., 2007). Eighty percent of sightings in the NARW catalog occur in consecutive years (Hamilton et al., 2007), compared to 3 years in SRW (Bannister, 2017; Charlton, 2017). In fall and spring, NPRW sightings and detections are much more diﬀuse across the species’ range from South Korea, Japan, and the Okhotsk Sea through British Columbia and California (Brownell et al., 2001; Clapham et al., 2004; Ford et al., 2016), and acoustic detections are less common (Munger et al., 2008; Wright, 2015, 2016). Various satellite telemetry systems have been developed for large whales and have been deployed on right whales since the mid 1980s (Mate et al., 1992, 1997; Goodyear, 1996; Andrews et al., 2008). The technology provides information at a higher temporal resolution compared to visual sightings and has been useful in conﬁrming that right whales can travel long distances over short periods of time between subsequent sightings in the same habitat (e.g., Mate et al., 1992), as well as tracking movements between known summer and winter habitat areas (Mate et al., 1997, 2011; Childerhouse et al., 2010; Zerbini et al., 2016). All three right whale species have been satellite tagged (Table 1). Tag systems have evolved over decades (Mate et al., 2007; Andrews et al., 2008) improving attachment mechanisms and increasing deployment duration, which can now last >200 days (Zerbini et al., 2016). However, tag impacts must be evaluated in the context of the value of the information, and continued improvement of attachment mechanisms is necessary to address health and welfare concerns (Moore et al., 2012; Best et al., 2015; Moore and Zerbini, 2017; Norman et al., 2018). Frontiers in Marine Science \| www.frontiersin.org Going From Sightings and Distributions to Individual Movements The ability to identify individuals (see section “Catalogs” above) and follow the change in location of subsequent sightings allows for the transition from describing general population movement, habitat use, and arrival times down to the resolution of speciﬁc individuals. Tracking individual movements provides a greater understanding of behaviors (Mayo and Marx, 1990), decisions, and corridors (Schick et al., 2009; Zerbini et al., 2016), and allows for movements to be predicted across a range of scales (Pendleton et al., 2012; Brillant et al., 2015, 2017). For example, Schick et al. (2013) characterized transition probabilities between broad geographic regions (i.e., feeding and breeding habitats) that diﬀered between male and female NARW. Do similar diﬀerences in movement exist based on health or entanglement status (van der Hoop et al., 2016a)? Predicting how and when individuals, especially of diﬀerent age, sex or life stage, distribute throughout the species’ range has implications for marine spatial planning (Vanderlaan, 2010; van der Hoop et al., 2014). How individual movements scale to reﬂect the overall distribution of populations and species are open questions in movement Many methods have been implemented to study the distribution of species and movement of individuals within and between populations (Table 1). These methods provide insight at diﬀerent spatial and temporal scales and rely on diﬀerent types of observations or samples. For example, biopsy, necropsy or sloughed skin can provide material for genetic analysis to determine identity, sex, parentage and population membership and stable isotope analysis to ascertain foraging location and trophic level. Baleen can also provide material for stable isotope analysis, providing more detailed histories of individual foraging locations (Best and Schell, 1996; Rowntree, 2001; Valenzuela et al., 2009). Acoustic detections provide the approximate location of a whale, but only when it is calling and within the detection range of the hydrophone, but then will provide evidence of presence of one or more whales (Mellinger et al., 2004, 2007, 2011; Wade et al., 2006; Van Parijs et al., 2009; Širovi´c et al., 2015) and seasonal distribution in space and in time (Morano et al., 2012; Davis et al., 2017). Behavior is the primary determinant of right whale calling January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 11 Perspectives on Right Whale Research and Conservation Harcourt et al. TABLE 2 \| Global right whale research priorities. Going From Sightings and Distributions to Individual Movements Action Information gained Threats addressed Technology Population/species Global catalog – total population assessment Demography, Calving interval, Philopatry, Population mixing, Population trends, Movements, Distribution, Body condition/health status Entanglement, Vessel strike, Resource availability, Loss of habitat (climate change, coastal development) Photo-ID, Machine learning, Genomics Existing for NARW, NPRW. Priority to coordinate SRW across ocean basins Global catalog-genomic Philopatry, Population mixing, Population trends, Pedigree, Genetic/genomic diversity (functional and neutral), Reproductive skew Emerging diseases, Climate change effects, Cumulative threats Genetic/genomic proﬁle matching, Metagenomics, Epigenomics (aging) Existing for NARW, NPRW. Priority to improve coordination SRW across ocean basins Global catalog-health Compromised recovery, Catalog diseases, Identify vectors, Morbidity/mortality causality, Genomic assessment of susceptibility Entanglement, Emerging diseases, Climate change effects Drones, Metagenomics, Microbes, Hormones, Biopsies, Post-mortems, Genomics Existing for NARW. Priority for NPRW, improve coordination SRW across ocean basins Coordinate acoustic monitoring repositories Population monitoring, distribution Acoustic exposure PAM, Gliders, Detection and classiﬁcation algorithms, Accessible repositories All ocean basins Develop universal population models Population recovery, prioritize threats Individual threats, Cumulative threats Cumulative, individual-based and population matrix models, stable isotopes All Develop models of individual drivers Individual movements, habitat requirements, foraging needs, energy acquisition and expenditure Resource availability, Climate change, Industry impacts Individual-based models, Sighting surveys, Biologging to improve empirical measurements of speciﬁc activities All Monitor low-recovery regions Individual movements, Habitat requirements, Patterns of recovery Resource availability, Climate change, Industry impacts Sightings surveys, Remote sensing, eDNA, PAM, Photo ID/genomic proﬁles All TABLE 2 \| Global right whale research priorities. Entanglement, Vessel strike, Resource availability, Loss of habitat (climate change, coastal development) Genetic/genomic proﬁle matching, Metagenomics, Epigenomics (aging) Emerging diseases, Climate change effects, Cumulative threats Drones, Metagenomics, Microbes, Hormones, Biopsies, Post-mortems, Genomics Entanglement, Emerging diseases, Climate change effects Resource availability, Climate change, Industry impacts prey (Baumgartner and Mate, 2003). Depth-related time budgets have been eﬀective in assessing the risk of entanglement in ﬁshing gear or vessel strike (Parks et al., 2012; Baumgartner et al., 2017) as well as assessing potential avenues to reduce those risks (Nowacek et al., 2004). Diving behavior and ﬁne- scale body orientation have been linked to diﬀerences in total buoyancy and changes in body condition (Nowacek et al., 2001; Nousek-McGregor et al., 2013), while tags have also been used to resolve changes in diving behavior, biomechanics and thrust production with entanglement in ﬁshing gear (van der Hoop et al., 2013b, 2017b; Table 1). Fine-Scale Movement and Orientation Detailed measurements of how right whales orient and behave, especially in relation to prey patches and anthropogenic threats, e.g., ﬁshing gear or near the surface, can be diﬃcult to obtain. Detailed ﬁeld observations (Mayo and Marx, 1990) have provided careful measurements of swimming paths and turning angles in traveling vs. foraging modes, and at a ﬁner scale than oﬀered by satellite telemetry. Similar measurements in other environments, populations or species are rare, yet these values are critical in informing modeling approaches for, e.g., spatio-temporal risk assessment (van der Hoop et al., 2012). We know that obstacle avoidance (Kraus et al., 2014), prey patch exploitation and decision making occur at these ﬁne scales. Frontiers in Marine Science \| www.frontiersin.org Going From Sightings and Distributions to Individual Movements The ﬁne-scale 3D movement, orientation, and behavior of right whales in the context of variable prey densities, conspeciﬁcs, or anthropogenic disturbances or threats is poorly known, especially for SRW and NPRW. For SRW, this is linked to poorly described and typically oﬀshore and somewhat inaccessible location of foraging grounds. ecology. For example, our understanding of SRW connectivity, and transition probabilities, would be highly informed by cross-catalog matching. Measuring large-scale and mesoscale movements with the approaches described above will help reveal individual residency and movement patterns and clarify how these scale to observed population distributions (Table 2). Passive Acoustic Monitoring shifts in distribution are diﬃcult to detect when annual sightings are incredibly low. For example, Tynan et al. (2001) reported changes in eastern NPRW distribution and prey selection with climate forcing; however, the prey species were no diﬀerent than those observed in the historical record of NPRW (Shelden et al., 2005), and the observed distribution of whales was likely the result of inadequate survey coverage (Wade et al., 2006, 2011b; Zerbini et al., 2015). Sightings presence is a poor indicator when annual sightings are fewer than ten individuals. The development and improvement of tools to collect, store, and analyze underwater acoustic recordings have expanded their scope and application. Archival and real-time systems can be low-cost solutions to monitor cetacean presence at the mesoscale (Van Parijs et al., 2009, Figure 3). Fixed, autonomous acoustic recorders can operate continuously for months to years, through poor weather and darkness, and have proven useful in detecting eastern NPRW in known and new habitats (Mellinger et al., 2004; Širovi´c et al., 2015; Wright et al., 2018). This greatly enhances aerial and vessel surveys (Rone et al., 2012; Baumgartner, 2014), helps identify changes in overall distribution, and arrival to and departure from speciﬁc habitats (Van Parijs et al., 2009; Davis et al., 2017). The coordination of many smaller-scale acoustic programs into larger-scale networks (Van Parijs et al., 2015) and post hoc analyses can describe changes at large scales, despite technical challenges in unifying datasets with varying duty cycles or deployment durations (Davis et al., 2017). However, it remains a challenge to estimate the number of individuals (and which individuals) or animal density from acoustic detections that reﬂect only pseudo-presence at variable detection ranges (but see Marques et al., 2011). Only recently have we begun to understand and quantify the plasticity of migration in NARW. In recent years, fewer individuals have been sighted on the calving grounds (Krzystan et al., 2018), whereas more individuals are overwintering in northern habitats, e.g., Cape Cod Bay and the Gulf of Maine (Bort et al., 2015; Davis et al., 2017; Mayo et al., 2018). Other causal factors have been identiﬁed and described across a range of scales. Passive Acoustic Monitoring For example, NARW were almost completely absent from their Roseway Basin feeding habitat in the mid-to-late 1990s (Brownell et al., 2001; Kenney et al., 2001; Figure 3 in van der Hoop et al., 2012), whereas over 100 individuals have been sighted in that habitat in other years. This multi- year abandonment has been linked to interannual variation in copepod abundance (Patrician and Kenney, 2010), attributed to the volume, strength, and density of water masses transported to the basin (Davies et al., 2013, 2014). Studies emphasize the importance of prey mapping, modeling, and in situ and satellite- derived oceanographic conditions (Baumgartner et al., 2003; Jiang et al., 2007; Hlista et al., 2009; Krumhansl et al., 2018) at resolutions that match eﬀort-corrected whale detections, as well as coordination between researchers in these disciplines. The development and implementation of detection and classiﬁcation algorithms has helped automate species identiﬁcation and ease data burden (Gillespie and Caillat, 2008; Baumgartner and Mussoline, 2011) and can be used for real-time monitoring for risk management and mitigation. For example, the Right Whale Listening Network created following the proposed building of deep-water ports for liqueﬁed natural gas oﬄoading near right whale critical habitat is comprised of 10 right whale automatic detection buoys, provided near-real time notiﬁcations to LNG vessels (Wiley et al., 2013). Other similar real- or near-real-time transmission systems have been deployed (e.g., in the New York Bight), and communicate by, e.g., twitter (@robots4whales) and the Whale Alert App. Real-time moored systems are being considered for other high-density areas in the NARW range with high vessel density (e.g., the Gulf of St. Lawrence, Canada) and could be implemented in similar habitat for SRW such as west of Melbourne, Victoria, Australia near the only known SRW calving habitat for the at-risk remnant south eastern population. Large scale changes in the overall distribution of NARW from 2004 to 2014 have also been resolved via acoustic monitoring. Davis et al. (2017) compiled acoustic records from Florida, United States to Greenland, and documented an increased presence of NARW in the mid-Atlantic and simultaneous decrease in the Gulf of Maine; these changes coincided with seasonal sightings data for the regions. Autonomous Platforms Incorporation of these types of acoustic monitoring devices and detection and classiﬁcation algorithms into autonomous ocean platforms (i.e., gliders Figure 3) has extended the duration, season, and track-line distance of sampling and surveying ability through poor weather (vs. shipboard surveys), while sampling simultaneous relevant in situ oceanographic information. Gliders have successfully detected ﬁn, sei, humpback, and NARW in habitats ranging from the Gulf of Maine to the Gulf of St. Lawrence1 (Baumgartner et al., 2013a; Davis et al., 2016). Gliders have also been deployed in the Chukchi sea area, primarily to detect bowhead, beluga, and ﬁn whales (Baumgartner et al., 2014b); however, this region is further north than the most Small-scale vertical and horizontal distributions of diving right whales correspond to the vertical and horizontal distributions of their prey. Collocated measurements of diving behavior (see sections “Automated Image Analysis” and “Biologgers”) and oceanographic and prey-ﬁeld measurements have been useful in characterizing inter- and intra-annual changes in foraging activity in NARW (Winn et al., 1995; Baumgartner et al., 2017), but little similar work has been carried out in SRW or NPRW. 1http://dcs.whoi.edu Passive Acoustic Monitoring From acoustic data, it is diﬃcult to assess how movements of individuals within the population have led to these measured changes, or the strength of various drivers; however, such studies emphasize how the combination of many smaller-scale acoustic projects or programs can provide a larger-scale assessment of seasonality and habitat use, and changes therein. Know? “Any eﬀective study of the ecology, behavior, or patterns of distribution of western North Atlantic right whales must include studies of C. ﬁnmarchicus.” – Winn et al. (1995). Bio-logging and bio-telemetry tags have been used to assess NARW diving behavior, speciﬁcally comparing diﬀerent diving modes that may be associated with foraging, as well as the tracking of diel vertically migrating prey layers (Winn et al., 1995; Goodyear, 1996, Table 1; see section “Biologgers”). Simultaneous use of echosounders and tags can record the depth to which animals dive with respect to the vertical distribution of their Predicting how movements at any scale will change with environmental change requires a mechanistic understanding of the processes that drive movement and distribution. It also requires suﬃcient sightings or detections to quantify change; January 2019 \| Volume 5 \| Article 530 12 Perspectives on Right Whale Research and Conservation Harcourt et al. Unmanned Aerial Systems Aerial photographs from manned aircraft have long been used to identify and measure NARW, SRW, and eastern NPRW (Best and Rüther, 1992; Miller et al., 2011; Clapham et al., 2012). The recent development of small unmanned aerial systems (UAS) with lightweight, high-resolution cameras, provides improved image resolution at lower cost and risk compared to manned ﬂight (Figure 3; Durban et al., 2015; Koski et al., 2015). UAS photogrammetry and sampling has recently been applied to SRW (Apprill et al., 2017; Christiansen et al., 2018) and NARW (Moore et al., 2017) to address a variety of research questions. One direct beneﬁt of UAS photography is obtaining high-resolution aerial images that can be used to identify individuals. UAS imagery represents a small but growing proportion of the NARW and SRW catalogs. This could be expanded in future if the technology is adapted to more challenging weather conditions seen in high latitudes. Technological Advances Application of new and developing methodologies will help answer questions regarding how right whales adapt to their changing environment. Here we discuss several of these. January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 13 Perspectives on Right Whale Research and Conservation Harcourt et al. FIGURE 3 \| Current and future threats to (left) and methods to study (right) global right whale populations. FIGURE 3 \| Current and future threats to (left) and methods to study (right) global right whale populations. in regions where other survey types (e.g., aerial) may be diﬃcult due to logistical considerations (Figure 3). frequented eastern NPRW habitats. Autonomous gliders could prove useful in surveying contemporary and historical right whale habitats around the world, especially in the context of oceanographic features. Satellite Imagery It is possible to detect and count individual animals in Very High Resolution (VHR) satellite imagery available either from open source platforms (e.g., Google Earth Imagery) or private companies (e.g., Digital globe; Cubaynes et al., 2018) (Figure 3). Fretwell et al. (2014) identiﬁed at least 55 SRW in a single image of the calving grounds of Golfo Nuevo, Península Valdéz, Argentina. This method can detect right whales where real-time sightings are available to validate detections, where weather conditions are favorable for cloud-free imagery, and where whales are frequently observed at or just below the surface. Satellite imagery suﬀers from availability biases in deeper waters, as dive times and surface detectability may diﬀer across locations, seasons, activity patterns and demographic groups. This method might also prove useful for detecting beach-cast mortalities in remote environments. Frontiers in Marine Science \| www.frontiersin.org Synthesis Despite our best eﬀorts to sight, photograph, record, sample, and track individual right whales, and to synthesize these observations into general consensus on the species’ ecology and distributions, individuals will continue to surprise us. NARW have given birth in Cape Cod, a generalized springtime foraging ground (Patrician et al., 2009), and NARW and SRW show out- of-season foraging (Mayo and Marx, 1990; Best and Schell, 1996; Kenney et al., 2001; Hoﬀmeyer et al., 2010). All species and populations show rare long-distance movements, and acoustic detections indicate that our understanding of all species’ seasonal distributions and contemporary ranges is always changing. This highlights the continued need to survey, identify, and acoustically monitor right whale habitats year-round, and in all weather conditions, expanding survey eﬀorts beyond current methods (Table 2). g g ( g ) Dolphins and seals are vulnerable to viral infection, with serious pandemics of inﬂuenza virus (Cox and Subbarao, 2000; Groth et al., 2014), poxvirus (Fiorito et al., 2015) and morbillivirus (Sierra et al., 2014, 2016). Although no pandemics have been reported for large whales, as right whales undertake large scale migrations they could be the source of, and vectors for, potentially serious zoonoses (Bogomolni et al., 2008). Their inshore movements into urbanized coastal waters further increase their vulnerability. As Gulland and Hall (2007) point out, morbidity and mortality in marine mammals have resulted from terrestrial pathogens spreading to the ocean, from harmful algal blooms to epidemics of virulent viruses and bacteria (Geraci et al., 1999; Miller et al., 2002). Brucellosis, a pathogen that aﬀects reproduction in mammals, has been reported in whales (Van Bressem et al., 2009), with potential for widespread population consequences. SRW adults and calves in Argentinian waters that suﬀer injuries from kelp gull attacks (Rowntree et al., 1998; Sironi et al., 2009), are exposed to pathogens such as Klebsiella, Salmonella, and Shigella conﬁrmed in kelp gull feces (Fazio et al., 2012). Necropsies with pathology studies and metagenomic surveys of viral and bacterial diversity are therefore important for developing a greater understanding of the impact of oceanic pathogens (Geoghegan et al., 2018) on right whales’ population biology. Quantifying changes in movement patterns at any temporal or spatial scale requires data at suﬃcient resolution, a mechanistic understanding, and long-term measures of potential drivers. Whaling records and sightings provide a reasonable foundation for many basic questions on the populations and their movements. Synthesis Combing these sightings with individual identiﬁcation (by genetics or visible marking), isotopes, acoustics, and simultaneous oceanographic measurements or models will increase our understanding of where animals are, why they are there, and where they might go next. Synthesizing data from across diﬀerent populations is particularly important in this context. There may be local (wintering or feeding ground) or global (climate change) drivers that can be inferred from such a comparative approach, particularly in SRW where there are multiple wintering grounds with long-term monitoring projects (Table 2). In NARW, there has been a long-standing visual health assessment program that uses individual identiﬁcation photographs to monitor individual whale health (Pettis et al., 2004). This tool has been applied to predictions of outcomes for animals that are entangled (Pettis et al., 2017), and has been used to model the health of the population and individual demographic groups (Rolland et al., 2016). These methods could easily be applied to other right whale populations where photo-identiﬁcation catalogs are maintained (Figure 3). Biologgers Instrumenting animals with high-resolution inertial sensors (e.g., accelerometers, magnetometers, and gyroscopes; Figure 3) has provided detailed information on their ﬁne-scale movements and biomechanics in foraging patches. Time-depth recorders have been deployed on NARW (Winn et al., 1995; Baumgartner and Mate, 2003; Baumgartner et al., 2017) and SRW (Argüelles et al., 2016), while tags with additional sensors (acoustic and/or inertial, e.g., DTAGs; Johnson and Tyack, 2003) have been deployed on NARW (Nowacek et al., 2001, 2004; Parks et al., 2011; Nousek- McGregor et al., 2013; van der Hoop et al., 2017b, Table 1). Eﬀorts to bridge the gap toward high-resolution, medium-duration tags with minimally invasive attachment are underway (Baumgartner et al., 2014a; Johnson et al., 2014; Mate et al., 2017). In aquatic mammals, contemporary assessments of marine mammal health are almost invariably biased toward animals whose health is already compromised, such as stranded animals (Gulland and Hall, 2007; Hunt et al., 2015), which in part reﬂects the diﬃculties in sampling aquatic environments. Yet identiﬁcation of chronic changes requires monitoring of free- living animals. There are several tools available and more are emerging (Figure 3). Environmental DNA (eDNA) Automated and semi-automated computer image analysis methods have been increasingly used to detect whales in images, and match whales to a catalog of known individuals (Hillman et al., 2008; Kniest et al., 2010) including SRW (Hiby and Lovell, 2001) and NARW (Bogucki et al., 2018). Bogucki et al. (2018) identiﬁed individual NARW with 87% accuracy using a series of convolutional neural networks in a fully automated pipeline with no need for user input. These computer vision identiﬁcation methods may increase the eﬃciency of matching, and enable the processing of larger volumes of data, especially from diﬀerent sources (still images and video taken from aircraft, UAS, and land and shipboard surveys). These methods could be especially helpful in coordinating catalogs from diﬀerent populations ( ) Ecological questions are now being addressed using environmental sampling of water in combination with high through-put sequencing. Work in marine mammals shows that the method can be used to identify species (Foote et al., 2012) and even genetically distinct killer whale ecotypes up to 2 h after the species was present in an area (Baker et al., 2018). Population-level data are also able to be obtained from aggregations of marine animals, as shown by work in whale sharks (Sigsgaard et al., 2016), which could be useful in the context of right whale wintering ground aggregations. In the future, eDNA could be used to detect the presence of right whales, in combination with other systematic survey methods, or January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 14 Perspectives on Right Whale Research and Conservation Harcourt et al. and right whale species to assess connectivity, ranging and movement. shortages, and climate change, particularly when subjected to extrinsic stressors such as ﬁshing gear entanglements, pollution, and acoustic disturbance (Figure 3). Both independently and in synergy these stressors may have an eﬀect on infectious disease transmission dynamics (Gulland and Hall, 2007), and they are also likely to aﬀect reproduction and the potential for mortality. Although stress is beneﬁcial in the short-term to respond to and recover from a stimulus or perceived threat, remaining in a stressed state in the long-term is maladaptive. Maintaining this unbalanced ‘allostatic state’ can impair future stress response and fecundity and result in immune and other physiological dysfunction (McEwen and Wingﬁeld, 2003; McEwen and Wingﬁeld, 2010; Houser et al., 2016). January 2019 \| Volume 5 \| Article 530 COMPARATIVE SYNTHESIS AND CUMULATIVE EFFECTS The lack of a cumulative eﬀects analytical approach impedes any understanding of multi-stressor conditions aﬀecting marine mammals, especially those in more industrialized sections of ocean (Figure 3). It also hampers eﬀective management, as it is impossible to know that managing stressor A is a better approach than managing stressor B, if stressor B is exacerbated by interactions with stressors C, D, and E. “If cumulative eﬀects cannot be accounted for, then unexpected adverse impacts from interactions between stressors pose a risk to marine mammal populations and the marine ecosystems on which people and marine mammals depend” (National Academies of Sciences, and Engineering and Medicine, 2017). Without such a framework every stressor is managed in isolation, wasting resources, and reducing eﬃcacy. There are many marine mammal examples. In harbor porpoise, Hall et al. (2006) showed that previous exposure to PCBs increased the risk of death from infectious diseases. In killer whales, Ford et al. (2010) suggest high POP concentrations may have acted synergistically with limited food resources, increasing mortality during times of low prey abundance. In baleen whales, the poster child for the cumulative eﬀect problem is the NARW. As discussed earlier, the NARW is subject to multiple anthropogenic stressors that are well studied, and multiple environmental stressors that are known, although the eﬀects are not fully understood (Figure 3). During whaling, body condition measurements included measures of girth and blubber thickness, with ﬁne-scale measurements of percent lipid, protein, and ash contents in diﬀerent parts of the body (Lockyer et al., 1985). More recently approximations to body condition have used indirect estimates of buoyancy (e.g., with bio-logging tags; Nowacek et al., 2001; Nousek-McGregor et al., 2013), direct measurements of blubber thickness with pole-mounted ultrasound (Miller et al., 2011) or body shape changes from photogrammetry, from aerial photographs (Best and Rüther, 1992; Miller et al., 2012) including from drones (Moore et al., 2017; Christiansen et al., 2018). y ( g ) Early work in this area was the development of a conceptual model of the Population Consequences of Acoustic Disturbance (PCAD), with subsequent models discussed by New et al. (2014); Fleishman et al. (2016), and Pirotta E. et al. (2018). Further developments in the ﬁeld resulted in the National Academies of Sciences, and Engineering and Medicine (2017) report, and another conceptual model called Population Consequences of Multiple Stressors (PcoMS: National Academies of Sciences, and Engineering and Medicine, 2017). METHODS FOR ASSESSING FACTORS AFFECTING RIGHT WHALE HEALTH Fecal collection is largely opportunistic and not viable for animals that are only available during fasting periods (e.g., most SRW). A potentially more accessible and repeatable method is to use whale respiratory vapor to assess hormone levels, disease organisms, and microbiome information. Burgess et al. (2018) demonstrated the feasibility of using pole- mounted collection of whale blow for hormone analyses with the caveat that sample concentrations must be normalized as an unknown amount of salt water is mixed with every exhalation (Burgess et al., 2018). UAS collection provides the opportunity to sample while keeping the boat further away from individual whales, Pirotta et al. (2017) used a drone to collect an extraordinary diversity and abundance of microbiota from the blow in migrating populations of humpback whales, clearly and unambiguously distinguishing microbial communities from surrounding control seawater and air samples (Pirotta et al., 2017). Drones such as these can sample whale blow and provide an assessment of respiratory bacteria, lipids, proteins, and even viruses (Apprill et al., 2017; Pirotta et al., 2017; Geoghegan et al., 2018). As with all new techniques, further reﬁnement is needed to determine how consistently representative bacterial and viral samples may be for the animal, and therefore its health. Nevertheless, this area shows great promise and is evolving rapidly. 2012b). Future research along this cascade includes clarifying the interactions between stress, energetic demand, nutrition, and metabolism, and ultimately the population consequences of those stressors (Figure 3). Frontiers in Marine Science \| www.frontiersin.org METHODS FOR ASSESSING FACTORS AFFECTING RIGHT WHALE HEALTH Another approach is the use of various matrices to assess reproductive and stress hormones in feces of individual right whales (Hunt et al., 2015). Hormonal proﬁles can reveal Marine mammals are vulnerable to a variety of intrinsic stressors, including infectious diseases, harmful algal blooms, prey January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 15 Perspectives on Right Whale Research and Conservation Harcourt et al. reproductive status (Rolland et al., 2005; Corkeron et al., 2017), increases in stress due to chronic entanglement (Rolland et al., 2017) and evidence of chronic stress due to shipping noise (NRC, 2005; Hunt et al., 2006; Rolland et al., 2012; Hunt et al., 2015). The ﬁndings of this program highlight the need to monitor free-ranging animals. Fecal collection is largely opportunistic and not viable for animals that are only available during fasting periods (e.g., most SRW). A potentially more accessible and repeatable method is to use whale respiratory vapor to assess hormone levels, disease organisms, and microbiome information. Burgess et al. (2018) demonstrated the feasibility of using pole- mounted collection of whale blow for hormone analyses with the caveat that sample concentrations must be normalized as an unknown amount of salt water is mixed with every exhalation (Burgess et al., 2018). UAS collection provides the opportunity to sample while keeping the boat further away from individual whales, Pirotta et al. (2017) used a drone to collect an extraordinary diversity and abundance of microbiota from the blow in migrating populations of humpback whales, clearly and unambiguously distinguishing microbial communities from surrounding control seawater and air samples (Pirotta et al., 2017). Drones such as these can sample whale blow and provide an assessment of respiratory bacteria, lipids, proteins, and even viruses (Apprill et al., 2017; Pirotta et al., 2017; Geoghegan et al., 2018). As with all new techniques, further reﬁnement is needed to determine how consistently representative bacterial and viral samples may be for the animal, and therefore its health. Nevertheless, this area shows great promise and is evolving rapidly. reproductive status (Rolland et al., 2005; Corkeron et al., 2017), increases in stress due to chronic entanglement (Rolland et al., 2017) and evidence of chronic stress due to shipping noise (NRC, 2005; Hunt et al., 2006; Rolland et al., 2012; Hunt et al., 2015). The ﬁndings of this program highlight the need to monitor free-ranging animals. CONCLUSION The name right whale was not a misnomer. The three right whale species were easy to exploit and all driven very close to extinction. Only a few populations of the SRW can be considered to have recovered well and the status of the eastern NPRW and NARW along with some populations of the SRW remain precarious. In this review we have outlined the manner in which the right whales were exploited and how we have monitored their recovery. The fragile status and in particular the recent reversal in recovery trajectory by NARW call for urgent action. Yet the close phylogenetic relationship and ecological equivalence between the three species may provide fruitful avenues for avoiding catastrophe. Insights from careful examination of the populations that have recovered, combined with approaches that draw on novel methods developed separately for diﬀerent populations may provide guiding principles for future conservation research, to hopefully make these the right whales to survive. The name right whale was not a misnomer. The three right whale species were easy to exploit and all driven very close to extinction. Only a few populations of the SRW can be considered to have recovered well and the status of the eastern NPRW and NARW along with some populations of the SRW remain precarious. In this review we have outlined the manner in which the right whales were exploited and how we have monitored their recovery. The fragile status and in particular the recent reversal in recovery trajectory by NARW call for urgent action. Yet the close phylogenetic relationship and ecological equivalence between the three species may provide fruitful avenues for avoiding catastrophe. Insights from careful examination of the populations that have recovered, combined with approaches that draw on novel methods developed separately for diﬀerent populations may provide guiding principles for future conservation research, to hopefully make these the right whales to survive. The ﬁndings of such cumulative impact models, and its components, could be integrated into international legislation to ensure the continued survival of the right whales. Today, the hunting of whales is controlled by the International Whaling Commission (IWC), and the trade in whale products is governed by the Convention on the International Trade of Endangered Species of wild fauna and ﬂora (CITES). REFERENCES Baker, C. S., Patenaude, N. J., Bannister, J. L., Robins, J., and Kato, H. (1999). Distribution and diversity of mtDNA lineages among southern right whales (Eubalaena australis) from Australia and New Zealand. Mar. Biol. 134, 1–7. doi: 10.1007/s002270050519 Aguayo-Lobo, A., Acevedo, J., Brito, J. L., Olavarría, C., Moraga, R., and Olave, C. (2008). La ballena franca del sur, Eubalaena australis (Desmoulins, 1822) en aguas chilenas: análisis de sus registros desde 1976 a 2008. Rev. Biol. Mar. Oceanogr. 43, 653–668. doi: 10.4067/S0718-19572008000300024 Baker, C. S., Steel, D., Nieukirk, S., and Klinck, H. (2018). Environmental DNA (eDNA) From the wake of the whales: droplet digital PCR for detection and species identiﬁcation. Front. Mar. Sci. 5:133. doi: 10.3389/fmars.2018. 00133 g Allee, W. C. (1931). Animal Aggregations A Study in General Sociology. Chicago, IL: University of Chicago press. doi: 10.5962/bhl.title.7313 University of Chicago press. doi: 10.5962/bhl.title.7313 Bannister, J. (2017). Project A7 - Monitoring Population Dynamics of ‘Western’ Right Whales oﬀSouthern Australia 2015–2018 Progress Report on activities for 2017. Muscatine, IA: Hon Associate. y g Allen, B. M., and Angliss, R. P. (2012). Alaska Marine Mammal Stock Assessments NPRW NOAA-TM-AFSC-245. Washington, DC: NOAA. NPRW NOAA-TM-AFSC-245. Washington, DC: NOAA. Allen, G. M. (1916). The whalebone whales of New England. Mem. Boston Soc. Nat. Hist. 8, 107–322. Bannister, J. L., Pastene, A., and Burnell, S. R. (1999). First record of movement of a southern right whale (Eubalaena australis) between warm water breeding grounds and the Antarctic Ocean, south of 60◦S. Mar. Mamm. Sci. 15, 1337–1342. doi: 10.1111/j.1748-7692.1999.tb00895.x Allendorf, F. W. (1986). Genetic drift and the loss of alleles versus heterozygosity. Zoo Biol. 5, 181–190. doi: 10.1002/zoo.1430050212 Andrews, R., Pitman, R., and Ballance, L. (2008). Satellite tracking reveals distinct movement patterns for type b and type c killer whales in the southern Ross Sea, Antarctica. Polar Biol. 31, 1461–1468. doi: 10.1007/s00300-008-0487-z Barendse, J., and Best, P. B. (2014). Shore-based observations of seasonality, movements, and group behavior of southern right whales in a nonnursery area on the South African west coast. Mar. Mamm. Sci. 30, 1358–1382. doi: 10.1111/mms.12116 Apprill, A., Miller, C. A., Moore, M. J., Durban, J. W., Fearnbach, H., and Barrett- Lennard, L. G. (2017). Extensive core microbiome in drone-captured whale blow supports a framework for health monitoring. mSystems 2:00119-17. doi: 10.1128/mSystems.00119-17 Barratclough, A., Jepson, P. D., Hamilton, P. K., Miller, C. A., Wilson, K., and Moore, M. J. (2014). CONCLUSION In addition, the Convention on the Conservation of Migratory Species of Wild Animals (CMS), which operates under the auspices of the United Nations Environmental Programme, provides a cross-border mechanism for protecting right whales throughout their migratory networks. All three species are listed under Appendix I of the CMS convention, and could be subject to a special agreement under the treaty. Such agreements should promote co-ordinated conservation and management plans across the range of species, in addition to promoting conservation and restoration plans for important habitat areas2. This would link together national AUTHOR CONTRIBUTIONS All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. 2https://www.cms.int/en/cms-instruments/agreements COMPARATIVE SYNTHESIS AND CUMULATIVE EFFECTS Both of these models are dependent upon a series of transfer functions that describe how exposure to stressors aﬀects individual behavior or health, which is deﬁned as all factors that comprise “ﬁtness” or homeostasis. Variations in health can aﬀect individual vital rates, and the variation in severity or level of exposure to a given stressor can be used to scale up to predictions of population level eﬀects. National Academies of Sciences, and Engineering and Medicine (2017) provided a decision tree for thinking about cumulative eﬀects analyses, and also recommended stepwise, hypothesis driven, adaptive management strategies to deal with the uncertainties inherent in most marine mammal/stressor interactions. The NAS Committee concluded that current scientiﬁc knowledge is not up to the task of predicting cumulative eﬀects of diﬀerent combinations of stressors on marine mammal populations. Harmful algal blooms and marine pollutants remain a concern. In NARW whales feeding close inshore are repeatedly exposed to multiple environmental neurotoxins produced by marine algae (Durbin et al., 2002; Doucette et al., 2012). Less well understood across the three species is the impact of marine pollutants: NARW are exposed to low but chronic levels. In this context, understanding the level of genomic diversity in populations and susceptibility of individuals to pathogens, harmful algal blooms and toxins will also help understand individual and population level impacts (Leroy et al., 2017). Finally, in the case of animals that have been entangled, the increased drag, altered behavior, and additional energy demand, leads to decreases in body condition (Cassoﬀet al., 2011; Moore et al., 2013; van der Hoop et al., 2013b; Barratclough et al., 2014; van der Hoop et al., 2016b), elevated stress hormones (Rolland et al., 2017), poor health condition (Rolland et al., 2016; Pettis et al., 2017) and reduced reproductive success (Knowlton et al., January 2019 \| Volume 5 \| Article 530 16 Perspectives on Right Whale Research and Conservation Harcourt et al. Nevertheless, there are many things that we can do to enhance future PcoMS or other cumulative eﬀects models. A comparative synthesis of biological parameters, known ecological parameters, and know anthropogenic threats, using data from many well- studied right whale populations, can inform ongoing and future cumulative eﬀect analyses on this taxon and we outline research priorities to move toward this goal as we see them (Table 2). level legislation to provide a web of protection for the species around the world. REFERENCES M. (1996). Stable isotopes in southern right whale (Eubalaena australis) baleen as indicators of seasonal movements, feeding and growth. Mar. Biol. 124, 483–494. doi: 10.1007/BF00351030 Carroll, E. L., Fewster, R., Childerhouse, S., Patenaude, N., Boren, L., and Baker, C. S. (2016). First direct evidence for natal wintering ground philopatry and estimate of juvenile survival in the New Zealand Southern right whale Eubalaena australis. PLoS One 11:e0146590. doi: 10.1371/journal.pone.0146590 Best, P. B. (1990). Natural markings and their use in determining calving intervals in right whales oﬀSouth Africa. S. Afr. J. Zool. 25, 114–123. doi: 10.1080/ 02541858.1990.11448199 Carroll, E. L., Patenaude, N., Alexander, A., Steel, D., Harcourt, R., Childerhouse, S., et al. (2011). Population structure and individual movement of southern right whales around New Zealand and Australia. Mar. Ecol. Prog. Ser. 432, 257–268. doi: 10.3354/meps09145 Best, P. B., Mate, B., and Lagerquist, B. (2015). Tag retention, wound healing, and subsequent reproductive history of southern right whales following satellite- tagging. Mar. Mamm. Sci. 31, 520–539. doi: 10.1111/mms.12168 Carroll, E. L., Rayment, W. J., Alexander, A. M., Baker, C. S., Patenaude, N. J., Steel, D., et al. (2014). Reestablishment of former wintering grounds by New Zealand southern right whales. Mar. Mamm. Sci. 30, 206–220. doi: 10. 1111/mms.12031 Best, P. B., Payne, R., Rowntree, V., Palazzo, J. T., and Both, M. D. C. (1993). Long- range movements of South Atlantic right whales Eubalaena australis. Mar. Mamm. Sci. 9, 227–234. doi: 10.1111/j.1748-7692.1993.tb00451.x Cassoﬀ, R. M., Moore, K. M., McLellan, W. A., Barco, S. G., Rotstein, D. S., and Moore, M. J. (2011). Lethal entanglement in baleen whales. Dis. Aquat. Organ. 96, 175–185. doi: 10.3354/dao02385 Bogomolni, A. L., Gast, R. J., Ellis, J. C., Dennett, M., Pugliares, K. R., Lentell, B. J., et al. (2008). Victims or vectors: a survey of marine vertebrate zoonoses from coastal waters of the Northwest Atlantic. Dis. Aquat. Org. 81:13. doi: 10.3354/dao01936 Charlton, C. M. (2017). Southern Right Whale (Eubalaena australis) Population Demographics in Southern Australia. Ph.D. thesis, Curtin University, Bentley. Bogucki, R., Cygan, M., Khan, C. B., Klimek, M., Milczek, J. K., and Mucha, M. (2018). Applying deep learning to right whale photo identiﬁcation. Conserv. Biol. doi: 10.1111/cobi.13226 [Epub ahead of print]. doi: 10.1111/cobi. 13226 Childerhouse, S., Double, M. C., and Gales, N. J. (2010). Satellite Tracking of Southern Right Whales (Eubalaena australis) at the Auckland Islands, New Zealand. Washington, DC: International Whaling Commission, 5. Bort, J., Van Parijs, S. REFERENCES How much does a swimming, underweight, entangled right whale (Eubalaena glacialis) weigh? Calculating the weight at sea, to facilitate accurate dosing of sedatives to enable disentanglement. Mar. Mamm. Sci. 30, 1589–1599. doi: 10.1111/mms.12132 Argüelles, M. B., Fazio, A., Fiorito, C., Pérez-Martínez, D., Coscarella, M., and Bertellotti, M. (2016). Diving behavior of Southern right whales (Eubalaena australis) in a maritime traﬃc area in Patagonia, Argentina. Aquat. Mamm. 42, 104–108. doi: 10.1578/AM.42.1.2016.104 Baumgartner, M., and Mussoline, S. E. (2011). A generalized baleen whale call detection and classiﬁcation system. J. Acoust. Soc. Am. 129, 2889–2902. doi: 10.1121/1.3562166 Arias, M., Coscarella, M., Romero, M. A., Sueyro, N., Svendsen, G. M., Crespo, E. A., et al. (2018). “Southern right whale Eubalaena australis recolonizes golfo san Matías (Patagonia, Argentina),” in Proceedings of Annual Meeting of the International Whaling Commission SC, Patagonia, 24. Baumgartner, M. F. (2014). Real-time passive acoustic monitoring of baleen whales from autonomous platforms. J. Acoust. Soc. Am. 136, 2116–2117. doi: 10.1121/ 1.4816406 Baker, C. S., and Clapham, P. J. (2004). Modelling the past and future of whales and whaling. Trends Ecol. Evol. 19, 365–371. doi: 10.1016/j.tree.2004. 05.005 Baumgartner, M. F., Cole, T. V. N., Clapham, P. J., and Mate, B. R. (2003). North Atlantic right whale habitat in the lower Bay of Fundy and on the January 2019 \| Volume 5 \| Article 530 17 Frontiers in Marine Science \| www.frontiersin.org Perspectives on Right Whale Research and Conservation Harcourt et al. SW Scotian Shelf during 1999-2001. Mar. Ecol. Prog. Ser. 264, 137–154. doi: 10.3354/meps264137 Brownell, R. J., and Mallette, S. D. (2018). Global Baleen Whale Bycatch: The Most Threatened Population. Ulsan: International Whaling Commission, 12. Browning, C. L., Rolland, R. M., and Kraus, S. D. (2010). Estimated calf and perinatal mortality in western North Atlantic right whales (Eubalaena glacialis). Mar. Mamm. Sci. 26, 648–662. Baumgartner, M. F., Fratantoni, D. M., Hurst, T. P., Brown, M. W., Cole, T. V., Van Parijs, S. M., et al. (2013a). Real-time reporting of baleen whale passive acoustic detections from ocean gliders. J. Acoust. Soc. Am. 134, 1814–1823. doi: 10.1121/1.4816406 Burdin, A., Nikulin, V., and Brownell, R. Jr. (2004). Cases of Entanglement of Western North-Paciﬁc Right Whales (Eubalaena japonica) in ﬁshing gear: serious threat for species survival Marine mammals of the Holarctic. Moscow: KMK Scientiﬁc Press, 95–97. Baumgartner, M. F., Lysiak, N. S. J., Carter Esch, H., Zerbini, A. N., Berchok, C. L., and Clapham, P. J. (2013b). REFERENCES Associations between North Paciﬁc right whales and their zooplanktonic prey in the southeastern Bering Sea. Mar. Ecol. Prog. Ser. 490, 267–284. doi: 10.3354/meps10457 Burgess, E. A., Hunt, K. E., Kraus, S. D., and Rolland, R. M. (2018). Quantifying hormones in exhaled breath for physiological assessment of large whales at sea. Sci. Rep. 8:10031. doi: 10.1038/s41598-018-28200-8 Baumgartner, M. F., Hammar, T., Robbins, J., and Kurle, C. (2014a). Development and assessment of a new dermal attachment for short-term tagging studies of baleen whales. Methods Ecol. Evol. 6, 289–297. doi: 10.1111/2041-210X.12325 Carroll, E. L., Alderman, R., Bannister, J., Bérubé, M., Best, P., Boren, L., et al. (2018). Incorporating non-equilibrium dynamics into demographic history inferences of a migratory marine species. Heredity 122, 53–68. doi: 10.1038/ s41437-018-0077-y Baumgartner, M. F., Staﬀord, K. M., Winsor, P., Statscewich, H., and Fratantoni, D. M. (2014b). Glider-based passive acoustic monitoring in the Arctic. Mar. Technol. Soc. J. 48, 40–51. doi: 10.4031/MTSJ.48.5.2 g D. M. (2014b). Glider-based passive acoustic monitoring in the Arctic. Mar. Technol. Soc. J. 48, 40–51. doi: 10.4031/MTSJ.48.5.2 Baumgartner, M. F., and Mate, B. R. (2003). Summertime foraging ecology of North Atlantic right whales. Mar. Ecol. Progress Ser. 264, 123–135. doi: 10.3354/ meps264123 Carroll, E. L., Baker, C. S., Watson, M., Alderman, R., Bannister, J., Gaggiotti, O. E., et al. (2015). Cultural traditions across a migratory network shape the genetic structure of southern right whales around Australia and New Zealand. Sci. Rep. 5:16182. doi: 10.1038/srep16182 Baumgartner, M. F., and Mate, B. R. (2005). Summer and fall habitat of North Atlantic right whales (Eubalaena glacialis) inferred from satellite telemetry. Can. J. Fish. Aquat. Sci. 62, 527–543. doi: 10.1139/f04-238 Carroll, E. L., Childerhouse, S., Fewster, R., Patenaude, N., Steel, D., Dunshea, G., et al. (2013). Accounting for female reproductive cycles in a superpopulation capture–recapture framework. Ecol. Appl. 23, 1677–1690. doi: 10.1890/12- 1657.1 Baumgartner, M. F., Wenzel, F. W., Lysiak, N. S. J., and Patrician, M. R. (2017). North Atlantic right whale foraging ecology and its role in human-caused mortality. Mar. Ecol. Prog. Ser. 581, 165–181. doi: 10.3354/meps12315 Carroll, E. L., Childerhouse, S. J., Christie, M., Lavery, S., Patenaude, N., Alexander, A., et al. (2012). Paternity assignment and demographic closure in the New Zealand southern right whale. Mol. Ecol. 21, 3960–3973. doi: 10.1111/ j.1365-294X.2012.05676.x Best, P., and Rüther, H. (1992). Aerial photogrammetry of southern right whales Eubalaena australis. J. Zool. 228, 595–614. doi: 10.1111/j.1469-7998.1992. tb04458.x Best, P., and Schell, D. REFERENCES M., Stevick, P. T., Summers, E., and Todd, S. (2015). North Atlantic right whale Eubalaena glacialis vocalization patterns in the central Gulf of Maine from October 2009 through October 2010. Endanger. Species Res. 26, 271–280. doi: 10.3354/esr00650 Christiansen, F., Vivier, F., Charlton, C., Ward, R., Amerson, A., Burnell, S., et al. (2018). Maternal body size and condition determine calf growth rates in southern right whales. Mar. Ecol. Prog. Ser. 592, 267–281. doi: 10.3354/ meps12522 Braham, H. W., and Rice, D. W. (1984). The right whale, Balaena glacialis. Mar. Fish. Rev. 46, 38–47. Clapham, P. J., Aguilar, A., and Hatch, L. T. (2008). Determining spatial and temporal scales for management: lessons from whaling. Mar. Mamm. Sci. 24, 183–201. doi: 10.1111/j.1748-7692.2007.00175.x Brandão, A., Vermeulen, E., Ross-Gillespie, A., Findlay, K., and Butterworth, D. S. (2018). Updated application of a Photo-Identiﬁcation Based Assessment Model to Southern Right Whales in South African Waters, Focussing on Inferences to be Drawn from a Series of Appreciably Lower Counts of Calving Females Over 2015 to 2017. Ulsan: International Whaling Commission. Clapham, P. J., Good, C., Quinn, S. E., Reeves, R. R., Scarﬀ, J. E., and Brownell, R. L. Jr. (2004). Distribution of North Paciﬁc right whales (Eubalaena japonica) as shown by 19th and 20th century whaling catch and sighting records. J. Cetacean Res. Manag. 6, 1–6. to 2017. Ulsan: International Whaling Commission. Clapham, P. J., Kennedy, A. S., Rone, B. K., Zerbini, A. N., Crance, J. L., and Berchok, C. L. (2012). North Paciﬁc right whales in the southeastern Bering Sea. Final Rep. Bureau Ocean Energy Manag. 144. Brillant, S. W., Vanderlaan, A. S. M., Rangeley, R. W., and Taggart, C. T. (2015). Quantitative estimates of the movement and distribution of North Atlantic right whales along the northeast coast of North America. Endanger. Species Res. 27, 141–154. doi: 10.3354/esr00651 Clapham, P. J., Young, S., and Brownell, R. J. (1999). Baleen whales: conservation issues and the status of the most endangered populations. Mamm. Rev. 29, 35–60. doi: 10.1046/j.1365-2907.1999.00035.x Brillant, S. W., Wimmer, T., Rangeley, R. W., and Taggart, C. T. (2017). A timely opportunity to protect North Atlantic right whales in Canada. Mar. Policy 81, 160–166. doi: 10.1016/j.marpol.2017.03.030 Cole, T. V. N., Hamilton, P., Henry, A. G., Duley, P., Pace, R. M., White, B. N., et al. (2013). Evidence of a North Atlantic right whale Eubalaena glacialis mating ground. Endanger. Species Res. 21, 55–64. doi: 10.3354/esr00507 Brownell, R. REFERENCES L., Svendsen, G. M., Degrati, M., and Coscarella, M. A. (2018). The southwestern Atlantic southern right whale, Eubalaena australis, population is growing but at a decelerated rate. Mar. Mamm. Sci. 35, 93–107. doi: 10.1111/mms.12526 activities. Mar. Mamm. Sci. 32, 1004–1021. doi: 10.1111/mms.1 Foote, A. D., Thomsen, P. F., Sveegaard, S., Wahlberg, M., Kielgast, J., Kyhn, L. A., et al. (2012). Investigating the potential use of environmental DNA (eDNA) for genetic monitoring of marine mammals. PLoS One 7:e41781. doi: 10.1371/ journal.pone.0041781 Cubaynes, H. C., Fretwell, P. T., Bamford, C., Gerrish, L., and Jackson, J. A. (2018). Whales from space: four mysticete species described using new VHR satellite imagery. Mar. Mamm. Sci. (in press). doi: 10.1111/mms.12544 Forcada, J., Trathan, P. N., Boveng, P. L., Boyd, I. L., Burns, J. M., Costa, D. P., et al. (2012). Responses of Antarctic pack-ice seals to environmental change and increasing krill ﬁshing. Biol. Conserv. 149, 40–50. doi: 10.1016/j.biocon.2012.02. 002 D’Agostino, V. C., Degrati, M., Santinelli, N., Sastre, V., Dans, S. L., and Hoﬀmeyer, M. S. (2018). The seasonal dynamics of plankton communities relative to the foraging of the southern right whale (Eubalaena australis) in northern Patagonian gulfs, Península Valdés, Argentina. Cont. Shelf Res. 164, 45–57. doi: 10.1016/j.csr.2018.06.003 Ford, J. K. B., Ellis, G. M., Olesiuk, P. F., and Balcomb, K. C. (2010). Linking killer whale survival and prey abundance: food limitation in the oceans’ apex predator? Biol. Lett. 6, 139–142. doi: 10.1098/rsbl.2009.0468 Dautenhahn, K., and Nehaniv, C. L. (2002). Imitation in Animals and Artifacts. Cambridge, MA: MIT Press. doi: 10.7551/mitpress/3676.001.0001 Ford, J. K. B., Pilkington, J. F., Gisborne, B., Frasier, T. R., Abernethy, R. M., and Ellis, G. M. (2016). Recent observations of critically endangered North Paciﬁc right whales (Eubalaena japonica) oﬀthe west coast of Canada. Mar. Biodivers. Rec. 9:50. doi: 10.1186/s41200-016-0036-3 Davies, K. T. A., Ross, T., and Taggart, C. T. (2013). Tidal and subtidal currents aﬀect deep aggregations of right whale prey, Calanus spp., along a shelf-basin margin. Mar. Ecol. Prog. Ser. 479, 263–282. doi: 10.3354/meps10189 Fortune, S. M. E., Trites, A. W., Mayo, C. A., Rosen, D. A. S., and Hamilton, P. K. (2013). Energetic requirements of North Atlantic right whales and the implications for species recovery. Mar. Ecol. Prog. Ser. 478, 253–272. doi: 10. 3354/meps10000 Davies, K. T. A., Taggart, C. T., and Smedbol, R. K. (2014). REFERENCES J., Clapham, P. J., Miyashita, T., and Kasuya, T. (2001). Conservation status of North Paciﬁc right whales. J. Cetacean Res. Manag. 2, 269–286. January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 18 Perspectives on Right Whale Research and Conservation Harcourt et al. Cooke, J., Rowntree, V., and Sironi, M. (2015). Southwest Atlantic Right Whales: Interim Updated Population Assessment from Photo-Id Collected at Península Valdéz, Argentina, Washington, DC: International Whaling Commission. Calanus ﬁnmarchicus. Harmful Algae 1, 243–251. doi: 10.1016/S1568-9883(02) 00046-X Elwen, S. H., and Best, P. (2004). Environmental factors inﬂuencing the distribution of southern right whales (Eubalaena australis) on the south coast of South Africa I: broad scale patterns. Mar. Mamm. Sci. 20, 567–582. doi: 10.1111/j.1748-7692.2004.tb01180.x Corkeron, P., Hamilton, P., Bannister, J., Best, P., Charlton, C., Groch, K. R., et al. (2018). The recovery of North Atlantic right whales, Eubalaena glacialis, has been constrained by human-caused mortality. R. Soc. Open Sci. 5:180892. doi: 10.1098/rsos.180892 Fazio, A., Bertellotti, M., and Villanueva, C. (2012). Kelp gulls attack Southern right whales: a conservation concern? Mar. Biol. 159, 1981–1990. doi: 10.1371/ journal.pone.0139291 Corkeron, P., Rolland, R. M., Hunt, K. E., and Kraus, S. D. (2017). A right whale pootree: classiﬁcation trees of faecal hormones identify reproductive states in North Atlantic right whales (Eubalaena glacialis). Conserv. Physiol. 5:cox006. doi: 10.1093/conphys/cox006 Figueiredo, G. C., Marcos, C. D. O., Siciliano, S., and Moura, J. F. (2017). Southern right whales (Eubalaena australis) in an urbanized area oﬀthe Southwestern Atlantic Ocean: Updated records and conservation issues. Aquat. Mamm. 43:52. doi: 10.1578/AM.43.1.2017.52 Corkeron, P. J., and Connor, R. C. (1999). Why do baleen whales migrate? Mar. Mamm. Sci. 15, 1228–1245. doi: 10.1111/j.1748-7692.1999.tb00887.x Courchamp, F., Clutton-Brock, T., and Grenfell, B. (1999). Inverse density dependence and the Allee eﬀect. Trends Ecol. Evol. 14, 405–410. doi: 10.1016/ S0169-5347(99)01683-3 Fiorito, C., Palacios, C., Golemba, M., Bratanich, A., Argüelles, M. B., Fazio, A., et al. (2015). Identiﬁcation, molecular and phylogenetic analysis of poxvirus in skin lesions of southern right whale. Dis. Aquat. Organ. 116, 157–163. doi: 10.3354/dao02918 Cox, N., and Subbarao, K. (2000). Global epidemiology of inﬂuenza: past and present. Annu. Rev. Med. 51, 407–421. doi: 10.1146/annurev.med.51.1.407 Fleishman, E., Costa, D. P., Harwood, J., Kraus, S., Moretti, D., New, L. F., et al. (2016). Monitoring population-level responses of marine mammals to human activities. Mar. Mamm. Sci. 32, 1004–1021. doi: 10.1111/mms.12310 Crespo, E. A., Pedraza, S. N., Dans, S. REFERENCES Water mass structure deﬁnes the diapausing copepod distribution in a right whale habitat on the Scotian Shelf. Mar. Ecol. Prog. Ser. 497, 69–85. doi: 10.3354/meps 10584 Davis, G. E., Baumgartner, M. F., Bonnell, J. M., Bell, J., Berchok, C., Bort Thornton, J., et al. (2017). Long-term passive acoustic recordings track the changing distribution of North Atlantic right whales (Eubalaena glacialis) from 2004 to 2014. Sci. Rep. 7:13460. doi: 10.1038/s41598-017-13 359-3 Fowler, C. W. (1984). Density Dependence in Cetacean Populations. Report of the International Whaling Commission. Silver Spring, MD: NOAA, 373–379. Frasier, T. R., Hamilton, P. K., Brown, M. W., Conger, L. A., Knowlton, A. R., Marx, M. K., et al. (2007). Patterns of male reproductive success in a highly promiscuous whale species: the endangered North Atlantic right whale. Mol. Ecol. 16, 5277–5293. doi: 10.1111/j.1365-294X.2007.03570.x Davis, R., Baumgartner, M., Comeau, A., Cunningham, D., Davies, K., Furlong, A., et al. (2016). “September tracking whales on the scotian shelf using passive acoustic monitoring on ocean gliders,” in Oceans 2016 MTS/IEEE Monterey, (Piscataway, NJ: IEEE), 1–4. Fretwell, P. T., Staniland, I. J., and Forcada, J. (2014). Whales from space: counting southern right whales by satellite. PLoS One 9:e88655. doi: 10.1371/journal. pone.0088655 Fujiwara, M., and Caswell, H. (2001). Demography of the endangered North Atlantic right whale. Nature 414, 537–541. doi: 10.1038/35107054 Dawbin, W. H. (1986). Right whales caught in waters around south eastern Australia and New Zealand during the nineteenth and early twentieth centuries. Rep. Int. Whaling Comm. 10, 261–267. Geoghegan, J. L., Pirotta, V., Harvey, E., Smith, A., Buchmann, J. P., Ostrowski, M., et al. (2018). Virological sampling of inaccessible wildlife with drones. Viruses 10:E300. doi: 10.3390/v10060300 De Oliveira, L. R., Ott, P. H., Flores, P. A., Siciliano, S., De Almeida, R. S., and Bonatto, S. L. (2009). First molecular estimate of sex-ratio of southern right whale calves, Eubalaena australis, for Brazilian waters. J. Mar. Biol. Assoc. U. K. 89, 1003–1007. doi: 10.1017/S0025315409003178 Geraci, J. R., Harwood, J., and Lounsbury, V. J. (1999). Conservation and Management of Marine Mammals, eds R. John, Jr. Twiss, and R. Randall Reeves (Washington, DC: Smithsonian Institution Press), 471. Doucette, G. J., Mikulski, C. M., King, K. L., Roth, P. B., Wang, Z., Leandro, L. F., et al. (2012). Endangered North Atlantic right whales (Eubalaena glacialis) experience repeated, concurrent exposure to multiple environmental neurotoxins produced by marine algae. Environ. Res. 112, 67–76. REFERENCES A note on an automated system for matching the callosity patterns on aerial photographs of southern right whales. J. Cetacean Res. Manag. 2, 291–295. Katona, S. K., and Kraus, S. D. (1999). “Eﬀorts to conserve the North Atlantic right whale,” in Conservation and Management of Marine Mammals, eds J. R. Twiss and R. R. Reeves (Washington, DC: Smithsonian Institution Press), 311–331. Hillman, G. R., Trask, K. D., Sweeney, K. L., Davis, A. R., Koski, W. R., Mocklin, J., et al. (2008). Photo-Identiﬁcation Software for Bowhead Whale Images SC/60/BRG24. Cambridge: International Whaling Commission Scientiﬁc Committee. Keith, S. A., and Bull, J. W. (2017). Animal culture impacts species’ capacity to realise climate-driven range shifts. Ecography 40, 296–304. doi: 10.1111/ecog. 02481 Hlista, B. L., Sosik, H. M., Martin Traykovski, L. V., Kenney, R. D., and Moore, M. J. (2009). Seasonal and interannual correlations between right-whale distribution and calving success and chlorophyll concentrations in the Gulf of Maine, USA. Mar. Ecol. Prog. Ser. 394, 289–302. doi: 10.3354/meps08267 Keller, C. A., Garrison, L., Baumstark, R., Ward-Geiger, L. I., and Hines, E. (2012). Application of a habitat model to deﬁne calving habitat of the North Atlantic right whale in the southeastern United States. Endanger. Species Res. 18, 73–87. doi: 10.3354/esr00413 Mar. Ecol. Prog. Ser. 394, 289–302. doi: 10.3354/meps08267 Hoﬀmeyer, M. S., Lindner, M. S., Carribero, A., Fulco, V. K., Meneìndez, M. C., Fernaìndez Severini, M. D., et al. (2010). Planktonic food and foraging of Eubalaena australis, on Peniìnsula Valdeìs (Argentina) nursery ground. Rev. Biol. Mar. Oceanogr. 45, 131–139. doi: 10.4067/S0718-19572010000 100013 Kemper, C., Coughran, D., Warneke, R., Pirzl, R., Watson, M., Gales, R., et al. (2008). Southern right whale (Eubalaena australis) mortalities and human interactions in Australia, 1950-2006. J. Cetacean Res. Manag. 10, 1–8. Kennedy, A. S., Salden, D. R., and Clapham, P. J. (2012). First high- to low-latitude match of an eastern North Paciﬁc right whale (Eubalaena japonica). Mar. Mamm. Sci. 28, E539–E544. doi: 10.1111/j.1748-7692.2011. 00539.x Houser, D. S., Champagne, C. D., Crocker, D. E., Kellar, N. M., Cockrem, J., Romano, T., et al. (2016). Natural Variation in Stress Hormones, Comparisons Across Matrices, and Impacts Resulting From Induced Stress in the Bottlenose Dolphin The Eﬀects of Noise on Aquatic Life II. New York, NY: Springer, 467–471. Kenney, R. D., Mayo, C. A., and Winn, H. E. (2001). Migration and foraging strategies at varying spatial scales in western North Atlantic right whales: a review of hypotheses. REFERENCES doi: 10.1016/ j.envres.2011.09.010 Gillespie, D., and Caillat, M. (2008). Statistical classiﬁcation of odontocete clicks. Can. Acoust. 36, 20–26. doi: 10.1121/1.4927695 Givens, G. H., Edmondson, S. L., George, J. C., Suydam, R., Charif, R. A., Rahaman, A., et al. (2016). Hawthorne-Thompson whale abundance estimation adjusting for uncertain recapture, temporal availability variation and intermittent eﬀort. Environmetrics 27, 134–146. doi: 10.1002/env.2379 j Du Pasquier, T. (1986). Catch history of French right whaling mainly in the South Atlantic. Rep. Int. Whaling Comm. 10, 268–274. Durban, J. W., Fearnbach, H., Barrett-Lennard, L. G., Perryman, W. L., and Leroi, D. J. (2015). Photogrammetry of killer whales using a small hexacopter launched at sea. J. Unman. Veh. Syst. 3, 131–135. doi: 10.1139/juvs-2015-0020 Goddard, P. D., and Rugh, D. J. (1998). A group of right whales seen in the Bering Sea in July 1996. Mar. Mamm. Sci. 14, 344–349. doi: 10.1111/j.1748-7692.1998. tb00725.x Good, C., and Johnston, D. (2010). “Spatial modeling of optimal North Paciﬁc right whale (Eubalaena japonica) calving habitats,” in Proceedings of the Alaska Marine Science Symposium, Hotel Captain Cook, Anchorage, 71. Durbin, E., Teegarden, G., Campbell, R., Cembella, A., Baumgartner, M. F., and Mate, B. R. (2002). North Atlantic right whales, Eubalaena glacialis, exposed to paralytic shellﬁsh poisoning (PSP) toxins via a zooplankton vector, January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 19 Perspectives on Right Whale Research and Conservation Harcourt et al. (Eubalaena glacialis) match known calving history. Conserv. Physiol. 4:cow014. doi: 10.1093/conphys/cow014 Goodyear, J. D. (1996). Signiﬁcance of Feeding Habitats of North Atlantic Right Whales Based on Studies of Diel Behaviour, Diving, Food Ingestion Rates, and Prey. Ph.D. thesis. University of Guelph, Guelph, ON. Hunt, K. E., Rolland, R. M., and Kraus, S. D. (2015). Conservation Physiology of an uncatchable animal: The North Atlantic Right Whale (Eubalaena glacialis). Integr. Comp. Biol. 55, 577–586. doi: 10.1093/icb/icv071 Greene, C. H., and Pershing, A. J. (2004). Climate and the conservation biology of North Atlantic right whales: the right whale at the wrong time? Front. Ecol. Environ. 2, 29–34. doi: 10.2307/3868292 g p Hunt, K. E., Rolland, R. M., Kraus, S. D., and Wasser, S. K. (2006). Analysis of fecal Hunt, K. E., Rolland, R. M., Kraus, S. D., and Wasser, S. K. (2006). Analysis of fecal glucocorticoids in the North Atlantic right whale (Eubalaena glacialis). Gen. Comp. Endocrinol. 148, 260–272. doi: 10.1016/j.ygcen.2006.03.012 Gregr, E. J., and Coyle, K. O. (2009). REFERENCES L., and Smith, T. D. (2016). An integrated approach to historical population assessment of the great whales. R. Soc. Open Sci. 3:150669. doi: 10.1098/rsos.150669 Hakamada, T., and Matsuoka, K. (2016). The number of blue, ﬁn, humpback, North Paciﬁc right whales in the western North Paciﬁc in the JARPN II oﬀshore survey area. Paper Presented at the SC/F16/JR13 Presented to the IWC SC JARPNII Review Workshop, Tokyo. Jacobs, E., Duﬀy, M., Magolan, J., Galletti Vernazzani, B., Cabrera, E., Landea, R., et al. (2018). First acoustic recordings of critically endangered eastern South Paciﬁc southern right whales (Eubalaena australis). Mar. Mamm. Sci. 35, 284–289. doi: 10.1111/mms.12519 Hall, A. J., Hugunin, K., Deaville, R., Law, R. J., Allchin, C. R., and Jepson, P. D. (2006). The risk of infection from polychlorinated biphenyl exposure in the harbor porpoise (Phocoena phocoena): a case–control approach. Environ. Health Perspect. 114:704. doi: 10.1289/ehp.8222 Jiang, M., Brown, M. W., Turner, J. T., Kenney, R. D., Mayo, C. A., Zhang, Z., et al. (2007). Springtime transport and retention of Calanus ﬁnmarchicus in Massachusetts and Cape Cod Bays, USA, and implications for right whale foraging. Mar. Ecol. Prog. Ser. 349, 183–197. doi: 10.3354/meps0 7088 Hamilton, P. K., Knowlton, A. R., and Marx, M. K. (2007). “Right whales tell their own stories: the photo-identiﬁcation catalog,” in The Urban Whale: North Atlantic right whales at the crossroads, eds S. D. Kraus and R. M. Rolland (Cambridge, MA: Harvard University Press), 75–104. Johnson, M., Tyack, P., Gillespie, D., and McConnell, B. (2014). A Multi-Week Behavioral Sampling Tag for Sound Eﬀects Studies: Design Trade-Oﬀs and Prototype Evaluation. St Andrews, KY: St Andrews University. Hamner, W. A., Stone, G. S., and Obst, B. S. (1988). Behavior of southern right whales, Eubalaena australis, feeding on the Antarctic krill, Euphausia superba. Fish. Bull. 86, 143–150. Johnson, M. P., and Tyack, P. L. (2003). A digital acoustic recording tag for measuring the response of wild marine mammals to sound. IEEE J. Ocean. Eng. 28, 3–12. doi: 10.1109/JOE.2002.808212 Hayes, S. A., Josephson, E., Maze-Foley, K., and Rosel, P. E. (2017). US Atlantic and Gulf of Mexico Marine Mammal Stock Assessments- 2016 NOAA Tech Memo NMFS-NE 241. Silver Spring, MD: NOAA. Kasuya, T. (1991). Density dependent growth in North Paciﬁc sperm whales. Mar. Mamm. Sci. 7, 230–257. doi: 10.1111/j.1748-7692.1991.tb00100.x Hiby, L., and Lovell, P. (2001). REFERENCES The biogeography of the North Paciﬁc right whale (Eubalaena japonica). Prog. Oceanogr. 80, 188–198. doi: 10.1016/ j.pocean.2008.12.004 j yg International Whaling Commission [IWC] (2001). Report of the workshop on comprehensive assessment of right whales: a worldwide comparison. International Whaling Commission [IWC] (2001). Report of the workshop on comprehensive assessment of right whales: a worldwide comparison. J. Cetacean Res. Manag. 2, 1–60. Groch, K. (2018). “Conservation advances for the southern right whales in brazil,” in Advances in Marine Vertebrate Research in Latin America: Technological Innovation and Conservation, eds M. R. Rossi-Santos and C. W. Finkl (Cham: Springer International Publishing), 441–475. J. Cetacean Res. Manag. 2, 1–60. Ivashchenko, Y. V., and Clapham, P. J. (2012). Soviet catches of right whales Eubalaena japonica and bowhead whales Balaena mysticetus in the North Paciﬁc Ocean and the Okhotsk Sea. Endanger. Species Res. 18, 201–217. doi: 10.3354/esr00443 Groch, K. R., Palazzo, J. T. Jr., Flores, P. A. C., Adler, F. R., and Fabian, M. E. (2005). Recent rapid increases in the right whale (Eubalaena australis) population oﬀ southern Brazil. Lat. Am. J. Aquat. Mamm. 4, 41–47. doi: 10.5597/lajam00068 Ivashchenko, Y. V., Clapham, P. J., and Brownell, R. Jr. (2013). Soviet catches of whales in the North Paciﬁc: revised totals. J. Cetacean Res. Manag. 13, 59–71. Groth, M., Lange, J., Kanrai, P., Pleschka, S., Scholtissek, C., Krumbholz, A., et al. (2014). The genome of an inﬂuenza virus from a pilot whale: Relation to inﬂuenza viruses of gulls and marine mammals. Infect. Genet. Evol. 24, 183–186. doi: 10.1016/j.meegid.2014.03.026 International Whaling Commission [IWC] (2011). Report of the southern right whale die-oﬀworkshop. J. Cetacean Res. Manag.t 12, 367–398. International Whaling Commission [IWC] (2012). Report of the workshop on the assessment of southern right whales, Buenos Aires, Argentina 13-16 September 2011. Cambridge: International Whaling Commission Scientiﬁc Committee. Gulland, F. M. D., and Hall, A. J. (2007). Is marine mammal health deteriorating? trends in the global reporting of marine mammal disease. Ecohealth 4, 135–150. doi: 10.1007/s10393-007-0097-1 Jackson, J., Patenaude, N., Carroll, E., and Baker, C. S. (2008). How few whales were there after whaling? Inference from contemporary mtDNA diversity. Mol. Ecol. 17, 236–251. doi: 10.1111/j.1365-294X.2007.03497.x Hain, J. H. W., Hampp, J. D., McKenney, S. A., Albert, J. A., and Kenney, R. D. (2013). Swim speed, behavior, and movement of North Atlantic right whales (Eubalaena glacialis) in coastal waters of Northeastern Florida, USA. PLoS One 8:e54340. doi: 10.1371/journal.pone.0054340 Jackson, J. A., Carroll, E. REFERENCES R., Gamage, G., Davis, A. R., Mathews, T., LeBlanc, B., and Ferguson, S. H. (2015). Evaluation of UAS for photographic re-identiﬁcation of bowhead whales, Balaena mysticetus. J. Unmanned Veh. Syst. 3, 22–29. doi: 10.1139/juvs- 2014-0014 Mate, B., Mesecar, R., and Lagerquist, B. (2007). The evolution of satellite- monitored radio tags for large whales: one laboratory’s experience. Deep Sea Res. Part II Top. Stud. Oceanogr. 54, 224–247. doi: 10.1016/j.dsr2.2006.11.021 Kraus, S., Fasick, J., Werner, T., and McFarron, P. (2014). Enhancing the Visibility of Fishing Ropes to Reduce Right Whale Entanglements Report to the Bycatch Reduction Engineering Program (BREP). Silver Spring, MD: Oﬃce of Sustainable Fisheries, 67–75. Mate, B., Neukirk, S. L., Mesecar, R., and Martin, T. (1992). Application of Remote Sensing Methods for Tracking Large Cetaceans: North Atlantic Right Whales (Eubalaena glacialis). Washington, DC: United States Department of the Interior, 183. Mate, B., Nieukirk, S. L., and Kraus, S. (1997). Satellite-monitored movements of the Northern right whale. J. Wildl. Manag. 61, 1393–1405. doi: 10.2307/3802143 Kraus, S. D., Hamilton, P. K., Kenney, R. D., Knowlton, A. R., and Slay, C. K. (2001). Reproductive parameters of the North Atlantic right whale. J. Cetacean Res. Manag. 2, 231–236. Mate, B. R., Best, P. B., Lagerquist, B. A., and Winsor, M. H. (2011). Coastal, oﬀshore, and migratory movements of South African right whales revealed by satellite telemetry. Mar. Mamm. Sci. 27, 455–476. doi: 10.1111/j.1748-7692. 2010.00412.x Kraus, S. D., and Hatch, J. J. (2001). Mating strategies in the North Atlantic right whale (Eubalaena glacialis). J. Cetacean Res. Manag. 2, 237–244. Kraus, S. D., Moore, K. E., Price, C. A., Crone, M. J., Watkins, W. A., Winn, H. E., et al. (1986). The Use of Photographs to Identify Individual North Atlantic Right Whales (Eubalaena glacialis). Cambridge: International Whaing Commnission, 145–151. Mate, B. R., Irvine, L. M., and Palacios, D. M. (2017). The development of an intermediate-duration tag to characterize the diving behavior of large whales. Ecol. Evol. 7, 585–595. doi: 10.1002/ece3.2649 Kraus, S. D., and Rolland, R. (2007). The Urban Whale: North Atlantic Right Whales at the Crossroads. Cambridge: Harvard University Press. Matthews, J., Brown, S., Gillespie, D., Johnson, M., McLanaghan, R., Moscrop, A., et al. (2001). Vocalisation rates of the North Atlantic right whale (Eubalaena glacialis). J. Cetacean Res. Manag. 3, 271–282. doi: 10.1121/1.356 2166 Krumhansl, K. A., Head, E. J., Pepin, P., Plourde, S., Record, N. R., Runge, J. A., et al. (2018). REFERENCES J. Cetacean Res. Manag. 2, 251–260. Hunt, K. E., Lysiak, N. S., Moore, M. J., and Rolland, R. M. (2016). Longitudinal progesterone proﬁles in baleen from female North Atlantic right whales Klumov, S. K. (1962). The right whales in the Paciﬁc Ocean. Biol. Mar. Stud. Trud. Institute Oceanogr. 58, 202–297. January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 20 Perspectives on Right Whale Research and Conservation Harcourt et al. Kniest, E., Burns, D., and Harrison, P. (2010). Fluke matcher: a computer-aided matching system for humpback whale (Megaptera novaeangliae) ﬂukes. Mar. Mamm. Sci. 26, 744–756. Lysiak, N. S. J., Trumble, S. J., Knowlton, A. R., and Moore, M. J. (2018). Characterizing the duration and severity of ﬁshing gear entanglement on a North Atlantic right whale (Eubalaena glacialis) using stable isotopes, steroid and thyroid hormones in baleen. Front. Mar. Sci. 5:168. doi: 10.3389/fmars. 2018.00168 Knowlton, A. R., Hamilton, P. K., Marx, M. K., Pettis, H. M., and Kraus, S. D. (2012a). Monitoring North Atlantic right whale Eubalaena glacialis entanglement rates: a 30 yr retrospective. Mar. Ecol. Prog. Ser. 466, 293–302. doi: 10.3354/meps09923 Malik, S., Brown, M. W., Kraus, S. D., Knowlton, A. R., Hamilton, P. K., and White, B. N. (1999). Assessment of mitochondrial DNA structuring and nursery use in the North Atlantic right whale (Eubalaena glacialis). Can. J. Zool. 77, 1217–1222. doi: 10.1139/z99-073 Knowlton, A. R., Hamilton, P. K., and Pettis, H. M. (2012b). Status of Reproductive Females in the North Atlantic Right Whale Population and Impacts of Human Activities on Their Reproductive Success. Silver Spring, MD: NOAA, 41. Marón, C. F., Beltramino, L., Di Martino, M., Chirife, A., Seger, J., Uhart, M., et al. (2015). Increased wounding of Southern right whale (Eubalaena australis) calves by kelp gulls (Larus dominicanus) at Península Valdés, Argentina. PLoS One 10:e0139291. doi: 10.1371/journal.pone.0139291 Knowlton, A. R., Kraus, S. D., and Kenney, R. D. (1994). Reproduction in North Atlantic right whales (Eubalaena glacialis). Can. J. Zool. 72, 1297–1305. doi: 10.1139/z94-173 Knowlton, A. R., Sigurjosson, J., Ciano, J. N., and Kraus, S. D. (1992). Long- distance movements of North Atlantic right whales (Eubalaena glacialis). Mar. Mamm. Sci. 8, 397–405. doi: 10.1111/j.1748-7692.1992.tb00054.x Marques, T. A., Munger, L., Thomas, L., Wiggins, S., and Hildebrand, J. A. (2011). Estimating North Paciﬁc right whale Eubalaena japonica density using passive acoustic cue counting. Endanger. Species Res. 13, 163–172. doi: 10.3354/ esr00325 Koski, W. REFERENCES Environmental drivers of vertical distribution in diapausing Calanus copepods in the Northwest Atlantic. Prog. Oceanogr. 162, 202–222. doi: 10.1016/j.pocean.2018.02.018 Mayo, C. A., Ganley, L., Hudak, C. A., Brault, S., Marx, M. K., Burke, E., et al. (2018). Distribution, demography, and behavior of North Atlantic right whales (Eubalaena glacialis) in Cape Cod Bay, Massachusetts, 1998-2013. Mar. Mamm. Sci. 34, 979–996. doi: 10.1111/mms.12511 Krzystan, A. M., Gowan, T. A., Kendall, W. L., Martin, J., Ortega-Ortiz, J. G., Jackson, K., et al. (2018). Characterizing residence patterns of North Atlantic right whales in the southeastern USA with a multistate open robust design model. Endanger. Species Res. 36, 279–295. doi: 10.3354/esr 00902 Mayo, C. A., Letcher, B. H., and Scott, S. (2001). Zooplankton ﬁltering eﬃciency of the baleen of a North Atlantic right whale, Eubalaena glacialis. J. Cetacean Res. Manag. 2, 225–229. Lanyon, J. M., and Janetzki, H. (2016). Mortalities of Southern right whales (Eubalaena australis) in a subtropical wintering ground, Southeast Queensland. Aquat. Mamm. 42:470. doi: 10.1578/AM.42.4.2016.470 g Mayo, C. A., and Marx, M. K. (1990). Surface foraging behaviour of the North Atlantic right whale, Eubalaena glacialis, and associated zooplankton characteristics. Can. J. Zool. 68, 2214–2220. doi: 10.1139/z90-308 Leaper, R., Cooke, J., Trathan, P., Reid, K., Rowntree, V., and Payne, R. (2006). Global climate drives southern right whale (Eubalaena australis) population dynamics. Biol. Lett. 2, 289–292. doi: 10.1098/rsbl.2005.0431 McEwen, B. S., and Wingﬁeld, J. C. (2003). The concept of allostasis in biology and biomedicine. Horm. Behav. 43, 2–15. doi: 10.1016/S0018-506X(02) 00024-7 McEwen, B. S., and Wingﬁeld, J. C. (2010). What’s in a name? Integrating homeostasis, allostasis and stress. Horm. Behav. 57:105. doi: 10.1016/j.yhbeh. 2009.09.011 LeDuc, R. G., Taylor, B. L., Martien, K. K., Robertson, K. M., Pitman, R. L., Salinas, J. C., et al. (2012). Genetic analysis of right whales in the eastern North Paciﬁc conﬁrms severe extirpation risk. Endangered Species Research 18, 163–167. doi: 10.3354/esr00440 McLeod, B. A., Brown, M. W., Frasier, T. R., and White, B. N. (2010). DNA proﬁle of a sixteenth century western North Atlantic right whale (Eubalaena glacialis). Conserv. Genet. 11, 339–345. doi: 10.1007/s10592-009-9811-6 Leiter, S., Stone, K., Thompson, J., Accardo, C., Wikgren, B., Zani, M., et al. (2017). North Atlantic right whale Eubalaena glacialis occurrence in oﬀshore wind energy areas near Massachusetts and Rhode Island, USA. Endanger. Species Res. 34, 45–59. doi: 10.3354/esr00827 McLeod, B. A., Brown, M. W., Moore, M. J., Stevens, W., Barkham, S. REFERENCES Marine Mammal Populations and Ocean Noise: Determining When Noise Causes Biologically Signiﬁcant Eﬀects. Washington, DC: National Academies Press. Moore, M., Berrow, S., Jensen, B., Carr, P., Sears, R., Rowntree, V., et al. (1999). Relative abundance of large whales around South Georgia (1979– 1998). Mar. Mamm. Sci. 15, 1287–1302. doi: 10.1111/j.1748-7692.1999.tb0 0891.x Omura, H., Ohsumi, S., Nemoto, T., Nasu, K., and Kasuya, T. (1969). Black right whales in the North Paciﬁc. Sci. Rep. Whales Res. Institute 21, 1–66. doi: 10. 1111/mms.12243 Moore, M. J., Durban, J., Fearnbach, H., Lynn, M., Pettis, H. M., and Perryman, W. L. (2017). “Photogrammetry from manned and unmanned aircraft reveals constrained growth of North Atlantic right whales,” in Proceedings of the 22nd Biennial Conference on the Biology of Marine Mammals, Halifax. Ovsyanikova, E., Fedutin, I., Belonovich, O., Burdin, A., Burkanov, V., Dolgova, E., et al. (2015). Opportunistic sightings of the endangered North Paciﬁc right whales (Eubalaena japonica) in Russian waters in 2003-2014. Mar. Mamm. Sci. 31, 1559–1567. doi: 10.1002/ece3.3406 Biennial Conference on the Biology of Marine Mammals, Halifax. Moore, M. J., Knowlton, A. R., Kraus, S. D., McLellan, W. A., and Bonde, R. K. (2004). Morphometry, gross morphology and available histopathology in North Atlantic right whale (Eubalaena glacialis) mortalities (1970–2002). J. Cetacean Res. Manag. 6, 199–214. Pace, R. M. III, Corkeron, P. J., and Kraus, S. D. (2017). State-space mark-recapture estimates reveal a recent decline in abundance of North Atlantic right whales. Ecol. Evol. 7, 8730–8741. doi: 10.1002/ece3.3406 Moore, M. J., van der Hoop, J. M., Barco, S. G., Costidis, A. M., Gulland, F. M., Jepson, P. D., et al. (2013). Criteria and case deﬁnitions for serious injury and death of pinnipeds and cetaceans caused by anthropogenic trauma. Dis. Aquat. Organ. 103, 229–264. doi: 10.3354/dao02566 Parks, S. E., Searby, A., Célérier, A., Johnson, M. P., Nowacek, D. P., and Tyack, P. L. (2011). Sound production behavior of individual North Atlantic right whales: implications for passive acoustic monitoring. Endanger. Species Res. 15, 63–76. doi: 10.1098/rsbl.2011.0578 Moore, M. J., and Zerbini, A. N. (2017). Dolphin blubber/axial muscle shear: implications for rigid transdermal intramuscular tracking tag trauma in whales. J. Exp. Biol. 220, 3717–3723. doi: 10.1242/jeb.165282 Parks, S. E., Warren, J. D., Stamieszkin, K., Mayo, C. A., and Wiley, D. (2012). Dangerous dining: surface foraging of North Atlantic right whales increases risk of vessel collisions. Biol. Lett. 8, 57–60. doi: 10.1098/rsbl.2011.0578 Morano, J. L., Rice, A. N., Tielens, J. REFERENCES T., Estabrook, B. J., Murray, A., Roberts, B. L., et al. (2012). Acoustically detected year-round presence of right whales in an urbanized migration corridor. Conserv. Biol. 26, 698–707. doi: 10.1111/j.1523- 1739.2012.01866.x Pastene, L., Taguchi, M., Lang, A., Goto, M., and Matsuoka, K. (2018). Population Genetic Structure and Historical Demography of North Paciﬁc Right Whales. Washington, D.C: International Whaling Commission, 18. Munger, L. M., Wiggins, S. M., Moore, S. E., and Hildebrand, J. A. (2008). North Paciﬁc right whale (Eubalaena japonica) seasonal and diel calling patterns from long-term acoustic recordings in the southeastern Bering Sea, 2000-2006. Mar. Mamm. Sci. 24, 795–814. doi: 10.1111/j.1748-7692.2008.00219.x Patenaude, N. J., and Baker, C. S. (2001). Population status and habitat use of southern right whales in the sub-Antarctic Auckland Islands of New Zealand. J. Cetacean Res. Manag. 2, 111–116. Patenaude, N. J., Baker, C. S., and Gales, N. J. (1998). Observations of southern right whales on New Zealand’s subantarctic wintering grounds. Mar. Mamm. Sci. 14, 350–355. doi: 10.1111/j.1748-7692.1998.tb00726.x Murison, L., and Gaskin, D. E. (1989). The distribution of right whales and zooplankton in the Bay of Fundy, Canada. Can. J. Zool. 67, 1411–1420. doi: 10.1139/z89-200 Patenaude, N. J., Portway, V. A., Schaeﬀ, C. M., Bannister, J. L., Best, P. B., Payne, R. S., et al. (2007). Mitochondrial DNA diversity and population structure among southern right whales (Eubalaena australis). J. Heredity 98, 147–157. doi: 10.1093/jhered/esm005 Muto, M. M., Helker, V. T., Angliss, R. P., Allen, B. A., Boveng, P. L., Breiwick, J. M., et al. (2017). Alaska Marine Mammal Stock Assessments, 2017. Washington, D.C: U.S. Dep. Commer. NARWC (2018). North Atlantic Right Whale Consortium Database (Anderson Cabot Center for Ocean Life at the New England Aquarium. Boston, MA: NARWC. Patrician, M. R., Biedron, I. S., Esch, H. C., Wenzel, F. W., Cooper, L. A., Hamilton, P. K., et al. (2009). Evidence of a North Atlantic right whale calf (Eubalaena glacialis) born in northeastern U.S. waters. Mar. Mamm. Sci. 25, 462–477. doi: 10.1111/j.1748-7692.2008.00261.x National Academies of Sciences, and Engineering and Medicine (2017). Approaches to Understanding the Cumulative Eﬀects of Stressors on Marine Mammals National Academies of Sciences, and Engineering and Medicine (2017). Approaches to Understanding the Cumulative Eﬀects of Stressors on Marine Mammals. Washington, DC: The National Academies Press. Patrician, M. R., and Kenney, R. D. (2010). REFERENCES Body shape changes associated with reproductive status, nutritive condition and growth in right whales Eubalaena glacialis and E. australis. Mar. Ecol. Prog. Ser. 459, 135–156. doi: 10.3354/meps09675 NOAA (2008). Diﬀerentiating Serious and Non-Serious Injury of Marine Mammals: Technical Memorandum NMFS-OPR-39. Silver Spring, MD: NOAA. condition and growth in right whales Eubalaena glacialis and E. australis. Mar. Ecol. Prog. Ser. 459, 135–156. doi: 10.3354/meps09675 NOAA (2017). 2017-2018 North Atlantic Right Whale Unusual Mortality Event. 2018: NOAA Fisheries. Silver Spring, MD: NOAA. doi: 10.1111/mms. 12443 Miller, C. A., Reeb, D., Best, P. B., Knowlton, A. R., Brown, M. W., and Moore, M. J. (2011). Blubber thickness in right whales Eubalaena glacialis and Eubalaena australis related with reproduction, life history status and prey abundance. Mar. Ecol. Prog. Ser. 438, 267–283. doi: 10.3354/meps09174 Norman, S. A., Flynn, K. R., Zerbini, A. N., Gulland, F., Moore, M. J., Raverty, S., et al. (2018). Assessment of wound healing of tagged gray (Eschrichtius robustus) and blue (Balaenoptera musculus) whales in the eastern North Paciﬁc using long-term series of photographs. Mar. Mamm. Sci. 34, 27–53. doi: 10.1086/ 671811 Miller, M., Gardner, I., Kreuder, C., Paradies, D., Worcester, K., Jessup, D., et al. (2002). Coastal freshwater runoﬀis a risk factor for Toxoplasma gondii infection of southern sea otters (Enhydra lutris nereis). Int. J. Parasitol. 32, 997–1006. doi: 10.1016/S0020-7519(02)00069-3 Nousek-McGregor, A. E., Miller, C. A., Moore, M. J., and Nowacek, D. P. (2013). Eﬀects of body condition on buoyancy in endangered North Atlantic right whales. Physiol. Biochem. Zool. 87, 160–171. doi: 10.1086/671811 Miyashita, T., and Kato, H. (1998). Recent Data on the Status of Right Whales in the New Paciﬁc Ocean SC/M98/AWL. Washington, DC: International Whaling Commission. Nowacek, D. P., Johnson, M. P., and Tyack, P. L. (2004). North Atlantic right whales (Eubalaena glacialis) ignore ships but respond to alerting stimuli. Proc. R. Soc. Lond. Series B Biol. Sci. 271, 227–231. doi: 10.1098/rspb.2001.1730 Moore, M., Andrews, R., Austin, T., Bailey, J., Costidis, A., George, C., et al. (2012). Rope trauma, sedation, disentanglement, and monitoring-tag associated lesions in a terminally entangled North Atlantic right whale (Eubalaena glacialis). Mar. Mamm. Sci. 29, E98–E113. doi: 10.1111/j.1748-7692.2012. 00591.x Nowacek, D. P., Johnson, M. P., Tyack, P. L., Shorter, K. A., McLellan, W. A., and Pabst, D. A. (2001). Buoyant balaenids: the ups and downs of buoyancy in right whales. Proc. R. Soc. B Biol. Sci. 268, 1811–1816. NRC (2005). REFERENCES H., Barkham, M., et al. (2008). Bowhead whales, and not right whales, were the primary target of 16th-to 17th-century Basque whalers in the western North Atlantic. Arctic 1, 61–75. Leroy, G., Carroll, E. L., Bruford, M. W., DeWoody, J. A., Strand, A., Waits, L., et al. (2017). Next-generation metrics for monitoring genetic erosion within populations of conservation concern. Evol. Appl. 11, 1066–1083. doi: 10.1111/ eva.12564 Mellinger, D. K., Nieukirk, S. L., Klinck, K., Klinck, H., Dziak, R. P., Clapham, P. J., et al. (2011). Conﬁrmation of right whales near a nineteenth-century whaling ground east of southern Greenland. Biol. Lett. 7, 411–413. doi: 10.1098/rsbl. 2010.1191 Lockyer, C. H., McConnell, L. C., and Waters, T. D. (1985). Body condition in terms of anatomical and biochemical assessment of body fat in North Atlantic ﬁn and sei whales. Can. J. Zool. 63, 2328–2338. doi: 10.1139/z85-345 Mellinger, D. K., Nieukirk, S. L., Matsumoto, H., Heimlich, S. L., Dziak, R. P., Haxel, J., et al. (2007). Seasonal occurrence of North Atlantic right whale (Eubalaena glacialis) vocalizations at two sites on the Scotian Shelf. Mar. Mamm. Sci. 23, 856–867. doi: 10.1111/j.1748-7692.2007.00144.x Lysiak, N. S. J. (2009). Investigating the Migration and Foraging Ecology of North Atlantic Right Whales with Stable Isotope Geochemistry of Baleen and Zooplankton. Ph.D. thesis, Boston University, Boston, MA. January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 21 Perspectives on Right Whale Research and Conservation Harcourt et al. New, L. F., Clark, J. S., Costa, D. P., Fleishman, E., Hindell, M. A., Klanjšcek, T., et al. (2014). Using short-term measures of behaviour to estimate long-term ﬁtness of southern elephant seals. Mar. Ecol. Prog. Ser. 496, 99–108. Mellinger, D. K., Staﬀord, K. M., Moore, S. E., Munger, L., and Fox, C. G. (2004). Detection of North Paciﬁc right whale (Eubalaena japonica) calls in the Gulf of Alaska. Mar. Mamm. Sci. 20, 872–879. doi: 10.1121/1.2799905 Meyer-Gutbrod, E. L., and Greene, C. H. (2018). Uncertain recovery of the North Atlantic right whale in a changing ocean. Glob. Change Biol. 24, 455–464. doi: 10.1111/gcb.13929 New Zealand Blue Book (1841). Blue Book of Statistics. 1841–1853 Australian Joint Copying Project Colonial Oﬃce (CO213) Pro reel. Canberra: National Library of Australia, 1502–1504. Miller, C. A., Best, P. B., Perryman, W. L., Baumgartner, M. F., and Moore, M. J. (2012). REFERENCES M., Hunt, K. E., Kraus, S. D., and Wasser, S. K. (2005). Assessing reproductive status of right whales (Eubalaena glacialis) using fecal hormone metabolites. Gen. Comp. Endocrinol. 142, 308–317. doi: 10.1016/j.ygcen.2005. 02.002 Pettis, H. M. (2009). North Atlantic Right Whale Consortium Annual Report Card International Whaling Commission SC/61/BRG11. Cambridge: IWC. Pettis, H. M., Rolland, R. M., Hamilton, P. K., Brault, S., Knowlton, A. R., and Kraus, S. D. (2004). Visual health assessment of North Atlantic right whales (Eubalaena glacialis) using photographs. Canadian Journal of Zoology 82, 8–19. doi: 10.1139/z03-207 Rolland, R. M., McLellan, W. A., Moore, M. J., Harms, C. A., Burgess, E. A., and Hunt, K. E. (2017). Fecal glucocorticoids and anthropogenic injury and mortality in North Atlantic right whales Eubalaena glacialis. Endange. Species Res. 34, 417–429. doi: 10.3354/esr00866 Pettis, H. M., Rolland, R. M., Hamilton, P. K., Knowlton, A. R., Burgess, E. A., and Kraus, S. D. (2017). Body condition changes arising from natural factors and ﬁshing gear entanglements in North Atlantic right whales Eubalaena glacialis. Endange. Species Res. 32, 237–249. doi: 10.3354/esr00800 Rolland, R. M., Parks, S. E., Hunt, K. E., Castellote, M., Corkeron, P. J., Nowacek, D. P., et al. (2012). Evidence that ship noise increases stress in right whales. Proc. R. Soc. Biol. Sci. B 279, 2363–2368. doi: 10.1098/rspb.2011.2429 ndange. Species Res. 32, 237–249. doi: 10.3354/esr00800 Rolland, R. M., Schick, R. S., Pettis, H. M., Knowlton, A. R., Hamilton, P. K., Clark, J. S., et al. (2016). Health of North Atlantic right whales Eubalaena glacialis over three decades: from individual health to demographic and population health trends. Mar. Ecol. Prog. Ser. 542, 265–282. doi: 10.3354/meps11547 Pirotta, E., Booth, C. G., Costa, D. P., Fleishman, E., Kraus, S. D., Lusseau, D., et al. (2018). Understanding the population consequences of disturbance. Ecol. Evol. 8, 9934–9946. doi: 10.1002/ece3.4458 Pirotta, V., Grech, A., Jonsen, I. D., Laurence, W., and Harcourt, R. (2018). Consequences of global shipping traﬃc for marine giants. Front. Ecol. Environ. doi: 10.1002/fee.1987 Rone, B. K., Berchok, C. L., Crance, J. L., and Clapham, P. J. (2012). Using air- deployed passive sonobuoys to detect and locate critically endangered North Paciﬁc right whales. Mar. Mamm. Sci. 28, E528–E538. doi: 10.1111/j.1748-7692. 2012.00573.x Pirotta, V., Smith, A., Ostrowski, M., Russell, D., Jonsen, I. D., Grech, A., et al. (2017). An economical custom-built drone for assessing whale health. Front. Mar. Sci. 4:425. doi: 10.3389/fmars.2017.00425 Rosenbaum, H. C., Brownell, R. L. REFERENCES Using the continuous plankton recorder to investigate the absence of North Atlantic right whales (Eubalaena glacialis) from the roseway basin foraging ground. J. Plankton Res. 32, 1685–1695. doi: 10.1093/plankt/fbq073 Nei, M. (1975). Molecular Population Genetics and Evolution. Amsterdam: North- Holland Publishing Company. doi: 10.3354/meps10547 January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 22 Perspectives on Right Whale Research and Conservation Harcourt et al. Payne, R. (1986). Long Term Behavioral Studies of the Southern Right Whale, Eubalaena australis. Washington, DC: International Whaling Commission 10, 161–167. Reeves, R. R., Smith, T. D., and Josephson, E. A. (2007). Near-Annihilation of a Species: Right Whaling in the North Atlantic. The Urban Whale: North Atlantic Right Whales at the Crossroads. Cambridge, MA: Harvard University Press, 39–74. Payne, R., Brazier, O., Dorsey, E. M., Perkins, J. S., Rowntree, V. J., and Titus, A. (1981). External Features in Southern Right Whales (Eubalaena australis) and their Use in Identifying individuals. Bethesda, MD: US Marine Mammal Commission. Reeves, R. R. and Smith, T. (2006). A Taxonomy of World Whaling. Whales, Whaling, and Ocean Ecosystems eds J. A. Estes, D. P. DeMaster, D. F. Doak, T. M. Williams, and R. L. Brownell (Berkeley: University of California Press), 82–101. Peel, D., Smith, J. N., and Childerhouse, S. (2018). Vessel strike of whales in Australia: the challenges of analysis of historical incident data. Front. Mar. Sci. 5:69. doi: 10.3389/fmars.2018.00069 Reid, K., Brierley, A. S., and Nevitt, G. A. (2000). An initial examination of relationships between the distribution of whales and Antarctic krill Euphausia superba at South Georgia. J. Cetacean Res. Manag. 2, 143–149. Pendleton, D. E., Sullivan, P. J., Brown, M. W., Cole, T. V. N., Good, C. P., Mayo, C. A., et al. (2012). Weekly predictions of North Atlantic right whale Eubalaena glacialis habitat reveal inﬂuence of prey abundance and seasonality of habitat preferences. Endanger. Species Res. 18, 147–161. doi: 10.3354/esr00433 Richards, R. (2009). Past and present distributions of southern right whales (Eubalaena australis). N. Z. J. Zool. 36, 447–459. doi: 10.1080/03014223.2009. 9651477 Rocha, R. C., Clapham, P. J., and Ivashchenko, Y. V. (2014). Emptying the oceans: a summary of industrial whaling catches in the 20th century. Mar. Fish. Rev. 76, 37–48. doi: 10.7755/MFR.76.4.3 Pettis, H., and Hamilton, P. (2015). North Atlantic Right Whale Consortium 2015 Annual Report card Report to the North Atlantic Right Whale Consortium. Available at: https://www.narwc.org/uploads/1/1/6/6/116623219/2015_report_ card.pdf Rolland, R. REFERENCES Jr., Brown, M. W., Schaeﬀ, C., Portway, V., White, B. N., et al. (2000). World-wide genetic diﬀerentiation of Eubalaena: questioning the number of right whale species. Mol. Ecol. 9, 1793–1802. doi: 10.1046/j.1365-294x.2000.01066.x Pirzl, R., Patenaude, N. J., Burnell, S., and Bannister, J. (2009). Movements of southern right whales (Eubalaena australis) between Australian and subantarctic New Zealand populations. Mar. Mamm. Sci. 25, 455–461. doi: 10.1111/j.1748-7692.2008.00276.x Rowntree, V., McGuinness, P., Marshall, K., Payne, R., Sironi, M., and Seger, J. (1998). Increased harassment of right whales (Eubalaena australis) by kelp gulls (Larus dominicanus) at Península Valdés, Argentina. Mar. Mamm. Sci. 14, 99–115. doi: 10.1111/j.1748-7692.1998.tb00693.x Rastogi, T., Brown, M. W., McLeod, B. A., Frasier, T. R., Grenier, R., Cumbaa, S. L., et al. (2004). Genetic analysis of 16th-century whale bones prompts a revision of the impact of Basque whaling on right and bowhead whales in the western North Atlantic. Can. J. Zool. 82, 1647–1654. doi: 10.1139/ z04-146 Rowntree, V. J. (2001). Foraging behaviour of southern right whales (Eubalaena australis) inferred from variation of carbon stable isotope ratios in their baleen. J. Cetacean Res. Manag.. Rayment, W., Davidson, A., Dawson, S., Slooten, E., and Webster, T. (2012). Distribution of southern right whales on the Auckland Islands calving grounds. N. Z. J. Mar. Freshwater Res. 46, 431–436. doi: 10.1080/00288330.2012.69 7072 Rowntree, V. J., Payne, R. S., and Schell, D. (2001). Changing patterns of habitat use by southern right whales (Eubalaena australis) on their nursery ground at Península Valdés, Argentina, and in their long-range movements. J. Cetacean Res. Manag. 2, 133–143. Rayment, W., Dawson, S., Webster, T., and Crame, A. (2015). Breeding status aﬀects ﬁne-scale habitat selection of southern right whales on their wintering grounds. J. Biogeogr. 42, 463–474. doi: 10.1111/jbi.12443 Rowntree, V. J., Uhart, M. M., Sironi, M., Chirife, A., Di Martino, M., La Sala, L., et al. (2013). Unexplained recurring high mortality of southern right whale Eubalaena australis calves at Penïnsula Valdïs, Argentina. Mar. Ecol. Progr. Ser. 493, 275–289. doi: 10.3354/meps10506 Rayment, W., Webster, T., Brough, T., Jowett, T., and Dawson, S. (2018). Seen or heard? A comparison of visual and acoustic autonomous monitoring methods for investigating temporal variation in occurrence of southern right whales. Mar. Biol. 165:12. doi: 10.1007/s00227-017-3264-0 Rowntree, V. J., Valenzuela, L. O., Fraguas, P. F., and Seger, J. (2008). Foraging Behaviour of Southern Right Whales (Eubalaena australis) Inferred f rom Variation of Carbon Stable Isotope Ratios in their Baleen. REFERENCES D., Reeves, R. R., Josephson, E. A., and Lund, J. N. (2012). Spatial and seasonal distribution of American whaling and whales in the age of sail. PLoS One 7:e34905. doi: 10.1371/journal.pone.0034905 Van Parijs, S. M., Baumgartner, M., Cholewiak, D., Davis, G., Gedamke, J., Gerlach, D., et al. (2015). NEPAN: A U.S. Northeast passive acoustic sensing network for monitoring, reducing threats and the conservation of marine animals. Mar. Technol. Soc. J. 49, 70–86. doi: 10.4031/MTSJ.49.2.16 Suisted, R., and Neale, D. M. (2004). Department of Conservation Marine Mammal Action Plan For 2005-2010. Te Aro: Department of Conservation Wellington. Thomas, P. O., Reeves, R. R., and Brownell, R. Jr. (2016). Status of the world’s baleen whales. Mar. Mamm. Sci. 32, 682–734. doi: 10.1111/mms.12281 Van Parijs, S. M., Clark, C. W., Sousa-Lima, R. S., Parks, S. E., Rankin, S., Risch, D., et al. (2009). Management and research applications of real-time and archival passive acoustic sensors over varying temporal and spatial scales. Mar. Ecol. Progr. Ser. 395, 21–36. doi: 10.3354/meps08123 Tormosov, D., Mikhaliev, Y., Best, P., Zemsky, V., Sekiguchi, K., and Brownell, R. Jr. (1998). Soviet catches of southern right whales Eubalaena australis, 1951–1971. Biol. Data Conserv. Implic. 86, 185–197. doi: 10.1016/S0006-3207(98)00008-1 Van Waerebeek, K., Santillán, L., and Suazo, E. (2009). On the native status of the southern right whale Eubalaena australis in Peru. Bol. Museo Nacional Hist. Nat. 58, 75–82. Torres, L. G., Rayment, W., Olavarría, C., Thompson, D. R., Graham, B., Baker, C. S., et al. (2017). Demography and ecology of southern right whales Eubalaena australis wintering at sub-Antarctic Campbell Island, New Zealand. Polar Biol. 40, 95–106. doi: 10.1007/s00300-016-1926-x Vanderlaan, A. S. M. (2010). Estimating Risk to the North Atlantic Right Whale (Eubalaena glacialis) from Ocean-going Vessels and Fishing Gear. Ph.D. thesis, Dalhousie University, Halifax. Torres, L. G., Smith, T. D., Sutton, P., MacDiarmid, A., Bannister, J., and Miyashita, T. (2013). From exploitation to conservation: habitat models using whaling data predict distribution patterns and threat exposure of an endangered whale. Divers. Distrib. 19, 1138–1152. doi: 10.1111/ddi.12069 Vernazzani, B. G., Cabrera, E., and Brownell, R. Jr. (2014). Eastern South Paciﬁc southern right whale photo-identiﬁcation catalog reveals behavior and habitat use patterns. Mar. Mamm. Sci. 30, 389–398. doi: 10.1111/mms.12030 whale. Divers. Distrib. 19, 1138–1152. doi: 10.1111/ddi.12069 Wada, S. (1973). The ninth memorandum on the stock assessment of whales in the North Paciﬁc. Int. Whaing Comm. 23, 164–169. REFERENCES Cambridge: International Whaling Commission. Reeves, R., Rosa, C., George, J. C., Sheﬃeld, G., and Moore, M. (2012). Implications of Arctic industrial growth and strategies to mitigate future vessel and ﬁshing gear impacts on bowhead whales. Mar. Policy 36, 454–462. doi: 10.1016/j. marpol.2011.08.005 Scarﬀ, J. E. (1986). Historic and present distribution of the right whale (Eubalaena glacialis) in the Eastern North Paciﬁc, South of 50 Nn and East Of 180 W. Int. Whaling Comm. 10, 43–63. Reeves, R. R., Breiwick, J. M., and Mitchell, E. D. (1992). “Pre-exploitation abundance of right whales oﬀthe eastern United States,” in The Right Whale in the Western North Atlantic: A Science and Management Workshop, ed. J. Hain (Silver Spring, MD: Northeast Fisheries Science Center), 5–7. Scarﬀ, J. E. (1991). Historic distribution and abundance of the right whale (Eubalaena glacialis) in the North Paciﬁc, Bering Sea, Sea of Okhotsk and Sea of Japan from the Maury whale charts. Int. Whaling Comm. 41, 467–489. Reeves, R. R., Breiwick, J. M., and Mitchell, E. D. (1999). History of whaling and estimated kill of right whales, Balaena glacialis, in the Northeastern United States, 1620–1924. Mar. Fish. Rev. 61, 1–36. Scarﬀ, J. E. (2001). Preliminary estimates of whaling-induced mortality in the 19th century North Paciﬁc right whale (Eubalaena japonicus) ﬁshery, adjusting for January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 23 Perspectives on Right Whale Research and Conservation Harcourt et al. struck-but-lost whales and non-American whaling. J. Cetacean Res. Manag. 2, 261–268. Tsoularis, A., and Wallace, J. (2002). Analysis of logistic growth models. Math. Biosci. 179, 21–55. doi: 10.1016/S0025-5564(02)00096-2 struck-but-lost whales and non-American whaling. J. Cetacean Res. Manag. 2, 261–268. Tynan, C. T., DeMaster, D. P., and Peterson, W. T. (2001). Endangered right whales on the southeastern Bering Sea shelf. Science 294, 1894–1894. doi: 10.1126/ science.1065682 Schaeﬀ, C. M., Kraus, S. D., Brown, M. W., and White, B. N. (1993). Assessment of the population structure of western North Atlantic right whales (Eubalaena glacialis) based on sighting and mtDNA data. Can. J. Zool. 71, 339–345. doi: 10.1139/z93-047 Valenzuela, L. O., Rowntree, V. J., Sironi, M., and Seger, J. (2018). Stable isotopes (δ15N, δ13C, δ34S) in skin reveal diverse food sources used by southern right whales Eubalaena australis. Mar. Ecol. Prog. Ser. 603, 243–255. doi: 10.3354/ meps12722 Schick, R. S., Halpin, P. N., Read, A. J., Slay, C. K., Kraus, S. D., Mate, B. REFERENCES (2017a). Predicting lethal entanglements as a consequence of drag from ﬁshing gear. Mar. Pollut. Bull. 115, 91–104. doi: 10.1016/j.marpolbul.2016.11.060 Sierra, E., Fernández, A., Suárez-Santana, C., Xuriach, A., Zucca, D., de Quirós, Y. B., et al. (2016). Morbillivirus and pilot whale deaths, Canary Islands, Spain, 2015. Emerg. Infect. Dis. 22, 740–742. doi: 10.3201/eid2204.150954 van der Hoop, J. M., Nowacek, D. P., Moore, M. J., and Triantafyllou, M. S. (2017b). Swimming kinematics and eﬃciency of entangled North Atlantic right whales. Endanger. Species Res. 32, 1–17. doi: 10.3354/esr00781 Sierra, E., Zucca, D., Arbelo, M., García-Álvarez, N., Andrada, M., Déniz, S., et al. (2014). Fatal systemic morbillivirus infection in bottlenose dolphin, Canary Islands, Spain. Emerg. Infect. Dis. 20:269–271. doi: 10.3201/eid2002.131463 van der Hoop, J. M., Moore, M. J., Barco, S. G., Cole, T. V., Daoust, P. Y., Henry, A. G., et al. (2013a). Assessment of management to mitigate anthropogenic eﬀects on large whales. Conserv. Biol. 27, 121–133. doi: 10.1111/j.1523-1739. 2012.01934.x Sigsgaard, E. E., Nielsen, I. B., Bach, S. S., Lorenzen, E. D., Robinson, D. P., Knudsen, S. W., et al. (2016). Population characteristics of a large whale shark aggregation inferred from seawater environmental DNA. Nat. Ecol. Evol. 1:0004. doi: 10.1038/s41559-016-0004 van der Hoop, J. M., Moore, M. J., Fahlman, A., Bocconcelli, A., George, C., Jackson, K., et al. (2013b). Behavioral impacts of disentanglement of a right whale under sedation and the energetic cost of entanglement. Mar. Mamm. Sci. 30, 282–307. doi: 10.1111/mms.12042 Sironi, M., Rowntree, V. J., Snowdon, C. T., Valenzuela, L., and Marón, C. (2009). Kelp Gulls (Larus dominicanus) Feeding on Southern Right Whales (Eubalaena australis) at Península Valdés, Argentina: Updated Estimates and Conservation Implications. Washington, DC: International Whaling Commission. van der Hoop, J. M., Vanderlaan, A. S. M., Cole, T. V. N., Henry, A. G., Hall, L., Mase-Guthrie, B., et al. (2014). Vessel strikes to large whales before and after the 2008 Ship Strike Rule. Conserv. Lett. 8, 24–32. doi: 10.1111/conl. 12105 Širovi´c, A., Johnson, S. C., Roche, L. K., Varga, L. M., Wiggins, S. M., and Hildebrand, J. A. (2015). North Paciﬁc right whales (Eubalaena japonica) recorded in the northeastern Paciﬁc Ocean in 2013. Mar. Mamm. Sci. 31, 800–807. doi: 10.1111/mms.12189 van der Hoop, J. M., Vanderlaan, A. S. M., and Taggart, C. T. (2012). Absolute probability estimates of lethal vessel strikes to North Atlantic right whales in Roseway Basin, Scotian Shelf. Ecol. Appl. 22, 2021–2033. doi: 10.1890/11-1841.1 Smith, T. REFERENCES R., et al. (2009). Striking the right balance in right whale conservation. Can. J. Fish. Aquat. Sci. 66, 1399–1403. doi: 10.1139/F09-115 Valenzuela, L. O., Sironi, M., and Rowntree, V. J. (2010). Interannual variation in the stable isotope diﬀerences between mothers and their calves in Southern Right Whales (Eubalaena australis). Aquat. Mamm. 36, 138–147. doi: 10.1578/ AM.36.2.2010.138 Schick, R. S., Kraus, S. D., Rolland, R. M., Knowlton, A. R., Hamilton, P. K., Pettis, H. M., et al. (2013). Using hierarchical Bayes to understand movement, health, and survival in the endangered North Atlantic right whale. PLoS One 8:e64166. doi: 10.1371/journal.pone.0064166 Schlaepfer, M. A., Runge, M. C., and Sherman, P. W. (2002). Ecological and evolutionary traps. Trends Ecol. Evol. 17, 474–480. doi: 10.1016/j.tree.2013. 04.004 Valenzuela, L. O., Sironi, M., Rowntree, V. J., and Seger, J. (2009). Isotopic and genetic evidence for culturally inherited site ﬁdelity to feeding grounds in southern right whales (Eubalaena australis). Mol. Ecol. 18, 782–791. doi: 10. 1111/j.1365-294X.2008.04069.x Sekiguchi, K., Onishi, H., Sasaki, H., Haba, S., Iwahara, Y., Mizuguchi, D., et al. (2014). Sightings of the western stock of North Paciﬁc right whales (Eubalaena japonica) in the far southeast of the Kamchatka Peninsula. Mar. Mamm. Sci. 30, 1199–1209. doi: 10.1111/mms.12105 Van Bressem, M.-F., Raga, J. A., Di Guardo, G., Jepson, P. D., Duignan, P. J., Siebert, U., et al. (2009). Emerging infectious diseases in cetaceans worldwide and the possible role of environmental stressors. Dis. Aquat. Organ. 86, 143– 157. doi: 10.3354/dao02101 Seyboth, E., Groch, K. R., Dalla Rosa, L., Reid, K., Flores, P. A., and Secchi, E. R. (2016). Southern right whale (Eubalaena australis) reproductive success is inﬂuenced by krill (Euphausia superba) density and climate. Sci. Rep. 6:28205. doi: 10.1038/srep28205 van der Hoop, J. M., Corkeron, P., and Moore, M. (2016a). Entanglement is a costly life-history stage in large whales. Ecol. Evol. 7, 92–106. doi: 10.1002/ece3.2615 van der Hoop, J. M., Corkeron, P., Kenney, J., Landry, S., Morin, D., Smith, J., et al. (2016b). Drag from ﬁshing gear entangling North Atlantic right whales. Mar. Mamm. Sci. 32, 619–642. doi: 10.1111/mms.12292 Shelden, K. E., Moore, S. E., Waite, J. M., Wade, P. R., and Rugh, D. J. (2005). Historic and current habitat use by North Paciﬁc right whales Eubalaena japonica in the Bering Sea and Gulf of Alaska. Mamm. Rev. 35, 129–155. doi: 10.1111/j.1365-2907.2005.00065.x van der Hoop, J. M., Corkeron, P., Henry, A. G., Knowlton, A. R., and Moore, M. J. REFERENCES Trathan, P., Brierley, A., Brandon, M., Bone, D., Goss, C., Grant, S., et al. (2003). Oceanographic variability and changes in Antarctic krill (Euphausia superba) abundance at South Georgia. Fish. Oceanogr. 12, 569–583. doi: 10.1046/j.1365- 2419.2003.00268.x Wade, P., Heide-Jorgensen, M. P., Shelden, K., Barlow, J., Carretta, J., Durban, J., et al. (2006). Acoustic detection and satellite-tracking leads to discovery of rare January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 24 Perspectives on Right Whale Research and Conservation Harcourt et al. concentration of endangered North Paciﬁc right whales. Biol. Lett. 2, 417–419. doi: 10.1098/rsbl.2006.0460 Wright, D., Crance, J., Woodrich, D., Brewer, A., and Berchok, C. L. (2017). Contemporary distribution of the eastern North Paciﬁc right whale in the Bering Sea. J. Acoust. Soc. Am. 142, 2503–2503. doi: 10.1121/1.501 4139 concentration of endangered North Paciﬁc right whales. Biol. Lett. 2, 417–419. doi: 10.1098/rsbl.2006.0460 Wade, P. R., De Robertis, A., Hough, K. R., Booth, R., Kennedy, A., LeDuc, R. G., Wade, P. R., De Robertis, A., Hough, K. R., Booth, R., Kennedy, A., LeDuc, R. G., et al. (2011a). Rare detections of North Paciﬁc right whales in the Gulf of Alaska, with observations of their potential prey. Endanger. Species Res. 13, 99–109. doi: 10.3354/esr00324 Wright, D. L. (2015). Simultaneous Identiﬁcation of Four Mysticete Species in the Bering Sea Using Passive Acoustic Monitoring Increases Conﬁdence in Acoustic Identiﬁcation of the Critically Endangered North Paciﬁc Right Whale (Eubalaena japonica) Final Report to the International Fund for Animal Welfare. Silver Spring, MD: NOAA, 63. Wade, P. R., Kennedy, A., LeDuc, R., Barlow, J., Carretta, J., Shelden, K., et al. (2011b). The world’s smallest whale population? Biol. Lett. 7, 83–85. doi: 10. 1098/rsbl.2010.0477 Waite, J. M., Wynne, K., and Mellinger, D. K. (2003). Documented sighting of a North Paciﬁc right whale in the Gulf of Alaska and post-sighting acoustic monitoring. Northwestern Nat. 84, 38–43. doi: 10.2307/3536721 Wright, D. L. (2016). Passive Acoustic Monitoring of the Critically Endangered Eastern North Paciﬁc Right Whale (Eubalaena japonica). Bethesda, MD: Marine Mammal Commission. 56. Waldick, R., Kraus, S., Brown, M., and White, B. (2002). Evaluating the eﬀects of historic bottleneck events: an assessment of microsatellite variability in the endangered, North Atlantic right whale. Mol. Ecol. 11, 2241–2249. doi: 10.1046/ j.1365-294X.2002.01605.x Wright, D. L., Berchok, C. L., Crance, J. L., and Clapham, P. J. (2018). REFERENCES Acoustic detection of the critically endangered North Paciﬁc right whale in the northern Bering Sea. Mar. Mamm. Sci. 35, 311–326. doi: 10.1111/mms. 12521 Webster, T. A., Dawson, S. M., Rayment, W. J., Parks, S. E., and Van Parijs, S. M. (2016). Quantitative analysis of the acoustic repertoire of southern right whales in New Zealand. J. Acoust. Soc. Am. 140, 322–333. doi: 10.1121/1.4955066 Zakardjian, B. A. (2003). Eﬀects of temperature and circulation on the population dynamics of Calanus ﬁnmarchicus in the Gulf of St. Lawrence and Scotian Shelf: study with a coupled, three-dimensional hydrodynamic, stage-based life history model. J. Geophys. Res. 108, 8016–8016. doi: 10.1029/2002JC0 01410 Whitehead, H. (2001). Analysis of animal movement using opportunistic individual identiﬁcations: application to sperm whales. Ecology 82, 1417–1432. doi: 10.1890/0012-9658(2001)082[1417:AOAMUO]2.0.CO;2 Zerbini, A. N., Baumgartner, M. F., Kennedy, A. S., Rone, B. K., Wade, P. R., and Clapham, P. J. (2015). Space use patterns of the endangered North Paciﬁc right whale Eubalaena japonica in the Bering Sea. Mar. Ecol. Prog. Ser. 532, 269–281. doi: 10.3354/meps11366 Whitehead, H. (2007). Learning, climate and the evolution of cultural capacity. J. Theor. Biol. 245, 341–350. doi: 10.1016/j.jtbi.2006.10.001 j j Whitehead, H. (2010). Conserving and managing animals that learn socially and share cultures. Learn. Behav. 38, 329–336. doi: 10.3758/LB.38.3.329 Zerbini, A. N., Mendez, M., Rosenbaum, H., Sucunza, F., Andriolo, A., Harris, G., et al. (2016). Tracking Southern Right Whales Through the Southwest Atlantic: New Insights Into Migratory Routes and Feeding Grounds. Washington, DC: International Whaling Commission, 1–15. Whitehead, H., and Payne, R. (1978). New techniques for assessing populations of right whales without killing them. Mamm. Seas Rep. 3:189. Wiley, D., Hatch, L., Thompson, M., Schwehr, K., and MacDonald, C. (2013). Marine sanctuaries and marine planning: protecting endangered marine life. Proc. Mar. Safety Secur. Council 70, 10–15. Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Winn, H. E., Goodyear, J. D., Kenney, R. D., and Petricig, R. O. (1995). Dive patterns of tagged right whales in the Great South Channel. Cont. Shelf Res. 15, 593–611. doi: 10.1016/0278-4343(94)00061-Q Copyright © 2019 Harcourt, van der Hoop, Kraus and Carroll. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). Frontiers in Marine Science \| www.frontiersin.org January 2019 \| Volume 5 \| Article 530 REFERENCES The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Winn, H. E., Price, C. A., and Sorensen, P. W. (1986). The distributional biology of the right whale (Eubalaena glacialis) in the Western North Atlantic. Rep. Int. Whaling Comm. 10, 129–138. Wishner, K. F., Schoenherr, J. R., Beardsley, R., and Chen, C. (1995). Abundance, distribution and population structure of the copepod Calanus ﬁnmarchicus in a springtime right whale feeding area in the southwestern Gulf of Maine. Cont. Shelf Res. 15, 475–507. doi: 10.1016/0278-4343(94)00057-T January 2019 \| Volume 5 \| Article 530 Frontiers in Marine Science \| www.frontiersin.org 25
W3106766293.txt	https://bmcmedinformdecismak.biomedcentral.com/track/pdf/10.1186/s12911-020-01339-z	en		Cognitive biomarker prioritization in Alzheimer’s Disease using brain morphometric data	BMC medical informatics and decision making	2,020	cc-by	8,625	(2020) 20:319 Peng et al. BMC Med Inform Decis Mak https://doi.org/10.1186/s12911-020-01339-z Open Access RESEARCH ARTICLE Cognitive biomarker prioritization in Alzheimer’s Disease using brain morphometric data Bo Peng1, Xiaohui Yao2, Shannon L. Risacher3, Andrew J. Saykin3, Li Shen2, Xia Ning1* and for the ADNI Abstract Background: Cognitive assessments represent the most common clinical routine for the diagnosis of Alzheimer’s Disease (AD). Given a large number of cognitive assessment tools and time-limited office visits, it is important to determine a proper set of cognitive tests for different subjects. Most current studies create guidelines of cognitive test selection for a targeted population, but they are not customized for each individual subject. In this manuscript, we develop a machine learning paradigm enabling personalized cognitive assessments prioritization. Method: We adapt a newly developed learning-to-rank approach PLTR to implement our paradigm. This method learns the latent scoring function that pushes the most effective cognitive assessments onto the top of the prioritization list. We also extend PLTR to better separate the most effective cognitive assessments and the less effective ones. Results: Our empirical study on the ADNI data shows that the proposed paradigm outperforms the state-of-the-art baselines on identifying and prioritizing individual-specific cognitive biomarkers. We conduct experiments in cross validation and level-out validation settings. In the two settings, our paradigm significantly outperforms the best baselines with improvement as much as 22.1% and 19.7%, respectively, on prioritizing cognitive features. Conclusions: The proposed paradigm achieves superior performance on prioritizing cognitive biomarkers. The cognitive biomarkers prioritized on top have great potentials to facilitate personalized diagnosis, disease subtyping, and ultimately precision medicine in AD. Keywords: Alzheimer’s Disease, Learning to rank, Bioinformatics, Machine learning Background Identifying structural brain changes related to cognitive impairments is an important research topic in Alzheimer’s Disease (AD) study. Regression models have been extensively studied to predict cognitive outcomes using morphometric measures that are extracted from structural magnetic resonance imaging (MRI) scans [1, 2]. These studies are able to advance our understanding on the neuroanatomical basis of cognitive impairments. However, they are not designed to have direct impacts on *Correspondence: ning.104@osu.edu 1 The Ohio State University, Columbus, USA Full list of author information is available at the end of the article clinical practice. To bridge this gap, in this manuscript we develop a novel learning paradigm to rank cognitive assessments based on their relevance to AD using brain MRI data. Cognitive assessments represent the most common clinical routine for AD diagnosis. Given a large number of cognitive assessment tools and time-limited office visits, it is important to determine a proper set of cognitive tests for the subjects. Most current studies create guidelines of cognitive test selection for a targeted population [3, 4], but they are not customized for each individual subject. In this work, we develop a novel learning paradigm that incorporate the ideas of precision medicine and customizes the cognitive test selection process to © The Author(s) 2020. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Peng et al. BMC Med Inform Decis Mak (2020) 20:319 the characteristics of each individual patient. Specifically, we conduct a novel application of a newly developed learning-to-rank approach, denoted as PLTR [5], to the structural MRI and cognitive assessment data of the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort [6]. Using structural MRI measures as the individual characteristics, we are able to not only identify individual-specific cognitive biomarkers but also prioritize them and their corresponding assessment tasks according to AD-specific abnormality. We also extend PLTR to PLTRh using hinge loss [7] to more effectively prioritize individual-specific cognitive biomarkers. The study presented in this manuscript is a substantial extension from our preliminary study [8]. Our study is unique and innovative from the following two perspectives. First, conventional regression-based studies for cognitive performance prediction using MRI data focus on identifying relevant imaging biomarkers at the population level. However, our proposed model aims to identify AD-relevant cognitive biomarkers customized to each individual patient. Second, the identified cognitive biomarkers and assessments are prioritized based on the individual’s brain characteristics. Therefore, they can be used to guide the selection of cognitive assessments in a personalized manner in clinical practice; it has the potential to enable personalized diagnosis and disease subtyping. Literature review Learning to rank Learning-to-Rank (LETOR ) [9] is a popular technique used in information retrieval [10], web search [11] and recommender systems [12]. Existing LETOR methods can be classified into three categories [9]. The first category is point-wise methods [13], in which a function is learned to score individual instance, and then instances are sorted/ranked based on their scores. The second category is pair-wise methods [14], which maximize the number of correctly ordered pairs in order to learn the optimal ranking structure among instances. The last category is list-wise methods [15], in which a ranking function is learned to explicitly model the entire ranking. Generally, pairwise and listwise methods have superior performance over point-wise methods due to their ability to leverage order structure among instances in learning [9]. Recently, LETOR has also been applied in drug discovery and drug selection [16–19]. For example, Agarwal et al. [20] developed a bipartite ranking method to prioritize drug-like compounds. He et al. [5] developed a joint push and learning-to-rank method to select cancer drugs for each individual patient. These studies demonstrate the great potential of LETOR in computational biology Page 2 of 11 and computational medicine, particularly for biomarker prioritization. Machine learning for AD biomarker discovery The importance of using big data to enhance AD biomarker study has been widely recognized [6]. As a result, numerous data-driven machine learning models have been developed for early AD detection and AD-relevant biomarker identification including cognitive measures. These models are often designed to accomplish tasks such as classification (e.g., [21]), regression (e.g., [1, 2, 22]) or both (e.g., [23, 24]), where imaging and other biomarker data are used to predict diagnostic, cognitive and/or other outcome(s) of interest. A drawback of these methods is that, although outcome-relevant biomarkers can be identified, they are identified at the population level and not specific to any individual subject. To bridge this gap, we adapt the PLTR method for biomarker prioritization at the individual level, which has greater potential to directly impact personalized diagnosis. Methods Materials The imaging and cognitive data used in our study were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database [6]. The ADNI was launched in 2003 as a public-private partnership, led by Principal Investigator Michael W. Weiner, MD. The primary goal of ADNI has been to test whether serial MRI, PET, other biological markers, and clinical and neuropsychological assessment can be combined to measure the progression of mild cognitive impairment (MCI, a prodromal stage of AD) and early AD. For up-to-date information, Please refer to [25] for more detailed, up-to-date information. Participants include 819 ADNI-1 subjects with 229 healthy control (HC), 397 MCI and 193 AD participants. We consider both MCI and AD subjects as patients, and thus we have 590 cases and 229 controls. We downloaded the 1.5T baseline MRI scans and cognitive assessment data from the ADNI website [25]. We processed the MRI scans using Freesurfer version 5.1 [26], where volumetric and cortical thickness measures of 101 regions relevant to AD were extracted to characterize brain morphometry. We focus our analysis on 151 scores assessed in 15 neuropsychological tests. For convenience, we denote these measures as cognitive features and these tests as cognitive tasks. The 15 studied tasks include Alzheimer’s Disease Assessment Scale (ADAS), Clinical Dementia Rating Scale (CDR), Functional Assessment Questionnaire (FAQ), Geriatric Depression Scale (GDS), Mini-Mental State Exam (MMSE), Modified Hachinski Scale (MODHACH), Neuropsychiatric Inventory Questionnaire (NPIQ), Boston Naming Test (BNT), Clock Drawing Test (CDT), Digit Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Page 3 of 11 Span Test (DSPAN), Digit Symbol Test (DSYM), Category Fluency Test (FLUENCY), Weschler’s Logical Memory Scale (LOGMEM), Rey Auditory Verbal Learning Test (RAVLT) and Trail Making Test (TRAIL). function (I(x) = 1 if x is true, otherwise 0). In Eq. (2), sp (fi ) is a scoring function defined as follows, Joint push and learning‑to‑rank using scores—PLTR that is, it calculates the score of feature fi on patient Pp using their respective latent vectors up and vi [29]. By ↑ minimizing Ps , PLTR learns to assign higher scores to relevant features than irrelevant features so as to rank the relevant features at the top of the final ranking list. Note that, PLTR learns different latent vectors and ranking lists for different subjects, and therefore enables personalized feature prioritization. In Problem (1), Os+ measures the ratio of mis-ordered feature pairs over the relevant features among all the subjects, defined as follows, We use the joint push and learning-to-rank method that we developed in He et al. [5], denoted as PLTR, for personalized cognitive feature prioritization. PLTR has also been successfully applied in our preliminary study [8]. We aim to prioritize cognitive features for each individual patient that are most relevant to his/her disease diagnosis. We will use patients’ brain morphometric measures that are extracted from their MRI scans for the cognitive feature prioritization. The cognitive features are in the form of scores or answers in the cognitive tasks that the patients take. The prioritization outcomes can potentially be used in clinical practice to suggest the most relevant cognitive features or tasks that can most effectively facilitate diagnosis of an individual subject. In order to prioritize MCI/AD cognitive features, PLTR learns and uses patient latent vector representations and their imaging features to score each cognitive feature for each individual patient. Then, PLTR ranks the cognitive features based on their scores. Patients with similar imaging feature profiles will have similar latent vectors and thus similiar ranking of cognitive features [27, 28]. During the learning, PLTR explicitly pushes the most relevant cognitive features on top of the less relevant features for each patient, and therefore optimizes the latent patient vectors and cognitive feature vectors in a way that they will reproduce the feature ranking structures [9]. In PLTR, these latent vectors are learned via solving the following optimization problem: min Ls = (1 − α)Ps↑ + αOs+ + U ,V β γ Ruv + Rcsim , (1) 2 2 where α, β and γ ∈ [0, 1] are coefficients of Os+, Ruv and Rcsim terms, respectively; U = [u1 , u2 , · · · , um ] and V = [v1 , v2 , · · · , vn ] are the latent matrices for patients and features, respectively (u and v are column latent patient vector and feature vector, respectively); Ls is the ↑ overall loss function. In Problem 1, Ps measures the average number of relevant cognitive features ranked below an irrelevant cognitive feature, defined as follows, Ps↑ = m 1 I(sp (fj n+ n− p=1 p p f − ∈Pp f + ∈Pp+ i j + ) ≤ sp (fi− )), (2) where m is the number of patients, fj+ and fi− are the − relevant and irrelevant features of patient Pp, n+ p and np are their respective numbers, and I(x) is the indicator sp (fi ) = upT vi , Os+ = m \|{fi+ p=1 (3) 1 ≻Pp fj+ }\| I(sp (fi+ ) < sp (fj+ )), fi+ ≻Pp fj+ (4) where fi ≻Pp fj represents that fi is ranked higher than ↑ fj for patient Pp. By minimizing Os , PLTR learns to push the most relevant features on top of the less relevant features. Thus, most relevant features are pushed to the very top of the ranking list. In Problem (1), Ruv is a regularizer on U and V to prevent overfitting, defined as, Ruv = 1 1 �U �2F + �V �2F , m n (5) where XF is the Frobenius norm of matrix X. Rcsim is a regularizer on patients to constrain patient latent vectors, defined as Rcsim = m m 1 wpq �up − uq �22 , m2 p=1 q=1 (6) where wpq is the similarity between subject Pp and Pq that is calculated using the imaging features of thesubjects. The assumption here is that patients who are similar in terms of imaging features could also be similar in terms of cognitive features. Joint push and learning‑to‑rank with marginalization—PLTRh The objective of PLTR is to score relevant features higher than less relevant features as shown in Eqs. 2 and 4. However, in some cases, the score of relevant features is expected to be higher than that of less relevant features by a large margin. For example, patients can be very sensitive to a few cognitive tasks but less sensitive to many others. In order to incorporate such information, we propose a new hinge loss [7] based PLTR , denoted as PLTRh. Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Page 4 of 11 In PLTRh, the overall loss function is very similar to Eq. 1, defined as follows, ↑ min Lh = (1 − α)Ph + αOh+ + U ,V β γ Ruv + Rcsim , (7) 2 2 where Lh is the overall loss function; U, V, Ruv and Rcsim ↑ are identical as those in Eq. 1. In PLTRh, Ph measures the average loss between the relevant features and irrelevant features using hinge loss as follows, ↑ Ph = m 1 n+ n− p=1 p p f − ∈Pp f + ∈Pp+ i j max(0, tp − (sp (fj+ ) − sp (fi− ))), (8) max(0, tp − (sp (fj ) − sp (fi− ))) + where is the hinge loss (max(0, x) = x if x > 0, otherwise 0) between the relevant feature fj+ and the irrelevant feature fi−, and tp is the pre-defined margin. Specifically, only when sp (fj+ ) − sp (fi− ) > tp will not induce any loss during optimization. Otherwise, the hinge loss will be positive and increase as sp (fj+ ) − sp (fi− ) gets smaller than tp. Thus, the hinge loss forces the scores of relevant features higher than those of irrelevant features by at least tp. By doing this, the relevant features are ranked higher than irrelevant features in the ranking list. Similarly, Oh+ measures the average loss among the relevant features also using hinge loss as follows, Oh+ = m 1 \|{fi+≻Pp fj+ }\| p=1 cognitive features remained and used in experiments. Additional file 1: Table S1 presents these 112 cognitive features. We conducted the same process as above on the imaging features. Additional file 1: Table S2 presents these imaging features used in experiments. Patient similarities from imaging features Through the normalization and filtering steps as in “Data normalization” section, we have 86 normalized imaging features remained. We represent each patient using a vector of these features, denoted as rp = [rp1 , rp2 , · · · , rp86 ], in which rpi (i = 1, · · · , 86) is an imaging feature for patient p. We calculate the patient similarity from imaging features using the radial2basis function (RBF) kernel, �r −r � that is, wpq = exp(− p2σ 2q ), where wpq is the patient similarity used in Rcsim. Results Baseline methods We compare PLTR and PLTRh with two baseline methods: the Bayesian Multi-Task Multi-Kernel Learning (BMTMKL) method [30] and the Kernelized Rank Learning (KRL) method [31]. Bayesian multi‑task multi‑kernel learning (BMTMKL) BMTMKL is a state-of-the-art baseline for biomarker prioritization. It was originally proposed to rank cell lines for drugs and won the DREAM 7 challenge [32]. In our max(0, to − (sp (fi+ ) − sp (fj− ))), fi+ ≻Pp fj+ where to is also the pre-defined margin. Data processing Data normalization Following the protocol in our preliminary study [8], we selected all the MCI and AD patients from ADNI and conducted the following data normalization for these patients. We first performed a t test on each cognitive feature between patients and controls, and selected those features if there is a significant difference between patients and controls on these features. Then, we converted the selected features into [0, 1] by shifting and scaling the feature values. We also converted all the normalized feature values according to the Cohen’s d of the features between patients and controls, and thus, smaller values always indicate higher AD possibility. After that, we filtered out features with values 0, 1 or 0.5 for more than 95% patients. This is to discard features that are either not discriminative, or extremely dominated by patients or controls. After the filtering step, we have 112 (9) study, BMTMKL uses the multi-task and multi-kernel learning within kernelized regression to predict cognitive feature values and learns parameters by conducting Bayesian inference. We use the patient similarity matrix calculated from FreeSurfer features as the kernels in BMTMKL. Kernelized rank learning (KRL) KRL represents another state-of-the-art baseline for biomarker prioritization. In our study, KRL uses kernelized regression with a ranking loss to learn the ranking structure of patients and to predict the cognitive feature values. The objective of KRL is to maximize the hits among the top k of the ranking list. We use the patient similarity matrix calculated from FreeSurfer features as the kernels in KRL. Peng et al. BMC Med Inform Decis Mak (2020) 20:319 patients training patients P5 P4 P3 P2 P1 f1 Page 5 of 11 testing patients f2 f3 f4 cognitive features f5 Fig. 1 Data split for cross validation (CV) Parameters We conduct grid search to identify the best parameters on each evaluation metric for each model. We use 0.3 and 0.1 as the value of tp and to , respectively. In the experimental results, we report the combinations of parameters that achieve the best performance on evaluation metrics. We implement PLTR and PLTRh using Python 3.7.3 and Numpy 1.16.2, and run the experiments on Xeon E5-2680 v4 with 128G memory. Evaluation metrics patients training patients P5 P4 P3 P2 P1 testing patients Metrics on cognitive feature level We use a metric named average feature hit at k (QH@k) as in our preliminary study [8] to evaluate the ranking performance, f1 f2 f3 f4 cognitive features f5 Fig. 2 Data split for leave-out validation (LOV) Training‑testing data splits Following the protocol in our preliminary study [8], we test our methods in two different settings: cross validation (CV ) and leave-out validation (LOV ). In CV , we randomly split each patient’s cognitive tasks into 5 folds: all the features of a cognitive task will be either split into training or testing set. We use 4 folds for training and the rest fold for testing, and do such experiments 5 times, each with one of the 5 folds as the testing set. The overall performance of the methods is averaged over the 5 testing sets. This setting corresponds to the goal to prioritize additional cognitive tasks that a patient should complete. In LOV , we split patients (not patient tasks) into training and testing sets, and a certain patient and all his/her cognitive features will be either in the training set or in the testing set. This corresponds to the use scenario to identify the most relevant cognitive tasks that a new patient needs to take, based on the existing imaging information of the patient, when the patient has not completed any cognitive tasks. Figures 1 and 2 demonstrate the CV and LOV data split processes, respectively. Please note that as presented in “Data normalization” section, for normalized cognitive features, smaller values always indicate more AD possibility. Thus, in both settings, we use the ranking list of normalized cognitive features of each patient as ground truth for training and testing. q q QH@k(τ , τ̃ ) = k q I(τ̃i ∈ τ q (1 : k)), (10) i=1 where τ q is the ground-truth ranking list of all the features in all the tasks, τ q (1 : k) is the top k features in the list, τ̃ q is the predicted ranking list of all the features, and q τ̃i is the ith ranked features in τ̃ q . That is, QH@k calculates the number of features among top k in the predicted feature lists that are also in the ground truth (i.e., hits). Higher QH@k values indicate better prioritization performance. We use a second evaluation metric weighted average feature hit at k (WQH@k) as follows: WQH@k(τ q , τ̃ q ) = k j=1 QH @j(τ q , τ̃ q )/k, (11) that is, WQH@k is a weighted version of QH@k that calculates the average of QH@j ( j = 1, · · · , k ) over top k. Higher WQH@k indicates more feature hits and those hits are ranked on top in the ranking list. Metrics on cognitive task level In in Peng et al. [8], we use the mean of the top-g normalized ground-truth scores/predicted scores on the features of each cognitive task for a patient as the score of that task for that patient. For each patient, we rank the tasks using their ground-truth scores and use the ranking as the ground-truth ranking of these tasks. Thus, these scores measure how much relevant to AD the task indicates for the patients. We use the predicted scores to rank cognitive tasks into the predicted ranking of the tasks. We define a third evaluation metric Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Page 6 of 11 task hit at k ( NHg @k) as follows to evaluate the ranking performance in terms of tasks, NHg @k(τgn , τ̃gn ) = k I(τ̃gin ∈ τgn (1 : k)), evaluation metrics on cognitive feature level (i.e., QH@5 and WQH@5). Specifically, PLTR outperforms the best baseline method BMTMKL at 9.1 ± 3.7% and 22.1 ± 9.5% on QH@5 and WQH@5, respectively. PLTRh also outperforms BMTMKL at 6.4 ± 4.3% and 19.2 ± 10.1% on QH@5 and WQH@5, respectively. These experimental results demonstrate that among the top 5 features in the ranking list, PLTR and PLTRh are able to rank more relevant features on top than the two state-of-the-art baseline methods and the positions of those hits are also higher than those in the baseline methods. (12) i=1 where τgn/τ̃gn is the ground-truth/predicted ranking list of all the tasks using top-g question scores. Experimental results Overall Performance on CV Table 1 presents the performance of PLTR , PLTRh and two baseline methods in the CV setting. Note that overall, PLTR and PLTRh have similar standard deviations; KRL and BMTMKL have higher standard deviations compared to PLTR and PLTRh. This indicates that PLTR and PLTRh are more robust than KRL and BMTMKL for the prioritization tasks. Comparison on cognitive feature level For cognitive features from all tasks, PLTR is able to identify on average 2.665 ± 0.07 out of the top-5 most relevant ground-truth cognitive features among its top-5 predictions (i.e., QH@5 = 2.665 ± 0.07). PLTRh achieves similar performance as PLTR , and identifies on average 2.599 ± 0.09 most relevant groundtruth cognitive features on its top-5 predictions (i.e., QH@5 = 2.599 ± 0.09). PLTR and PLTRh significantly outperform the baseline methods in terms of all the Comparison on cognitive task level For the scenario to prioritize cognitive tasks that each patient should take, PLTR and PLTRh are able to identify the top-1 most relevant task for 72.5 ± 6.0% and 74.3 ± 4.0% of all the patients when using 3 features to score cognitive tasks, respectively (i.e., NH3 = 0.725 ± 0.06 for PLTR and NH3 = 0.743 ± 0.04 for PLTRh). This indicates the strong power of PLTR and PLTRh in prioritizing cognitive features and in recommending relevant cognition tasks for real clinical applications. We also find that PLTR and PLTRh are able to outperform baseline methods on most of the metrics on cognitive task level (i.e., NHg @1). PLTR outperforms the best baseline method at 11.6 ± 5.6%, 16.7 ± 6.1% and 14.2 ± 6.6% on NH1 @1, NH2 @1 and NH3 @1, respectively. PLTRh performs even better than PLTR on NH1 @1 and NH3 @1, in addition to that it outperforms the Table 1 Overall performance in CV Method PLTR PLTRh Parameters Feature level Task level d QH@5 WQH@5 NH1 @1 NH2 @1 NH3 @1 NH5 @1 NHall @1 10 – 2.665 ± 0.07 3.136 ± 0.12 0.605 ± 0.03 0.701 ± 0.04 0.713 ± 0.05 0.725 ± 0.05 0.683 ± 0.04 10 – 2.647 ± 0.08 3.191 ± 0.14 0.599 ± 0.03 0.677 ± 0.04 0.707 ± 0.04 0.725 ± 0.05 0.677 ± 0.04 10 – 2.569 ± 0.08 2.957 ± 0.11 0.635±0.03 0.707 ± 0.04 0.689 ± 0.05 0.719 ± 0.04 0.653 ± 0.04 10 – 2.623 ± 0.06 3.073 ± 0.09 0.623 ± 0.03 0.713 ± 0.05 0.707 ± 0.04 0.719 ± 0.04 0.671 ± 0.04 50 – 2.467 ± 0.07 2.992 ± 0.11 0.605 ± 0.03 0.695 ± 0.04 0.725 ± 0.06 0.725 ± 0.04 0.653 ± 0.04 30 – 2.491 ± 0.07 3.080 ± 0.14 0.563 ± 0.04 0.689 ± 0.05 0.713 ± 0.04 0.749 ± 0.04 0.689 ± 0.03 10 – 10 – 2.575 ± 0.08 2.599 ± 0.09 3.111 ± 0.12 0.623 ± 0.02 0.671 ± 0.03 0.713 ± 0.03 0.719 ± 0.04 0.707 ± 0.03 10 – 2.419 ± 0.09 2.827 ± 0.12 30 – 2.138 ± 0.10 2.583 ± 0.18 0.647 ± 0.03 0.629 ± 0.02 0.695 ± 0.03 0.701 ± 0.02 0.671 ± 0.03 0.707 ± 0.03 3.115 ± 0.13 0.623 ± 0.03 0.677 ± 0.03 0.737 ± 0.04 0.695 ± 0.03 50 – 2.102 ± 0.07 2.470 ± 0.10 0.533 ± 0.03 0.677 ± 0.03 30 – 2.281 ± 0.07 2.768 ± 0.18 0.563 ± 0.03 0.689 ± 0.03 KRL – 2 – 1.5 2.102 ± 0.26 2.167 ± 0.37 0.569 ± 0.03 0.611 ± 0.05 0.635 ± 0.04 BMTMKL – – 2.443 ± 0.12 2.614 ± 0.20 0.413 ± 0.07 0.491 ± 0.08 0.593 ± 0.05 2.078 ± 0.15 2.143 ± 0.25 0.503 ± 0.04 0.575 ± 0.05 0.743 ± 0.04 0.707 ± 0.04 0.617 ± 0.05 0.749 ± 0.03 0.695 ± 0.04 0.695 ± 0.03 0.635 ± 0.03 0.593 ± 0.05 0.754 ± 0.03 0.629 ± 0.05 0.683 ± 0.03 0.689 ± 0.07 0.760 ± 0.05 0.677 ± 0.04 0.784 ± 0.05 0.701 ± 0.05 0.760 ± 0.06 0.749 ± 0.05 The column “d” corresponds to the latent dimension. The numbers in the form of x ± y represent the mean (x) and standard deviation (y). The best performance of each method is in italic. The best performance under each evaluation metric is underlined Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Page 7 of 11 best performance of baseline methods at 13.7 ± 5.3%, 14.7 ± 4.8% and 17.0 ± 8.8% on NH1 @1, NH2 @1 and NH3 @1, respectively. PLTR and PLTRh perform slightly worse than baseline methods on NH5 @1 and NHall @1 (0.760 ± 0.05 vs 0.784 ± 0.05 on NH5 @1 and 0.707 ± 0.03 vs 0.760 ± 0.06 on NHall @1). These experimental results indicate that PLTR and PLTRh are able to push the most relevant task to the top of the ranking list than baseline methods when using a small number of features to score cognitive tasks. Note that in CV , each patient has only a few cognitive tasks in the testing set. Therefore, we only consider the evaluation at the top task in the predicted task rankings (i.e., only NHg @1 in Table 1). Table 1 also shows that PLTRh outperforms PLTR on most of the metrics on cognitive task level (i.e., NHg @1). PLTRh outperforms PLTR at 1.9 ± 0.5%, 2.5 ± 1.2%, 1.5 ± 0.3% and 2.6 ± 0.9% on NH1@1, NH3@1, NH5@1 and NHall@1, respectively. This indicates that generally PLTRh is better than PLTR on ranking cognitive tasks in CV setting. The reason could be that the hinge-based loss functions with pre-defined margins can enable significant difference between the scores of relevant features and irrelevant features, and thus effectively push relevant features upon irrelevant features. Overall performance on LOV Tables 2 and 3 present the performance of PLTR , PLTRh and two baseline methods in the LOV setting. Due to space limit, we did not present the standard deviations in the tables, but they have similar trends as those in Table 1. We first hold out 26 (Table 2) and 52 (Table 3) AD patients as testing patients, respectively. We determine these hold-out AD patients as the ones that have more than 10 similar AD patients in the training set with corresponding patient similarities higher than 0.67 and 0.62, respectively. Comparison on cognitive feature level Tables 2 and 3 show that PLTR and PLTRh significantly outperform the baseline methods in terms of all the evaluation metrics on cognitive feature level (i.e., QH@5 and WQH@5), which is consistent with the experimental results in CV setting. When 26 patients are hold out for testing, with parameters α = 0.5, β = 1.5, γ = 1.0 and d = 30, PLTR outperforms the best baseline method KRL at 13.4% and 1.3% on QH@5 and WQH@5, respectively. The performance of PLTRh is very comparable with that of PLTR ” PLTRh outperforms KRL at 13.4% and 0.5% on QH@5 and WQH@5, respectively. When 52 patients are hold out for testing, with parameters α = 0.5, β = 0.5, γ = 1.0 and d = 50, PLTR outperforms the best baseline method KRL at 18.1% and 7.8% on QH@5 and WQH@5, respectively. PLTRh even performs better than PLTR in this setting. In addition, PLTRh outperforms KRL at 19.7% and 9.5% on QH@5 and WQH@5, respectively. These experimental results demonstrate that for new patients, PLTR and PLTRh are able to rank more relevant features to the top of the ranking list than the two baseline methods. They also indicate that for new patients, ranking based methods (e.g., PLTR and PLTRh) are more effective than regression based methods (e.g., KRL and BMTMKL) for biomarker prioritization. Table 2 Overall performance in LOV on 26 testing patients Method PLTR PLTRh KRL BMTMKL Feature level Task level QH@5 WQH@5 NH 1@1 NH 1@5 1.615 1.906 0.846 3.231 1.500 1.778 0.846 3.269 1.538 1.856 0.846 3.192 1.577 1.851 0.846 3.192 NH2@1 NH2@5 NH3@1 NH3@5 NH 5@1 NH 5@5 NHall@1 NHall@5 0.577 3.385 0.231 3.654 0.308 3.346 0.808 3.692 0.577 3.538 0.269 3.654 0.269 3.269 0.808 3.577 0.577 3.423 0.308 3.731 0.346 3.346 0.808 3.615 0.577 3.462 0.308 3.654 0.346 3.462 0.808 3.654 1.615 1.906 0.846 3.231 0.577 3.385 0.231 3.654 0.308 3.346 0.808 3.692 1.615 1.836 0.846 3.192 0.577 3.500 0.269 3.731 0.346 3.731 0.808 4.154 1.538 1.891 0.846 3.192 0.577 3.500 0.269 3.731 0.346 3.615 0.808 4.038 1.538 1.856 0.769 3.308 0.577 3.462 0.269 3.615 0.308 3.385 0.808 3.500 1.538 1.712 0.846 3.115 0.577 3.423 0.154 3.731 0.308 3.808 0.808 4.269 1.423 1.656 0.615 2.615 0.577 3.308 0.038 3.577 0.346 3.962 0.808 4.269 1.346 1.881 0.577 2.615 0.577 3.308 0.038 3.577 0.346 3.962 0.808 4.269 1.346 1.435 0.808 3.423 0.538 3.500 0.346 3.731 0.154 3.423 0.808 3.538 0.423 0.212 0.846 2.615 0.577 3.308 0.038 3.577 0.346 3.769 0.808 4.269 The column “n” corresponds to the number of hold-out testing patients. The bset performance of each method is in italic.The best performance under each evaluation metric is underlined Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Page 8 of 11 Table 3 Overall Performance in LOV on 52 testing patients Method PLTR PLTRh KRL BMTMKL Feature level Task level QH@5 WQH@5 NH 1@1 1.385 1.668 0.788 3.212 0.423 3.654 0.115 3.750 0.288 3.423 0.788 3.423 1.327 1.616 0.808 3.269 0.423 3.654 0.115 3.731 0.173 3.423 0.788 3.404 1.327 1.652 0.788 3.212 0.423 3.712 0.115 3.750 0.269 3.423 0.788 3.404 1.308 1.616 0.788 3.154 0.423 3.654 0.115 3.712 0.288 3.481 0.788 3.615 1.288 1.581 0.808 3.173 0.423 3.596 0.115 3.750 0.192 3.519 0.788 3.635 1.269 1.616 0.808 3.115 0.423 3.635 0.115 3.731 0.250 3.481 0.788 3.635 NH 1@5 NH2@1 NH2@5 NH3@1 NH3@5 NH 5@1 NH 5@5 NHall@1 NHall@5 1.404 1.656 0.750 2.827 0.404 3.250 0.173 3.481 0.385 3.596 0.788 4.154 1.365 1.695 0.731 2.808 0.365 3.308 0.173 3.462 0.365 3.596 0.788 4.154 1.327 1.562 0.808 3.077 0.404 3.365 0.135 3.577 0.250 3.673 0.788 4.115 1.327 1.605 0.769 3.154 0.385 3.596 0.135 3.712 0.212 3.519 0.788 3.577 1.308 1.609 0.769 2.904 0.385 3.308 0.192 3.442 0.365 3.654 0.788 4.154 1.327 1.605 0.769 3.154 0.385 3.596 0.135 3.712 0.212 3.519 0.788 3.577 1.288 1.545 0.788 3.000 0.404 3.385 0.154 3.558 0.308 3.712 0.788 4.154 1.173 1.548 0.096 2.577 0.385 3.231 0.077 3.385 0.346 3.808 0.788 4.154 1.173 1.534 0.154 2.615 0.250 3.192 0.077 3.385 0.346 3.712 0.788 4.154 1.096 1.437 0.077 2.577 0.462 3.231 0.077 3.385 0.346 3.808 0.788 4.154 0.423 0.504 0.019 2.019 0.038 2.500 0.115 2.481 0.115 2.712 0.019 2.673 0.403 0.255 0.808 2.577 0.481 3.231 0.077 3.385 0.346 3.596 0.788 4.154 The column “n” corresponds to the number of hold-out testing patients. The best performance of each model is in italic. The best performance under each evaluation metric is upon underline. Comparison on cognitive task level Table 2 also shows that when 26 patients are hold out for testing, PLTR and PLTRh are both able to identify the top most relevant questionnaire for 84.6% of the testing patients (i.e., 22 patients) under NH1 @1. Table 3 shows that when 52 patients are hold out for testing, PLTR and PLTRh are both able to identify for 80.8% of the testing patients (i.e., 42 patients) under NH1 @1. Note that the hold-out testing patients in LOV do not have any cognitive features. Therefore, the performance of PLTR and PLTRh as above demonstrates their strong capability in identifying most AD related cognitive features based on imaging features only. We also find that PLTR and PLTRh are able to achieve similar or even better results compared to baseline methods in terms of the evaluation metrics on cognitive task level (i.e., NHg @1 and NHg @5). When 26 patients are hold out for testing, PLTR and PLTRh outperform the baseline methods in terms of NHg @1 (i.e., g = 1, 2 . . . 5). They are only slightly worse than KRL on ranking relevant tasks on their top-5 of predictions when g = 1 or g = 5 (3.308 vs 3.423 on NH1@5 and 3.808 vs 3.962 on NH5 @5). When 52 patients are hold out for testing, PLTR and PLTRh also achieve the best performance on most of the evaluation metrics. They are only slightly worse than KRL on NH2@1, NH5 @5 (0.423 vs 0.481 on NH2 @1 and 3.712 vs 3.808 on NH5@5). These experimental results demonstrate that among top 5 tasks in the ranking list, PLTR and PLTRh rank more relevant task on top than KRL. It’s notable that in Tables 2 and 3, as the number of features used to score cognitive tasks (i.e., g in NHg @k ) increases, the performance of all the methods in NHg @1 first declines and then increases. This may indicate that as g increases, irrelevant features which happen to have relatively high scores will be included in scoring tasks, and thus degrade the model performance on NHg @1. However, generally, the scores of irrelevant features are considerably lower than those of relevant ones. Thus, as more features are included, the scores for tasks are more dominated by the scores of relevant features and thus the performance increases. We also find that BMTMKL performs poorly on NH3 @1 in both Tables 2 and 3. This indicates that BMTMKL, a regression-based method, could not well rank relevant features and irrelevant features. It’s also notable that generally the best performance for the 26 testing patients is better than that for 52 testing patients. This may be due to that the similarities between the 26 testing patients and their top 10 similar training patients are higher than those for the 52 testing patients. The high similarities enable accurate latent vectors for testing patients. Tables 2 and 3 also show that PLTRh is better than PLTR on ranking cognitive tasks in LOV setting. When Peng et al. BMC Med Inform Decis Mak (2020) 20:319 26 patients are hold out for testing, PLTRh outperforms PLTR on NH1@5, NH5@5 and NHall @5 and achieves very comparable performance on the rest metrics. When 52 patients are hold out for testing, PLTRh is able to achieve better performance than PLTR on QH@5, WQH@5, NH3@1, NH5@1, NH5@5 and NHall @5 and also achieves very comparable performance on the rest metrics. Generally, PLTRh outperforms PLTR in terms of metrics on cognitive task level. This demonstrates the effectiveness of hinge loss-based methods in separating relevant and irrelevant features during modeling. Discussion Our experimental results show that when NH1 @1 achieves its best performance of 0.846 for the 26 testing patients in the LOV setting (i.e., the first row block in Table 2), the task that is most commonly prioritized for the testing patients is Rey Auditory Verbal Learning Test (RAVLT), including the following cognitive features: (1) trial 1 total number of words recalled; (2) trial 2 total number of words recalled; (3) trial 3 total number of words recalled; (4) trial 4 total number of words recalled; (5) trial 5 total number of words recalled; (6) total Score; (7) trial 6 total number of words recalled; (8) list B total number of words recalled; (9) 30 min delay total; and (10) 30 min delay recognition score. RAVLT is also the most relevant task in the ground truth if tasks are scored correspondingly. RAVLT assesses learning and memory, and has shown promising performance in early detection of AD [33]. A number of studies have reported high correlations between various RAVLT scores with different brain regions [34]. For instance, RAVLT recall is associated with medial prefrontal cortex and hippocampus; RAVLT recognition is highly correlated with thalamic and caudate nuclei. In addition, genetic analysis of APOE ε4 allele, the most common variant of AD, reported its association with RAVLT score in an early-MCI (EMCI) study [26]. The fact that RAVLT is prioritized demonstrates that PLTR is powerful in prioritizing cognitive features to assist AD diagnosis. Similarly, we find the top-5 most frequent cognitive tasks corresponding to the performance at NH3 @5 = 3.731 for the 26 hold-out testing patients. They are: Functional Assessment Questionnaire (FAQ), Clock Drawing Test (CDT), Weschler’s Logical Memory Scale (LOGMEM), Rey Auditory Verbal Learning Test (RAVLT), and Neuropsychiatric Inventory Questionnaire (NPIQ). In addition to RAVLT discussed above, other top prioritized cognitive tasks have also been reported to be associated with AD or its progression. In an MCI to AD conversion study, FAQ, NPIQ and RAVLT showed significant difference between MCI-converter and MCIstable groups [35]. We also notice that for some testing Page 9 of 11 subjects, PLTR is able to very well reconstruct their ranking structures. For example, when NH3 @5 achieves its optimal performance 3.731, for a certain testing subject, her top-5 predicted cognitive tasks RAVLT, LOGMEM, FAQ, NPIQ and CDT are exactly the top-5 cognitive tasks in the ground truth. These evidences further demonstrate the diagnostic power of our method. Conclusions We have proposed a novel machine learning paradigm to prioritize cognitive assessments based on their relevance to AD at the individual patient level. The paradigm tailors the cognitive biomarker discovery and cognitive assessment selection process to the brain morphometric characteristics of each individual patient. It has been implemented using newly developed learning-to-rank method PLTR and PLTRh. Our empirical study on the ADNI data has produced promising results to identify and prioritize individual-specific cognitive biomarkers as well as cognitive assessment tasks based on the individual’s structural MRI data. In addition, PLTRh shows better performance than PLTR on ranking cognitive assessment tasks. The resulting top ranked cognitive biomarkers and assessment tasks have the potential to aid personalized diagnosis and disease subtyping, and to make progress towards enabling precision medicine in AD. Supplementary information Supplementary information accompanies this paper at https://doi. org/10.1186/s12911-020-01339-z. Additional file 1. Supplementary materials. Abbreviations AD: Alzheimer’s Disease; MRI: Magnetic resonance imaging; ADNI: Alzheimer’s Disease Neuroimaging Initiative; LETOR: Learning-to-Rank; PET: Positron emission tomography; MCI: Mild Cognitive Impairment; HC: Healthy control; ADAS: Alzheimer’s Disease Assessment Scale; CDR: Clinical Dementia Rating Scale; FAQ: Functional Assessment Questionnaire; GDS: Geriatric Depression Scale; MMSE: Mini-Mental State Exam; MODHACH: Modified Hachinski Scale; NPIQ: Neuropsychiatric Inventory Questionnaire; BNT: Boston Naming Test; CDT: Clock Drawing Test; DSPAN: Digit Span Test; DSYM: Digit Symbol Test; FLUENCY: Category Fluency Test; LOGMEM: Weschler’s Logical Memory Scale; RAVLT: Rey Auditory Verbal Learning Test; TRAIL: Trail Making Test; RBF: Radial Basis Function; PLTR: Joint Push and Learning-to-Rank Method; PLTRh: Joint Push and Learning-to-Rank Method using Hinge Loss; BMTMKL: Bayesian Multi-Task Multi-Kernel Learning; KRL: Kernelized Rank Learning; CV: Cross Validation; LOV: Leave-Out Validation; QH@k: Average Feature Hit at k; WQH@k: Weighted Average Feature Hit at k; NHg@k: Task Hit at k; APOE: Apolipoprotein E; EMCI: Early-MCI. Acknowledgements Data collection and sharing for this project was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Peng et al. BMC Med Inform Decis Mak (2020) 20:319 Biogen; Bristol-Myers Squibb Company; CereSpir, Inc.; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Disease Cooperative Study at the University of California, San Diego. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California. For the ADNI: Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni. usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at: https://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ ADNI_Data_Use_Agreement.pdf. Authors’ contributions XN and LS designed the research study. BP and XY contributed to the conduct of the study: XY extracted and processed the data from ADNI; BP conduced the model development and data analysis. The results were analyzed, interpreted and discussed by BP, XY, SL, AJ, LS and XN. BP and XN drafted the manuscript, all co-authors revised the manuscript, and all authors read and approved the final manuscript. Funding This work was supported in part by NIH R01 EB022574, R01 AG019771, and P30 AG010133; NSF IIS 1837964 and 1855501. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the funding agencies. Availability of data and materials The dataset supporting the conclusions of this article is available in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) [25]. Ethics approval and consent to participate The dataset supporting the conclusions of this article is available in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) [25]. ADNI data can be requested by all interested investigators; they can request it via the ADNI website and must agree to acknowledge ADNI and its funders in the papers that use the data. There are also some other reporting requirements; the PI must give an annual report of what the data have been used for, and any publications arising. More details are available at http://adni.loni.usc.edu/data-samples/acces s-data/. Page 10 of 11 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Consent for publication Not applicable. 17. Competing interests The authors declare that they have no competing interests. 18. Author details 1 The Ohio State University, Columbus, USA. 2 University of Pennsylvania, Philadelphia, USA. 3 Indiana University, Indianapolis, USA. Received: 27 July 2020 Accepted: 17 November 2020 19. 20. References 1. Wan J, Zhang Z, et al. Identifying the neuroanatomical basis of cognitive impairment in Alzheimer’s disease by correlation- and nonlinearity-aware 21. sparse Bayesian learning. IEEE Trans Med Imaging. 2014;33(7):1475–87. https://doi.org/10.1109/TMI.2014.2314712. Yan J, Li T, et al. Cortical surface biomarkers for predicting cognitive outcomes using group l2,1 norm. Neurobiol Aging. 2015;36(Suppl 1):185–93. https://doi.org/10.1016/j.neurobiolaging.2014.07.045. Cordell CB, Borson S, et al. Alzheimer’s Association recommendations for operationalizing the detection of cognitive impairment during the medicare annual wellness visit in a primary care setting. Alzheimers Dement. 2013;9(2):141–50. https://doi.org/10.1016/j.jalz.2012.09.011. Scott J, Mayo AM. Instruments for detection and screening of cognitive impairment for older adults in primary care settings: a review. Geriatr Nurs. 2018;39(3):323–9. https://doi.org/10.1016/j.gerinurse.2017.11.001. He Y, Liu J, Ning X. Drug selection via joint push and learning to rank. IEEE/ACM Trans Comput Biol Bioinform. 2020;17(1):110–23. Weiner MW, Veitch DP, et al. The Alzheimer’s disease neuroimaging initiative 3: continued innovation for clinical trial improvement. Alzheimers Dement. 2017;13(5):561–71. Gentile C, Warmuth MK. Linear hinge loss and average margin. In: Proceedings of the 11th International Conference on Neural Information Processing Systems. NIPS’98. MA, USA: MIT Press, Cambridge; 1999. p. 225–31. Peng B, Yao X, Risacher SL, Saykin AJ, Shen L, Ning X. Prioritization of cognitive assessments in alzheimer’s disease via learning to rank using brain morphometric data. In: Proceedings of 2019 IEEE EMBS International Conference on Biomedical Health Informatics. New York, NY: IEEE; 2019. p. 1–4 . https://doi.org/10.1109/BHI.2019.8834618. Liu T-Y. Learning to rank for information retrieval. 1st ed. Berlin: Springer; 2011. p. 1–285. https://doi.org/10.1007/978-3-642-14267-3 Li, H. Learning to rank for information retrieval and natural language processing. 1st ed. In: Synthesis lectures on human language technologies, p. 114. San Rafael, California USA: Morgan & Claypool Publishers; 2011. https://doi.org/10.2200/S00607ED2V01Y201410HLT026. Agichtein E, Brill E, Dumais S, Brill E, Dumais S. Improving web search ranking by incorporating user behavior. In: Proceedings of SIGIR 2006; 2006. Karatzoglou A, Baltrunas L, Shi Y. Learning to rank for recommender systems. In: Proceedings of the 7th ACM conference on recommender systems. RecSys’13. New York: ACM; 2013. p. 493–4. https://doi. org/10.1145/2507157.2508063. Cao Z, Qin T, Liu T-Y, Tsai M-F, Li H. Learning to rank: from pairwise approach to listwise approach. In: Proceedings of the 24th international conference on machine learning. ACM; 2007. p. 129–36. Burges CJC, Ragno R, Le QV. Learning to rank with nonsmooth cost functions. In: Proceedings of the 19th international conference on neural information processing systems. NIPS’06. Cambridge: MIT Press; 2006. p. 193–200 Lebanon G, Lafferty J. Cranking: Combining rankings using conditional probability models on permutations. In: ICML, 2002; vol. 2, p. 363–70. Citeseer. Liu J, Ning X. Multi-assay-based compound prioritization via assistance utilization: a machine learning framework. J Chem Inf Model. 2017;57(3):484–98. Zhang W, Ji L, Chen Y, Tang K, Wang H, Zhu R, Jia W, Cao Z, Liu Q. When drug discovery meets web search: learning to rank for ligand-based virtual screening. J Cheminform. 2015;7(1):5. Liu J, Ning X. Differential compound prioritization via bi-directional selectivity push with power. In: Proceedings of the 8th ACM international conference on bioinformatics, computational biology, and health informatics. ACM-BCB’17. New York: ACM; 2017. p. 394–9. https://doi. org/10.1145/3107411.3107486. Liu J, Ning X. Differential compound prioritization via bi-directional selectivity push with power. J Chem Inf Model. 2017;57(12):2958–75. https:// doi.org/10.1021/acs.jcim.7b00552. Agarwal S, Dugar D, Sengupta S. Ranking chemical structures for drug discovery: a new machine learning approach. J Chem Inf Model. 2010;50(5):716–31. Wang X, Liu K, Yan J, Risacher SL, Saykin AJ, Shen L, Huang H et al. Predicting interrelated alzheimer’s disease outcomes via new self-learned structured low-rank model. In: International conference on information processing in medical imaging. Springer; 2017. p. 198–209. Peng et al. BMC Med Inform Decis Mak (2020) 20:319 22. Yan J, Deng C, Luo L, Wang X, Yao X, Shen L, Huang H. Identifying imaging markers for predicting cognitive assessments using wasserstein distances based matrix regression. Front Neurosci. 2019;13:668. https:// doi.org/10.3389/fnins.2019.00668. 23. Wang H, Nie F, Huang H, Risacher SL, Saykin AJ, Shen L. Alzheimer’s disease neuroimaging I. Identifying disease sensitive and quantitative trait-relevant biomarkers from multidimensional heterogeneous imaging genetics data via sparse multimodal multitask learning. Bioinformatics. 2012;28(12):127–36. https://doi.org/10.1093/bioinformatics/bts228. 24. Brand L, Wang H, Huang H, Risacher S, Saykin A, Shen L et al. Joint highorder multi-task feature learning to predict the progression of alzheimer’s disease. In: International conference on medical image computing and computer-assisted intervention. Springer; 2018. p. 555–62. 25. Weiner MW, Veitch DP et al. Alzheimer’s Disease Neuroimaging Initiative. https://adni.loni.usc.edu. Accessed 22 July 2020 26. Risacher S, Kim S, et al. The role of apolipoprotein e (apoe) genotype in early mild cognitive impairment (e-mci). Front Aging Neurosci. 2013;5:11. 27. Sarwar B, Karypis G, Konstan J, Riedl J. Item-based collaborative filtering recommendation algorithms. In: Proceedings of the 10th international conference on world wide web. WWW’01. New York: Association for Computing Machinery; 2001. p. 285–95. https://doi.org/10.1145/37192 0.372071. 28. Wang J, De Vries AP, Reinders MJ. Unifying user-based and item-based collaborative filtering approaches by similarity fusion. In: Proceedings of the 29th annual international ACM SIGIR conference on research and development in information retrieval; 2006. p. 501–8. Page 11 of 11 29. Koren Y, Bell R, Volinsky C. Matrix factorization techniques for recommender systems. Computer. 2009;42(8):30–7. 30. Costello JC, Heiser LM, Georgii E, Gönen M, Menden MP, Wang NJ, Bansal M, Hintsanen P, Khan SA, Mpindi J-P, et al. A community effort to assess and improve drug sensitivity prediction algorithms. Nat Biotechnol. 2014;32(12):1202. 31. He X, Folkman L, Borgwardt K. Kernelized rank learning for personalized drug recommendation. Bioinformatics. 2018;34(16):2808–16. 32. Challenge D. DREAM 7 NCI-DREAM drug sensitivity prediction challenge. http://dreamchallenges.org/project/dream-7-nci-dream-drug-sensitivit y-prediction-challenge/. Accessed 23 July 2020 33. Moradi E, Hallikainen I, et al. Rey’s Auditory Verbal Learning Test scores can be predicted from whole brain MRI in Alzheimer’s disease. NeuroImage: Clinical. 2017;13:415–27. 34. Balthazar MLF, Yasuda CL, et al. Learning, retrieval, and recognition are compromised in aMCI and mild AD: Are distinct episodic memory processes mediated by the same anatomical structures? J Int Neuropsychol Soc. 2010;16(1):205–9. 35. Risacher SL, Saykin AJ, et al. Baseline MRI predictors of conversion from MCI to probable AD in the ADNI cohort. Curr Alzheimer Res. 2009;6(4):347–61. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions
https://openalex.org/W2969806566	https://sajbm.org/index.php/sajbm/article/download/1079/1020	English	null	Effective safety propaganda: A study in South African industry	South African journal of business management	1,985	cc-by	4,459	Introduction In South Africa the National Occupational Safety Association (NOSA) conservatively estimates the cost of industrial accidents to be at least R400 million each year, (NOSA, 1979). In human terms 17 789 industrial injuries were reported to the Department of Manpower in 1982 and 16 489 in 1983. This figure includes 493 people who were fatally injured in 1982 and 489 in 1983, (Department of Manpower, 1984). ( p p ) Reduction in the severity and frequency of industrial accidents is clearly desirable in human, organizational, and financial terms. One route for the reduction of unsafe actions which may lead to accidents is to promote awareness of alternative safe actions through safety propaganda. In hierdie studie is 'n veiligheidspropagandaprogram ontwerp, voorgestel, gemonitor en geanaliseer in samehang met huidige industri~le aktiwiteite. Vorige eksperimentele ondersoeke het areas van onveilige aktiwiteite onthul wat toe onderhewig gemaak is aan die eksperimentele veiligheidspropagandamateriaal. Resultate in 'n aantal werl<winkelstudies het bewys dat die propaganda 'n posi- tiewe effek getoon het in terme van 'n vermindering van die aantal onveilige voorvalle. Ondergeskikte bevindinge het 'n aantal verskille in die formele en informele houdings van beide bestuurs- en toesighoudende vlakke getoon ten opsig- te van eenvoudige veiligheidsregulasies. Sell (1977) has provided a comprehensive review of the use of safety propaganda. He suggests that propaganda is: Sell (1977) has provided a comprehensive review of the use of safety propaganda. He suggests that propaganda is: ' ... an attempt to change the attitudes and/or behaviour of people who are not specifically in a learning situation, and who do not expect or may not even be aware of the information or instructions being given to them'. (Sell, 1977: 203); while safety propaganda is: ' . . . an organized attempt to change attitudes towards safety, and to cause people to act in a safe way, but excluding those situations where the target population is undergoing a period of formal instruction.' (Sell, 1977: 203). The more common types of safety propaganada include posters, full length films, short-length films (e.g. television and cinema advertising), verbal propaganda (radio, loudspeakers, etc.), safety competitions, and other incentives and leaflets applicable to non-training situations. pp g Much of the important research into the use and effec- tiveness of safety propaganda is directed towards evaluating large-scale campaigns on issues such as road safety and fire prevention. Effective safety propaganda: A study in South African industry J. Fisher and R. Parry Division of Industrial Psychology, University of the Witwatersrand, Johannesburg In this study a safety propaganda programme was designed, introduced, monitored, and analysed in the context of ongoing industrial activity. Pre-experimental investigations revealed areas of unsafe activity which were then subjected to the experimental safety propaganda mate- rial. Results showed that in a number of workshops studied the propaganda had a positive effect in terms of a reduction in the number of unsafe incidents. Ancillary findings showed wide differences in the formal and informal attitudes expressed by both managerial and supervisory grades towards simple safety situations. It is suggested that such inconsistency is unlikely to be conducive to a stable behavioural commitment to safety procedures. S. Afr. J. Bus. Mgmt. 1985, 16: 92- 97 J. Fisher* and R. Parry School of Ps~ chology, University of the Witwatersrand, I Jan Smuts Avenue, Johannesburg, 2001 Republic of South Africa •To whom correspondence should be addressed Accepted February 1985 Procedure The study was conducted at a large electrical component manufacturing plant situated in the East Rand in the Transvaal. The organization comprised a labour-intensive structure with a mainly 'european' executive and supervisory management group controlling a 'non-european', semi-skilled, and unskilled labour force. Although it could not be established whether the meaning of the two example posters had ever been explained to those workers interviewed, it should be apparent from the kind of response quoted above that as visual reminders this material is quite devoid of helpful association. The general nature of the poster in Figure l (b) is, as Sell (1977) suggests, of little value. Artwork was found to be confusing on both posters, and slogans were not understood by a work-force fluent in neither English nor Afrikaans. In view of the minimal under- standing displayed no attempt was made to take accurate ratings of meaning for these posters. Evaluating existing safety propaganda 2. Be negative, because this can show the wrong way of acting when what is required is the correct way. Safety propaganda during the preliminary period was sparse. It consisted of a small number of standard posters distributed by NOSA. These were often placed on notice boards, a NOSA recommendation, but not in strategic positions. Two examples of these posters are given in Figure 1. 3. Be general, because almost all people think they act safely. This type of propaganda is thus seen as only relevant to other people.' (Sell, 1977: 213) In more general terms safety propaganda should be purpose- fully directed towards those who are seen to be at risk. Moreover, in the industrial context where the attitudes symbolized by the safety programme are different from those of management, safety propaganda will be of little value. Safety propaganda must cover topics over which workers have some control, and management support for the proposed action is necessary. The effectiveness of these two posters, in terms of being visual reminders to encourage safe behaviour, was established through informal interviews conducted with small groups of workers. As few of the shopfloor work-force were fluent in either English or Afrikaans an interpreter proved to be essential. These interviews revealed a minimum understanding of the posters, often leading to incorrect and humorously erroneous interpretations, as illustrated by the following quotations: Aims of the present study Respondent: Shopfloor worker. Target: Poster, Figure 1 (a). 'The goggles look like they can be used for swimming, but I don't know what the dog is doing there ... What does the dog have to do with swimming?' The effectiveness of the recommendations for the construction and use of safety propaganda proposed by Sell (l 977) has not, as far as we are aware, been put to the test in the context of South African industry. The aim of the present study was to monitor and analyse a safety propaganda programme in the context of ongoing industrial activities, closely following the guidelines and recommendations set out by Sell (l 977) for the use of effective safety propaganda. Respondent: Shopfloor worker. Target: Poster, Figure l (b). 'With all these things around that man's head it looks like he is going to protect himself in a battle ... So it looks like he is a soldier in the army preparing himself for war.' All NOSA posters carry, in very small print, six points on how to use the material effectively. These suggest inter alia: 'Explain the meaning of this poster to employees through 5 minute "safety tip talks".' Introduction In the case of large-scale television campaigns evaluation of possible behavioural changes is made difficult because of the lack of control over the exact proportion of the population exposed to the propaganda. Numbers of viewers exposed can never be established for certain. However in more controlled, restricted circumstances such as industrial environments there is evidence to suggest that safety propaganda techniques can successfully lead to a decrease in unsafe acts, (Lander and Sell, 1960). The use of safety propaganda in South African industry has not been widely researched. Apart from a study reported by Winter (1963), little formal evaluation has been published. Winter examined the understanding shown by black South •To whom correspondence should be addressed •To whom correspondence should be addressed Accepted February 1985 S. Afr. J. Bus. Mgmt. 1985, 16(2) 93 African workers towards the content of a series of job-related safety posters. The low level of comprehension found could be attributable to such factors as the portrayal of behavioural variables through artistic conventions which were not a part of the subject sample's tradition. Table 1 Showing details of unsafe phenomena re- vealed from the pre-experimental investigation Phenomenon Failure to provide and use goggles at grinder workstations Failure to use ear muffs at riveting and other workstations Failure to use protective gloves at workstations when welding, spot-welding and cutting metal Workshop A,8,C A,D A,E Table 1 Showing details of unsafe phenomena re- vealed from the pre-experimental investigation j p Sell ( 1977) concludes that to be really effective safety propaganda must: p g 'l. Be specific to a task or situation. 2. Back up a training programme. 3. Give a positive instruction. 4. Be placed close to where the desired action is to take place. 5. Build on existing attitudes and knowledge. stations in workshop A. General workstations in workshop A were recorded at 90- 95 dBA. All of these levels are well above the recommended and statutory levels for accoustic safety. 6. Emphasise non safety aspects.' (Sell, 1977: 213) and should not: 'I. Involve horror because in the present state of our knowledge this appears to bring in defence mechanisms in the people at whom the propaganda is most directed. Preliminary inspection A preliminary inspection of the plant was carried out over a four-week period in search of unsafe phenomena. Inspection was guided by an analysis of accident reports for the preceding six months. These investigations revealed that a significant number of incidents occurred from three phenomena spread over five large workshops within the plant. The observed pattern is set out in Table I. In order to preserve the anonimity of the participating company these workshops are identified by the letters 'A' to 'E'. Evaluating managerial and supervisory disposition to safety 13 (7) The use of visual aids and reminders would benifit existing safety measures. Do you agree? 13 (8) Shopfloor safety is primarily the responsibility of the shopfloor workers. Do you agree? 3 11 'If the workers bugger up my machines I kick their arses for them. If they injure themselves I do the same'. 'The workers are given gloves to wear but they just don't wear them. I don't know why. If they get hurt then I feel that it is just their problem.' Content analysis of these statements indicated that both managers and supervisors largely viewed their own role in the promotion and maintenance of safety practices as being limited to an instructional one; the determinants of unsafe activity were largely attributed to a mysterious inevitability over which they had no control. Evaluating managerial and supervisory disposition to safety Evaluating managerial and supervisory disposition to safety During this preliminary period a questionnaire survey was conducted to assess the formal disposition of supervisors and executive and shopfloor managers towards safety in the work- place. The findings of this survey are set out in Table 2 and reveal a strong, positive commitment to safety issues, safety propaganda and the need to improve safety practices. Measurement of noise levels in workshops A and D, where failure to wear ear muffs was felt to be a problem, revealed levels of 92 - 95 dBA at riveting workstations in workshop D, and 100-105 dBA at sandblasting and fabricating work- Such a finding is not surprising in view of the direct nature 94 S.-Afr. Tydskr. Bedryfsl. 1985, 16(2) (b) Protect Yourself Beskertn U self Figure 1 Showing two examples of existing (NOSA) safety posters: (a) The need to protect eyesight, (b) General protective equipment use (b) Figure 1 Showing two examples of existing (NOSA) safety posters: Protect Yourself Beskertn U self Protect Yourself Beskertn U self Figure 1 Showing two examples of existing (NOSA) safety posters: (a) The need to protect eyesight, (b) General protective equipment use Table 2 Showing results of the pre-experimental sur- vey of management (shopfloor and executive) and su- pervisory grades' dispostion towards safety measures in the workplace, (N = 14) the noise.' (A response to the suggestion that supervisors could set an example by wearing ear protection when working in high-noise areas). Respondent: Shop manager. 'We give the workers protective clothing . · .. If they won't wear it what can we do?' Response Question Yes No (1 - 3) Biographical data (4) Time spent on safety matters is time wasted. Do you agree? nil 14 (5) Most shopfloor workers comply with shopfloor safety measures. Do you agree? 13 (6) Shopfloor safety measures and practices can be improved. Do you agree? 13 (7) The use of visual aids and reminders would benifit existing safety measures. Do you agree? 13 (8) Shopfloor safety is primarily the responsibility of the shopfloor workers. Do you agree? 3 11 Response Question Yes No (1 - 3) Biographical data (4) Time spent on safety matters is time wasted. Do you agree? nil 14 (5) Most shopfloor workers comply with shopfloor safety measures. Do you agree? 13 (6) Shopfloor safety measures and practices can be improved. Do you agree? Data collection A repeated measures observational design was conducted over three four-week periods - a control, an experimental, and a post-experimental period. These periods are set out in Table 3. The dependent variable was thus the number of unsafe acts observed per observation hour. The independent variable was ~e experimental safety posters. A susidiary dependent variable 1s the degree of support given to safety procedures by mana- gerial and supervisory grades. These initial sketches were then shown to individual shop- ~oor workers ~ho were selected at random from the popula- 10ns most at nsk to each safety situation. Each respondent Nas requested, via a translator, to rank the sketches relating o each of the three target phenomena in order of significance ~d meanin~ in terms of instructional value about the right :hin_g to do m each case. The leading designs were then pro- fessionally drawn up by graphics specialists into full poster format. The final designs are shown in Figure 2. Table 3 Showing details of the experimental design Time Period Conn-ol period Experimental period Post-experimental period Experimental procedure Existing safety behaviour is observed and recorded Shopfloor workers exposed to experimental safety propaganda material. Safety be- haviour is observed and recorded Experimental safety propaganda material is withdrawn. Safety behaviour is observed and recorded Table 3 Showing details of the experimental design Time Period Conn-ol period Experimental procedure Existing safety behaviour is observed and recorded Table 3 Showing details of the experimental design (a) (b) (c) Figure 2 Showing the final designs for experimental safety posters: (a) The need to wear protective goggles, (b) The need to wear protective gloves, (c) The need to wear protective ear muffs The observational procedure was shaped by the diverse nature of the target phenomena. Certain jobs (e.g. riveting) are conducted with the operator remaining at one permanent, fixed workstation. Other target phenomena (e.g. use of grindstones) were observable at fixed points in the workshops but used only infrequently by various members of the work- force. A pseudo-randomized observational schedule was necessary to sample these different situations. The observation schedule covered all three of the four-week periods. A variety of non-safety related tasks (e.g. inspection and stock control), around the factory premises provided cover for the experi- mental observations. Preparation of experimental safety posters of the questions asked. However, later in this preliminary period these same respondents were interviewed on an infor- mal basis about the same issues. The positive disposition displayed in the questionnaire survey was not reflected by the informal comments recorded, particularly at supervisory levels. The following quotations are illustrative of this contradictory position: Paying close attention to the impressionistic comments .of shopfloor workers towards existing poster material and with due regard to the wealth of information in psychological, ergonomics, and safety literature regarding the perception and interpretation of international safety symbols (Dreyfus, 1972; Stillerman, 1976; Anon, 1979 and Anon, 1981), a number of draft posters were sketched for each of the three safety phenomena (see Table 1). Effort was made to keep each sketch artistically simple and to concentrate on presenting the Respondent: Shop foreman. Respondent: Shop foreman. 'I've been in this game for 33 years now. I've got used to S. Afr. J. Bus. Mgmt. 1985, 16(2) 9S undamental elements of each target safety phenomenon. [bus for incidents involving the use of grindstones the ~tial iactors arising from the initial interviews with shopfloor Norkers were the grindstones, goggles, and eyes. The questionnaire Respondents completed the questionnaire in exactly the same way as they had done during the preliminary investigation some four months previously. There is thus no change from Table 2. No further statistical analysis was conducted on these data. Figure 2 Showing the final designs for experimental safety posters: (a) The need to wear protective goggles, (b) The need to wear protective gloves, (c) The need to wear protective ear muffs Figure 2 Showing the final designs for experimental safety posters: (a) The need to wear protective goggles, (b) The need to wear protective gloves, (c) The need to wear protective ear muffs The observations These posters were then printed and used as experimental safety propaganda material in the workshops under investiga- tion. Care was taken to position the experimental posters in prorninant positions at the workstations to be monitored. No other safety instruction or advice was given to either managers, supervisors, or workers. Results set out in Table 4 were obtained from observational monitoring of the behavioural effect of safety propaganda related to the three target phenomena, viz. non-use of goggles, gloves, and ear muffs. The data are expressed as units of observed unsafe incidents per hour of observation time. Fluc- tuations in numbers of workers per workshop were monitored and were found to be negligable for the purposes of this study. Data are displayed on a phenomenon/workshop basis. Data collection The questionnaire survey conducted in the pre-experimental period to establish managerial and supervisory disposition towards safety was repeated following the end of the post- experimental period in order to record any changes in formally expressed disposition which may have arisen as a result of the safety propaganda campaign. Experimental hypothesis Safety propaganda cannot combat the uner- gonomic design of essential equipment. Table 5 Showing unsafe acts recorded per hour of ob- servation time expressed as ratios of control period figures for target phenomena and workshop location. Experimental period Target Work- Post- phenomenon shop Control Experimental experimental Goggles A l,00 0,76 0,62 B 1,00 1,07 0,96 C l,00 0,41 0,45 Gloves A l,00 0,53 0,59 D l,00 1,00 1,00 Ear muffs A 1,00 1,16 2,21 E 1,00 0,46 0,76 Lastly the constant levels of non-use of goggles in workshop B contrast with the substantial reductions found in workshops A and C. This is in part attributable to the relative infrequent use of the grindstones in workshop B and in part to their most prominent users being the workshop supervisors who refused to use the goggles provided. g gg p This apparent ambivalence between formal commitment to safety and observable unconcerned, often obstructive, beha- viour is clearly of major concern above and beyond its effects upon safety propaganda impact. There is clearly much scope for further study into its determinants. Psychology is of course rich in theoretical constructs devised for the explanation of such disparate manifestations. However, an important feature of this situation is the genuine lack of awareness on the part of supervisory staff of both the nature of safety hazards and their own role in the promotion of safe behaviour. At the very least this points to an immediate training need (at all levels of the work-force) and may indicate a fundamental psycholo- gical problem. Discussion of results To take the follow-up questionnaire aspect first, it should not be surprising to find no shift in respondents' manifest position of a strong positive commitment to safety and safety propa- ganda. However the apparent ambivalence between manifest statement and observable action, particularly on the part of supervisory grades, has much influence on the results of the observational study. The observation results indicate a moderate trend in favour of the experimental hypothesis. For four of the seven targets the introduction of the experimental safety propaganda was followed by a prominant reduction in the number of unsafe acts observed. Experimental hypothesis The introduction of experimental safety propaganda will have the effect of reducing the number of observed unsafe incidents related to the target phenomena to below that found in a control period but these levels will rise again following the withdrawal of the experimental propaganda material. p y p p A Page's L non-parametric trend test (Page, 1963) showed X:L = 1,786; p = 0,09. The data recorded in Table 4 are re-expressed in Table 5 as ratios using the control period as unity. S.-Afr. Tydskr. Bedryfsl. 198S, 16(l) 96 Table 4 Showing the number of unsafe acts recorded per hour of observation time as a function of target phenomena and workshop location Table 4 Showing the number of unsafe acts recorded per hour of observation time as a function of target phenomena and workshop location occurrence is in part attributable to the flamboyant behaviour on the part of the shop supervisors, who themselves refused to wear ear protection or to enforce its wearing by others and in part to the comparative mobility of workers around the workshop. The need to use goggles and gloves to perform specific tasks was recognized by all in this workshop (though unsafe acts were still in evidence) but the insidious haz.ard of high levels of continuous noise was largely not recognized despite the positioning of experimental posters at strategic high-noise locations. Experimental period Target Work- Post- phenomena shop Control Experimental experimental Goggles A 8,33 6,33 5,20 B l,00 1,07 0,96 C 2,13 0,87 0,96 Gloves A 33,67 18,00 20,00 D 83,00 83,00 83,20 Ear muffs A 6,33 7,33 14,00 E 33,67 16,00 25,60 Experimental period In contrast workers in workshop E attended more static workstations with high levels of noise generated as an integral part of their primary task performance, (riveting). The non- use of ear muffs was halved by the prominent positioning of experimental posters. Similarly the induced reduction in non-use of gloves in workshop A contrasted with the apparent indifference to the experimental propaganda recorded in workshop D. Again this difference may be attributed to the nature of the task characteristics. The higher degree of manual dexterity required for most tasks in workshop D was not facilitated by the type of protective gloves provided. Non-use of gloves in this instance was not a safety choice so much as a task perfor- mance necessity. Experimental hypothesis This reduction remained substantial even during the post-experimental period, though towards the end of this last four-week period the rate of unsafe acts observed was beginning to increase to the level recorded during the control period, suggesting that the awareness and concern generated by the poster campaign had finally ebbed. Conclusion Lander, S. & Sell, R. 1960. An experiment on the effect of specially designed safety posters. Occup. Psycho/., vol. 34, 153- 169. Lander, S. & Sell, R. 1960. An experiment on the effect of The study thus adds support to the view that safety posters can be effective, but cannot be regarded as alternatives to devising safety procedures and promoting safety awareness in a broader organizational context, (Sell, 1977). They may however serve as reminders to reinforce other methods, (Lander & Sell, 1960). National Occupational Safety Association (NOSA). 1979. The MBO system leading to a 5 star roting. NOSA. Pretoria. Page, E.B. 1963. Ordered hypothesis for multiple treatments; a significance test for linear ranks. Am. Stat. Ass-oc. J., 216-230. Robinson G.H. 1982. Accidents and sociotechnical systems; principles for design. Accid. Anal. Prev., vol. 14, (2), 121 - 130. General discussion The study demonstrates that through the application of some basic exploration and evaluation techniques from the human sciences the meaning of safety posters can be increased from zero to significance. Indeed the results show that provided safety propaganda posters are constructed in a way which emphasizes the relevance of unsafe phenomena to those who are most at risk, their use will lead to the promotion of safe actions. However, where there is a general nonrealization of a hazard (e.g. the insidious nature of noise), and/or where there are unsupportive managerial or supervisory dispositions, and/or where there is evidence of the unergonomic nature of safety equipment, the likelihood of safety propaganda producing a substantial change in safety behaviour will be negligible. The pattern of changes in observed unsafe acts is best seen in Table 5. Of the three target phenomena which did not show reduc- tion two remained constant across the experimental periods (non-use of goggles in workshop B and non-use of gloves in workshop D) while the third (non-use of ear muffs in work- shop A) increased twofold. No clear associations emerged between rate of unsafe act, workshop, or nature and frequency of use of workstations. An investment of both time and effort in the observance of safety procedures is necessary for the elimination of unsafe acts in the workplace. So too is a broader understanding of the relationship between unsafe behaviour and the supportiv_e factors of task structure and procedure, equipment appropn- p q y Workshop A for example shows substantial reductions in the non-use of goggles and gloves but provides a substantial increase (twofold) in the non-use of ear muffs. The latter S. Afr. J. Bus. Mgmt. 1985, 16(2) 97 ateness and the general operating environment, (Surry, 1969; Adams, Barlow & Hiddlestone, 1981; Edwards, 1981; Robinson, 1982 and Fisher, 1983). ateness and the general operating environment, (Surry, 1969; Adams, Barlow & Hiddlestone, 1981; Edwards, 1981; Robinson, 1982 and Fisher, 1983). procedure. Appl. Ergon., vol. 12, (2), 111 - 115. Fisher, J. 1983. Reducing the frequency and severity of industrial accidents - an ergonomics approach. S. Afr. J. Labour Re/at., vol. 7, (4), 4-17. References Anon 1979. Standardised workplace safety signs. Appl. Ergon., vol. 10, (3), 190- 191. Siegal, S. 1956. Nonparametric statistics for the behavioural sciences. Kogakusha, Tokyo: McGraw-Hill. Anon 1981. Standard safety signs. Appl. Ergon., vol. 12, (1), 61 - 62. g y Sell, R.G. 1977. What does safety propaganda do for safety? A review. Appl. Ergon., vol. 8, (4), 203 - 214. Adams, N.L., Barlow A. and Hiddlestone J. 1981. Obtaining ergonomics information about industrial injuries - a five year analysis. Appl. Ergon., vol. 12, (2), 71 - 81. Stillerman, E.D. 1976. Comprehension of ISO industrial safety signs among white, black literate and black illiterate workers. National Institute for Personnel Research. Johannesburg. Department of Manpower. 1984. Report of the Director General for the year ended 31 December 1983. Government printer. Pretoria. g Surry, J. 1969. Industrial accident research. A human engineering appraisal. Ontario Ministry of Labor. Toronto. p Dreyfus, H. 1972. Symbol sourcebook. New York: McGraw-Hill. Winter, W. 1963. The perception of safety posters by bantu industrial workers. Psycho/. Afr., vol. 10, 127 - 135. Edwards, M. 1981. The design of an accident investigation
https://openalex.org/W1505669681	https://europepmc.org/articles/pmc3991836?pdf=render	English	null	School situation for specially abled children – How can we as qualitative researchers contribute?	International journal of qualitative studies on health and well-being	2,014	cc-by	829	EDITORIAL School situation for specially abled children How can we as qualitative researchers contribute? specially abled children are teased or even bullied. Bullying in childhood can lead to decreased self- esteem, social isolation, anxiety, and suicidal thoughts both in the short and long term. This means that creating friendships in children with and without disabilities is not done automatically; instead it requires an active engagement from adults around the children. Specially abled children often absent from school because of their disabilities. It is difficult for the teachers, who often lack special training, to help these children to make up for their absence which means such children often underperform compared with their normal schoolmates. This also affects their self-esteem. School is and should be an important part in the lives of all children and youth. Much time is spent in school, and it serves not only as an educational arena but also as a social arena that fosters us as human beings. Lifelong friendships are ultimately developed in school, which serve as a protection when life gets rough. Our achievements in school predict our later possibilities to attend higher education and in turn possibilities to find a job, which has an influence on our social and financial standard, as also our health and wellbeing. Many countries in the Western world are actively working toward the inclusion of specially abled children in schools primarily planned and built for children without disabilities. This is a political stand- point based on every individual’s equal right to par- ticipate in the society on equal terms. There is an ongoing debate in the research society as to whether this is in the best interest of the child with a disability. We know today that there are both benefits and risks in including these children in ‘‘regular’’ schools. We also know that if the child’s individual needs are not fulfilled, then inclusion in ‘‘regular’’ schools is not fruitful for any individual involved, especially not for the child with a disability. On the contrary, specially abled children who attend schools built specifically to cater to their spe- cial needs often have better self-esteem. They mirror themselves with other children who also have special needs which mean that they, to some extent, are happily unaware of their individual shortcomings, which in turn serve as a protection from decreased self-esteem. # 2014 U. Hallberg. This is an Open Access article distributed under the terms of the Creative Commons CC-BY 4.0 License (http:// creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license. 1 Citation: Int J Qualitative Stud Health Well-being 2014, 9: 24396 - http://dx.doi.org/10.3402/qhw.v9.24396 (page number not for citation purpose International Journal of Qualitative Studies on Health and Well-being International Journal of Qualitative Studies on Health and Well-being International Journal of Qualitative Studies on Health and Well-being EDITORIAL School situation for specially abled children How can we as qualitative researchers contribute? Their teachers often have special training to understand and appreciate their needs, thereby willing to impart the lessons in accordance with the individual child’s level of understanding and needs. The intention of including specially abled children in ‘‘regular’’ schools is primarily to give these children a chance to socialize and build long-lasting relation- ships with children without disabilities. Having friends is crucial for development in all children. It is within these relationships that socialization takes place, and important values and approaches such as respect toward others are shaped and tried out. Long- lasting friendships also serve as a protection when children are going through rough times in life. Another benefit in the inclusion of specially abled children in ‘‘regular’’ schools is to acknowledge them in society, which in the long term leads to increased acceptance for this group. We know that specially abled children themselves regard negative attitudes from schoolmates as the biggest obstacle for inclusion in ‘‘regular’’ schools. But we need more knowledge to help these children to increase their self-esteem and their achievements in school. Here qualitative researchers have an impor- tant role to play. High-quality qualitative studies are an excellent instrument in expanding the knowledge about these children’s school situations and thereby improving their lives both in the short and long term. To let the children themselves speak and tell their stories can give us valuable insight into their situation and can guide us in our attempts to improve the health and wellbeing of all specially abled children. Sadly, many specially abled children who are included in ‘‘regular’’ schools are excluded from the possibilities of creating friendships with other children without disabilities. Often, their schoolmates have a neutral or indifferent attitude toward them, some- times even a negative attitude. In worst cases the Ulrika Hallberg Associate Professor Nordic School of Public Health (NHV) Gothenburg, Sweden g/10.3402/qhw.v9.24396 (page number not for citation purpose)
https://openalex.org/W2325001694	https://europepmc.org/articles/pmc4817047?pdf=render	English	null	Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram-Scale Production of SnO2 Hollow Nanospheres via Spray Drying System	Scientific reports	2,016	cc-by	6,828	Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram-Scale Production of SnO2 Hollow Nanospheres via Spray Drying System Jung SangCho, Hyeon Seok Ju &YunChan Kang Hollow nanospheres are typically prepared using easily removable organic or inorganic nanopowders as template mate- rials8–13. The template nanopowders are uniformly coated with metal precursors by hydrothermal or precipita- tion methods. The post-treatment process forms a metal oxide layer and eliminates the template nanopowder, producing a metal oxide hollow nanosphere. Monodisperse organic polymer nanopowders, such as polystyrene (PS) and poly(methyl methacrylate) (PMMA), which are mainly prepared by emulsion methods, have been used because they are easily removed by combustion under an oxygen atmosphere. In addition, monodisperse silica nanopowders prepared by a modified Stöber method have been used as inorganic templates because silica can be eliminated with a HF or NaOH solution6,14,15. However, the processes for producing hollow nanospheres using organic and inorganic nanopowders as templates are time consuming and not cost-effective. Moreover, the yield is often on the scale of milligrams to grams, limiting their application for large-scale production of homogeneous metal oxide hollow nanospheres. In recent years, conversion chemical reactions employing nanoscale Kirkendall diffusion have received sig- nificant attention as a method for producing hollow nanospheres16–22. Metal nanopowders formed by a reduction reaction in aqueous media are transformed into hollow nanospheres by nanoscale Kirkendall diffusion; however, that preparation of homogeneous nanopowders is not easily carried out without aggregation. As a result, oxi- dation of metal nanopowders by the nanoscale Kirkendall diffusion process is typically performed by bubbling oxygen through the aqueous media16,17. Alternatively, metal nanopowders are oxidized over the carbon-coated Cu Department of Materials Science and Engineering, Korea University, Anam-Dong, Seongbuk-Gu, Seoul 136-713, Republic of Korea. Correspondence and requests for materials should be addressed to Y.C.K. (email: yckang@korea. ac.kr) Metal-oxide hollow nanospheres with a high surface area and a sufficiently void internal volume can be used in various applications, including energy storage devices, solar cells, catalysts, sensors, and drug delivery1–9. Hollow nanospheres are typically prepared using easily removable organic or inorganic nanopowders as template mate- rials8–13. The template nanopowders are uniformly coated with metal precursors by hydrothermal or precipita- tion methods. The post-treatment process forms a metal oxide layer and eliminates the template nanopowder, producing a metal oxide hollow nanosphere. Monodisperse organic polymer nanopowders, such as polystyrene (PS) and poly(methyl methacrylate) (PMMA), which are mainly prepared by emulsion methods, have been used because they are easily removed by combustion under an oxygen atmosphere. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Department of Materials Science and Engineering, Korea University, Anam-Dong, Seongbuk-Gu, Seoul 136-713, Republic of Korea. Correspondence and requests for materials should be addressed to Y.C.K. (email: yckang@korea. ac.kr) Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram-Scale Production of SnO2 Hollow Nanospheres via Spray Drying System Jung SangCho, Hyeon Seok Ju &YunChan Kang received: 24 November 2015 accepted: 16 March 2016 Published: 01 April 2016 Jung Sang Cho, Hyeon Seok Ju & Yun Chan Kang A commercially applicable and simple process for the preparation of aggregation-free metal oxide hollow nanospheres is developed by applying nanoscale Kirkendall diffusion to a large-scale spray drying process. The precursor powders prepared by spray drying are transformed into homogeneous metal oxide hollow nanospheres through a simple post-treatment process. Aggregation-free SnO2 hollow nanospheres are selected as the first target material for lithium ion storage applications. Amorphous carbon microspheres with uniformly dispersed Sn metal nanopowder are prepared in the first step of the post-treatment process under a reducing atmosphere. The post-treatment of the Sn-C composite powder at 500 °C under an air atmosphere produces carbon- and aggregation-free SnO2 hollow nanospheres through nanoscale Kirkendall diffusion. The hollow and filled SnO2 nanopowders exhibit different cycling performances, with their discharge capacities after 300 cycles being 643 and 280 mA h g−1, respectively, at a current density of 2 A g−1. The SnO2 hollow nanospheres with high structural stability exhibit superior cycling and rate performances for lithium ion storage compared to the filled ones. 1 Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 A commercially applicable and simple process for the preparation of aggregation-free metal oxide hollow nanospheres is developed by applying nanoscale Kirkendall diffusion to a large-scale spray drying process. The precursor powders prepared by spray drying are transformed into homogeneous metal oxide hollow nanospheres through a simple post-treatment process. Aggregation-free SnO2 hollow nanospheres are selected as the first target material for lithium ion storage applications. Amorphous carbon microspheres with uniformly dispersed Sn metal nanopowder are prepared in the first step of the post-treatment process under a reducing atmosphere. The post-treatment of the Sn-C composite powder at 500 °C under an air atmosphere produces carbon- and aggregation-free SnO2 hollow nanospheres through nanoscale Kirkendall diffusion. The hollow and filled SnO2 nanopowders exhibit different cycling performances, with their discharge capacities after 300 cycles being 643 and 280 mA h g−1, respectively, at a current density of 2 A g−1. The SnO2 hollow nanospheres with high structural stability exhibit superior cycling and rate performances for lithium ion storage compared to the filled ones. Metal-oxide hollow nanospheres with a high surface area and a sufficiently void internal volume can be used in various applications, including energy storage devices, solar cells, catalysts, sensors, and drug delivery1–9. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram-Scale Production of SnO2 Hollow Nanospheres via Spray Drying System Jung SangCho, Hyeon Seok Ju &YunChan Kang In addition, monodisperse silica nanopowders prepared by a modified Stöber method have been used as inorganic templates because silica can be eliminated with a HF or NaOH solution6,14,15. However, the processes for producing hollow nanospheres using organic and inorganic nanopowders as templates are time consuming and not cost-effective. Moreover, the yield is often on the scale of milligrams to grams, limiting their application for large-scale production of homogeneous metal oxide hollow nanospheres.f p In recent years, conversion chemical reactions employing nanoscale Kirkendall diffusion have received sig- nificant attention as a method for producing hollow nanospheres16–22. Metal nanopowders formed by a reduction reaction in aqueous media are transformed into hollow nanospheres by nanoscale Kirkendall diffusion; however, that preparation of homogeneous nanopowders is not easily carried out without aggregation. As a result, oxi- dation of metal nanopowders by the nanoscale Kirkendall diffusion process is typically performed by bubbling oxygen through the aqueous media16,17. Alternatively, metal nanopowders are oxidized over the carbon-coated Cu Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 1 www.nature.com/scientificreports/ Figure 1. Schematic diagram for the formation mechanism of the aggregation-free SnO2 hollow nanospheres. Figure 1. Schematic diagram for the formation mechanism of the aggregation-free SnO2 hollow nanospheres. grids used for TEM measurements19,20. Therefore, the development of a simple process for large-scale production of metal oxide hollow nanospheres remains a large challenge if widespread application is to take place. In this study, a simple and commercially applicable process to produce aggregation-free metal oxide hollow nanospheres has been developed, which uses large-scale spray drying. The precursor powders prepared by the spray drying process are transformed into homogeneous metal oxide hollow nanospheres by a simple two-step post-treatment process. Transition metal oxide hollow nanospheres can be successfully used as anode materials for lithium ion batteries (LIBs) as a result of their short Li-ion diffusion length and good accommodation of volume changes resulting from repeated insertion and extraction of Li4–6,23–28. Nanostructured tin oxide (SnO2) materials, which have a wide band gap of 3.54 eV, are widely applied in various fields, such as gas sensors and catalysts as well as energy storage devices6,29–38. In this study, the LIB anode was selected as the first target applica- tion for aggregation-free SnO2 hollow nanospheres. Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram-Scale Production of SnO2 Hollow Nanospheres via Spray Drying System Jung SangCho, Hyeon Seok Ju &YunChan Kang The detailed formation mechanism of aggregation-free SnO2 hollow nanospheres by the nanoscale Kirkendall diffusion process was studied by investigating morphological changes of the spray dried powders during the post-treatment process under reducing and oxidizing atmospheres. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 Results and Discussion The characteristics of the precursor powders produced by the spray drying process are shown in Figure S3. The x-ray diffraction (XRD) pattern of the precursor powders reveals the broad crystalline peaks of the tetragonal SnO2 phase. The thermogravimetric (TG) curve shown in Figure S3b shows a three-step weight loss of the precursor powders below 500 °C. The distinct weight losses at temperatures around 170, 260 and 380 °C were attributed to decomposition of tin oxalate, carbonization of PVP, and combustion of the carbon component, respectively. The decomposition of some amount of tin oxalate into SnO2 occurred in the spray drying process, as shown by XRD and TG analysis. The precursor powders showed the typical collapsed structure of hollow powders. The formation of a gas-impermeable layer during an early drying stage of the droplets resulted in hollow powders, with further gas evolution by water evaporation expanding the balloon–like structures. The explosion of expanded powder resulted in collapsed structures with holes, as shown by arrows in Figure S3c.h p , y g The characteristics of the Sn-C composite powders obtained by reduction of the spray dried precursor pow- ders are shown in Fig. 3 and Figure S4a. The XRD pattern shown in Figure S4a revealed the marked crystalline peaks of the metallic Sn phase. Complete decomposition of tin oxalate into tin oxide and subsequent reduction to metallic Sn occurred during this stage. The low-resolution scanning electron microscope (SEM) image shown in Fig. 3a reveals a similar morphology to that of the spray dried precursor powder. However, the high-resolution SEM and TEM images in Fig. 3b–d reveal a unique morphology of the powders prepared by the reduction pro- cess. Ultrafine nanopowders with a narrow size distribution were uniformly distributed over the transparent powder. The amorphous carbon formed by carbonization of PVP appears as a transparent matrix supporting the nanopowders in the SEM and TEM images. The early stage of the reduction process formed the homogeneous amorphous carbon-tin oxide composite powder as an intermediate product. Reduction of tin oxide into metal- lic Sn changed the morphology of the powders, and segregation of Sn metal with its low melting temperature occurred within the amorphous carbon matrix to form aggregation-free Sn nanopowders. The Sn nanopowder located close to the surface of the Sn-C composite powder also shows a carbon layer, indicated by an arrow in Fig. 3e. Results and Discussion A diagram illustrating the formation of aggregation-free SnO2 hollow nanospheres is given in Fig. 1. The Sn oxalate-PVP composite powder prepared by a pilot-scale spray drying process showed particles several tens of micrometers in size (Fig. 1①). Post-treatment of the spray-dried powder at 300 °C under a H2/Ar gas mix- ture produced the Sn-C composite powder (Fig. 1②). Carbonization of PVP during the post-treatment process produced an amorphous carbon matrix. Decomposition of Sn oxalate into tin oxide occurred, with subsequent reduction to Sn metal. The segregation of Sn during reduction, with its low melting temperature, resulted in aggregation-free Sn metal nanopowders. Next, post-treatment of the Sn-C composite powder at 500 °C under an air atmosphere produced carbon-free and aggregation-free SnO2 hollow nanospheres through the process of well-known nanoscale Kirkendall diffusion (Fig. 1③,④,⑤). The Kirkendall effect, a vacancy flux and subsequent void formation process resulting from diffusivity differences at inorganic interfaces, was described in detail in Fig. 2. The Kirkendall effect results in the formation of a thin SnO2 layer on the Sn metal surface, followed by simultaneous outward diffusion of Sn cations through the oxide layer and inward diffusion of oxygen into the nanospheres, creating an intermediate Sn@SnO2 core–shell structure (Fig. 2②). Sn cations diffused outward more quickly than oxygen diffused inward, which is consistent with the larger ionic radius of oxygen anions (140 pm) than Sn cations (Sn2+ is 93 pm, Sn4+ is 69 pm). Accordingly, Kirkendall voids were generated near the Sn/SnO2 interface during vacancy-assisted exchange of the material via bulk interdiffusion (Fig. 2③), which gave rise to coarsening and enhancement of pore growth in the spheres (Fig. 2④). Eventually, both complete con- version of Sn metal into SnO2 by Kirkendall-type diffusion and complete combustion of the amorphous carbon Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 2 www.nature.com/scientificreports/ Figure 2. Possible formation mechanism of a hollow SnO2 nanosphere by Kirkendall diffusion effect and its chemical conversion process in the surface region of a sphere. Figure 2. Possible formation mechanism of a hollow SnO2 nanosphere by Kirkendall diffusion effect an its chemical conversion process in the surface region of a sphere. material surrounding the Sn metal spheres resulted in the carbon-free and aggregation-free SnO2 hollow nano- spheres (Fig. 2⑤).h The formation mechanism of the carbon-free and aggregation-free SnO2 hollow nanospheres was investigated on the basis of morphology changes induced by post-treatment in the Sn oxalate-PVP composite powder. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 Results and Discussion The line profiling and elemental mapping images revealed pure Sn metal nanopowder embedded within the amorphous carbon matrix (Fig. 3f,g). The mean particle size of the Sn metal nanopowders measured from the SEM images was 95 nm. The TG curve of the Sn-C composite powders in Figure S5a revealed a slight weight increase beginning at 150 °C and distinct weight loss between 350 and 380 °C. Weight increase from oxidation of metallic Sn nanopowder was reduced by combustion of amorphous carbon. The amorphous carbon content of the Sn-C composite powders estimated from the TG analysis was 50 wt%.h p p y The characteristics of the tin oxide powders obtained by oxidation of the Sn-C composite powders at 500 °C are shown in Fig. 4 and Figure S4b. The XRD pattern shown in Figure S4b reveals a mixed crystal structure of tetragonal and orthorhombic SnO2 phases; complete oxidation of the metallic Sn nanopowders into SnO2 Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 3 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 3. Morphologies, line profiling analysis, and elemental mapping images of the Sn-C composi obtained by reduction of the spray dried precursor powders at 300 °C under 10% H2/Ar gas: (a–c) SE (d,e) TEM images, (f) line profiling analysis, and (g) elemental mapping images. Figure 3. Morphologies, line profiling analysis, and elemental mapping images of the Sn-C composite powders obtained by reduction of the spray dried precursor powders at 300 °C under 10% H2/Ar gas: (a–c) SEM images, (d,e) TEM images, (f) line profiling analysis, and (g) elemental mapping images. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 4 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 4. Morphologies, SAED pattern, and elemental mapping images of the SnO2 hollow nanosph obtained by oxidation of reduced Sn-C composite powders at 500 °C under air: (a) SEM, (b–d) TEM (e) HR-TEM image, (f) SAED pattern, and (g) elemental mapping images. Figure 4. Morphologies, SAED pattern, and elemental mapping images of the SnO2 hollow nanospheres obtained by oxidation of reduced Sn-C composite powders at 500 °C under air: (a) SEM, (b–d) TEM imag (e) HR-TEM image, (f) SAED pattern, and (g) elemental mapping images. Figure 4. Morphologies, SAED pattern, and elemental mapping images of the SnO2 hollow nanospheres obtained by oxidation of reduced Sn-C composite powders at 500 °C under air: (a) SEM, (b–d) TEM images (e) HR-TEM image, (f) SAED pattern, and (g) elemental mapping images. Figure 4. www.nature.com/scientificreports/ occurred during the process. Morphology of the powders has changed drastically with oxidation. The Sn-C composite powder with particles several tens of micrometers in size has changed into SnO2 nanopowder with nanometer-sized particles, as a result of combustion of the amorphous carbon matrix. The TEM images revealed the hollow structure of the SnO2 nanopowders. The Sn nanopowder, which had a dense structure, was trans- formed into hollow SnO2 nanospheres by the well-known nanoscale Kirkendall diffusion process elucidated in Fig. 4. A clear void space was observed inside the SnO2 nanopowders as indicated by arrows in Fig. 4c. The shell thickness and diameter of the SnO2 nanospheres shown in Fig. 4d were 32 and 200 nm, respectively. Ultrafine SnO2 nanocrystals below 8 nm constituted the hollow thin shell. The enlarged TEM image in Fig. 4e shows clear lattice fringes separated by 0.34 nm, corresponding to the (110) lattice plane of tetragonal SnO2. The selected area electron diffraction (SAED) pattern shown in Fig. 4f reveals the highly crystalline structure of the SnO2 hollow nanospheres. The elemental mapping images and TG curve shown in Fig. 4g and Figure S5b, respectively, show a trace amount of the carbon component in the SnO2 hollow nanospheres. Further evidence for the oxidation of metallic Sn into SnO2 during the post-treatment in air is provided by the XPS analysis in Figure S6. Deconvolution of the XPS Sn 3d peaks at binding energies of 487.7 eV (Sn 3d5/2) and 496.0 eV (Sn 3d3/2) shows a SnO2 layer in addition to the metallic Sn nanopowder (Figure S6a); this oxide resulted from partial surface oxidation of Sn nanocrystals by exposure to air. However, the XPS spectrum of the SnO2 hollow nanospheres (Figure S6b) shows Sn peaks for only the oxide form, with binding energies of 487.0 eV (Sn 3d5/2) and 495.5 eV (Sn 3d3/2). This provides evidence that the Sn-C composite powder was completely transformed into aggregation-free SnO2 hol- low nanospheres by a simple oxidation process. p y p p For lithium ion storage devices, electrochemical properties were compared between SnO2 hollow nanospheres and SnO2 nanopowders with a filled structure, as prepared by one-pot flame spray pyrolysis. The nanopowder for- mation mechanisms in flame spray pyrolysis have been described in our previous publications39. The drying and decomposition of a droplet inside the diffusion flame formed a micron-sized SnO2 powder. www.nature.com/scientificreports/ Complete evaporation of SnO2 powder in the high-temperature diffusion flame formed the vapors of SnO2. The SnO2 nanopowders were formed by nucleation and growth mechanisms from the SnO2 vapors. The TEM images shown in Figure S7b–d reveal the filled structure of highly crystalline SnO2 nanopowders. The mean crystallite and particle sizes of the SnO2 nanopowders were 27 and 34 nm, respectively, measured from the XRD pattern and TEM image. The Brunauer-Emmett-Teller (BET) surface area of the SnO2 nanopowders was 22 m2 g−1 for both the filled and hollow structures (Figure S8).hi g The cyclic voltammogram (CV) curves of the two samples for the first 5 cycles at a scan rate of 0.07 mV s−1 are shown in Figure S9. The CV curves of the two samples had similar shapes. However, the relative intensity of the reduction peak observed at around 0.8 V in Figure S9a was lower than in Figure S9b. In the first cathodic step, the apparent reduction peak observed at around 0.8 V was mainly associated with the formation of metallic Sn nanograins and amorphous Li2O through reduction of SnO2 38,40. The ultrafine crystallite size of the SnO2 hollow nanospheres broadens the reduction peak observed at around 0.8 V. The broad reduction and oxidation peaks at around 0.2 and 0.5 V, which were attributed to the alloying and de-alloying reactions of metallic Sn with lithium, respectively, were observed in the two samples from the second cycle onward32–35. The good overlapping of the CV curves from the second cycle onward revealed good reversibility of the electrochemical reactions during the first 5 cycles in the two samples.h i y p The discharge and charge curves of SnO2 nanopowders at a constant current density of 2.0 A g−1 are shown in Figure S10. The SnO2 nanopowders exhibited similarly shaped initial discharge and charge curves irrespective of their morphologies. The clear plateaus at around 0.80 and 0.73 V, which were attributed to the formation of metal- lic Sn nanograins and amorphous Li2O through reduction of SnO2, were observed in the initial discharge curves of the hollow and filled SnO2 nanopowders, respectively. The initial discharge and charge capacities of the SnO2 hollow nanospheres were 1762 and 680 mA h g−1, respectively, and its corresponding Coulombic efficiency was 39%. The SnO2 nanopowders with filled structure had similar discharge and charge capacities to the hollow nano- spheres. Results and Discussion Morphologies, SAED pattern, and elemental mapping images of the SnO2 hollow nanospheres obtained by oxidation of reduced Sn-C composite powders at 500 °C under air: (a) SEM, (b–d) TEM images, (e) HR-TEM image, (f) SAED pattern, and (g) elemental mapping images. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 5 www.nature.com/scientificreports/ Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 www.nature.com/scientificreports/ However, the two types of SnO2 nanopowders had different cycling performances as shown in Fig. 5a. The discharge capacities of the SnO2 hollow and filled nanopowders after 300 cycles were 643 and 280 mA h g−1, respectively. The hollow SnO2 nanopowders could accommodate the large volume variation during repeated lith- ium insertion and extraction and decrease Li+ diffusion length, thus leading to improved cycling stability even at the high current density of 2 A g−1. g y g Electrochemical impedance spectroscopy (EIS) measurements were carried out to explain the superior cycling performance of the SnO2 hollow nanospheres compared to the filled nanopowders. The Nyquist impedance plots of the two samples obtained before and after cycling under a fully charged state are shown in Fig. 5b,c. The medium-frequency semicircles in the Nyquist plots of the electrode were assigned to the charge-transfer resist- ance (Rct)31,41,42. The SnO2 hollow nanospheres with ultrafine crystallite size had a lower charge transfer resistance than the filled nanopowders before cycling (Fig. 5b). The charge transfer resistances of the two samples decreased strictly after the first cycle as a result of formation of ultrafine nanocrystals during the first discharging and charg- ing process. The low charge transfer resistance of the SnO2 hollow nanospheres remained constant even after 100 cycles (Fig. 5c); however, the charge transfer resistance of the electrode with SnO2 filled nanopowders increased with cycle number, as a result of structural destruction of the SnO2 filled nanopowders during cycling. In contrast, the hollow structure of the SnO2 nanospheres accommodated the large volume change during repeated lithium insertion and extraction. The high structural stability of SnO2 hollow nanospheres during cycling improved their cycling performance even at the high current density of 2 A g−1. The rate performance of SnO2 hollow nano- spheres is shown in Fig. 5d, with the current density increasing stepwise from 0.5 A g−1 to 7.0 A g−1 with 10 cycles performed at each step. The final rate capacities of the SnO2 hollow nanospheres were 780, 714, 653, 621 and 597 mA h g−1 at current densities of 0.5, 1.5, 3.0, 5.0 and 7.0 A g−1, respectively. The discharge capacity of the SnO2 hollow nanospheres recovered well to 783 mA h g−1, when the current density was returned to 0.5 A g−1 after the 50 cycle test sequence. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 6 www.nature.com/scientificreports/ Figure 5. www.nature.com/scientificreports/ Electrochemical properties of the SnO2 hollow nanospheres formed by applying nanoscale Kirkendall diffusion process and filled-structured SnO2 nanoparticles formed by conventional flame spray pyrolysis process: (a) Cycling performances at a constant current density of 2.0 A g−1 and Coulombic efficiencies of the SnO2 hollow nanospheres, (b) Nyquist impedance plots before cycling, (c) Nyquist impedance plots after cycling, and (d) rate performance of the SnO2 hollow nanospheres. Figure 5. Electrochemical properties of the SnO2 hollow nanospheres formed by applying nanoscale Kirkendall diffusion process and filled-structured SnO2 nanoparticles formed by conventional flame spray pyrolysis process: (a) Cycling performances at a constant current density of 2.0 A g−1 and Coulombic efficiencies of the SnO2 hollow nanospheres, (b) Nyquist impedance plots before cycling, (c) Nyquist impedance plots after cycling, and (d) rate performance of the SnO2 hollow nanospheres. The morphologies of the SnO2 nanospheres formed by applying nanoscale Kirkendall diffusion process and filled-structured SnO2 nanopowders prepared by flame spray pyrolysis process obtained after 300 cycles are shown in Figure S11. The hollow SnO2 nanospheres formed by applying nanoscale Kirkendall diffusion process maintained their morphologies quite well even after repeated lithium insertion and desertion processes as shown by TEM images in Figure S11a. However, the filled structured SnO2 nanopowders prepared by flame spray pyrol- ysis process were broken into several pieces and aggregated after repeated cycling (Figure S11b).i t To confirm the possibility of the hollow SnO2 nano spheres for commercial application, the hollow SnO2 nano- spheres anode was prelithiated and combined with a high voltage LiMn2O4 cathode to construct a full Li-ion cell. Yolk–shell structured LiMn2O4 powders were prepared as a cathode active material by spray pyrolysis process43,44. The morphologies and phase of the yolk–shell structured LiMn2O4 powders prepared by spray pyrolysis are shown in Figure S12. The electrochemical performances of the yolk–shell structured LiMn2O4 powders were shown in Figure S13. The charge and discharge curves and cycling performance of hollow SnO2-nanospheres/LiMn2O4 yolk–shell full cells with a cut-off voltage range of 3.0–4.3 V are shown in Fig. 6. As shown in Fig. 6a, these mate- rials can exhibit charge and discharge capacities of about 630 and 423 mA h g−1, respectively, at the first cycle at a current density of 1.0 A g−1, based on the mass of hollow SnO2-nanospheres anode. In Fig. www.nature.com/scientificreports/ 6b, the Coulombic efficiency of the cell in the initial cycle was 67%, and it increased quickly to close to an average value of 99% in the following cycles. The discharge and charge capacities of the cell after 200 cycles were 289 and 287 mA h g−1, respectively. Materials and Methods ate a s a d et ods Sample preparation. Aggregation-free SnO2 nanoparticles with a hollow structure were prepared using a commercial spray-drying system (Figure S1), followed by a simple two step heat-treatment. Spray solution for the synthesis of tin(II) oxalate- polyvinylpyrrolidone (PVP) composite precursor powders was prepared by dissolv- ing 0.1 M tin(II) oxalate (Sn(Oct)2, 99.9%, Aldrich) and 15 g PVP [(C6H9NO)n, Mw-1,300,000, Aldrich] in 1 L of distilled water. The prepared spray solution was pumped by an atomizing device (20 mL min−1) with a two-fluid nozzle operated at a pressure of 0.2 bar, which generated numerous droplets in a stream of hot air. The spray-dried powder was separated from the humid air centrifugally in a cyclone system. Temperatures at the inlet and outlet of the spray dryer were maintained at 300 °C and 130 °C, respectively. To produce SnO2 hollow nanospheres, the Sn(Oct)2-PVP composite powders were post-treated at 300 °C in a 10% H2/Ar gas mixture for 3 h and subse- quently held at 500 °C in air for 5 h. For comparison to the hollow nanospheres, SnO2 nanoparticles with a filled structure were prepared from a 0.1 M tin oxalate spray solution without PVP, using a flame spray pyrolysis system (Figure S2) consisting of a droplet generator, flame nozzle, powder collector, and blower. A 1.7 MHz ultrasonic spray generator with 6 resonators was used to generate droplets, which were carried into a high-temperature dif- fusion flame by carrier gas (oxygen). Propane and oxygen were the fuel and oxidizer, respectively, for the diffusion flame. The flow rates of fuel, oxidizer, and carrier gas were 5, 40 and 5 L min−1, respectively. Characterization. Microstructures of the prepared powders were observed by field-emission scanning elec- tron microscopy (FE-SEM, Hitachi, S-4800) and field-emission transmission electron microscopy (FE-TEM, JEOL, JEM-2100F). Crystal phases were identified by X-ray diffractometry (XRD, X’Pert PRO MPD), using Cu Kα radiation (λ = 1.5418 Å) at the Korea Basic Science Institute (Daegu). X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha) with focused monochromatic Al Kα radiation, operating at 12 kV and 20 mA, was used to analyze specimen composition. Surface areas of the aggregation-free SnO2 hollow nanospheres were meas- ured by the Brunauer–Emmett–Teller (BET) method, using N2 as adsorbate gas. Thermogravimetric analysis was performed (Pyris 1 TGA, Perkin Elmer) within the temperature range 25–650 °C at a heating rate of 10 °C min−1 under a static air atmosphere. Conclusions In this study, a simple and commercially viable process for large scale production of aggregation-free metal oxide hollow nanospheres has been described. Application of the nanoscale Kirkendall diffusion process in a large-scale spray drying process enabled the preparation of metal oxide hollow nanospheres. The key idea was the prepara- tion of an amorphous carbon matrix with a uniform dispersion of metal nanopowders as an intermediate prod- uct. Metal nanopowders were transformed into metal oxide hollow nanopowders by the nanoscale Kirkendall diffusion process. The amorphous carbon matrix enabled the formation of aggregation-free metal oxide hollow nanospheres. The aggregation-free SnO2 hollow nanospheres had superior electrochemical properties for lithium ion storage compared to the SnO2 nanopowders with filled structure. The simple process applied in this study could be applied in the preparation of metal oxide hollow nanospheres with various compositions for numerous applications, including energy storage devices. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 7 www.nature.com/scientificreports/ Figure 6. Electrochemical properties of a full cell of anode (hollow SnO2 nanospheres)/cathode (LiMn2O4): (a) charge–discharge curves at a current density of 1.0 A g−1 and (b) cycling performance at a current density at 1.0 A g−1 based on the anode (hollow SnO2 nanospheres) mass. Figure 6. Electrochemical properties of a full cell of anode (hollow SnO2 nanospheres)/cathode (LiMn2O4): (a) charge–discharge curves at a current density of 1.0 A g−1 and (b) cycling performance at a current density at 1.0 A g−1 based on the anode (hollow SnO2 nanospheres) mass. References 1. Ma, F. X. et al. Formation of uniform Fe3O4 hollow spheres organized by ultrathin nanosheets and their excellent lithium storage properties. Adv. Mater. 27, 4097–4101 (2015). p p ( ) 2. Yao, Y. et al. Interconnected silicon hollow nanospheres for lithium-ion battery anodes with long cycle life. Nano Lett. 11, 2949–2954 (2011). ( ) 3. Lai, X. et al. General synthesis and gas‐sensing properties of multiple‐shell metal oxide hollow microspheres. Angew. Chem. 123 2790–2793 (2011). 4. Zhou, L. et al. Cheap and scalable synthesis of α -Fe2O3 multi-shelled hollow spheres as high-performance anode materials for lithium ion batteries. Chem. Commun. 49, 8695–8697 (2013). 5. Zhou, X., Yin, Y. X., Wan, L. J. & Guo, Y. G. A robust composite of SnO2 hollow nanospheres enwrapped by graphene as a high capacity anode material for lithium-ion batteries. J. Mater. Chem. 22, 17456–17459 (2012). 6. Niu, C. et al. VO2 nanowires assembled into hollow microspheres for high-rate and long-life lithium batteries. Nano Lett. 14 2873–2878 (2014). ( ) 7. Wu, H., Xu, M., Wang, Y. & Zheng, G. Branched Co3O4/Fe2O3 nanowires as high capacity lithium-ion battery anodes. Nano Res. 6, 167–173 (2013). r, L. A. & Yang, Z. Hollow micro‐/nanostructures: Synthesis and app Lou, X. W., Archer, L. A. & Yang, Z. Hollow micro‐/nanostructures: 9. Xia, X. H., Tu, J. P., Wang, X. L., Gu, C. D. & Zhao, X. B. Mesoporous Co3O4 monolayer hollow-sphere array as electrochemical pseudocapacitor material. Chem. Commun. 47, 5786–5788 (2011). 10. Hu, J., Chen, M., Fang, X. & Wu, L. Fabrication and application 10. Hu, J., Chen, M., Fang, X. & Wu, L. Fabrication and application of inorganic hollow spheres. Chem. Soc. Rev. 40, 5472–5491 (2011). 11. Lai, X., Halpert, J. E. & Wang, D. Recent advances in micro-/nano-structured hollow spheres for energy applications: From simple to complex systems. Energy Environ. Sci. 5, 5604–5618 (2012). p y gy ( ) 12. An, K. & Hyeon, T. Synthesis and biomedical applications of hollow nanostructures. Nano Today 4, 359–373 (2009). 12. An, K. & Hyeon, T. Synthesis and biomedical applications of ho 13. Zhang, Q., Wang, W., Goebl, J. & Yin, Y. Self-templated synthesis of hollow nanostructures. Nano Today 4, 494–507 (2009). 14. Li, W. & Zhao, D. Extension of the Stöber method to construct mesoporous SiO2 and TiO2 shells for uniform multifunctional core–shell structures. Adv. Mater. 25, 142–149 (2013). 5. Stöber, W., Fink, A. & Bohn, E. References Controlled growth of monodisperse silica spheres in the micron size range. J. Colloid. Interface Sci 26, 62–69 (1968). ( ) 16. Yen, C., Mahmoud, M. & El-Sayed, M. Photocatalysis in gold nanocage nanoreactors. J. Phys. Chem. A 113, 4340–4345 (2009) y y g g y 17. Liu, B. & Zeng, H. C. Fabrication of ZnO “dandelions” via a modified Kirkendall process. J. Am. Chem. Soc. 126, 16744–16746 (2004).i 8. Cho, J. S., Hong, Y. J. & Kang, Y. C. Design and synthesis of bubble-nanorod-structured Fe2O3–carbon nanofibers as advanced anode material for Li-Ion Batteries. ACS Nano 9, 4026–4035 (2015).f 19. Railsback, J. G., Johnston-Peck, A. C., Wang, J. & Tracy, J. B. Size-dependent nanoscale Kirkendall effect during the oxidation of nickel nanoparticles. ACS Nano 4, 1913–1920 (2010).f p ( ) 20. Yin, Y. et al. Formation of hollow nanocrystals through the nanoscale Kirkendall effect. Science 304, 711–714 (2004).f y gf , ( ) 21. Knez, M. et al. Monocrystalline spinel nanotube fabrication based on the Kirkendall effect. Nat. Mater. 5, 627–631 (2006) al. Monocrystalline spinel nanotube fabrication based on the Kirke 21. Knez, M. et al. Monocrystalline spinel nanotube fabrication ba y pf 22. Cho, J. S., Hong, Y. J., Lee, J. H. & Kang, Y. C. Design and synthesis of micron-sized spherical aggregates composed of hollow nanospheres for use in lithium-ion batteries. Nanoscale 7, 8361–8367 (2015). nanospheres for use in lithium-ion batteries. Nanoscale 7, 836 p 3. Wang, X. et al. Synthesis and lithium storage properties of Co3O4 nanosheet‐assembled multishelled hollow spheres. Adv. Funct Mater. 20, 1680–1686 (2010). 24. Zhu, J. et al. Hierarchical hollow spheres composed of ultrathin Fe2O3 nanosheets for lithium storage and photocatalytic water oxidation. Energy Environ. Sci. 6, 987–993 (2013). gy 5. Yin, X. M. et al. One-step synthesis of hierarchical SnO2 hollow nanostructures via self-assembly for high power lithium ion batteries. J. Phys. Chem. C 114, 8084–8088 (2010). y 26. Koo, B. et al. Hollow iron oxide nanoparticles for application in lithium ion batteries. Nano Lett. 12, 2429–2435 (2012 27. Mahmood, N. & Hou, Y. Electrode nanostructures in lithium based batteries. Adv. Sci. 1, 1400012 (2014). 28. Mahmood, N., Zhang, C., Liu, F., Zhu, J. & Hou, Y. Hybrid of Co3Sn2@Co nanoparticles and nitrogen-doped graphene as a lithium ion battery anode. ACS Nano 7, 10307–10318 (2013). y 29. Chang, Y. E. et al. Fabrication and gas sensing properties of hollow SnO2 hemispheres. Chem. Commun. References 4019–4021 (2009). D Z l Q i l h ll d S O h ll h h l h f h h f d d 30. Dong, Z. et al. Quintuple‐shelled SnO2 hollow microspheres with superior light scattering for high‐performance dye‐sensitized cells. Adv. Mater. 26, 905–909 (2014). 1. Hong, Y. J., Son, M. Y. & Kang, Y. C. One‐pot facile synthesis of double‐shelled SnO2-yolk‐shell‐structured powders by continuou process as anode materials for Li‐ion batteries. Adv. Mater. 25, 2279–2283 (2013). p 32. Huang, X. et al. A facile one‐step solvothermal synthesis of SnO2/graphene nanocomposite and its application as an anode material for lithium‐ion batteries. ChemPhysChem 12, 278–281 (2011). for lithium‐ion batteries. ChemPhysChem 12, 278–281 (2011). y ( ) 3. Wen, Z., Wang, Q., Zhang, Q. & Li, J. In situ growth of mesoporous SnO2 on multiwalled carbon nanotubes: A novel composite with porous‐tube structure as anode for lithium batteries. Adv. Funct. Mater. 17, 2772–2778 (2007). g Q g Q J g p 2 p porous‐tube structure as anode for lithium batteries. Adv. Funct. Mater. 17, 2772–2778 (2007). 34. Zhou, X., Dai, Z., Liu, S., Bao, J. & Guo, Y. G. Ultra‐uniform SnOx/carbon nanohybrids toward advanced lithium‐ion battery anodes. porous‐tube structure as anode for lithium batteries. Adv. Funct. Mater. 17, 2772–2778 (2007). 34. Zhou, X., Dai, Z., Liu, S., Bao, J. & Guo, Y. G. Ultra‐uniform SnOx/carbon nanohybrids toward advanced lithium‐ion battery anodes. 4. Zhou, X., Dai, Z., Liu, S., Bao, J. & Guo, Y. G. Ultra‐uniform SnOx/carbon nanohybrids toward advanced lithium‐ion battery anodes Adv. Mater. 26, 3943–3949 (2014). 5. Sharma, Y., Sharma, N., Subba Rao, G. V. & Chowdari, B. V. R. Studies on nano-CaO·SnO2 and nano-CaSnO3 as anodes for Li-ion batteries. Chem. Mater. 20, 6829–6839 (2008). 36. Oh, H. S., Nong, H. N. & Strasser, P. Preparation of mesoporous Sb‐, F‐, and In‐doped SnO2 bulk p use as catalyst supports in electrolytic cells. Adv. Funct. Mater. 25, 1074–1081 (2015).hi Nong, H. N. & Strasser, P. Preparation of mesoporous Sb‐, F‐, and In‐doped SnO2 bulk powder with high surface area for lyst supports in electrolytic cells. Adv. Funct. Mater. 25, 1074–1081 (2015).hi 6. Oh, H. S., Nong, H. N. & Strasser, P. Preparation of mesoporous Sb‐, F‐, and In‐doped SnO2 bulk powder with high surface area fo use as catalyst supports in electrolytic cells. Adv. Funct. Mater. 25, 1074–1081 (2015).hi y pp y 37. Shin, J. et al. Materials and Methods An image analyzer (ImageJ, NIH) was used to determine particle size of the nanospheres. Electrochemical measurements. Electrochemical properties of the aggregation-free SnO2 hollow nano- spheres were analyzed by constructing a 2032-type coin cell. The anode was prepared by mixing the active mate- rial, carbon black, and sodium carboxymethyl cellulose (CMC) in a mass ratio of 7:2:1. Li metal and microporous polypropylene film were used as the counter electrode and separator, respectively. The electrolyte was 1 M LiPF6 dissolved in a mixture of fluoroethylene carbonate and dimethyl carbonate (FEC/DMC; 1:1 v/v). The discharge/ charge characteristics of the samples were investigated by cycling in the 0.001–1.0 V potential range at various current densities. Cyclic voltammograms were measured at a scan rate of 0.07 mV s−1. The negative electrode using SnO2 nanoparticles was of dimensions 1 cm × 1 cm and the mass loading was approximately 1.2 mg cm−2. The cathode was prepared by mixing the active material (yolk–shell structured LiMn2O4), carbon black, and sodium carboxymethyl cellulose (CMC) in a weight ratio of 8:1:1. For full cell assembly, the LiMn2O4 yolk–shell electrode with a loading mass of 3 mg cm−2 was used as a cathode, whereas the anode mass loading was kept at 0.4 mg cm−2. For the full cell, the electrolyte was 1 M LiPF6 dissolved in a mixture of ethylene carbonate/diethyl carbonate (EC/DEC;1:1 v/v). The electrochemical properties of the 2032-type coin full cells were examined at 1.0 A g−1 in voltage windows between 3.0 and 4.3 V. The electrode capacity was calculated according to the weight of the anode materials. 8 Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 www.nature.com/scientificreports/ References Thin‐wall assembled SnO2 fibers functionalized by catalytic Pt nanoparticles and their superior exhaled‐breath‐sensing properties for the diagnosis of diabetes. Adv. Funct. Mater. 23, 2357–2367 (2013). hi properties for the diagnosis of diabetes. Adv. Funct. Mater. 23, 23 g 8. Zhang, L. et al. Integrated SnO2 nanorod array with polypyrrole coverage for high-rate and long-life lithium batteries. Phys. Chem Chem. Phys. 17, 7619–7623 (2015).l 9. Choi, S. H. & Kang, Y. C. Characteristics of ZnMn2O4 nanopowders prepared by flame spray pyrolysis for use as anode material in lithium ion batteries. Int. J. Electrochem. Sci, 8, 6281–6290 (2013). 40. Park, M. S. et al. Preparation and electrochemical properties of SnO2 nanowires for application in lithium-ion batteries. Angew. Chem. 119, 764–767 (2007). 41. Li, N. et al. Battery performance and photocatalytic activity of mesoporous anatase TiO2 nanospheres/graphene composites by template‐free self‐assembly. Adv. Funct. Mater. 21, 1717–1722 (2011).iii p y 2. Cho, J. S., Hong, Y. J. & Kang, Y. C. Electrochemical properties of fiber‐in‐tube‐and filled‐structured TiO2 nanofiber anode material for lithium‐ion batteries. Chem. -Eur. J. 21, 11082–11087 (2015). 43. Sim, C. M., Choi, S. H. & Kang, Y. C. Superior electrochemical properties of LiMn2O4 yolk–shell powders prepared by a simple spray pyrolysis process. Chem. Commun. 49, 5978–5980 (2013). 44. Park, G. D., Cho, J. S. & Kang, Y. C. Novel cobalt oxide-nanobubble-decorated reduced graphene oxide sphere with superior electrochemical properties prepared by nanoscale Kirkendall diffusion process. Nano Energy 17, 17–26 (2015). Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 9 www.nature.com/scientificreports/ Acknowledgementsh g This work was supported by the Energy Efficiency & Resources Core Technology Program of the Korea Institute of Energy Technology Evaluation and Planning (KETEP), granted financial resource from the Ministry of Trade, Industry & Energy, Republic of Korea (201320200000420). This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (NRF-2015R1A2A1A15056049). Author Contributions J.S.C. and Y.C.K. devised the concept, designed the experiment, and wrote the manuscript. J.S.C. and H.S.J. performed the experiments and analyzed the data. Y.C.K. supervised the project. All authors discussed the results and contributed in this manuscript. Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 Additional Information Supplementary information accompanies this paper at http://www.nature.com/srep Supplementary information accompanies this paper at http://www.nature.com/srep Competing financial interests: The authors declare no competing financial interests. How to cite this article: Cho, J. S. et al. Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram- Scale Production of SnO2 Hollow Nanospheres via Spray Drying System. Sci. Rep. 6, 23915; doi: 10.1038/ srep23915 (2016). How to cite this article: Cho, J. S. et al. Applying Nanoscale Kirkendall Diffusion for Template-Free, Kilogram- Scale Production of SnO2 Hollow Nanospheres via Spray Drying System. Sci. Rep. 6, 23915; doi: 10.1038/ srep23915 (2016) This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 6:23915 \| DOI: 10.1038/srep23915 10
https://openalex.org/W2599384442	https://seer.ufu.br/index.php/biosciencejournal/article/download/34494/20005	English	null	Antimicrobial activity of endophytic fungi from coffee plants	Bioscience journal	2,017	cc-by	4,967	381 381 Original Article ANTIMICROBIAL ACTIVITY OF ENDOPHYTIC FUNGI FROM COFFEE PLANTS ATIVIDADE ANTIMICROBIANA DE FUNGOS ENDOFÍTICOS DE PLANTAS DE CAFÉ Mônica Cristina Pereira MONTEIRO1; Natálie Martins ALVES1; Marisa Vieira de QUEIROZ2; Danilo Batista PINHO³; Olinto Liparini PEREIRA4; Sara Maria Chalfoun de SOUZA5; Patrícia Gomes CARDOSO1 1. Department of Biology, Federal University of Lavras, Lavras, MG, Brazil; 2. Department of Microbiology, Federal University of Viçosa,Viçosa, MG, Brazil; 3. Department of Plant Pathology, University of Brasília, Distrito Federal, DF, Brazil; 4. Department of Plant Pathology, Federal University of Viçosa,Viçosa, MG, Brazil; 5. Agricultural Research Company of Minas Gerais, Tech Center South Minas, Lavras, MG, Brasil. patricia@dbi.ufla.br ABSTRACT: Endophytic fungi are a promising source for discovery of compounds with biotechnological potential. The aim of this study was to select and identify endophytic fungi from Coffea arabica that produce volatile organic compounds (VOCs), evaluate the effect of the VOCs produced by endophytic fungi on the growth of Rhizoctonia solani, Fusarium oxysporum, Phoma sp., Botrytis cinerea, Fusarium solani, Fusarium verticillioides, Cercospora coffeicola and Pestalotia longisetula, and select endophytic fungi with potential for biological control of Aspergillus ochraceus inoculated in coffee beans and F. verticillioides inoculated in corn seeds. An isolate of Muscodor albus was used as selection tool for VOC producing fungi. Among the 400 endophytic fungi isolates, 11 were able to grow in the presence of VOCs produced by M. albus. These fungi were identified as Muscodor spp. (9) and Simplicillium sp. according to searches in UNITE database using DNA sequences of internal transcribed spacer (ITS). The VOC’s produced by endophytic fungi inhibited the growth the phytopathogenic fungi with different efficacies, compared to the control. The VOCs produced by Muscodor coffeanum (COAD 1842) showed fungicidal effect against A. ochraceus on coffee beans. Six endophytic fungi completely inhibited growth of F. verticillioides inoculated in corn seeds. This study demonstrates that the volatile-compound producing endophytic fungi, isolated from Coffea arabica, are promising sources of bioactive compounds. KEYWORDS: Aspergillus ochraceus. Fusarium verticillioides. Inhibition. Muscodor spp. Volatiles compounds. KEYWORDS: Aspergillus ochraceus. Fusarium verticillioides. Inhibition. Muscodor spp. Volatiles compounds. Bioassay for volatile antimicrobials The inhibitory antimicrobial activity of VOCs produced by endophytic fungi from coffee was tested against the following phytopathogenic fungi: Rhizoctonia solani (LAPS 369), Fusarium oxysporum (LAPS 152), Phoma sp. (DFP 01), Botrytis cinerea (LAPS 300), Fusarium solani (LAPS 298), Fusarium verticillioides (CML 1896), Cercospora coffeicola (CML 2984) and Pestalotia longisetula (DFP 02). Endophytes were cultivated in PDA medium in Petri dishes and incubated at 25°C for 7 days. After this period, the phytopathogenic fungi were transferred to the other side of the plate. The plates were incubated at 25°C for 7 days. Phytopathogenic fungal growth was measured and compared with control plates without endophytic fungi. The colony diameters (Cm), were measured and classified the according to following scale: T- Total inhibition (0); P-Partial inhibition (1-2.0); N- No inhibition (≥2.1). The experiment was repeated twice with three replicates. Screening of VOC producing isolates Screening of endophytic fungi t g p g Screening of endophytic fungi that produce VOCs was made as described by Strobel et al., (2001) with modifications. A culture of the original isolate of M. albus (strain CZ620) was used as selection tool for VOC producing fungi. Muscodor albus was placed and grown on one side of the plate for 7 days at 25ºC. A mycelial disk of each endophytic isolate (5 mm diameter) was deposited on the opposite side. Each isolate was tested in three replicates. The plate was wrapped with Parafilm® and incubated at 25ºC for one week. The experiment was performed twice and only isolates able to grow in the presence of VOCs produced by M. albus were selected for identification. INTRODUCTION 2010; SUWANNARACH et al., 2012; SAXENA et al., 2015). Muscodor species produce a mixture of volatile organic compounds that open new possibilities for the biological control of microbial decay in food and agriculture by biofumigation (STROBEL et al., 2001; DAISY et al., 2002; MERCIER; SMILANICK 2005; GRIMME et al., 2007). 2010; SUWANNARACH et al., 2012; SAXENA et al., 2015). Muscodor species produce a mixture of volatile organic compounds that open new possibilities for the biological control of microbial decay in food and agriculture by biofumigation (STROBEL et al., 2001; DAISY et al., 2002; MERCIER; SMILANICK 2005; GRIMME et al., 2007). Endophytic fungi colonize living tissues of various plants, establishing mutualistic relationship without causing any symptom of disease (PETRINI, 1991; AZEVEDO et al., 2000; HYDE; SOYTONG, 2008). Their distribution within plants is ubiquitous but varies according to plant tissue (root, leaf, stems and fruits) and from strain to strain (TAN; ZOU, 2001). Endophytes have received considerable attention because of their ability to produce several novel compounds including terpenoids, alkaloids, phenylpropanoids, polyketides, aminoacids, and phytohormones (STROBEL et al., 2001; TEJESVI et al., 2007). Studies have been conducted with endophytes using species of plants that have economic significance, especially coffee crops (SANTAMARIA; BAYMAN, 2005; VEGA et al., 2005a, 2006b, SETTE et al., 2006a; SAUCEDO- GARCIA et al., 2014). Due to the economic importance of this crop and biotechnological potential of endophytic fungi, the aims of this study were to isolate and identify endophytic fungi from Coffea arabica in Brazil that produce volatile organic compounds (VOCs), evaluate the effect of the VOCs produced by endophytic fungi on the growth of phytopathogenic fungi and select endophytic fungi with potential for biological Some metabolites produced by endophytic fungus can help the host plant to tolerate biotic and abiotic stress, protect plants against diseases and from insect and nematode attack, as well as favor the growth of crop plants (KOGEL et al., 2006). In addition, endophytic fungi have been reported to reduce the growth of the different phytopathogenic fungi (EZRA; STROBEL, 2003; ZHANG et al., Biosci. J., Uberlândia, v. 33, n. 2, p. , Mar./Apr. 2017 Received: 25/05/16 Accepted: 05/11/16 382 Antimicrobial activity… MONTEIRO, M. C. P. et al. control of A. ochraceus in coffee beans and F. verticillioides in corn seeds. Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 Biofumigation with endophytic fungi from coffee Biofumigation with endophytic fungi from coffee The VOCs produced by endophytic fungi were tested against A. ochraceus (SCM 1.15), producer of sclerotia and ochratoxin A in coffee beans (coffee in the dried bean and hulled coffee), isolated belonging to the Culture Collection of the Departament of Food Sciences (CDCA; Federal University of Lavras, Minas Gerais, Brazil), and F. verticillioides (CML1896), in corn seeds. Coffee beans and corn seeds were surface disinfested by immersion in 70% ethanol for 3 min, sodium hypochlorite at 2.5% for 5 min and three times with sterile distilled water. After air-dying the beans and corn seeds, they were inoculated by immersion in a suspension of A. ochraceus (1.5 x 105 conidia/mL) and F. verticillioides (2.0 x 105conidia/mL) spores, respectively. In bipartite Petri dishes containing PDA medium, the endophytes were cultivated for 7 days at 25°C. After this period, coffee beans inoculated with A. ochraceus and corn seeds inoculated with F. verticillioides were placed on the other side of the plate. The effect of volatile compounds produced was evaluated by the presence or absence of growth in grains inoculated with A. ochraceus and F. verticillioides. The control treatment consisted of grains inoculated with plant pathogens without the presence of the endophytic fungi. To assess the fungistatic and fungicidal action of the volatile compounds the coffee grains and corn seeds, they were transferred to PDA medium after 7 days of exposure to volatile compounds. The experiment was repeated twice with three replicates. Sample collection and isolation Field surveys were carried out during 2011 in the Zona da Mata region, Viçosa municipality, Minas Gerais, Brazil to obtain endophytic fungi on organic coffee plantations. Coffee tissue parts were rinsed in sterile distilled water for 1 min and dried. Small pieces (4–5 mm) of apparently healthy tissue were then disinfected in 70% ethanol for 1 min followed by 2.5% sodium hypochlorite for 3 min and washed in sterile distilled water. Fragments were placed in Petri dishes with Potato Dextrose Agar (PDA - Acumedia®) amended with chloramphenicol 100 ppm and incubated at 25°C. Hyphal tips of fungal colonies emerging from plant tissue pieces were transferred to PDA dishes and incubated at 25°C. The cultures were stored in tubes on PDA at 10°C. Endophytic fungi from Coffea arabica A total of 620 fragments were obtained from stems (391), leaves (113) and fruits (116) of the Coffea arabica. Among the 400 endophytic fungi isolated from stems (261), fruits (97), and leaves (42), eleven (stems 7 and leaves 4) were able to grow in the presence of VOCs produced by M. Table 1. Identification of the isolated endophytic fungi producing volatile compounds Isolate Origin Accession no. M. coffeanum (COAD 1842) Leaf KM514680 M.coffeanum (COAD 1899) Leaf KM514681 M.coffeanum (COAD 1900) Leaf KP862879 M. vitigenus (C20) Stem KU094049 M. vitigenus (HZM10) Stem KU094053 M. vitigenus (HZM39) Stem KU094054 KU094055 M. vitigenus (HZM41) Stem KU094055 M. yucatanensis (HZM60) Leaf KU094056 M. yucatanensis (HZM64) Leaf KU094052 Simplicillium sp. (C18) Stem KU094050 Simplicillium sp. (C12) Stem KU094051 Table 1. Identification of the isolated endophytic fungi producing volatile compounds intense musty odor. Colonies, when grown on PDA, usually form a whitish, flocculose colony with an uncolored reverse and a mycelium that grows slowly (GONZÁLEZ et al., 2009). Muscodor is a genus of sterile endophytic fungi, all species of this genus were characterized by the production of volatile organic compounds (VOCs) that inhibit the growth of other microorganisms (STROBEL et al., 2001a; STROBEL, 2006b; STROBEL, 2011c; STINSON et al., 2003; MERCIER; JIMENEZ, 2004; MERCIER; MANKER 2005; MERCIER et al., 2007; WORAPONG; STROBEL, 2009; ZHANG et al., 2010; SUWANNARACH et al., 2012; KUDALKAR et al., 2012; SAXENA et al., 2015). The specie M. vitigenus, identified in our study was first isolated from Paullinia paullinioides by Daisy et al. (2002). These authors report that this specie produces compounds such as styrene, benzaldehyde, butylated hydroxytoluene, toluene, naphthalene and a number of minor benzene derivatives and that the compound produced, naphthalene, causes modifications in insect behaviour. Muscodor is a genus of sterile endophytic fungi, all species of this genus were characterized by the production of volatile organic compounds (VOCs) that inhibit the growth of other microorganisms (STROBEL et al., 2001a; STROBEL, 2006b; STROBEL, 2011c; STINSON et al., 2003; MERCIER; JIMENEZ, 2004; MERCIER; MANKER 2005; MERCIER et al., 2007; WORAPONG; STROBEL, 2009; ZHANG et al., 2010; SUWANNARACH et al., 2012; KUDALKAR et al., 2012; SAXENA et al., 2015). Two isolates of the genus Simplicillium also were identified in our study. The species S. lonosoniveum and S. Endophytic fungi from Coffea arabica lamellicola were isolated from coffee plants but not as endophytes and they have been exploited as biological control agent (ZARE et al., 2001; WARD et al., 2010). Muscodor coffeanum reported in this study (COAD 1842, COAD 1899 and COAD 1900) is a new species isolated from leaves and stems from coffee plants in Brazil (HONGSANAN et al., 2015). The specie M. vitigenus, identified in our study was first isolated from Paullinia paullinioides by Daisy et al. (2002). These authors report that this specie produces compounds such as styrene, benzaldehyde, butylated hydroxytoluene, toluene, naphthalene and a number of minor benzene derivatives and that the compound produced, naphthalene, causes modifications in insect behaviour. Molecular identification The genomic DNA was extracted from pure cultures grown on PDA using a Wizard® Genomic DNA Purification Kit (Promega Corporation, WI, U.S.A). The internal transcribed spacer (ITS) was amplified using primers ITS1 and ITS4 (WHITE et al., 1990). PCR products were purified and sequenced by Macrogen, South Korea. The sequences were edited using BioEdit software (HALL, 1999). A BLAST search was performed to check for similarity with other sequences and identification was performed according to searches in UNITE database (NILSSON et al., 2014). Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 383 MONTEIRO, M. C. P. et al. Antimicrobial activity… Antimicrobial activity… RESULTS AND DISCUSSION albus. Colonies of 11 endophytic fungi on PDA were white, cottony with slow growth and absence of sporulation. The fungi were identified as Muscodor coffeanum (3), Muscodor vitigenus (4), Muscodor yucatanensis (2) and Simplicilium sp. (2), according to searches in UNITE database using DNA sequences of internal transcribed spacer (ITS) (Table 1). Endophytic fungi from Coffea arabica Activity of VOCs against seed pathogens y g p g The endophytic fungus M. coffeanum (COAD 1842) showed fungicidal activity since it completely inhibited the mycelial growth of A. ochraceus. The endophytic fungi Simplicillium sp. (C12), M. coffeanum (COAD 1900) and M. coffeanum (COAD 1899) showed fungistatic activity. Endophytic fungi also showed growth inhibition of F. verticillioides in corn seeds. Among the eleven evaluated endophyte fungi the isolates M. coffeanum (COAD 1842), M. coffeanum (COAD 1899), M. coffeanum (COAD 1900), M. vitigenus (C20), Simplicillium sp. (C12) and Simplicillium sp. (C18) showed total growth inhibition of F. verticillioides (Table 3). Previous works with VOCs produced by Muscodor albus presented antimicrobial potential against fungi, oomycetes and bacteria. The growth of Botrytis cinerea, Aspergillus fumigatus, Rhizoctonia solani, Sclerotinia sclerotiorum, Pythium ultimum, Verticillium dahliae, Phytophthora cinnamomi, Candida albicans, Escherichia coli, Bacillus subtilis and Staphylococcus aureus was inhibited or the fungi died after exposure to VOCs of M. albus (WORAPONG et al., 2001; STROBEL et al., 2001). Moreover, other Muscodor species have been described to inhibit the growth of fungi associated with post-harvest decay (MITCHELL et al., 2008). Coffee is an important commercial product, the fungus A. ochraceus is reported as producer of ochratoxin A (OTA) in coffee beans, and its presence, as well as the production of OTA in coffee, is undesirable because it may be used as a trade barrier, affecting the economies of producing countries (SUAREZ-QUIROZ et al., 2004). The fungi, producers of volatile compounds with broad antimicrobial activity, isolated from C. arabica have potential for biotechnological applications. None of the endophytic fungi showed total inhibition against F. oxysporum. Fusarium species may be less susceptible to VOCs (FIALHO et al., 2010). The VOCs produced by endophytic fungus M. yucatanensis were lethal to Colletotrichum sp., Phomopsis sp., Guignardia mangiferae, Phythophthora capsici, P. parasitica, Rhizoctonia sp., and Alternaria solani but there was no complete growth inhibition of F. oxysporum when compared with the control (MACÍAS-RUBALCAVA et al., 2010). These findings open new possibilities for developing mycofumigation as a post-harvest treatment, since, Muscodor spp. and Simplicilium stand out as potential candidates as biocontrol agents in post-harvest technology constituting an alternative to replace chemical fungicides. Characterization studies on the bioactive metabolites of the potent fungal strains from C. arabica and their use as biocontrol agents are in progress. Biological activity of the VOC’s produced by endophytic fungi The action of volatile organic compounds produced by endophytic fungi was tested against a spectrum of phytopathogenic fungi and fungi associated postharvest diseases (Table 2). The VOCs produced by endophytic fungi exhibited Muscodor yucatanensis, one of the species herein identified, is a recognized producer of an Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 384 Antimicrobial activity… MONTEIRO, M. C. P. et al. antifungal activity with different efficiency. The phytopatogenic fungi R. solani, C. coffeicola and Phoma sp., were completely suppressed by VOCs produced by most endophytic fungi, whereas fungi like B. cinerea, A. ochraceus and F. verticillioides showed sensitivity to VOCs. Activity of VOCs against seed pathogens Strobel et al., (2001) also found similar results, among several tested fungi the phytopathogenic fungi Fusarium solani was more resistant to the VOCs produced by M. albus. In addition, the artificial mixtures of VOCs produced by Gliocladium sp. partially inhibited F. oxysporum (STINSON et al., 2003). Fusarium verticillioides is one of the most commonly reported soil-borne fungal pathogens infecting maize (Zea mays L.), one of the most important cereal grains grown worldwide. This fungus produces secondary metabolites such as fumonisins (FB), especially fumonisin B1 (FB1), which affects human and animal health (BACON et al., 1996). Since F. verticillioides is endophytic in maize and is almost universally associated with maize and maize products, it is very important to control this species in this agriculturally important commodity. Furthermore, root colonization by F. verticillioides has been considered the initiator of systemic infection that eventually results in the fungus producing fumonisins in kernels. Seed treatment with biocontrol agents is an appropriate method for biocontrol of soil-borne plant pathogens in the spermosphere and rhizosphere (KERRY, 2000). ( ) In our study, the VOCs produced by Simplicillium sp. (C12, C18) also exhibited antifungal activity, the isolate C12 completely inhibit the growth of R. solani, C. coffeicola and Phoma sp., whereas isolate C18 inhibited total growth of C. coffeicola. To our knowledge, our study is one of the few that has reported the antifungal activity of Simplicillium sp. through production of VOCs. Thus, these isolates may be candidates for more detailed studies involving biological control. Moreover, further studies are needed to understand how these compounds act and to know their effect on these organisms. Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 P Biosci. J., Uberlândia, v. 33, n. 2, p. , Mar./Apr. 2017 growth inhibition of phytopathogenic fungi T-Total inhibition (0); P-Partial inhibit enic fungi F. verticillioides C. coffeicola F. oxysporum Phoma sp. P. lo T T N T T P N T T P P P T T N T N T P T N T N T N T P T P T P P P T N T P T N T N T N P Biosci. J., Uberlândia, v. 33, n. 2, p. , Mar. growth inhibition of phytopathogenic fungi T-Total inhibition (0); P-Par enic fungi F. verticillioides C. coffeicola F. Activity of VOCs against seed pathogens oxysporum Phoma s T T N T T P N T T P P P T T N T N T P T N T N T N T P T P T P P P T N T P T N T N T N P 386 Antimicrobial activity… MONTEIRO, M. C. P. et al. Table 3. Aspergillus ochraceus and Fusarium verticillioides inhibited by volatile organic compounds produced by endophytic fungi. Endophytic fungi A. ochraceus in Coffee in the dried bean A. ochraceus in Hulled coffee F. verticillioides in corn seed M. coffeanum (COAD 1842) + + + M. coffeanum (COAD 1899) ± ± + M. coffeanum (COAD 1900) ± ± + M. vitigenus (C20) - - + M.vitigenus (HZM10) - - ± M. vitigenus (HZM39) - - ± M. vitigenus (HZM41) - - ± M. yucatanensis (HZM60) - - ± M. yucatanensis (HZM64) - - ± Simplicillium sp. (C12) - - ± Simplicillium sp. (C18) ± ± + Total inhibition (+); Partial inhibition (±); No inhibition (-). gillus ochraceus and Fusarium verticillioides inhibited by volatile organic compounds produced dophytic fungi. Total inhibition (+); Partial inhibition (±); No inhibition (-). CONCLUSIONS antimicrobial action, since, plant pathogens were inhibited or killed by endophytic fungi, producers of volatile organic compounds. Volatile compound producing endophytic fungi were isolated from C. arabica, among the 400 fungi, 12 isolates were able to grow in the presence of VOCs produced by M. albus. PALAVRAS-CHAVE: Aspergillus ochraceus. Fusarium verticillioides. Inibição. Muscodor spp. Compostos oláteis. REFERENCES AZEVEDO, J. L.; MACCHERONI JR, W.; PEREIRA, J. O.; DE ARAÚJO, W. L. Endophytic microorganisms: a review on insect control and recent advances on tropical plants. Electronic Journal of Biotechnology, Valparaíso, v. 3, n. 1, p. 15-16, 2000. https://doi.org/10.2225/vol3-issue1-fulltext-4 BACON, C. W.; HINTON, D. M. Symptomless endophytic colonization of maize by Fusarium moniliforme. Canadian Journal of Botany, Canada. v. 74, n. 8, p. 1195-1202, 1996. https://doi.org/10.1139/b96-144 DAISY, B. H.; STROBEL, G. A.; CASTILLO, U.; EZRA, D.; SEARS, J.; WEAVER, D. K.; RUNYON, J. Naphthalene, an insect repellent, is produced by Muscodor vitigenus, a novel endophytic fungus. Microbiology, London, v. 148, n. 11, p. 3737-3741, 2002. https://doi.org/10.1099/00221287-148-11- 3737 EZRA, D.; STROBEL, G. A. Effect of substrate on the bioactivity of volatile antimicrobials produced by Muscodor albus. Plant Science, Ireland. v. 165, n. 6, p. 1229-1238, 2003. https://doi.org/10.1016/S0168- 9452(03)00330-3 FIALHO, M. B.; TOFFANO, L.; PEDROSO, M. P; AUGUSTO, F.; PASCHOLATI, S. F. Volatile organic compounds produced by Saccharomyces cerevisiae inhibit the in vitro development of Guignardia citricarpa, the causal agent of citrus black spot. World Journal of Microbiology and Biotechnology, Netherlands. v. 26, n. 5, p. 925-932, 2010. https://doi.org/10.1007/s11274-009-0255-4 GONZÁLEZ, M. C.; ANAYA, A. L.; GLENN, A. E.; MACÍAS-RUBALCAVA, M. L.; HERNÁNDEZ- BAUTISTA, B. E.; HANLIN, R. T. Muscodor yucatanensis, a new endophytic ascomycete from Mexican chakah, Bursera simaruba. Mycotaxon, Portland, v. 110, n. 1, p. 363-372, 2009. https://doi.org/10.5248/110.363 GRIMME, E.; ZIDACK, N. K., SIKORA, R. A.; STROBEL, G. A.; JACOBSEN, B. J. Comparison of Muscodor albus volatiles with a biorational mixture for control of seedling diseases of sugar beet and root-knot nematode on tomato. Plant Disease, St. Paul, v. 91, n. 2, p. 220-225, 2007. HALL, T. A. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. In: Nucleic acids symposium series, Oxford, 1999. p. 95-98. HONGSANAN, S.; HYDE, K. D.; BAHKALI, A. H.; CAMPORESI, E.; CHOMNUNTI, P.; EKANAYAKA, H.; PEREIRA, O. L. Fungal Biodiversity Profiles 11–20. Cryptogamie, Mycologie, Washington, v. 36, n. 3, p. 355-380, 2015. https://doi.org/10.7872/crym/v36.iss3.2015.355 HYDE, K. D.; SOYTONG, K. The fungal endophyte dilemma. Fungal Divers, Thailand. v. 33, n. 163173, p. 2, 2008. KERRY, B. R. Rhizosphere interactions and the exploitation of microbial agents for the biological control of plant-parasitic nematodes. Annual review of phytopathology, Palo Alto, v. 38, n. 1, p. 423-441, 2000. KOGEL, K.H.; FRANKEN, P.; HUCKELHOVEN, R. Endophyte or parasite–what decides?. Current opinion in plant biology, London, v. 9, n. ACKNOWLEDGMENT The authors would like to thank the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and the Fundação de Amparo à pesquisa do Estado de Minas Gerais (FAPEMIG) for financial support and scholarships. The VOC producing fungi belong the genus Muscodor (9) Simplicillium (2) and Acremonium (1). The volatile compounds produced by M. coffeanum (COAD 1842) showed fungicidal activity against A. ochraceus and six isolates inhibited the growth of F. verticillioides. The results demonstrate the potential of fungal endophytes from C. arabica with RESUMO: Fungos endofíticos são uma fonte promissora para a descoberta de compostos com potencial biotecnológico. O objetivo deste estudo foi selecionar e identificar fungos endofíticos de Coffea arabica que produzem compostos orgânicos voláteis (COVs), avaliar o efeito dos compostos orgânicos voláteis produzido por fungos endofíticos sobre o crescimento de Rhizoctonia solani, Fusarium oxysporum, Phoma sp., Botrytis cinerea, Fusarium solani, Fusarium verticillioides, Cercospora coffeicola e Pestalotia longisetula e selecionar fungos endofíticos com potencial para controle biológico de Aspergillus ochraceus inoculado em grãos de café e F. verticillioides inoculado em sementes de milho. Um isolado de Muscodor albus foi utilizado como ferramenta de seleção para fungos endofíticos produtores de COVs. Dentre os 400 fungos endofíticos isolados, 11 foram capazes de crescer na presença de COVs produzidos por M. albus. Estes fungos foram identificados como Muscodor spp. (9) e Simplicillium sp. de acordo com pesquisas na base de dados UNITE usando sequências de DNA do espaçador transcrito interno (ITS). Os COVs produzidos por fungos endofíticos inibiram o crescimento dos fungos fitopatogênicos em comparação com o controle com diferentes eficácias. Os COVs produzidos por Muscodor coffeanum (COAD 1842) apresentou efeito fungicida contra A. ochraceus em grãos de café. Seis fungos endofíticos inibiram completamente o crescimento de F. verticillioides inoculado em sementes de milho. Este estudo demonstra que os fungos endofíticos produtores de compostos voláteis isolados de Coffea arabica são fontes promissoras de compostos bioativos. voláteis. Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 387 MONTEIRO, M. C. P. et al. Antimicrobial activity… Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 REFERENCES 4, p. 358-363, 2006. KUDALKAR, P.; STROBEL, G.; RIYAZ-UL-HASSAN, S.; GEARY, B.; SEARS, J. Muscodor sutura, a novel endophytic fungus with volatile antibiotic activities. Mycoscience, Japan, v. 53, n. 4, p. 319-325, 2012. https://doi.org/10.1007/S10267-011-0165-9 388 MONTEIRO, M. C. P. et al. Antimicrobial activity… MACÍAS-RUBALCAVA, M. L.; HERNÁNDEZ-BAUTISTA, B. E.; OROPEZA, F.; DUARTE, G.; GONZÁLEZ, M. C.; GLENN, A. E.; ANAYA, A. L. Allelochemical effects of volatile compounds and organic extracts from Muscodor yucatanensis, a tropical endophytic fungus from Bursera simaruba. Journal of chemical ecology, Florida, v. 36, n. 10, p. 1122-1131, 2010. MERCIER, J.; JIMÉNEZ, J. I. Control of fungal decay of apples and peaches by the biofumigant fungus Muscodor albus. Postharvest Biology and Technology, Leuven, v. 31, n. 1, p. 1-8, 2004. https://doi.org/10.1016/j.postharvbio.2003.08.004 MERCIER, J.; MANKER, D. C. Biocontrol of soil-borne diseases and plant growth enhancement in greenhouse soilless mix by the volatile-producing fungus Muscodor albus. Crop Protection, Cotonou, v. 24, n. 4, p. 355-362, 2005. MERCIER, J.; SMILANICK, J. L. Control of green mold and sour rot of stored lemon by biofumigation with Muscodor albus. Biological Control, Amsterdam, v. 32, n. 3, p. 401-407, 2005. MERCIER, J.; JIMENEZ, J. I. Potential of the volatile-producing fungus Muscodor albus for control of building molds. Canadian journal of microbiology, Ottawa, v. 53, n. 3, p. 404-410, 2007. https://doi.org/10.1016/j.biocontrol.2004.12.002 MITCHELL, A. M.; STROBEL, G. A.; HESS, W. M.; VARGAS, P. N.; EZRA, D. Muscodor crispans, a novel endophyte from Ananas ananassoides in the Bolivian Amazon. Fungal Diversity, Thailand, v. 31, p. 37-43, 2008. NILSSON, R. H.; HYDE, K. D.; PAWŁOWSKA, J.; RYBERG, M.; TEDERSOO, L., A. B.; ARNOLD, A. E. Improving ITS sequence data for identification of plant pathogenic fungi. Fungal diversity, Thailand. v. 67, n. 1, p. 11-19, 2014. PETRINI, O. Fungal endophytes of tree leaves. In: Microbial ecology of leaves. Springer New York, USA.1991. p. 179-197. https://doi.org/10.1007/978-1-4612-3168-4_9 SANTAMARÍA, J.; BAYMAN, P. Fungal epiphytes and endophytes of coffee leaves (Coffea arabica). Microbial Ecology, Rockville, v. 50, n. 1, p. 1-8, 2005. https://doi.org/10.1007/s00248-004-0002-1 SAUCEDO-GARCÍA, A.; ANAYA, A. L.; ESPINOSA-GARCÍA, F. J.; GONZÁLEZ, M. C. Diversity and communities of foliar endophytic fungi from different agroecosystems of Coffea arabica L. in two regions of Veracruz, Mexico. PloS one, California, v. 9, n. 6, p. e98454, 2014. SAXENA, S.; MESHRAM, V.; KAPOOR, N. Muscodor tigerii sp. nov.-Volatile antibiotic producing endophytic fungus from the Northeastern Himalayas. Annals of Microbiology, Milan, v. 65, n. 1, p. Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 STROBEL, G. Muscodor species-endophytes with biological promise. Phytochemistry Reviews, Dordrecht, v. 10, n. 2, p. 165-172, 2011. https://doi.org/10.1007/s11101-010-9163-3 SUÁREZ‐QUIROZ, M.; GONZÁLEZ‐RIOS, O.; BAREL, M.; GUYOT, B.; SCHORR‐GALINDO, S.; GUIRAUD, J. P. Study of ochratoxin A‐producing strains in coffee processing. International journal of food science & technology, Christchurch, v. 39, n. 5, p. 501-507, 2004. SUWANNARACH, N.; KUMLA, J.; BUSSABAN, B.; LUMYONG, S. Biocontrol of Rhizoctonia solani AG- 2, the causal agent of damping-off by Muscodor cinnamomi CMU-Cib 461. World Journal of Microbiology and Biotechnology, Hull, v. 28, n. 11, p. 3171-3177, 2012. https://doi.org/10.1007/s11274-012-1127-x SUWANNARACH, N.; KUMLA, J.; BUSSABAN, B.; NUANGMEK, W.; MATSUI, K.; LUMYONG, S. Biofumigation with the endophytic fungus Nodulisporium spp. CMU-UPE34 to control postharvest decay of citrus fruit. Crop protection, Cotonou, v. 45, p. 63-70, 2013. TAN, R. X.; ZOU, W. X. Endophytes: a rich source of functional metabolites. Natural product reports, Cambridge, v. 18, n. 4, p. 448-459, 2001. TEJESVI, M. V.; NALINI, M. S.; MAHESH, B.; PRAKASH, H. S.; KINI, K. R.; SHETTY, H. S.; SUBBIAH, V. New hopes from endophytic fungal secondary metabolites. Boletín de la Sociedad Química de México, Del Benito Juárez, v. 1, n. 1, p. 19-26, 2007. VEGA, F. E.; PAVA‐RIPOLL, M.; POSADA, F.; BUYER, J. S. Endophytic bacteria in Coffea arabica L. Journal of basic microbiology, Jena, v. 45, n. 5, p. 371-380, 2005. https://doi.org/10.3852/mycologia.98.1.31 VEGA, F. E.; POSADA, F.; PETERSON, S. W.; GIANFAGNA, T. J.; CHAVES, F. Penicillium species endophytic in coffee plants and ochratoxin A production. Mycologia, Tucson, v. 98, n. 1, p. 31-42, 2006. WARD, N. A.; SCHNEIDER, R. W.; ROBERTSON, C. L. Field evaluations of Simplicillium lanosoniveum as a biological control agent for Phakopsora pachyrhizi. Phytopathology, Salinas, 100. (6): 134-134. 2010. WHITE, T. J.; BRUNS, T.; LEE, S. J. W. T.; TAYLOR, J. W. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. PCR protocols: a guide to methods and applications, San Diego, v. 18, n. 1, p. 315-322, 1990. https://doi.org/10.1016/b978-0-12-372180-8.50042-1 WORAPONG, J.; STROBEL, G. A. Biocontrol of a root rot of kale by Muscodor albus strain MFC2. BioControl, Sophia Antipolis Cedex, v. 54, n. 2, p. 301-306, 2009. https://doi.org/10.1007/s10526-008- 9175-8 ZARE, R.; GAMS, W. A revision of Verticillium section Prostrata. IV. The genera Lecanicillium and Simplicillium gen. nov. Nova Hedwigia, Stuttgart, v.73, n. 1, p. 1-50, 2001. ZHANG, C. L.; WANG, G. P.; MAO, L. J.; KOMON-ZELAZOWSKA, M.; YUAN, Z. L.; LIN, F. C.; KUBICEK, C. P. Muscodor fengyangensis sp. nov. from southeast China: morphology, physiology and production of volatile compounds. REFERENCES 47-57, 2015. https://doi.org/10.1007/s13213-014-0834-y SETTE, L. D.; PASSARINI, M. R. Z.; DELARMELINA, C.; SALATI, F.; DUARTE, M. C. T. Molecular characterization and antimicrobial activity of endophytic fungi from coffee plants. World Journal of Microbiology and Biotechnology, Hull, v. 22, n. 11, p. 1185-1195, 2006. https://doi.org/10.1007/s11274-006- 9160-2 STINSON, M.; EZRA, D.; HESS, W. M.; SEARS, J.; STROBEL, G. An endophytic Gliocladium sp. of Eucryphia cordifolia producing selective volatile antimicrobial compounds. Plant Science, Amsterdam, v. 165, n. 4, p. 913-922, 2003. https://doi.org/10.1016/s0168-9452(03)00299-1 STROBEL, G. A.; DIRKSE, E.; SEARS, J.; MARKWORTH, C. Volatile antimicrobials from Muscodor albus, a novel endophytic fungus. Microbiology, London, v. 147, n. 11, p. 2943-2950, 2001. https://doi.org/10.1099/00221287-147-11-2943 Biosci. J., Uberlândia, v. 33, n. 2, p. 381-389, Mar./Apr. 2017 389 Antimicrobial activity… MONTEIRO, M. C. P. et al. STROBEL, G. Muscodor albus and its biological promise. Journal of Industrial Microbiology and Biotechnology, Houston, v. 33, n. 7, p. 514-522, 2006. https://doi.org/10.1007/s10295-006-0090-7 STROBEL, G. Muscodor species-endophytes with biological promise. Phytochemistry Reviews, Dordrecht, v. 10, n. 2, p. 165-172, 2011. https://doi.org/10.1007/s11101-010-9163-3 STROBEL, G. Muscodor species-endophytes with biological promise. Phytochemistry Reviews, Dordrecht, v. 10, n. 2, p. 165-172, 2011. https://doi.org/10.1007/s11101-010-9163-3 Fungal biology, Manchester, v. 114, n. 10, p. 797-808, 2010.
https://openalex.org/W1991052078	https://www.biodiversitylibrary.org/partpdf/64161	English	null	LIV.—<i>On a new hare from the Transvaal</i>	Annals & magazine of natural history	1,907	public-domain	1,546	* Subspecific name from Rand, popular abbreviation for Witwaters- rand, the name of the range of hills on which Johannesburg is situated. 404 404 L. Jameson on a dark line, margined on each side by a series of pale yellow spots, like beads.” According to Moreau, the pale spots bordering the dark undulated stripes may frequently be absent, or in large speci- mens they may be greyish white in colour and the stripes may be less distinctly marked than in the young. LIV.—On a new Hare from the Transvaal. By H. Lyster JAMESON. In June of this year I noticed a large grey Pronolagus on the Observatory Kopje adjoining the town of Johannesburg. After some difficulty I succeeded in trapping this specimen, a full-grown and pregnant female, which proved to belong to a hitherto undescribed form. Pending a fuller study of the genus Pronolagus I shall describe this hare as a High-Veld race of Pronolagus Ruddi (Thomas and Schwann, Abstract Po ZS. no, 18) p. 23, April 25, 1905; and ‘PZ. 1905, vol. i. p. 272). p ) The Witwatersrand Pronolagus is specially interesting because it extends considerably the range of the P. Ruddi group westwards, hitherto known only from Zululand and the Kastern Transvaal. This form differs trom the type species (so far as I can gather by comparing it with Thomas’s description, for I have not had an opportunity of comparing it with the type) in its generally grey colour, in the dark tail, black for its distal third, in the absence of slaty bases to the fur, and in having black soles to its feet. ) The Witwatersrand Pronolagus is specially interesting because it extends considerably the range of the P. Ruddi group westwards, hitherto known only from Zululand and the Kastern Transvaal. This form differs trom the type species (so far as I can gather by comparing it with Thomas’s description, for I have not had an opportunity of comparing it with the type) in its generally grey colour, in the dark tail, black for its distal third, in the absence of slaty bases to the fur, and in having black soles to its feet. Pronolagus Ruddi randensis , subsp. n. Size as in P. Ruddi (much larger than in P. crassi- caudatus). Coat harsh, as in P. Ruddi. Ground-colour buff, heavily pencilled with black, giving the impression of rather dark grey when seen at a distance. Colour a little lighter on rump. The long stiff hairs are black, with a subterminal buff zone, as in P. Ruddi, and are about 18 mm. long. Among these, especially along the back, are a few very long hairs (40 mm. or more), which are white at the base and black distally, sometimes with a subterminal white new Hare from the 405 ring. These hairs give the middle part of the back a parti- cularly dark grey appearance. Wool-hairs buff, without the slaty-grey bases characteristic of P. Ruddi. Under surface buff, passing into grey on the flanks, so that there is no sharp line between back and belly. This grada- tion is due to the amount of black in the long coarse hairs becoming less towards the ventral surface, until, in the mid- ventral region, these hairs are entirely buff-coloured. Head bluish grey, with a slight rufous tinge on the fore- head. Nape-patch small and inconspicuous, rather more rufous than back. Throat and chest as back. Ears grey, as in P, Ruddi, but without the black edge. Limbs buffy rufous, not any lighter towards digits. Soles black. Tail very large and bushy, about 140 mm. long, including the hairs ; dark reddish brown, pencilled with black, at base ; distal third of tail black. Skull as in P, Rudd?, but slightly larger and rather narrower in the nasal region. The palatal foramina are not quite so much narrowed posteriorly, their inwardly directed edges, so conspicuous in P. Ruddi, being reduced to mere narrowed ridges of bone, so that the walls of the nasal chamber are exposed, as in P, crassicaudatus. The notch in the incisors is more distinct than in Thomas’s figure of the skull of P. Ruddi (P. Z. 8S. 1905, i. pl. xvi. fig. 4). The upper molars have the uncrenulated anterior enamel-wall of the posterior lamina extending considerably less than halfway across the tooth, thus presenting conditions intermediate between P. Ruddi and P. crassicaudatus. ‘The anterior lower premolars also show an intermediate condition, their anterior walls having a single notch on the left side and a double one on the right. P. Z. 8. 1906, ii. p. 348. + “ Dryomys parvulus, Tschudi, Fauna Peruana, p. 179, lam. xiii. fig. 1.” Philippi, An. Mus. Chile, Murideos de Chile, p. 20 (1900). Although this is, no doubt, merely an erroneous rendering of Dry- momys, yet, as it occurs with a specific name attached and a reference both to a description and figure, it seems to be technically too valid as a name to be used again for another animal. Another of my generic names, Neotomys, was used as a misprint for Jectomys by Wallace some years before I published it, but there the misprinted term was without any mention of a species or reference to a description, and consequently, viewed simply by itself, was a mere nomen nudum, which was not the case with Philippi’s Dryomys. Pronolagus Ruddi randensis , subsp. n. Dimensions of the type (measured in the flesh) :— Dimensions of the type (measured in the flesh) :— Head and body 480 mm.; tail 110; hind foot 100; ear 97; ear-opening 87. Skull dimensions: greatest length 94; basilar length 74; zygomatic breadth 41; nasals 47 x 20; interorbital breadth 18, intertemporal breadth 15; palatal foramina 29 x 8°5; palatal bridge 10; diastema 32°53 greatest breadth of nasal region, t l l f t i l 23 F t Dimensions of the type (measured in the flesh) :— Head and body 480 mm.; tail 110; hind foot 100; ear 97; ear-opening 87. Skull dimensions: greatest length 94; basilar length 74; Head and body 480 mm.; tail 110; hind foot 100; ear 97; ear-opening 87. Skull dimensions: greatest length 94; basilar length 74; zygomatic breadth 41; nasals 47 x 20; interorbital breadth 18, intertemporal breadth 15; palatal foramina 29 x 8°5; palatal bridge 10; diastema 32°53 greatest breadth of nasal region, at level of anterior upper premolar, 23. For measurements of P. Ruddi see ‘Thomas and Schwann (loc. cit.) ; for dimen- sions of two races of P. crassicaudatus see Thomas, Ann, & Mag. Nat. Hist. (7) vol. x. pp. 244-246 (1902). Hab. Witwatersrand Range, Transvaal. Hab. Witwatersrand Range, Transvaal. Type. Female. No, 108 in my collection. Observatory, Johannesburg, 5900 ft., 25th June, 1907. Type. Female. No, 108 in my collection. Observatory, Johannesburg, 5900 ft., 25th June, 1907. In Aabits this hare resembles P. crassicaudatus and P, Ruddi, a new Dormouse from Asia Minor. 406 It frequents the stony kopjes of the Witwatersrand, spending the day in holes under rocks and boulders, and coming out to feed at dusk. I have only once found it lying out in the daytime. y Like other Pronolag?, it frequently returns to the same spot to defecate ; hence its presence in a locality can often be detected by the piles of droppings on the hill-sides. Itisa very retiring species, and seems little known among local sportsmen, although it occurs right up to the outskirts of Johannesburg. Its short ears, heavy build, and general pose at once recall the European rabbit. It is very fleet on foot; in fact, the “red hare’’ (a collective name for the members of the genus Pronolagus) is regarded by sportsmen as the fastest of the South-African hares. The uterus of the type specimen contained a single large leveret. P. Z. 8. 1906, ii. p. 348. “ Jameson, H Lyster. 1907. "LIV.—On a new hare from the Transvaal." The Annals and magazine of natural history; zoology, botany, and geology 20, 404–406. https://doi.org/10.1080/00222930709487358. View This Item Online: https://www.biodiversitylibrary.org/item/85040 DOI: https://doi.org/10.1080/00222930709487358 Permalink: https://www.biodiversitylibrary.org/partpdf/64161 Pronolagus Ruddi randensis , subsp. n. University College, Johannesburg, September 1907. LV.—On a new Dormouse from Asia Minor, with Remarks on the Subgenus “ Dryomys.” By OLDFIELD 'THOMAS. LV.—On a new Dormouse from Asia Minor, with Remarks on the Subgenus “ Dryomys.” By OLDFIELD 'THOMAS. In 1906 I formeda special subgenus to contain the dormouse previously known as Hliomys nitedula, Pallas (syn. E. dryas, Schr.), and gave it the name of Dryomys; but I now find that this name is preoccupied {, and would propose to replace it by Dyromys, an anagram of the same word. Since I formed the subgenus there has been discovered the large but nearly related Central-Asian species to which I applied the specific name angelus, and Mr. Gerrit Miller has drawn my attention to additional points of distinction between Holding Institution Smithsonian Libraries and Archives This file was generated 31 March 2024 at 23:29 UTC Copyright & Reuse Copyright & Reuse Copyright Status: Public domain. The BHL considers that this work is no longer under copyright protection. This document was created from content at the Biodiversity Heritage Library, the world's largest open access digital library for biodiversity literature and archives. Visit BHL at https://www.biodiversitylibrary.org. This file was generated 31 March 2024 at 23:29 UTC
https://openalex.org/W1991718146	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0053397&type=printable	English	null	Structural Insight into DFMO Resistant Ornithine Decarboxylase from Entamoeba histolytica: An Inkling to Adaptive Evolution	PloS one	2,013	cc-by	10,766	Abstract Background: Polyamine biosynthetic pathway is a validated therapeutic target for large number of infectious diseases including cancer, giardiasis and African sleeping sickness, etc. a-Difluoromethylornithine (DFMO), a potent drug used for the treatment of African sleeping sickness is an irreversible inhibitor of ornithine decarboxylase (ODC), the first rate limiting enzyme of polyamine biosynthesis. The enzyme ODC of E. histolytica (EhODC) has been reported to exhibit resistance towards DFMO. Methodology/Principal Finding: The basis for insensitivity towards DFMO was investigated by structural analysis of EhODC and conformational modifications at the active site. Here, we report cloning, purification and crystal structure determination of C-terminal truncated Entamoeba histolytica ornithine decarboxylase (EhODCD15). Structure was determined by molecular replacement method and refined to 2.8 A˚ resolution. The orthorhombic crystal exhibits P212121 symmetry with unit cell parameters a = 76.66, b = 119.28, c = 179.28 A˚. Functional as well as evolutionary relations of EhODC with other ODC homologs were predicted on the basis of sequence analysis, phylogeny and structure. Conclusions/Significance: We determined the tetrameric crystal structure of EhODCD15, which exists as a dimer in solution. Insensitivity towards DFMO is due to substitution of key substrate binding residues in active site pocket. Additionally, a few more substitutions similar to antizyme inhibitor (AZI), a non-functional homologue of ODCs, were identified in the active site. Here, we establish the fact that EhODC sequence has conserved PLP binding residues; in contrast few substrate binding residues are mutated similar to AZI. Further sequence analysis and structural studies revealed that EhODC may represent as an evolutionary bridge between active decarboxylase and inactive AZI. Citation: Preeti, Tapas S, Kumar P, Madhubala R, Tomar S (2013) Structural Insight into DFMO Resistant Ornithine Decarboxylase from Entamoeba histolytica: An Inkling to Adaptive Evolution. PLoS ONE 8(1): e53397. doi:10.1371/journal.pone.0053397 Editor: Dan Zilberstein, Technion-Israel Institute of Technology, Israel Editor: Dan Zilberstein, Technion-Israel Institute of Technology, Israel Received October 10, 2012; Accepted November 28, 2012; Published January 11, 2013 Received October 10, 2012; Accepted November 28, 2012; Published January 11, 2013 Copyright: 2013 Preeti et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: Preeti thanks CSIR (Council of Scientific & Industrial Research), S. Abstract Tapas thanks DRDO (Defence Research and Development Organisation, Government of India), PK thanks DST (Department of Science and Technology, Government of India), and S. Tomar thanks ICMR (Indian Council of Medical Research) for financial support. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: shailfbt@iitr.ernet.in Structural Insight into DFMO Resistant Ornithine Decarboxylase from Entamoeba histolytica: An Inkling to Adaptive Evolution Preeti1, Satya Tapas1, Pravindra Kumar1, Rentala Madhubala2, Shailly Tomar1* 1 Department of Biotechnology, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, India, 2 School of Life Sciences, Jawaharlal Nehru University, New Delhi, India Preeti1, Satya Tapas1, Pravindra Kumar1, Rentala Madhubala2, Shailly Tomar1* Cloning of C-terminal truncated EhODC Polymerase Chain Reaction (PCR) amplification was carried out using forward primer 59-ATATCCATATGAAACAAA- CATCTCTAGAAG-39 and reverse primer 59- GAACCTC- GAGTCATTCAATTGACTTAGGGATTTGAAT-39 with NdeI and XhoI restriction enzyme sites respectively to obtain DNA fragment encoding the C-terminal 15 residues truncated EhODC (EhODCD15). The previously cloned full-length EhODC was used as a template in the PCR reaction [29,36]. PCR was performed in a 50 ml reaction mixture containing 10 ml of 56HF phusion buffer supplied with the enzyme, 300 mM of dNTP mix, 6.25 pmol of each of forward and reverse primers, 10 ng of template DNA, 1 ml of 2.5 U/ml phusion polymerase and water. The reaction was performed with initial denaturation at 95uC for 30 s, followed by 30 PCR cycles of denaturation at 95uC for 30 s, annealing at 51uC for 60 s and extension at 72uC for 1 min and 15 s. A final extension was carried out at 72uC for 15 min. The resultant PCR product was subcloned into NdeI and XhoI sites of pET-28c with His6-tag preceding the N-terminal and tobacco etch virus (TEV) protease cleavage site to allow the removal of tag from recombinant protein. Ligated product was transformed into freshly prepared E. coli DH5a competent cells. Kanamycin resistant transformants were selected and grown in LB broth supplemented with 50 mg/ml kanamycin. The pET28- EhODCD15 plasmid was isolated and right size insert in the construct was confirmed by DNA sequencing from TCGA, New Delhi. Furthermore, in mammals, the activity of antizyme is negatively regulated by a protein called antizyme inhibitor (AZI). AZI binds to antizyme and blocks the binding of antizyme to ODC which down regulates ODC degradation as well as leads to ODC activation. AZI has higher binding affinity for antizyme as compared to ODC which results in antizyme sequestration and elevation of ODC levels [18,19,20,21,22,23]. Previously, it has been reported that AZI is homologous to ODC and the major residues involved in catalytic activity of ODC are conserved in AZI [24]. However, AZI does not possess enzymatic activity due to changes in the sequence that lead to protein inability to bind cofactor PLP along with the failure in decarboxylation activity [24,25,26]. In E. histolytica, ODC is the only enzyme of polyamine biosynthetic pathway that has been reported to exist in the organism [27]. The analysis of polyamine content shows that considerable amount of putrescine is present in E. histolytica. Introduction which are involved in various cellular processes that govern cell growth and proliferation [9]. Subsequently, the actively prolifer- ating cells have higher concentrations of polyamines. The intracellular concentrations of polyamines are tightly regulated by different mechanisms including biosynthesis, inter-conversion, degradation, and uptake from the surrounding through polyamine transporter. The failure in regulation of polyamine levels in cells has been linked to various cancers. Hence, polyamine metabolic pathway is also a potential target for cancer treatment [10,11,12]. Consequently, not only the polyamine biosynthetic pathway but also the key components of polyamine homeostasis are potential therapeutic targets [8]. The two enzymes of polyamine biosyn- thesis pathway, ornithine decarboxylase (ODC) and S-adenosyl- methionine decarboxylase (SAMDC) are highly-regulated and have a very short half-life by which cells quickly alter the levels of polyamines [13]. Entamoeba histolytica is responsible for causing amoebiasis, amoebic liver abscess and amoebic colitis in humans. It is the third major and a dangerous public health problem in the world [1,2]. Though a small number of drugs including metronidazole, emetine, tinidazole, chloroquine and nitrazoxanide are used for the treatment of the disease, most of them are associated with numerous side effects. In some cases, frequent use of these drugs has led to the development of clinical drug resistance in the pathogen [3,4]. Thus, it is crucial to identify and elucidate a potent metabolic pathway in E. histolytica which could be set as a therapeutic target for development of new anti-amoebic drugs. In last few decades, the polyamine metabolic pathway in protozoan diseases including African sleeping sickness [5], giardiasis [6] and leishmaniasis [7] has emerged as a potential therapeutic target [8]. The polyamines such as putrescine, spermidine and spermine are essential polycationic compounds, Ornithine decarboxylase catalyzes the first and rate-limiting step of polyamine biosynthetic pathway. L-ornithine is decarbox- January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 1 January 2013 \| Volume 8 \| Issue 1 \| e53397 Crystal Structure of ODC from E. histolytica ylated by ODC enzyme in the presence of cofactor pyridoxal-59- phosphate (PLP) to produce putrescine. The enzymatic activity of ODC is tightly regulated by a distinct mechanism in which polyamines induce the expression of a regulatory protein called antizyme (AZ) by +1 ribosomal frameshifting [14]. AZ inhibits ODC enzyme activity by binding and disrupting active ODC homodimers, and subsequently marks the enzyme for ubiquitin- independent degradation by the 26S proteasome [15,16]. Expression and purification p p The pET28-EhODCD15 plasmid containing truncated EhODC gene was transformed into E. coli BL21 (DE3) competent cells. For protein expression, transformed BL21 (DE3) cells were grown at 37uC to an optical density of ,0.6 at 600 nm (OD600) and induced with 0.5 mM isopropyl-ß-thiogalactopyranoside (IPTG). Induced cultures were transferred to 18uC and cells were grown for ,14 h. Cells were harvested by centrifugation at 5,000 rpm at 4uC and cell pellets were stored at 220uC until further use. For protein purification, cell pellets from 1 litre culture were re- suspended in 20 ml of ice cold binding buffer containing 50 mM Tris HCl (pH 7.5), 40 mM imidazole, 250 mM sodium chloride, 2 mM phenylmethylsuphonyl fluoride (PMSF) and 5% glycerol (v/v). Lysozyme was added to a final concentration of 100 mg/ml and kept on rocking platform at 4uC for 45 min. Cells were disrupted by sonication on ice with 50% amplitude and a pulse of 20 sec on and 60 sec off for 15 min. The lysate was centrifuged at 18,000 rpm for 45 min at 4uC to separate supernatant from cell debris. The supernatant was loaded onto 5 ml HisTrap HP affinity column pre-equilibrated with the binding buffer. Protein was eluted by running a linear gradient of 40–1000 mM imidazole in 60 ml of buffer A [50 mM Tris HCl (pH 7.5), 1 M imidazole, 250 mM sodium chloride and 5% glycerol (v/v)] at a flow rate of 1 ml/min. Eluted fractions were analyzed on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and frac- tions containing pure protein were pooled together. To remove the N-terminal His-tag, TEV protease was added to the sample with protein to TEV ratio 1:20 and incubated for ,12 h at 4uC and simultaneously dialyzed against buffer A without imidazole. To remove uncleaved His-tag protein and His-tag TEV protease, the sample was again loaded onto 5 ml HisTrap HP column. Flow- through containing EhODCD15 without His-tag was collected and In this study, we have determined the crystal structure of EhODC to elucidate the structural features responsible for DFMO insensitivity and low substrate binding affinity. Furthermore, detailed comparative sequence and structural analysis was performed with functional ODCs and non-functional ODC homologue i.e. AZI to investigate the evolutionary status of EhODC. Introduction Addi- tionally, AZ negatively regulates the uptake of polyamines by repressing polyamine transporter [17]. Thus, polyamine homeo- stasis is maintained in a cell through polyamines themselves via a negative feedback system, by governing the synthesis of AZ protein. screens were obtained from Hampton Research (Hampton Research Inc. Aliso Viejo, CA). The plasmid pET30a containing full length of EhODC was taken as template for sub-cloning [29]. Cloning of C-terminal truncated EhODC While, very low levels of spermidine and no spermine is detected supporting the absence of other genes of polyamine biosynthetic pathway in E. histolytica genome [28,29]. Interestingly, the comparison of EhODC kinetic parameters with other well characterized ODCs indicates that it has low substrate affinity and catalytic efficiency [29]. Moreover, DFMO, a suicide substrate inhibitor of ODC is used for the treatment of African sleeping sickness, a protozoan disease caused by Trypanosome brucei gambiense [30,31]. Interestingly, DFMO being an effective drug against T. brucei gambiense is reported to have relatively poor effect on the more virulent strain T. brucei rhodesiense [32,33]. Further- more, the ODC of E. histolytica, being a pathogenic strain from protozoa kingdom, is insensitive to DFMO due to sequence divergence in the substrate binding residues [29,34,35,36]. Natural resistance to DFMO within the same Trypanosome species as well as within the protozoa kingdom draws attention towards the sequence and structural divergence for their evolutionary adap- tation. Reagents Restriction enzymes NdeI, XhoI, T4-DNA ligase and phusion polymerase were purchased from NEB. Primers were ordered from Integrated DNA Technology. HisTrap HP Ni Sepharose column and Hiload 16/60 Superdex 200 pg size exclusion column were obtained from GE healthcare. For crystallization, PEG ION January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 2 Crystal Structure of ODC from E. histolytica method using Molrep program of CCP4-6.0 suite [39]. The model was generated using previously reported crystal structure of human ODC (PDB ID: 2ON3) [40]. Non-crystallographic symmetry restraints were applied throughout the refinement stages using four EhODCD15 molecules in the asymmetric unit. Structure refinement was performed using CNS v.1.2, Phenix v 1.7.2-869, and REFMAC 5.2 refinement tools [41,42,43]. Rounds of model building were carried out using program Coot v 0.6.2 [44]. The quality of the model was evaluated by PROCHECK [45]. concentrated using a 10 kDa cut-off Amicon Ultra-15 concentra- tor (Millipore, Bedford, Massachusetts, USA). The concentrated protein was loaded onto HiLoad 16/60 prep grade Superdex 200 size-exclusion chromatography column pre-equilibrated with buffer B containing 30 mM HEPES-Na (pH 7.5), 250 mM NaCl, 1 mM EDTA, 10% (v/v) glycerol and 1 mM DTT. The major peak fractions containing pure protein were pooled and concen- trated to 5 mg/ml. Homogeneity of purified EhODCD15 protein was analysed on 12% SDS-PAGE. Protein concentration and yield were determined using the Bio-Rad protein assay kit with bovine serum albumin (BSA) as a standard. Gel filtration analysis The average molecular weight of EhODCD15 was determined using size exclusion chromatography and compared with previ- ously characterized full-length EhODC [36]. In brief, the purified protein was concentrated to 5 mg/ml and was injected onto HiLoad 16/60 Superdex 200 gel filtration chromatography column using A¨ KTA purification system (GE Healthcare). Protein was allowed to pass through the column at a rate of 0.5 ml/min. For the molecular weight estimation of EhODCD15, the elution profile of the major peak containing purified protein was compared with the elution profile of the standard Gel Filtration HMW Calibration kit molecular weight markers (GE healthcare). Table 1. Statistical representation of data collection and structure refinement parameters along with quality of the model accessed by Ramachandran plot. Data collection Space group P212121 Unit cell parameters a (A˚), b (A˚), c (A˚) 76.66, 119.28, 179.28 Resolution (A˚) 99.5–2.87 (2.92–2.87)a Number of reflections 35570 Completeness (%) 92.1(59.0)a Mean redundancy 3.4 (2.1)a I/s 4.82 (2.0)a Rmerge b (%) 0.150 (0.670)a Refinement Resolution (A˚) 99.5–2.87 (2.92–2.87)a Number of non-H atoms in asymmetric unit Protein 10484 Water molecules 101 R-factor (%) 25.3 Rfree c value (%) 29.9 Average B-factor (A˚2) 54.4 Rms deviations bond lengths (A˚) 0.005 bond angles (u) 0.831 Ramachandran plot Residues in favored region (%) 88.7 Residues in allowed region (%) 10 Residues in generously allowed region (%) 0.9 Residues in outlier region (%) 0.4 avalue in parentheses are for the highest resolution shell. bRmerge = S \| I2I¯ \|/S I \| where I = observed intensity and I¯= average intensity. cRfree = S (\|F\|obs2\|F\|calc\|)/S \|F\|obs where \|F\|obs are observed structure factor amplitudes for a given reflection and \|F\|calc are calculated structure factor amplitude. doi:10.1371/journal.pone.0053397.t001 Model generation for active site analysis In the crystal structure of EhODCD15, the flexible loops missing in one subunit were present in the other subunits. Therefore, coordinates for missing loops near active site in the structure were EhODCD15 enzymatic activity The sequence of EhODC, along with other functional ODCs and AZI were retrieved from NCBI database [46]. Multiple sequence alignment and phylogenetic tree of these sequences were obtained using ClustalW [47] for evolutionary variation analysis. To confirm that the truncation of 15 residues from C-terminus does not inactivate EhODC, ornithine decarboxylation activity of purified protein and production of putrescine was spectrophoto- metrically determined using the method developed by Badolo et al [37]. Enzymatic activity of the purified EhODCD15 protein was compared with full-length EhODC [36]. Crystallization For crystallization, purified EhODCD15 protein was concen- trated to 12.5 mg/ml in 30 mM HEPES-NaOH buffer (pH 7.5) containing 1 mM EDTA, 0.25 M NaCl, 1 mM DTT and 10% (v/v) glycerol. Crystallization trials were performed using the sitting drop vapour diffusion method in 96 well plates (Hampton Research) at 20uC and 4uC. The drops were prepared by mixing 2 ml of protein solution with 1 ml of reservoir solution and equilibrated against 80 ml reservoir solution. Hampton Research crystallization screens Crystal screen, Crystal screen 2 and PEG/ ION screen (Hampton Research, USA) were used to explore the initial crystallization conditions. Crystals were obtained in PEG ION screen containing 20% PEG 3350 in 0.2 M LiCl solution maintained at pH 6.8. Diamond shaped crystals of EhODCD15 appeared in four months at 20uC. Prior to data collection, crystal was cryo-protected by bathing it in mother liquor containing 3% (v/v) ethylene glycol for 10 s. The crystal was flash-frozen under cryogenic conditions at 100 K using liquid nitrogen stream to prevent radiation damage during data collection. C-terminal Truncation and Purification of EhODC C-terminal Truncation and Purification of EhODC The ODC enzyme from E. histolytica belongs to fold type III group IV decarboxylase of a B6-dependent family, having eukaryotic ornithine decarboxylase characteristics [50,51]. Under this classification, crystal structures are only available from three different sources including human, mouse and Trypanosome brucei ODC [52,53,54]. For crystal structure determination of EhODC, full-length protein was purified using the previously established protocol [36] and was used for crystallization experiments. However, extensive crystallization trials of full-length EhODC were unsuccessful. In order to decrease the conformational heterogeneity, it is a common practice to truncate the flexible N and/or C-terminal residues to facilitate the crystallization process. Therefore, EhODC sequence was examined to identify disordered regions using bioinformatics tools DisEMBL and GlobPlot [55,56]. These programs predicted a fragment of approximately 13–17 residues at the C-terminus of EhODC to be flexible. Additionally, it has been reported that the truncation of 37 residues from the C-terminus of mouse ODC resulted in protein stability and has been crystallized successfully for structure determination [53,57,58]. The C-terminal sequence of EhODC shows similarity with mouse ODC in having a PEST like sequence [36]. Based on these observations, 15 residues were deleted from the C-terminus of EhODC. Expression and solubility of EhODCD15 construct was optimized by varying induction temperature (37uC, 25uC, and 18uC). Maximum solubility was observed at 18uC when induced with 0.5 mM IPTG for ,14 h. Recombinant EhODCD15 was purified in three sequential purification steps, with yield of ,5 mg per liter of E. coli culture. Elution profiles from the gel filtration column demonstrated that EhODCD15 exists in the dimeric form similar to full-length EhODC [36]. The purified protein exhibited a single band of approximately ,45 kDa in 12% SDS-PAGE gel (Figure 1). The enzymatic activities of EhODCD15 and wild-type proteins were compared using previously established protocol [36]. The comparative analysis of both the full-length and EhODCD15 forms didn’t show any notable difference in the activity indicating that the truncation of 15 residues from the C-terminus of EhODC does not affect its activity. In the tetrameric structure, residues Phe91 and Leu87 of chain A interact with Ser388 of chain C through a water molecule. In addition, Glu90 of chain A interacts to Ser388 of chain C through polar interaction. Similarly, Asp88 of chain A is forming direct interaction with residue Leu386 of chain C and vice versa. Overall structure and folding g Each monomer consists of b/a barrel and b-sheet domain which are arranged identical to previously known ODC structures (Figure 4). However, the tetramer arrangement displays a number of unusual features. Residues from barrel involved in contact and dimer formation are located at the surface or in proximity to sheet domain of opposite monomer. Interface residues of helices a5, a7, a8 and a9 of chain A barrel form extensive contacts with sheet domain S2 of chain B. All four chains in asymmetric unit showed similar structures and are involved in similar interactions. The analysis of dimer-dimer interactions exhibited large intermolecular distances of ,4.0 A˚ . In a monomer, helix a1 is connected to sheet b1 (Gly27-Phe31) through a loop and enters the barrel. The barrel is composed of eight helices i.e. a2 (T33-N46), a3 (P62-L71), a4 (L80-L89), a5 (Y105-L114), a6 (I124-Y133), a7 (D163-K175), a8 (E194-F213) and a9 (F232-L246) followed by eight alternate b- strands b2 (R51-A55), b3 (G74-C77), b4 (I96-Y98), b5 (H118- V121), b6 (G138-R142), b7 (V182-F184), b8 (L219-D221) and b9 (R253-A256). The sheet domain comprises of eight randomly arranged b-strands which can be further divided into S1 and S2 b- sheets that are perpendicular to each other. Sheet S1 consisted of four sheets b10 (F267-S271), b15 (L355-F357), b16 (I381-T383) in addition to b1, which are roughly perpendicular to S2 containing b11 (H274-Q281), b12 (K284-S291), b13 (Y325-Y330) and b14 (A341-L345) (Figure 4). However, both domains are connected by two loops in between b1-a2 and a10-b10. The barrel and b-sheet domains of the monomeric subunits are associated in head to tail manner in the dimer. In addition, various polar interactions at the dimer interface including salt bridges and hydrophobic interac- tions are involved in the formation of dimer. The structure of EhODC has several highly mobile loop moieties that are depicted by dashed lines in Figure 4. Crystal Structure of ODC from E. histolytica Crystal Structure of ODC from E. histolytica generated by MODELLER 9.10 [48] using the solved crystal structure of one subunit of EhODC as a template. Evaluation of the steriochemical properties of obtained structure having built-in loops was performed using PROCHECK [45]. All the figures of structure and active sites were generated using PyMol [49]. The four monomers in the asymmetric unit of crystal are arranged as two separate dimers (subunits A, B and subunits C, D) facing each other at the convex surfaces. Each monomer in a dimer makes side to side contacts with each other forming an overall bent structure. Further, as the loops in a dimer interface are disordered and clear density was not observed, the central part of dimer forms a hollow structure. In the dimer, chain A and chain B are arranged in head to tail manner at origin (0,0,0) of orthorhombic unit cell (Figure 2). The b/a barrel of chain A and b sheet of chain B pose at origin and their counterpart extends along X-direction. Dimer of AB is situated along with X-axis by an angle of 30u approximately; whereas other dimer CD is situated at rotation angle of 180u with a screw distance of 19.1 A˚ that occupies approximately one quarter of unit cell. The crystallized structure of EhODCD15 consists of a tetramer. The asymmetric unit contains two dimers comprising of chain A, B, C, D. The total area of the molecule of EhODC containing four molecules was estimated to be 61227.6 A˚ 2. Each dimer interacts with its symmetry mate to form dimer-dimer interfaces as A–B dimer interacts with C–D dimer (Figure 2). Interface area evaluated by PISA web server was averaged to 1373.4 A˚ 2 which was 1599.8 A˚ 2 and 1147.0 A˚ 2 between B, A and D, C respectively [59]. C-terminal Truncation and Purification of EhODC Residues Glu110 and His113 of chain B are at a distance of 3.1 A˚ and 3.3 A˚ from Lys84 and Asp88 of chain D showing polar interactions and vice versa (Figure 3). Data Collection and structure determination The diffraction data of EhODCD15 were collected at 100 K using Cu Ka radiation generated by a Bruker Microstar-H rotating-anode generator assembled with MAR 345 imaging-plate system. The data were collected at 1.54 A˚ with a crystal-to- detector distance of 200 mm and 1u oscillation per image with 20 min exposure per frame. Crystal diffracted to 2.8 A˚ resolution. The data were indexed, integrated and scaled using HKL2000 program [38]. Table 1 summarizes data collection and processing statistics. The structure was solved by molecular replacement January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 3 January 2013 \| Volume 8 \| Issue 1 \| e53397 Crystal packing Crystallization of purified EhODCD15 was performed using sitting drop vapor diffusion method. Crystals were obtained at 20uC in Hampton PEG ion screen 4 containing 20% (v/v) PEG 3350, 0.2 M LiCl maintained at pH 6.8. Crystals belonged to the orthorhombic space group exhibiting P212121 symmetry with unit cell parameters a = 76.66, b = 119.28, c = 179.28 A˚ and a = b = c = 90u. The crystal diffracted to 2.8 A˚ resolution, possess- ing four molecules per asymmetric unit and the solvent content was calculated to be 46.69% with a Matthews coefficient of 2.2 A˚ 3 Da21. Quality of the obtained structure was assessed with the PROCHECK program showing 88.7% of the residues in the favored region, whereas 10% in allowed, 0.9% in generously allowed and only 0.4% residues are observed in the disallowed region of Ramachandran Plot (Table 1). January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 4 Crystal Structure of ODC from E. histolytica Figure 1. Purification and gel filtration profile of EhODCD15. A) 12% SDS-PAGE gel showing the affinity purified protein Lane 1: Molecular weight markers shown in kDa. Lane 2–3: Protein purified by affinity chromatography. B) Elution profile of the EhODC1D15 protein. The protein was eluted at a volume of 74 ml corresponding to molecule weight of ,87 kDa. Insert shows the purified protein in 12% SDS-PAGE after gel filtration chromatography. doi:10.1371/journal.pone.0053397.g001 Crystal Structure of ODC from E. histolytica Figure 1. Purification and gel filtration profile of EhODCD15. A) 12% SDS-PAGE gel showing the affinity purified protein Lane 1: Molecular weight markers shown in kDa. Lane 2–3: Protein purified by affinity chromatography. B) Elution profile of the EhODC1D15 protein. The protein was eluted at a volume of 74 ml corresponding to molecule weight of ,87 kDa. Insert shows the purified protein in 12% SDS-PAGE after gel filtration chromatography. doi:10.1371/journal.pone.0053397.g001 Figure 1. Purification and gel filtration profile of EhODCD15. A) 12% SDS-PAGE gel showing the affinity purified protein Lane 1: Molecular weight markers shown in kDa. Lane 2–3: Protein purified by affinity chromatography. B) Elution profile of the EhODC1D15 protein. The protein was eluted at a volume of 74 ml corresponding to molecule weight of ,87 kDa. Insert shows the purified protein in 12% SDS-PAGE after gel filtration chromatography. doi:10.1371/journal.pone.0053397.g001 Figure 2. Crystal packing Schematic representation of overall structure of the model obtained after molecular replacement. A) Cartoon diagram of tetrameric model of EhODC showing AB-CD, dimer-dimer interface; B) Active site of EhODC at the interface of dimer where (9) denotes the residues from the other subunit. doi:10.1371/journal.pone.0053397.g002 Figure 2. Schematic representation of overall structure of the model obtained after molecular replacement. A) Cartoon diagram of tetrameric model of EhODC showing AB-CD, dimer-dimer interface; B) Active site of EhODC at the interface of dimer where (9) denotes the residues from the other subunit. doi:10.1371/journal.pone.0053397.g002 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 5 Crystal Structure of ODC from E. histolytica Figure 3. Tetrameric structure with dimer-dimer interaction. A–C) shows the interaction between chain A and chain C. B–D) indicates the interaction between chain B and chain D. Pink dashes shows the interaction of residues through water molecule and green dashes indicates the polar interactions. Symbol (0) and (9) denotes the residues of chain C and chain D, respectively. doi:10.1371/journal.pone.0053397.g003 Figure 3. Tetrameric structure with dimer-dimer interaction. A–C) shows the interaction between chain A and chain C. B–D) indicates the interaction between chain B and chain D. Pink dashes shows the interaction of residues through water molecule and green dashes indicates the polar interactions. Symbol (0) and (9) denotes the residues of chain C and chain D, respectively. doi:10.1371/journal.pone.0053397.g003 Comparative analysis of active site architecture Furthermore, alpha-carbon backbone consisting residues Tyr331 and Asp332 in TbODC shows direct interactions with bound DFMO and these interactions play a role in proper DFMO molecule orientation in the active site pocket. However, these residues are mutated to Phe305 and Glu306 respectively in EhODC. Also, EhODC crystal structure reveals that the loop consisting of Phe305 and Glu306 residues is not located close to the active site thus may not contribute to DFMO binding. Moreover, residue Tyr’323 from other subunit of TbODC also supports the favourable orientation of DFMO by side chain hydroxyl group interaction with DFMO through a water molecule. Tyr’323 is replaced with His’296 in EhODC and the loop containing His’296 residue is positioned away from the active site (Figure 5). In contrast, the cofactor PLP and substrate L- ornithine are accommodated in EhODC active site with polar interactions to facilitate the catalysis (Figure 2) [29,35]. These structural details indicate that amino acid substitutions in the active site of EhODC create a novel architecture which not only makes it resistant to DFMO but also lowers its catalytic efficiency by weakening substrate binding, as the reported Km values for L- ornithine for active but DFMO sensitive T. brucei and mouse ODC binding residues (His197, Gly276, Arg277 and Tyr389 respec- tively) of TbODC (54, PDB ID: 2TOD). The side chains of these conserved residues interacting with PLP through polar interactions in TbODC are also expected to bind PLP and correctly orientation it into the active site of EhODC. Interestingly, Ser200 of TbODC is present in a loop and is seen interacting with PLP in TbODC structure (54, PDB ID: 2TOD). However, this residue (Ser188) is also conserved in EhODC, but shows small displacement from its expected position and has opposite orientation in the crystal structure of EhODC. Though, at this point it cannot be ruled out that the flexible loop of EhODC possessing Ser188 may approach the active site and may orient Ser188 in favourable position in the presence of PLP, whereas apo-enzyme might not restrict its position. Addition to this, Gly225 of EhODC shares exact position of Gly237 of TbODC, whose backbone carbon chain contributes to polar interactions. Apart from this, few residues interact with PLP through water molecules which include residues Phe238, Tyr278, Arg154 and Ala111 in TbODC. Comparative analysis of active site architecture In contrast, Cys’334 the conserved residue of EhODC that is expected to form a covalent bond with DFMO is slightly displaced from its position and has distinct orientation that is structurally unfavourable for covalent linkage with DFMO. In addition, the residue Asp’361 of TbODC is substituted by Asn’335 in EhODC, which is not expected to interact with DFMO. Furthermore, alpha-carbon backbone consisting residues Tyr331 and Asp332 in TbODC shows direct interactions with bound DFMO and these interactions play a role in proper DFMO molecule orientation in the active site pocket. However, these residues are mutated to Phe305 and Glu306 respectively in EhODC. Also, EhODC crystal structure reveals that the loop consisting of Phe305 and Glu306 residues is not located close to the active site thus may not contribute to DFMO binding. Moreover, residue Tyr’323 from other subunit of TbODC also supports the favourable orientation of DFMO by side chain hydroxyl group interaction with DFMO through a water molecule. Tyr’323 is replaced with His’296 in EhODC and the loop containing His’296 residue is positioned away from the active site (Figure 5). In contrast, the cofactor PLP and substrate L- ornithine are accommodated in EhODC active site with polar interactions to facilitate the catalysis (Figure 2) [29,35]. These structural details indicate that amino acid substitutions in the active site of EhODC create a novel architecture which not only makes it resistant to DFMO but also lowers its catalytic efficiency by weakening substrate binding, as the reported Km values for L- ornithine for active but DFMO sensitive T. brucei and mouse ODC subunit plays the most critical role in DFMO binding by making a permanent covalent bond with the enzyme. However, covalent bond formation of Cys’360 with substrate L-ornithine has not been reported for any ODC enzyme. In addition to this, the next residue of TbODC Asp’361 helps to position the Cys’360 residue in proper orientation and also interacts with DFMO through a water molecule. In contrast, Cys’334 the conserved residue of EhODC that is expected to form a covalent bond with DFMO is slightly displaced from its position and has distinct orientation that is structurally unfavourable for covalent linkage with DFMO. In addition, the residue Asp’361 of TbODC is substituted by Asn’335 in EhODC, which is not expected to interact with DFMO. Comparative analysis of active site architecture both the subunits. For comparative analysis of active site, we superimposed the EhODC structure over TbODC complexed with DFMO. The super-imposition of monomers shows root mean square deviation (rmsd) of 1.18 A˚ whereas super-imposition of dimer shows rmsd of 1.7 A˚ . Active site superimposition of TbODC and EhODC shows that most of the conserved residues in active site of EhODC share same positions as in TbODC, however few residues pose in different orientation (Figure 5). His185, Gly259, Arg260 and Tyr363, the well conserved PLP binding residues of EhODC share the position and have orientation similar to the PLP The ODC enzyme is an obligate homodimer with two symmetry-related active sites located at the dimer interface. According to our previous report, EhODC enzyme is functionally active in the dimeric form [36]. As expected, the crystal structure of EhODC contains two equivalent active site pockets at the dimer interface formed by residues that are contributed from both the subunits (Figure 2). The proper orientation of active site formed by two subunits is highly essential for functionality of the enzyme. The active site in EhODC is mainly contributed by loops from Figure 4. Crystal structure of EhODC monomeric subunit. A) Cartoon diagram of the monomer showing arrangement of barrel and sheet domain. B) Topology diagram of monomer of EhODC where helices are represented with cylinder and sheets with the arrows connected with loops, dashed line indicates the sequence missing in the structure. doi:10.1371/journal.pone.0053397.g004 Figure 4. Crystal structure of EhODC monomeric subunit. A) Cartoon diagram of the monomer showing arrangement of barrel and sheet domain. B) Topology diagram of monomer of EhODC where helices are represented with cylinder and sheets with the arrows connected with loops, dashed line indicates the sequence missing in the structure. doi:10.1371/journal.pone.0053397.g004 January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 6 Crystal Structure of ODC from E. histolytica subunit plays the most critical role in DFMO binding by making a permanent covalent bond with the enzyme. However, covalent bond formation of Cys’360 with substrate L-ornithine has not been reported for any ODC enzyme. In addition to this, the next residue of TbODC Asp’361 helps to position the Cys’360 residue in proper orientation and also interacts with DFMO through a water molecule. EhODC sequence and structural comparison with ODC homologs From various mutational studies, it is reported that a conserved Lys (Lys57 in EhODC) is important as it forms Schiff base with PLP which is later displaced by L-ornithine that undergoes decarboxylation through nucleophilic attack via a conserved Cys (Cys3349 in EhODC) [36,63,64,65]. However, both the residues are well conserved in both functional ODCs and AZI (except in few AZIs). In all functional ODCs including ODCs from T. brucei, Homo sapiens (HsODC) and mouse, residue Ala111 and Arg154 (HsODC) are highly conserved and interact with PLP through water. Interestingly, in EhODC, though Arg142 is conserved, however Ala111 is uniquely substituted with Ser99. AZI possesses substitution at both the positions with Ala to Thr/Ile/Ser and Arg to His/Gln that make AZI incapable of binding to PLP. Out of sixteen PLP binding residues, AZIs have major mutations in five positions whereas EhODC possesses a single mutation at position 99 with substitution of Ala to Ser (Figure 7 and Figure 8). The novel active site architecture revealed from the crystal structure of EhODC and previously reported low catalytic efficiency of the enzyme hints towards possible adaptive evolution which lead to DFMO insensitive. AZI is an inactive ODC homolog that possesses a broader active site due to unusual packing of AZI dimers [61]. The architecture of AZI active site does not favour the accommodation of substrate as well as the co- factor for enzyme catalysis, which makes it an inactive homolog of ODC. Recently, it has been proposed that several homologs of ODCs including putative antizyme inhibitors apparently arise independently through evolution [46]. Robust sequence analysis and active site structure comparisons were performed to explore the evolutionary relationship of EhODC with respect to ODC homologs including AZI and to uncover the possibility of additional EhODC functions. Multiple sequence alignment was done and phylogenetic tree was generated for ODC homologues including functional ODC and nonfunctional AZI (Figure 6). In HsODC, five residues Tyr323, Tyr331, Asp332, Cys360 and Asp361 are reported to be key active site residues which interact with L-ornithine and these residues are highly conserved in all functional ODCs (Figure 7). However, EhODC is an exception where only Cys334 is conserved while both Tyr and both Asp residues are substituted by His296, Phe305, Glu306 and Asn335 respectively. Crystal Structure of ODC from E. histolytica identified 27 residues from sequence alignment of ODCs from different organisms responsible for the formation of active site pocket and in ODC enzyme dimerization (Figure 7). Out of 27 residues, 16 residues contribute to the active site formation by interacting with cofactor PLP, 5 residues for substrate binding, 3 residues for salt bridge formation, 3 residues as critical interface residues and 1 residue for dimerization (Figure 7). enzymes are 0.24 mM and 0.09 mM respectively, whereas for DFMO resistant EhODC it is 1.5 mM [29,60]. Not only substituted residues, but also the displacement of loops (His’296 loop/Phe305 and Glu306 loop) away from EhODC active site seems to contribute towards DFMO insensitivity. Comparative analysis of active site architecture These residues are present at the same position in EhODC active site except Ala111 where it is substituted by Ser99 in EhODC. Overall, the architecture of EhODC for binding to PLP is similar to that of TbODC. DFMO, a substrate analogue makes a stable covalent bond with conserved Cys residue in the active site of ODC enzyme and inhibits its catalytic reaction. Binding of DFMO in the proper orientation for covalent bond formation with its active site is also supported by its interaction with other residues that are there in the substrate binding pocket. To extricate the intricate structural details of EhODC responsible for low substrate affinity and/or DMFO insensitivity, structural comparison of EhODC active site architecture for substrate/DFMO binding was done with the active site of DFMO bound TbODC crystal structure (Figure 5) (54, PDB ID: 2TOD). In TbODC, Cys’360 from counterpart Figure 5. Superimposition of active site of EhODC with TbODC bound to DFMO. Residues of active site at the dimer interface are represented in sticks. TbODC residues are colored with green, EhODC residues are colored with orange. PLP and DFMO are colored with blue and polar interactions were indicated by black dashes; water molecule in shown in red sphere. Residues with (9) symbol are of opposite monomer. doi:10.1371/journal.pone.0053397.g005 Figure 5. Superimposition of active site of EhODC with TbODC bound to DFMO. Residues of active site at the dimer interface are represented in sticks. TbODC residues are colored with green, EhODC residues are colored with orange. PLP and DFMO are colored with blue and polar interactions were indicated by black dashes; water molecule in shown in red sphere. Residues with (9) symbol are of opposite monomer. doi:10.1371/journal.pone.0053397.g005 January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 7 Crystal Structure of ODC from E. histolytica EhODC sequence and structural comparison with ODC homologs Colour indication: Violet columns signifies the mutation in AZI; Orange columns signifies the mutated residues in E. histolytica ODC which are similar to AZI; Gray shows the unique mutations in EhODC which is neither conserved in ODC nor in AZI; Blue indicate the mutation in EhODC which are rarely found in AZI and functional ODC; Olive color point out the mutations in EhODC which are similar to some ODC. Sequence analysis and numbering has been done according to EhODC. Residues which are not conserved are shown by single letter, the conserved residues are indicted by – and D indicates the deleted amino acids. % identity indicates the identity of EhODC sequence with other homologous ODC sequences [46]. doi:10.1371/journal.pone.0053397.g007 Asn335 positions are unique to EhODC as these two residues are found to be conserved as Tyr323 and Asp361 in both AZI as well as in functional ODC. Interestingly, substitution at Glu306 instead of Asp332 is similar to AZIs, the inactive homologs of ODCs. However, in AZIs only Asp332 is substituted by Glu, whereas other four residues are mostly conserved (Figure 7). Asn335 positions are unique to EhODC as these two residues are found to be conserved as Tyr323 and Asp361 in both AZI as well as in functional ODC. Interestingly, substitution at Glu306 instead of Asp332 is similar to AZIs, the inactive homologs of ODCs. However, in AZIs only Asp332 is substituted by Glu, whereas other four residues are mostly conserved (Figure 7). of AZI as classified on the basis of conserved key amino acid residues was found to be a functional ODC [67]. In contrast to this, ODC from Aedes aegypti is found to be enzymatically non- functional [46]. Thus, mutations in and around active site, ranging from substitution of one residue to substitution of fourteen residues in single polypeptide may cause enzyme inactivation. In T. nigroviridis, 11 residues are altered in the active site whereas in mammals 4 residues are altered to convert a functional ODC to a nonfunctional homolog [46]. However, in Drosophila melanogaster, though all 18 key residues of active site are conserved, but a single mutation of Asp332Tyr hinders dimer formation in ODC in addition to cofactor and substrate binding, which makes it a nonfunctional ODC. Furthermore, in case of TbODC and other functional ODCs, Tyr331 contributes to form an aromatic zipper responsible for complementary packing in two monomers [53,54]. EhODC sequence and structural comparison with ODC homologs The substitution of active site residues at His296 and AZI is an inactive homolog of ODC which has lost decarbox- ylation activity due to mutation of critical residues in the active site [24,25,26,46,62]. However, they are important in mammals as they are responsible for antizyme down regulations, thus regulate the ODC activity in cell system [21]. In this study, we have Figure 6. Multiple sequence alignment of ornithine decarboxylase and its homologues antizyme inhibitor to determine the conservation of sequence and mutation of active site and substrate binding residues. Circles indicate the residues important for enzymatic activity. Numbering is according to EhODC. doi:10.1371/journal.pone.0053397.g006 Figure 6. Multiple sequence alignment of ornithine decarboxylase and its homologues antizyme inhibitor to determine the conservation of sequence and mutation of active site and substrate binding residues. Circles indicate the residues important for enzymatic activity. Numbering is according to EhODC. doi:10.1371/journal.pone.0053397.g006 January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 8 Crystal Structure of ODC from E. histolytica Figure 7. Sequence analysis of ODC and antizyme inhibitor, comparing the active site residues of ODC/AZI from various organisms. Abbreviation denoted: Cf for cofactor binding; Bs salt bridge formation; S substrate binding residues; If dimer interface residues; Di important for dimer formation. Species with the name of protein are shown on left side. Colour indication: Violet columns signifies the mutation in AZI; Orange columns signifies the mutated residues in E. histolytica ODC which are similar to AZI; Gray shows the unique mutations in EhODC which is neither conserved in ODC nor in AZI; Blue indicate the mutation in EhODC which are rarely found in AZI and functional ODC; Olive color point out the mutations in EhODC which are similar to some ODC. Sequence analysis and numbering has been done according to EhODC. Residues which are not conserved are shown by single letter, the conserved residues are indicted by – and D indicates the deleted amino acids. % identity indicates the identity of EhODC sequence with other homologous ODC sequences [46]. doi:10.1371/journal.pone.0053397.g007 Figure 7. Sequence analysis of ODC and antizyme inhibitor, comparing the active site residues of ODC/AZI from various organisms. Abbreviation denoted: Cf for cofactor binding; Bs salt bridge formation; S substrate binding residues; If dimer interface residues; Di important for dimer formation. Species with the name of protein are shown on left side. EhODC sequence and structural comparison with ODC homologs In AZI, this is mutated to Ser rendering a loose contact between monomers [61]. But in EhODC, same residue is substituted by aromatic amino acid Phe305 that is expected to perform same job in aromatic zipper. The mutation of Tyr to Phe is also reported in Plasmodium falciparum, Leishmania donovani and Glycine max ODCs (Figure 7, Figure 7) [46,66]. The evolutionary relationship of ODC and AZI can be evaluated by considering the root of phylogenic tree which connects the branch of both homologs. Evidences indicated that both the homologs are from same subfamily and have evolved and AZI genes have accumulated mutations in key residues that are important for ODC activity. In Petromyzon marinus, the homologue January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org January 2013 \| Volume 8 \| Issue 1 \| e53397 9 Crystal Structure of ODC from E. histolytica Figure 8. Active sites comparison of functional ODC, antizyme inhibitor and EhODC. A) Human ODC active site residues colored in blue. B) EhODC active site residues identical to human ODC colored blue, residues identical to AZI colored green and unique to EhODC colored red. C) AZI interface region showing residues identical to human ODC in blue and those are mutated colored green. doi:10.1371/journal.pone.0053397.g008 Figure 8. Active sites comparison of functional ODC, antizyme inhibitor and EhODC. A) Human ODC active site residues colored in blue. B) EhODC active site residues identical to human ODC colored blue, residues identical to AZI colored green and unique to EhODC colored red. C) AZI interface region showing residues identical to human ODC in blue and those are mutated colored green. doi:10.1371/journal.pone.0053397.g008 that contributes to dimerization. Though, same residues i.e. Lys169, Asp364, Asp134, and Lys294 are conserved in AZI (mouse), still residues do not approach to form the salt bridge [61]. Furthermore, AZI is inefficient to bind to PLP consequently unable to carry out decarboxylation reaction. The structure of AZI (mouse) reveals that the active site is too wide to make suitable pocket for substrate and PLP binding. However, EhODC binds to PLP and catalyzes decarboxylation of L-ornithine and relatively less active as compared to other active ODCs. It is interesting to note that though EhODC possesses similar property with other ODC on the basis of structure and function, it shares some similarity with AZI based on amino acid sequence. EhODC sequence and structural comparison with ODC homologs Firstly, the substrate binding residue Asp332 (HsODC) is conserved in functional ODCs where in EhODC same residue is altered to Glu306 and Glu is well conserved in AZI. Secondly, PLP binding residue Ala is altered to Ser in EhODC and such alternation is reported in AZI of Danio rerio, Tetraodon nigroviridis and Anolis crolinensis. Thirdly, EhODC possesses unique mutations at His296 and Asn335 those are neither reported in any functional ODC or AZI. Such alternation of critical residues particularly in protozoa provides the evidences of adaptive evolution of ODC. AZI dependent ODC regulation is only reported in higher organisms and absent from lower organisms. Even such regulation is not reported in protozoa till date. However, it can be hypothesized that ODC in protozoa takes the modification towards AZI though it functions less efficently as an active ODC and its function as AZI needs to be investigated. diverged according to their function. In phylogenetic tree, the group of AZI and ODC make different clusters according to the sequence alignment. Interestingly, EhODC is clustering to the ODC group just beneath the Aedes aegypti which is a nonfunctional ODC due to His197Asn and Asp332Arg substitutions as shown in Figure 7 and Figure 9. ODC of Aedes aezypti, being a non- functional ODC, represents the border line of functional ODCs and nonfunctional AZI. EhODC, the enzyme with low catalytic efficiency is found to be more evolutionarily related to nonfunc- tional ODC of Aedes aezypti and AZIs. These evidences from sequence alignment and phylogeny profile of EhODC allow us to establish the fact that during the course of evolution it gained DFMO resistance by acquiring critical alternation in its sequence similar to both functional ODCs and non-functional AZI. Though, the evolutionary changes in the sequence also influenced its catalytic efficiency. However, the possibility of additional biological role of EhODC such as antizyme inhibitory activity needs to be investigated. References 9. Pegg AE (2006) Regulation of ornithine decarboxylase. J Biol Chem 281: 14529–14532. 1. Rosas-Arreguı´n P, Arteaga-Nieto P, Reynoso-Orozco R, Villago´mez-Castro JC, Sabanero-Lo´pez M, et al. (2008) Bursera fagaroides, effect of an ethanolic extract on ornithine decarboxylase (ODC) activity in vitro and on the growth of Entamoeba histolytica. Exp Parasitol 119: 398–402. 10. Evageliou NF, Hogarty MD (2009) Disrupting polyamine homeostasis as a therapeutic strategy for neuroblastoma. Clin Cancer Res 15: 5956–5961. 2. Lo´pez-Vallejo F, Castillo R, Ye´pez-Mulia L, Medina-Franco JL (2011) Benzotriazoles and indazoles are scaffolds with biological activity against Entamoeba histolytica. J Biomol Screen 16: 862–868. 2. Lo´pez-Vallejo F, Castillo R, Ye´pez-Mulia L, Medina-Franco JL (2011) Benzotriazoles and indazoles are scaffolds with biological activity against Entamoeba histolytica. J Biomol Screen 16: 862–868. 11. Oredsson SM (2003) Polyamine dependence of normal cell-cycle progression. Biochem Soc Trans 31: 366–370. 12. Fleidervish IA, Libman L, Katz E, Gutnick MJ (2008) Endogenous polyamines regulate cortical neuronal excitability by blocking voltage-gated Na+ channels. Proc Natl Acad Sci U S A 105: 18994–18999. y J 3. Tanyuksel M, Petri WA Jr (2003) Laboratory diagnosis of amebiasis. Clin Microbiol Rev 16: 713–729. y 3. Tanyuksel M, Petri WA Jr (2003) Laboratory diagnosis of amebiasis. Clin Microbiol Rev 16: 713–729. 13. Krause T, Lu¨ersen K, Wrenger C, Gilberger TW, Mu¨ller S, et al. (2000) The ornithine decarboxylase domain of the bifunctional ornithine decarboxylase/S- adenosylmethionine decarboxylase of Plasmodium falciparum: recombinant expres- sion and catalytic properties of two different constructs. Biochem J 352: 287– 292. 4. Bansal D, Sehgal R, Chawla Y, Mahajan RC, Malla N (2004) In vitro activity of antiamoebic drugs against clinical isolates of Entamoeba histolytica and Entamoeba dispar. Ann Clin Microbiol Antimicrob 3: 27. 4. Bansal D, Sehgal R, Chawla Y, Mahajan RC, Malla N (2004) In vitro activity of antiamoebic drugs against clinical isolates of Entamoeba histolytica and Entamoeba dispar. Ann Clin Microbiol Antimicrob 3: 27. p 5. Heby O, Roberts SC, Ullman B (2003) Polyamine biosynthetic enzymes as drug targets in parasitic protozoa. Biochem Soc Trans 31: 415–419. 14. Palanimurugan R, Scheel H, Hofmann K, Dohmen RJ (2004) Polyamines regulate their synthesis by inducing expression and blocking degradation of ODC antizyme. EMBO J 23: 4857–4867. 6. Gillin FD, Reiner DS, McCann PP (1984) Inhibition of growth of Giardia lamblia by a-difluoromethylornithine, a specific inhibitor of polyamine biosynthesis. J Protozool 31: 161–163. 15. Conclusion In the present report, we successfully determined the 3D structure of EhODC to elucidate its intricate active site architec- ture that made it DFMO insensitive. Further on the basis of sequence analysis, we unveiled many unique characteristics of EhODC that show similarity with both functional ODC and non- functional AZI. EhODC exists as a dimer like other functional ODCs and in contrast AZI is monomer in solution due to weaker interaction between two monomers. Interface of EhODC shows 45 contacts and 16 hydrogen bonds in addition to salt bridges which stabilize the dimer. As studied in AZI structure (mouse AZI), only 43 contacts and 15 hydrogen bonds are reported which lack salt bridge formation and make interface less interactive as compared to ODC [61]. Structure of EhODC at 2.8 A˚ revealed two salt bridges between Lys157-Asp238 at a distance 2.9 A˚ and Asp122- Arg277 at 3.2 A˚ . Same salt bridges are also reported in HsODC Our study will facilitate to investigate the molecular evolution of ODCs and AZI. It also suggests additional functional properties for EhODC such as it may also play a role similar to that of AZI in E. histolytica. Additionally, availability of EhODC crystal structure will be helpful in development of structure based anti-amoebiasis drugs. January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 10 Crystal Structure of ODC from E. histolytica Figure 9. Phylogenetic relationship of EhODC with antizyme inhibitor and ODC. Sequence of ODC and their evolutionary related homologous were retrieved from various sources. Antizyme inhibitor of Homo sapiens (BAA23593.1), Nomascus leucogenys (XP_003256127.1), Macaca mulatta (XP_002805501.1), Mus musculus (AAB87464.1), Rattus norvegicus (BAA23594.1), Monodelphis domestica (XP_001369332.1), Xenopus laevis (NP_001087584.1), Danio rerio (BAB84695.1), Tetraodon nigroviridis (ENSTNIT00000008148.1), Anolis carolinensis (XP_003219500.1), Gallus gallus (NP_001008729.1), Ornithorhynchus anatinus (XP_001506230.1), Canis familiaris (XP_849306.1), Bos Taurus (NP_001076080.1), Loxodonta africana (XP_003408472.1). Ornithine decarboxylase sequence from Aedes aegypti (EAT48998.1), Entamoeba histolytica (AAX35675.1), Plasmodium falciparum (AAF14518.1), Leishmania donovani (AAA29259.1), Datura stramonium (CAA61121.1), Solanum lycopersicum (NP_001234616.1), Glycine max (CAD91349.1), Chlamydomonas reinhardtii (CAE46409.1), Monodelphis domestica (XP_001371947.1), Bos Taurus (AAA92339.1), Macaca mulatta (NP_001185615.1), Homo sapiens (NP_002530.1), Mus musculus (NP_038642.2), Anolis carolinensis (XP_003215471.1), Trypanosoma brucei (AAA30219.1), Xenopus laevis (CAA39760.1). doi:10.1371/journal.pone.0053397.g009 Figure 9. Phylogenetic relationship of EhODC with antizyme inhibitor and ODC. Sequence of ODC and their evolutionary related homologous were retrieved from various sources. Acknowledgments The authors thank Macromolecular Crystallographic Unit (MCU), IIC, IIT Roorkee for providing protein expression, purification, data collection Accession number Structure factors and final refined atomic coordinates for EhODC have been deposited in the Protein Data Bank (http:// www.rcsb.org) with accession number 4AIB. Author Contributions Conceived and designed the experiments: Preeti S. Tomar RM. Performed the experiments: Preeti S. Tapas PK S. Tomar. Analyzed the data: Preeti S. Tapas RM PK S. Tomar. Contributed reagents/materials/analysis tools: S. Tomar RM PK. Wrote the paper: Preeti S. Tapas S. Tomar. Conclusion Antizyme inhibitor of Homo sapiens (BAA23593.1), Nomascus leucogenys (XP_003256127.1), Macaca mulatta (XP_002805501.1), Mus musculus (AAB87464.1), Rattus norvegicus (BAA23594.1), Monodelphis domestica (XP_001369332.1), Xenopus laevis (NP_001087584.1), Danio rerio (BAB84695.1), Tetraodon nigroviridis (ENSTNIT00000008148.1), Anolis carolinensis (XP_003219500.1), Gallus gallus (NP_001008729.1), Ornithorhynchus anatinus (XP_001506230.1), Canis familiaris (XP_849306.1), Bos Taurus (NP_001076080.1), Loxodonta africana (XP_003408472.1). Ornithine decarboxylase sequence from Aedes aegypti (EAT48998.1), Entamoeba histolytica (AAX35675.1), Plasmodium falciparum (AAF14518.1), Leishmania donovani (AAA29259.1), Datura stramonium (CAA61121.1), Solanum lycopersicum (NP_001234616.1), Glycine max (CAD91349.1), Chlamydomonas reinhardtii (CAE46409.1), Monodelphis domestica (XP_001371947.1), Bos Taurus (AAA92339.1), Macaca mulatta (NP_001185615.1), Homo sapiens (NP_002530.1), Mus musculus (NP_038642.2), Anolis carolinensis (XP_003215471.1), Trypanosoma brucei (AAA30219.1), Xenopus laevis (CAA39760.1). doi:10 1371/journal pone 0053397 g009 and computational facilities. We also thank Sonali Dhindwal for her help in data collection and refinement. and computational facilities. We also thank Sonali Dhindwal for her help in data collection and refinement. and computational facilities. We also thank Sonali Dhindwal for her help in data collection and refinement. Crystal Structure of ODC from E. histolytica DeLano WL (2002) The PyMol molecular graphics system. San Carlos, CA, , USA: DeLano. Scientific. Available: http://www.pymol.org. 49. DeLano WL (2002) The PyMol molecular graphics system. San Carlos, CA, , USA D L S i tifi A il bl htt // l USA: DeLano. Scientific. Available: http://www.pymol.org. 50. Grishin NV, Phillips MA, Goldsmith EJ (1995) Modeling of th 50. Grishin NV, Phillips MA, Goldsmith EJ (1995) Modeling of the spatial structure of eukaryotic ornithine decarboxylases. Protein Sci 4: 1291–1304. 24. Murakami Y, Ichiba T, Matsufuji S, Hayashi S (1996) Cloning of antizyme inhibitor, a highly homologous protein to ornithine decarboxylase. J Biol Chem 271: 3340–3342. of eukaryotic ornithine decarboxylases. Protein Sci 4: 1291–13 51. Sandmeier E, Hale TI, Christen P (1994) Multiple evolutionary origin of Pyridoxal-59-dependent amino acid decarboxylases. Eur J Biochem 221: 997– 1002. 25. Murakami Y, Suzuki J, Samejima K, Kikuchi K, Hascilowicz T, et al. (2009) The change of antizyme inhibitor expression and its possible role during mammalian cell cycle. Exp Cell Res 315: 2301–2311. 52. Almrud JJ, Oliveira MA, Kern AD, Grishin NV, Phillips MA, et al. (2000) Crystal structure of human ornithine decarboxylase at 2.1 A˚ resolution: structural insights to antizyme binding. J Mol Biol 295: 7–16. y p 26. Kahana C (2009) Regulation of cellular polyamine levels and cellular proliferation by antizyme and antizyme inhibitor. Essays Biochem 46: 47–61. 53. Kern AD, Oliveira MA, Coffino P, Hackert ML (1999) Structure of mammalian ornithine decarboxylase at 1.6 A˚ resolution: stereochemical implications of PLP dependent amino acid decarboxylases. Structure 7: 567–581. p y y y y 27. Anderson IJ, Loftus BJ (2005) Entamoeba histolytica: observations on metabolism based on the genome sequence. Exp Parasitol 110: 173–177. based on the genome sequence. Exp Parasitol 110: 173–177. 28. Bakker-Grunwald T, Martin JB, Klein G (1995) Characterization of glycogen and amino acid pool of Entamoeba histolytica by 13C-NMR spectroscopy. J Eukaryot Microbiol 42: 346–349. 54. Grishin NV, Osterman AL, Brooks HB, Phillips MA, Goldsmith EJ (1999) X-ray structure of ornithine decarboxylase from Trypanosoma brucei: the native structure and the structure in complex with alpha-difluoromethylornithine. Biochemistry 38: 15174–15184. 29. Jhingran A, Padmanabhan PK, Singh S, Anamika K, Bakre AA, et al. (2008) Characterization of Entamoeba histolytica ornithine decarboxylase-like enzyme. PLoS Negl Trop Dis 2: e115. doi:10.1371/journal.pntd.0000115. 55. Linding R, Jensen LJ, Diella F, Bork P, Gibson TJ, et al. (2003) Protein disorder prediction: implications for structural proteomics. Structure 11: 1453–1459. Crystal Structure of ODC from E. histolytica 30. Poulin R, Ackermann LB, Bey P, Pegg AE (1992) Mechanism of the irreversible inactivation of mouse ornithine decarboxylase by a-difluoromethylornithine. Characterization of sequences at the inhibitor and coenzyme binding sites. J Biol Chem 267: 150–158. 56. Linding R, Russell RB, Neduva V, Gibson TJ (2003) GlobPlot: Exploring protein sequences for globularity and disorder. Nucleic Acids Res 31: 3701– 3708. 57. Kern A, Oliveira MA, Chang NL, Ernst SR, Carroll DW, et al. (1996) Crystallization of a mammalian ornithine decarboxylase. Proteins 24: 266–268. 31. Pegg AE, Madhubala R, Kameji T, Bergeron RJ (1988) Control of ornithine decarboxylase activity in alpha-difluoromethylornithine-resistant L1210 cells by polyamines and synthetic analogues. J Biol Chem 263: 11008–11014. 58. Ghoda L, Van Daalen Wetters T, Macrae M, Ascherman D, Coffino P (1989) Prevention of rapid intracellular degradation of ODC by a carboxyl-terminal truncation. Science 243: 1493–1495. 32. Iten M, Matovu E, Brun R, Kaminsky R (1995) Innate lack of susceptibility of Ugandan Trypanosoma brucei rhodesiense to DL-alpha-difluoromethylornithine (DFMO). Trop Med Parasitol 46: 190–194. 59. Krissinel E, Henrick K (2007) Inference of macromolecular assemblies from crystalline state. J Mol Biol 372: 774–797. 33. Iten M, Mett H, Evans A, Enyaru JC, Brun R, et al. (1997) Alterations in ornithine decarboxylase characteristics account for tolerance of Trypanosoma brucei rhodesiense to D,L-alpha-difluoromethylornithine. Antimicrob Agents Che- mother 41: 1922–1925. 60. Osterman AL, Kinch LN, Grishin NV, Phillips MA (1995) Acidic residues important for substrate binding and cofactor reactivity in eukaryotic ornithine decarboxylase identified by alanine scanning mutagenesis. J Biol Chem 270: 11797–11802. 34. Arteaga-Nieto P, Villago´mez-Castro JC, Calvo-Me´ndez C, Lo´pez-Romero E (1996) Partial purification and characterization of ornithine decarboxylase from Entamoeba histolytica. Int J Parasitol 26: 253–260. 61. Albeck S, Dym O, Unger T, Snapir Z, Bercovich Z, et al. (2008) Crystallographic and biochemical studies revealing the structural basis for antizyme inhibitor function. Protein Sci 17:793–802. y J 35. Arteaga-Nieto P, Lopez-Romero E, Teran-Figueroa Y, Cano-Canchola C, Luna Arias JP, et al. (2002) Entamoeba histolytica: purification and characterization of ornithine decarboxylase. Exp Parasitol 101: 215–222. 62. Hascilowicz T, Murai N, Matsufuji S, Murakami Y (2002) Regulation of ornithine decarboxylase by antizymes and antizyme inhibitor in zebrafish (Danio rerio). Biochim Biophys Acta 1578: 21–28. y p 36. Preeti, Tapas S, Kumar P, Madhubala R, Tomar S (2012) Biochemical, mutational and in silico structural evidence for a functional dimeric form of the ornithine decarboxylase from Entamoeba histolytica. Crystal Structure of ODC from E. histolytica 17. Sakata K, Kashiwagi K, Igarashi K (2000) Properties of a polyamine transporter regulated by antizyme. Biochem J 347: 297–303. 42. Adams PD, Afonine PV, Bunko´czi G, Chen VB, Davis IW, et al. (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr 66: 213–221. g y y J 18. Fujita K, Murakami S, Hayashi S (1982) A macromolecular inhibitor of the antizyme to ornithine decarboxylase. Biochem J 204: 647–653. g y y J 18. Fujita K, Murakami S, Hayashi S (1982) A macromolecular i 43. Vagin AA, Steiner RA, Lebedev AA, Potterton L, McNicholas S, et al. (2004) REFMAC5 dictionary: organization of prior chemical knowledge and guidelines for its use. Acta Crystallogr D Biol Crystallogr 60: 2184–2195. y y J 19. Kitani T, Fujisawa H (1989) Purification and characterization of antizyme inhibitor of ornithine decarboxylase from rat liver. Biochim Biophy Acta 991: 44–49. for its use. Acta Crystallogr D Biol Crystallogr 60: 2184–2195. 44. Emsley P, Cowtan K (2004) Coot: model-building tools for mole 44. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 20. Liu YC, Hsu DH, Huang CL, Liu YL, Liu GY, et al. (2011) Determinants of the differential antizyme-binding affinity of ornithine decarboxylase. PLoS One 6: e26835. Acta Crystallogr D Biol Crystallogr 60: 2126–2132. 45. Luthy R, Bowie JU, Eisenberg D (1992) Assessment of protein models with three- dimensional profiles. Nature 356: 83–85. 21. Nilsson J, Grahn B, Heby O (2000) Antizyme inhibitor is rapidly induced in growth-stimulated mouse fibroblasts and releases ornithine decarboxylase from antizyme suppression. Biochem J 346: 699–704. 46. Ivanov IP, Firth AE, Atkins JF (2010) Recurrent emergence of catalytically inactive ornithine decarboxylase homologous forms that likely have regulatory function. J Mol Evol 70: 289–302. 22. Cohavi O, Tobi D, Schreiber G (2009) Docking of antizyme to ornithine decarboxylase and antizyme inhibitor using experimental mutant and double mutant cycle data. J Mol Biol 390: 503–515. J 47. Higgins DG, Thompson JD, Gibson TJ (1996) Using CLUSTAL for multiple sequence alignments. Methods Enzymol 266: 383–402. 48. Sali A, Blundell TL (1993) Comparative protein modelling by satisfaction of spatial restraints. J Mol Biol 234: 779–815. 23. Keren-Paz A, Bercovich Z, Porat Z, Erez O, Brener O, et al. (2006) Overexpression of antizyme-inhibitor in NIH3T3 fibroblasts provides growth advantage through neutralization of antizyme functions. Oncogene 25: 5163– 5172. p J 49. References Li X, Coffino P (1992) Regulated degradation of ornithine decarboxylase requires interaction with the polyamine-inducible protein antizyme. Mol Cell Biol 12: 3556–3562. 15. Li X, Coffino P (1992) Regulated degradation of ornithine decarboxylase requires interaction with the polyamine-inducible protein antizyme. Mol Cell Biol 12: 3556–3562. 7. Heby O, Persson L, Rentala M (2007) Targeting the polyamine biosynthetic enzymes: a promising approach to therapy of African sleeping sickness, Chagas disease, and leishmaniasis. Amino Acids 33: 359–366. 8. Birkholtz LM, Williams M, Niemand J, Louw AI, Persson L, et al. (2011) Polyamine homoeostasis as a drug target in pathogenic protozoa: peculiarities and possibilities. Biochem J 438: 229–244. 16. Murakami Y, Matsufuji S, Kameji T, Hayashi S, Igarashi K, et al. (1992) Ornithine decarboxylase is degraded by the 26S proteasome without ubiquitination. Nature 360: 597–599. 8. Birkholtz LM, Williams M, Niemand J, Louw AI, Persson L, et al. (2011) Polyamine homoeostasis as a drug target in pathogenic protozoa: peculiarities and possibilities. Biochem J 438: 229–244. January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 January 2013 \| Volume 8 \| Issue 1 \| e53397 11 PLOS ONE \| www.plosone.org Crystal Structure of ODC from E. histolytica Crystal Structure of ODC from E. histolytica PLoS Negl Trop Dis 6: e1559. 63. Tsirka SE, Turck CW, Coffino P (1993) Multiple active conformers of mouse ornithine decarboxylase. Biochem J 293: 289–295. 64. Osterman AL, Brooks HB, Jackson L, Abbott JJ, Phillips MA (1999) Lysine-69 plays a key role in catalysis by ornithine decarboxylase through acceleration of the Schiff base formation, decarboxylation, and product release steps. Biochemistry 38: 11814–11826. y y g p 37. Badolo L, Berlaimont V, Helson-Cambier M, Hanocq M, Dubois J (1999) Simple and rapid enzymatic assay of ornithine decarboxylase activity. Talanta 48: 127–134. 38. Otwinowski Z, Minor W (1997) Processing of X-ray Diffraction Data Collected in Oscillation Mode. Methods Enzymol 276: 307–326. y 65. Myers DP, Jackson LK, Ipe VG, Murphy GE, Phillips MA (2001) Long-range interactions in the dimer interface of ornithine decarboxylase are important for enzyme function. Biochemistry 40: 13230–13236. 39. Vagin A, Teplyakov A (1997) MOLREP: an automated program for molecular replacement. J Appl Cryst 30: 1022–1025. 66. Jackson LK, Brooks HB, Myers DP, Phillips MA (2003) Ornithine decarboxylase promotes catalysis by binding the carboxylate in a buried pocket containing phenylalanine 397. Biochemistry 42: 2933–2940. 40. Dufe VT, Ingner D, Heby O, Khomutov AR, Persson L (2007) A structural insight into the inhibition of human and Leishmania donovani ornithine decarboxylases by 1-amino-oxy-3-aminopropane. Biochem J 2007, 405: 261– 268. 67. Ivanov IP, Loughran G, Atkins JF (2008) uORFs with unusual translational start codons autoregulate expression of eukaryotic ornithine decarboxylase homologs. Proc Natl Acad Sci USA 105:10079–10084 41. Brunger AT (2007) Version 1.2 of the Crystallography and NMR System. Nature Protocols 2: 2728–2733. January 2013 \| Volume 8 \| Issue 1 \| e53397 PLOS ONE \| www.plosone.org 12
W4387029659.txt	https://masujournal.org/store_file/archive/73-2-2-101-104.pdf	en		HEAT UNIT EFFICIENCY IN PIGEONPEA	Madras Agricultural Journal	1,986	cc-by	1	https://doi.org/10.29321/MAJ.10.A02237
https://openalex.org/W2038699893	https://europepmc.org/articles/pmc3669024?pdf=render	English	null	Pharmacological Characterization of a 5-HT1-Type Serotonin Receptor in the Red Flour Beetle, Tribolium castaneum	PloS one	2,013	cc-by	9,864	Abstract Serotonin (5-hydroxytryptamine, 5-HT) is known for its key role in modulating diverse physiological processes and behaviors by binding various 5-HT receptors. However, a lack of pharmacological knowledge impedes studies on invertebrate 5-HT receptors. Moreover, pharmacological information is urgently needed in order to establish a reliable classification system for invertebrate 5-HT receptors. In this study we report on the molecular cloning and pharmacological characterization of a 5- HT1 receptor from the red flour beetle, Tribolium castaneum (Trica5-HT1). The Trica5-HT1 receptor encoding cDNA shows considerable sequence similarity with members of the 5-HT1 receptor class. Real time PCR showed high expression in the brain (without optic lobes) and the optic lobes, consistent with the role of 5-HT as neurotransmitter. Activation of Trica5-HT1 in mammalian cells decreased NKH-477-stimulated cyclic AMP levels in a dose-dependent manner, but did not influence intracellular Ca2+ signaling. We studied the pharmacological profile of the 5-HT1 receptor and demonstrated that a- methylserotonin, 5-methoxytryptamine and 5-carboxamidotryptamine acted as agonists. Prazosin, methiothepin and methysergide were the most potent antagonists and showed competitive inhibition in presence of 5-HT. This study offers important information on a 5-HT1 receptor from T. castaneum facilitating functional research of 5-HT receptors in insects and other invertebrates. The pharmacological profiles may contribute to establish a reliable classification scheme for invertebrate 5-HT receptors. Citation: Vleugels R, Lenaerts C, Baumann A, Vanden Broeck J, Verlinden H (2013) Pharmacological Characterization of a 5-HT1-Type Serotonin Receptor in the Red Flour Beetle, Tribolium castaneum. PLoS ONE 8(5): e65052. doi:10.1371/journal.pone.0065052 Editor: Hubert Vaudry, University of Rouen, France Editor: Hubert Vaudry, University of Rouen, France Received January 17, 2013; Accepted April 22, 2013; Published May 31, 2013 Copyright: 2013 Vleugels et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: The authors thank the Interuniversity Attraction Poles program (Belgian Science Policy Grant P7/40) and the KU Leuven Research Foundation (GOA/11/ 02) for financial support. RV and CL were supported by the Agency for Innovation by Science and Technology (IWT). HV was supported by the Research Foundation of Flanders(FWO). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: jozef.vandenbroeck@bio.kuleuven.be Pharmacological Characterization of a 5-HT1-Type Serotonin Receptor in the Red Flour Beetle, Tribolium castaneum Rut Vleugels1, Cynthia Lenaerts1, Arnd Baumann2, Jozef Vanden Broeck1*, Heleen Verlinden1 1 Department of Animal Physiology and Neurobiology, Zoological Institute, KU Leuven, Leuven, Belgium, 2 Institute of Complex Systems (ICS4), Research Centre Ju¨lich, Ju¨lich, Germany Introduction signaling pathways in all eukaryote organisms. The vertebrate 5- HT GPCRs (5-HT1,2,4–7) were classified based on their sequence similarities, gene organization, downstream signaling pathways and pharmacological properties [17–20]. 5-HT1 and 5-HT5 receptors couple preferentially to Gi/o proteins and thus inhibit cyclic AMP (cAMP) synthesis. 5-HT2 receptors couple preferen- tially to Gq/11 proteins which cause an increase in cytosolic Ca2+ levels. 5-HT4, 5-HT6 and 5-HT7 receptors are all preferentially linked to Gs proteins and promote cAMP production. Biogenic amines play an important role in very diverse physiological processes and behaviors. In insects, the six major biogenic amines are serotonin (5-hydroxytryptamine, 5-HT), dopamine, tyramine, octopamine, acetylcholine and histamine. 5-HT is known to play a crucial role in the regulation of important processes in most, if not all, animal phyla. Alterations in 5-HT neurotransmission are associated with several human disorders, such as migraine, depression, schizophrenia and anxiety [1]. Normal human processes, such as sleep, mood level, appetite, sexual activity and learning abilities are also modulated by 5-HT. In insects, 5-HT signaling controls nutrition [2], modulation of heart rate [3], secretory processes in the salivary gland [4–7], development [8], circadian rhythms and sleep regulation [9,10], aggression [11], behavioral gregarization in locusts [12,13], phototactic behavior in honeybees [14] and learning and memory in fruit flies [15,16]. The major classes, 5-HT1, 5-HT2, and 5-HT6, probably evolved from a primordial 5-HT receptor over 750 million years ago. The 5-HT5 and 5-HT7 receptor classes diverged from 5-HT1 650 to 700 million years ago [21,22]. Since these events even predate the estimated divergence of protostomes and deutero- stomes about 600 to 650 million years ago [23], the invertebrate and vertebrate serotonergic systems are believed to possess roughly the same main receptor classes [21,24]. However, evolution allowed further differentiation in various subtypes within each main class, and these subtypes are believed to have evolved independently in vertebrates and invertebrates [21,25]. Thus far, only four types of 5-HT receptors are characterized in insects, namely 5-HT1A, 5-HT1B, 5-HT2, and 5-HT7 [8,14,26–29]. Classification of invertebrate 5-HT receptors according to the To mediate such a variety of processes, 5-HT acts through multiple 5-HT receptor types. In vertebrates, 5-HT receptors are divided in seven main classes. Six of these are G protein-coupled receptors (GPCRs) and the sole exception, 5-HT3, is a ligand- gated ion channel. qRT-PCR Study of Transcript Levels y For determination of expression levels of the receptor, tissues from sexually mature T. castaneum were dissected in phosphate buffered saline (PBS) (NaCl 137 mM, KCl 2.7 mM, Na2HPO4 10 mM, KH2PO4 1.76 mM; pH 7.2) and snap-frozen in liquid nitrogen. For all samples, tissues of at least fifteen animals were pooled. Tissues were homogenized and RNA was extracted using the RNAqueous Micro Kit (Ambion) according to the protocol recommended by the kit. The protocol included an additional DNase treatment to digest remaining DNA. Total RNA was reverse transcribed into cDNA using SuperScriptIII reverse transcriptase (Invitrogen) as recommended by the manufacturer, and diluted ten-fold prior to use. Transcript levels were quantified using the Fast Sybr Green assay kit (Applied Biosystems) in a StepOne Plus detection system (ABI Prism, Applied Biosystems). Primers (sense primer 59-GCCCTCTGGCTGGGCTAT-39 and antisense primer 59-CGGGTTGAAGATCGTGTAAATGA-39) (Sigma-Aldrich) were used in final concentrations of 500 nM. Other conditions were as recommended by the manufacturer. Reactions were run in duplicate and incubated for 2 min at 50uC, followed by 10 min at 95uC, followed by 40 cycles of [15 s at 95uC and 1 min at 60uC]. The specificity of the PCR products was assessed generating a dissociation curve (95uC for 15 s, 60uC for 1 min, and increase in temperature in 0.7uC increments from 60uC to 95uC). Agarose gel electrophoresis of the PCR products confirmed the presence of a single band of the expected size and sequencing confirmed their identity. The relative quantity of target cDNA was quantified using the DDCT-method including normal- ization to a calibrator on all PCR plates and an endogenous control. From a list of seven housekeeping genes (Table S1; [34]), the combination of genes for this endogenous control was determined using GeNorm [35]. Expression was most stable for RPs3 (ribosomal protein 3) and RPs18 with respect to sex and tissue and these transcripts were thus selected for further use as endogenous controls (results not shown). In the present study, we will discuss the characterization of a 5- HT1 receptor from the red flour beetle, Tribolium castaneum (Trica5- HT1). The genome of T. castaneum has been completely sequenced (Tribolium Genome Sequencing Consortium) [30]. Use of annota- tion software led to the discovery of twenty Tribolium genes that code for putative biogenic amine GPCRs. All these proteins have orthologues in Drosophila melanogaster and Apis mellifera [31]. Animal Rearing Conditions Beetles were reared in a dark incubator at 30uC on wheat flour and brewer’s yeast in Petri dishes. Adult beetles were sexed based on the presence of a small patch of short bristles on the inside of the first pair of legs in males, according to the T. castaneum rearing protocol (http://bru.gmprc.ksu.edu/proj/tribolium/wrangle.asp) [32]. qRT-PCR Study of Transcript Levels Four receptors could be assigned as putative 5-HT receptors based on sequence similarity to 5-HT receptors of D. melanogaster and A. mellifera [8,14,26–28,31]. After cloning the Trica5-HT1 cDNA, we analyzed its tissue distribution by quantitative real-time PCR (qRT-PCR) and elucidated its downstream signaling pathway. In cells expressing Trica5-HT1, application of 5-HT inhibited NKH- 477 (a water-soluble forskolin analog) stimulated cAMP synthesis. The pharmacological profile of the receptor was established after application of several synthetic 5-HT receptor agonists and antagonists. These results will facilitate future in vivo studies aiming to unravel the contribution of individual 5-HT receptors to the animals’ physiology and behavior. Pharmacology of a Beetle 5-HT1 Receptor Pharmacology of a Beetle 5-HT1 Receptor Cycle Sequencing Kit (Applied Biosystems). Bacterial cells known to contain the correct receptor insert were grown at large scale in 100 ml Luria–Bertani broth medium. The expression vectors were subsequently isolated from these cells using the EndoFree Plasmid Maxi Kit (Qiagen) according to the protocol recommended by the kit. existing vertebrate classes is mainly based on well conserved amino acid sequences and activated second-messenger systems for 5-HT receptors across all species. On the other hand, pharmacological profiles from insect (and other invertebrate) receptors seem to differ significantly from those of vertebrates [25]. Since only very few data about pharmacological properties of invertebrate 5-HT receptors are available, there is no general classification system for invertebrate 5-HT receptors based on pharmacological properties yet. This explains the need for detailed pharmacological studies on insect and other invertebrate 5-HT receptors. Cloning of Trica5-ht1 and Construction of pcTrica5-ht1 Expression Vector The full length sequence encoding the receptor was amplified with PCR using whole body T. castaneum cDNA, Taq polymerase (REDTaqHReadyMixTMPCR Reaction Mix, Sigma-Aldrich), and 10 mM of sense primer 59-ATGGGGACAGTAAA- TAATCCCTCCTG-39 and antisense primer 59-TTATC- TAATTTTGCCCGAGCGG-39 (Sigma-Aldrich). Primers were designed based on sequences available in Beetlebase (Tcas_3.0; http://www.beetlebase.org/) [33] released by the Human Ge- nome Sequencing Center. PCR started with initial denaturation for 2 min at 95uC, followed by 35 cycles of [30 s at 94uC, 30 s at 62uC, 2 min at 68uC], followed by final elongation for 2 min at 68uC. Amplification products were run on a 1.2% agarose gel and purified with the GenEluteTM Gel extraction Kit (Sigma–Aldrich). The DNA fragments were cloned into a pcDNA3.1/V5-His- TOPOHTA expression vector via TA TOPO cloning (Invitrogen) and transformed into One Shot TOP10 chemically competent Escherichia coli cells (Invitrogen). Bacteria were grown according to the protocol recommended by the kit. Plasmids were isolated via the GenEluteTM HP Plasmid Miniprep kit (Sigma-Aldrich) and DNA sequences were determined by means of the ABI PRISM 3130 Genetic Analyzer (Applied Biosystems) following the protocol outlined in the ABI PRISM BigDye Terminator Ready Reaction Materials and Methods Animal Rearing Conditions Introduction GPCRs play a vital role in many essential PLOS ONE \| www.plosone.org May 2013 \| Volume 8 \| Issue 5 \| e65052 1 Aequorin-luminescence Assay Cloning and Sequence Analysis of Trica5-ht1 A cDNA fragment encoding a 5-HT1 receptor from T. castaneum was amplified by PCR. The open reading frame of Trica5-ht1 contains 1,644 nucleotides (Figure S1) encoding the Trica5-HT1 protein of 547 amino acids (Figure 1) with a calculated molecular weight of 60.8 kDa. Transmembrane topology prediction revealed the presence of seven putative transmembrane domains (TM1-7), characteristic of all GPCRs. Consensus motifs for N-linked glycosylation (N-x-[S/T]) are found in the extracellular N- terminus, and consensus sites for phosphorylation by protein kinase C (PKC) ([S/T-x-[R/K]) are located within the third intracellular loop (Figure 1). The C7.69 residue (numbering according to the Ballesteros-Weinstein system [40]) in the intracellular C-terminus is a putative palmitoylation site. The large third intracellular loop and the short intracellular C-terminal region are consistent with other known 5-HT1 and biogenic amine receptors that couple via Gi. Other typical biogenic amine and 5- HT receptor characteristics are present as well. The DRY tripeptide (D3.49R3.50Y3.51) located in the second intracellular loop is the key to the conformational changes necessary for receptor activation [41]. The combination of the D3.32 in TM3 with the conserved W7.40 in TM7 is considered a unique fingerprint for biogenic amine and trace amine GPCRs. The charged D3.32 residue is thought to interact with the protonated amine moiety of amine ligands [42,43]. As in other 5-HT receptors, Trica5-HT1 typically has a conserved group of hydrophobic amino acids (W3.28, F5.47, W6.48, F6.51, F6.52, W7.40, Y7.43) that form the hydrophobic ligand-binding pocket within the tertiary structure. This binding pocket may be stabilized by a disulfide bridge formed between C2.55 in TM2 and C3.25 in extracellular loop 1 [42,44,45]. In addition, the consensus sequence of non-peptide receptors in TM6, F6.44-x-x-x-W6.48-x-P6.50, is followed by a pair of Phe residues (F6.51 and F6.52) unique to aminergic receptors. Also the N7.49P7.50-x-x-Y7.53 motif in TM7 is conserved, which may participate in agonist mediated receptor sequestration and resensitization [46]. Aequorin luminescence Assay The transfected CHO cells were detached with phosphate buffered saline (PBS) containing 0.2% EDTA and collected in DMEM/F-12 (without phenol red, with L-glutamine and 10 mM HEPES) (Gibco). The amount of viable cells was determined using the NucleoCounter NC-100+TM (Chemometic). Cells were pelleted for 4 min at 800 rpm at room temperature and resuspended in BSA-medium (DMEM/F12 without phenol red, with L-glutamine and 10 mM HEPES, supplemented with 0.1% bovine serum albumin) to a concentration of 56106 cells/ml. Aequorin-luminescence Assay Coelenterazine H (Invitrogen) was added to a final concentration of 5 mM, and cells were gently shaken for 4 h at room temperature in the dark. After a 10-fold dilution in BSA-medium, cells were incubated another 30 min. The pharmacological ligands were dissolved in BSA-medium. For agonists, 50 ml containing the final ligand concentration was added to appropriate wells of a 96-well plate. For antagonists, 25 ml of the antagonist solution was supplemented with 25 ml of a 5-HT solution. Receptor activity was measured as the light emission after adding 50 ml of the cell suspension. Light emission was measured for 30 s using a Mithras LB940 (Berthold Technologies). Subsequently, cells were lysed with Triton X-100 (0.1% in BSA-medium) and light emission was recorded for another 8 s. BSA-medium was used as a negative control. Light emission from each well was calculated relative to the total response (ligand+Triton X-100) using the output file of Mikrowin2000 software (Mikrotek). Further analysis was done in Graphpad Prism 5. Drugs The pharmacological ligands 3-hydroxytyramine (dopamine) hydrochloride, 5-carboxamidotryptamine maleate (5-CT), 5-HT hydrochloride (5-HT), 5-methoxytryptamine (5-MT), (6)-8-hy- droxy-2-(dipropylamino)tetralin hydrobromide (8-OH-DPAT), a- methylserotonin maleate (am-5-HT), (+)-butaclamol hydrochlo- ride, ketanserin (+)-tartrate, methiothepin mesylate, methysergide maleate, mianserin hydrochloride, prazosin hydrochloride, DL- octopamine hydrochloride, SB-269970 hydrochloride, tyramine hydrochloride, WAY-100635 maleate (WAY = N-{2-[4-(2-meth- oxyphenyl)-1-piperazinyl]-ethyl}-N-(2-pyridinyl) cyclohexanecar- boxamide), and yohimbine hydrochloride were purchased from Sigma-Aldrich. Pharmacology of a Beetle 5-HT1 Receptor Pharmacology of a Beetle 5-HT1 Receptor dispensed. After incubation for 3–4 h in a CO2 incubator at 37uC, 100 ml of SteadyLite Plus (Perkin-Elmer) was added to each well and the plate was gently shaken for 15 min in the dark. Light emission was measured for 5 s per well using a Mithras LB940 (Berthold Technologies). Medium containing IBMX was used as a negative control. Data were analyzed as described for CHO cells. expressing apoaequorin (CHO-PAM28) [36], or both apoaequorin and Ga16 (CHO-WTA11 cells) [37] and are thus still used as additional antibiotics in the appropriate screens. For CHO cells, the medium was supplemented with 10% fetal calf serum (inactivated at 65uC) (Sigma-Aldrich). For HEK293 cells, the medium was supplemented with 2% Ultroser G serum substitute (Pall Life Sciences). expressing apoaequorin (CHO-PAM28) [36], or both apoaequorin and Ga16 (CHO-WTA11 cells) [37] and are thus still used as additional antibiotics in the appropriate screens. For CHO cells, the medium was supplemented with 10% fetal calf serum (inactivated at 65uC) (Sigma-Aldrich). For HEK293 cells, the medium was supplemented with 2% Ultroser G serum substitute (Pall Life Sciences). ( ) Cells were cultured in vitro as monolayers at 37uC, 5% CO2 and high relative humidity, and subcultivated twice a week. For transfection of CHO cells, 2.5 ml Opti-MEMHI (Invitrogen), 5 mg plasmid DNA and 12.5 ml PlusTMreagent (Invitrogen) were mixed, stored at room temperature for 5 min and next repleted with 30 ml LipofectamineTMLTX (Invitrogen). After 30 min incubation at room temperature, this transfection mixture was added dropwise to the cells together with 5 ml medium. HEK293 cells were cotransfected with receptor construct (4 mg) and CRE-luciferase construct (2 mg), consisting of the open reading frame of the reporter gene, luciferase, downstream of a multimerized cAMP- response-element (CRE6x) [38,39]. All transfections were per- formed in T75 flasks with a confluency of about 60 percent. Cells were grown overnight, after which 15 ml of medium was added for an additional overnight incubation period. Cell Culture and Transfection General binding studies were performed in Chinese hamster ovary (CHO) WTA11 cells stably coexpressing apoaequorin and the promiscuous Ga16. This allowed us to measure the pharma- cology independent of the downstream signaling. CHO-PAM28 cells stably expressing apoaequorin, but not the promiscuous Ga16, and human embryonic kidney cells (HEK) 293 cells were used to measure downstream signaling via Ca2+ and cAMP, respectively. CHO cell lines were provided by Prof. Marc Parmentier (University of Brussels, Belgium) and Dr. Michel Detheux (Euroscreen S.A., Belgium). HEK293 cells were a gift from Prof. Arnd Baumann (Research Centre Ju¨lich, Germany. All cells were cultured in Dulbecco’s Modified Eagles Medium nutrient mixture F12-Ham (DMEM/F12) (Invitrogen) supple- mented with 1% penicillin/streptomycin (10000 units/ml penicil- lin and 10 mg/ml streptomycin in 0.9% NaCl) (Invitrogen) to prevent bacterial contamination of gram-positive and gram- negative bacteria, respectively. For CHO-WTA11 cells, 250 mg/ ml zeocin (Invitrogen) was added and for CHO-PAM28 cells, 5 mg/ml puromycin dihydrochloride (Invitrogen) was added. Puromycin and zeocin were initially used to select for cells stably May 2013 \| Volume 8 \| Issue 5 \| e65052 May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org 2 Luciferase Reporter-gene Assay Cotransfected HEK293 cells were detached and the amount of viable cells was determined as described for CHO cells. Cells were pelleted for 4 min at 800 rpm at room temperature and resuspended in DMEM/F12 medium (without phenol red, with L-glutamine and 10 mM HEPES) containing 200 mM 3-isobutyl- 1-methylxanthine (IBMX, Sigma-Aldrich) to a concentration of 106 cells/ml. The pharmacological ligands were dissolved in DMEM/F12 medium without phenol red containing 200 mM IBMX. The water-soluble forskolin analog NKH-477 was added to a concentration of 20 mM to measure Trica5-ht1-mediated effects on cellular cAMP signaling. In each well of a 96-well plate, 50 ml of ligand suspension and 50 ml of cell suspension were BLASTx (http://blast.ncbi.nlm.nih.gov/blast/) searches indi- cate similarities of the receptor to other insect 5-HT1 receptors (Pea5-HT1, 60% identity; Am5-HT1A, 55% identity; Dm5-HT1A, 46% identity; Dm5-HT1B, 42% identity). When compared to the computationally predicted sequence in Beetlebase [31], a stretch of May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org 3 Figure 1. Amino acid sequence alignment of T. castaneum 5-HT1 sequence (Trica5-HT1, GenBank accession no. KC196076). Alignment against sequences of orthologous receptors from Apis mellifera (Am5-HT1A, no. CBI75449), Drosophila melanogaster (Dm5-HT1A, no. CAA77570 and Dm5-HT1B, no. CAA77571), and Periplaneta americana (Pea5-HT1, no. CAX65666). Identical residues between the receptors are shown as white characters against black background. Conservatively substituted residues are shaded. Putative transmembrane domains are indicated by grey bars (TM1-7). Dots indicate putative phosphorylation sites for PKC, and stars indicate putative N-linked glycosylation sites. Inverted triangles indicate differences between the current sequence derived from cloned cDNA and the annotated sequence from Beetlebase. doi:10.1371/journal.pone.0065052.g001 Pharmacology of a Beetle 5-HT1 Receptor Pharmacology of a Beetle 5-HT1 Receptor 75 bp is missing in our cloned receptor cDNA sequence. Since this stretch is flanked by splice sites and not present in cDNA sequences of other species, it is presumed to be an intron. After multiple sequencing runs on different cDNA samples, also five single base mismatches (three silent and two missense mutations) were found between the amplified and the annotated sequence. Furthermore, two consecutive nucleotides were different, resulting in an amino acid change (Figure S1). Figure 1. Amino acid sequence alignment of T. castaneum 5-HT1 sequence (Trica5-HT1, GenBank accession no. KC196076). Alignment against sequences of orthologous receptors from Apis mellifera (Am5-HT1A, no. CBI75449), Drosophila melanogaster (Dm5-HT1A, no. CAA77570 and Dm5-HT1B, no. CAA77571), and Periplaneta americana (Pea5-HT1, no. CAX65666). Transcript Level Study An initial tissue distribution screen of Trica5-HT1 mRNA in sexually mature beetles was performed using qRT-PCR (Figure 2A). Transcripts of the Trica5-ht1 gene were detected in considerable amounts in the head and the gut, consistent with the prevalence of 5-HT in central nervous system (CNS) and the digestive tract of most animal species. However, expression in the head appeared significantly more abundant. Differences between males and females were not significant. In the fat body and the reproductive tract of sexually mature beetles, only very low levels of receptor transcript were observed. When transcript levels in the brain (without the optic lobes) and the optic lobes were measured separately, expression in both tissues seemed significantly higher than in the gut (Figure 2B). Moreover, expression in the brain (without the optic lobes) was 3.5 times higher compared to the optic lobes. Although there are indications for 5-HT receptor expression in the salivary gland of some insect species, we detected almost no Trica5-HT1 transcripts in the salivary gland of adult beetles. Potential antagonists were tested by simultaneously applying 5- HT (100 nM) and a high dose of antagonist (100 mM) to Trica5- HT1 expressing cells (Figure 4A). In addition, the dose- dependence of the antagonistic effects were measured with antagonist concentrations ranging from 10 nM to 1 mM (Figure 4B). From these experiments, we can conclude that ketanserin and mianserin, two selective antagonists of mammalian 5-HT2, and butaclamol, a dopamine receptor antagonist, displayed no detectable inhibition of 5-HT induced responses in cells expressing Trica5-HT1. Mianserin and butaclamol even seemed to have some agonistic effects at high concentration (100 mM). On the other hand prazosin, a selective a1-adrenergic receptor antagonist in mammals, was shown to be the most potent antagonist. It decreased the effect of 5-HT on the receptor in a dose-dependent manner with a half maximal inhibitory concen- tration (IC50) of 1.39 mM (logIC50 = 25.8660.18, mean 6 SEM). The IC50 values for all antagonists are shown in Table 2. Methiothepin and methysergide, two non-selective antagonists of mammalian 5-HT receptors also showed dose-dependent inhibi- tion. Moderate inhibition was achieved with SB-269970, a selective antagonist of mammalian 5-HT7. The level of inhibition induced by WAY-100635 didn’t drop below 50% and was variable which might be due to complex effects of the compound on the receptor and/or other targets in the assay. Transcript Level Study Also yohimbine, known to behave as both agonist and antagonist on some mammalian 5- HT-receptors, showed only about 30% of inhibition, even at Pharmacology of a Beetle 5-HT1 Receptor Pharmacology of a Beetle 5-HT1 Receptor set at 100% (Figure 3B). Surprisingly, the most potent agonist was the mammalian 5-HT2 receptor agonist, am-5-HT, with an EC50 value of 10.74 mM (logEC50 = 24.9760.18, mean 6 SEM). However, this is more than 100-fold less potent than 5-HT itself. Also the 5-HT analog, 5-CT, a selective agonist for mammalian 5- HT1 and 5-HT7 receptors, and the non-selective 5-HT receptor agonist, 5-MT, acted as partial agonists in a dose-dependent manner. 5-CT was more potent but had a lower efficacy than 5- MT. The EC50 values for all agonists are shown in Table 1. 8-OH- DPAT, a partial and selective agonist for mammalian 5-HT1 and 5-HT7 receptors evoked responses only at concentrations $100 mM. The biogenic amines, dopamine, octopamine and tyramine did not generate any detectable responses at concentra- tions #100 mM (results not shown). Luciferase Reporter-gene Assay Identical residues between the receptors are shown as white characters against black background. Conservatively substituted residues are shaded. Putative transmembrane domains are indicated by grey bars (TM1-7). Dots indicate putative phosphorylation sites for PKC, and stars indicate putative N-linked glycosylation sites. Inverted triangles indicate differences between the current sequence derived from cloned cDNA and the annotated sequence from Beetlebase. doi:10.1371/journal.pone.0065052.g001 Figure 1. Amino acid sequence alignment of T. castaneum 5-HT1 sequence (Trica5-HT1, GenBank accession no. KC196076). Alignment against sequences of orthologous receptors from Apis mellifera (Am5-HT1A, no. CBI75449), Drosophila melanogaster (Dm5-HT1A, no. CAA77570 and Dm5-HT1B, no. CAA77571), and Periplaneta americana (Pea5-HT1, no. CAX65666). Identical residues between the receptors are shown as white characters against black background. Conservatively substituted residues are shaded. Putative transmembrane domains are indicated by grey bars (TM1-7). Dots indicate putative phosphorylation sites for PKC, and stars indicate putative N-linked glycosylation sites. Inverted triangles indicate differences between the current sequence derived from cloned cDNA and the annotated sequence from Beetlebase. doi:10.1371/journal.pone.0065052.g001 Figure 1. Amino acid sequence alignment of T. castaneum 5-HT1 sequence (Trica5-HT1, GenBank accession no. KC196076). Alignment against sequences of orthologous receptors from Apis mellifera (Am5-HT1A, no. CBI75449), Drosophila melanogaster (Dm5-HT1A, no. CAA77570 and Dm5-HT1B, no. CAA77571), and Periplaneta americana (Pea5-HT1, no. CAX65666). Identical residues between the receptors are shown as white characters against black background. Conservatively substituted residues are shaded. Putative transmembrane domains are indicated by grey bars (TM1-7). Dots indicate putative phosphorylation sites for PKC, and stars indicate putative N-linked glycosylation sites. Inverted triangles indicate differences between the current sequence derived from cloned cDNA and the annotated sequence from Beetlebase. doi:10.1371/journal.pone.0065052.g001 75 bp is missing in our cloned receptor cDNA sequence. Since this stretch is flanked by splice sites and not present in cDNA sequences of other species, it is presumed to be an intron. After multiple sequencing runs on different cDNA samples, also five single base mismatches (three silent and two missense mutations) were found between the amplified and the annotated sequence. Furthermore, two consecutive nucleotides were different, resulting in an amino acid change (Figure S1). May 2013 \| Volume 8 \| Issue 5 \| e65052 May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 4 Pharmacological Characterization and Downstream Signaling Properties of Trica5-HT1 In order to pharmacologically characterize Trica5-HT1, we used CHO-WTA11 cells stably expressing apoaequorin and the promiscuous Ga16 subunit. No 5-HT evoked signal was observed in non-transfected cells or in cells transfected with an empty vector (results not shown). Significant responses were obtained when cells expressing Trica5-HT1 were incubated with 100 mM 5-HT or the 5-HT receptor agonists am-5-HT, 5-CT, 5-MT and 8-OH-DPAT (Figure 3A). The dose-response relationship for 5-HT and these synthetic agonists was examined at concentrations ranging from 1 pM to 1 mM (Figure 3B). The resulting sigmoidal dose-response curve of 5-HT shows receptor activation in a dose-dependent and saturable manner. Half-maximal activation (EC50) was achieved at 5-HT concentrations of 95.15 nM (logEC50 = 27.0160.043, mean 6 SEM). The maximal response was attained at 5-HT concentrations of $10 mM. Since the efficacy achieved by any agonist depends on the number of receptors expressed, we measured a dose-response curve for 5-HT in every experiment and normalized all agonist effects to the maximum 5-HT response, Figure 2. Expression profile of transcripts encoding Trica5-HT1 in sexually mature beetles. The data represent mean values of (A) three independent samples of 306heads, 506guts, 206fat body and 506reproductive system; and (B) three independent samples of 15 beetles each; run in duplicate 6 SEM, normalized relative to RPs3 (ribosomal protein 3) and RPs18 transcript levels. Statistically significant differences are indicated by asterisks above the bars (p#0.05) (Kruskal-Wallis, IBM SPSS Statistics 20). Abbreviations: FB, fat body; RS, reproductive system; Brain, brain without the optic lobes; OL, optic lobes; SalGl, salivary glands. doi:10.1371/journal.pone.0065052.g002 Figure 2. Expression profile of transcripts encoding Trica5-HT1 in sexually mature beetles. The data represent mean values of (A) three independent samples of 306heads, 506guts, 206fat body and 506reproductive system; and (B) three independent samples of 15 beetles each; run in duplicate 6 SEM, normalized relative to RPs3 (ribosomal protein 3) and RPs18 transcript levels. Statistically significant differences are indicated by asterisks above the bars (p#0.05) (Kruskal-Wallis, IBM SPSS Statistics 20). Abbreviations: FB, fat body; RS, reproductive system; Brain, brain without the optic lobes; OL, optic lobes; SalGl, salivary glands. doi:10.1371/journal.pone.0065052.g002 May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org 5 Pharmacology of a Beetle 5-HT1 Receptor Figure 3. Effect of various agonists on Trica5-HT1 in CHO-WTA11 cells. Pharmacological Characterization and Downstream Signaling Properties of Trica5-HT1 The nature of inhibition obtained with the three most potent antagonists: prazosin, methiothepin and methysergide was further examined by studying the dose- response relationship of 5-HT in presence of different concentra- tions of antagonist (10 nM to 1 mM) (Figure 5). The higher the concentration of antagonist, the higher was the resulting EC50 value for 5-HT activity. Since their efficacy didn’t change, these compounds behaved as truly competitive antagonists. We used the Gaddum/Schild plot to compare the inhibitory potencies of the antagonists by their pA2 values (i.e. the logarithm of the concentration of antagonist that doubles the amount of 5-HT required for obtaining the same effect) [47]. The pA2 values (6 SEM) for prazosin, methiothepin and methysergide were, respec- tively, 7.18 (60.13), 6.17 (60.11) and 5.96 (60.21), confirming that prazosin has the highest affinity, followed by methiothepin and methysergide. concentrations up to 100 mM. The nature of inhibition obtained with the three most potent antagonists: prazosin, methiothepin and methysergide was further examined by studying the dose- response relationship of 5-HT in presence of different concentra- tions of antagonist (10 nM to 1 mM) (Figure 5). The higher the concentration of antagonist, the higher was the resulting EC50 value for 5-HT activity. Since their efficacy didn’t change, these compounds behaved as truly competitive antagonists. We used the Gaddum/Schild plot to compare the inhibitory potencies of the antagonists by their pA2 values (i.e. the logarithm of the concentration of antagonist that doubles the amount of 5-HT required for obtaining the same effect) [47]. The pA2 values (6 SEM) for prazosin, methiothepin and methysergide were, respec- tively, 7.18 (60.13), 6.17 (60.11) and 5.96 (60.21), confirming that prazosin has the highest affinity, followed by methiothepin and methysergide. Pharmacological Characterization and Downstream Signaling Properties of Trica5-HT1 (A) Receptor activation after stimulation with 100 mM of agonist, shown as the percentage of activation achieved with 100 mM of 5-HT (set at 100%). (B) Dose-dependent activation of Trica5-HT1 with synthetic 5-HT receptor agonists, shown as the percentage of activation achieved with 1 mM 5-HT (maximum response = 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Data represent the mean 6 SEM of (A) three independent measurements (each performed in duplicate) and (B) seven independent measurements (each performed in duplicate) for 5-HT, or four independent measurements (three performed in triplicate, one performed in duplicate) for the synthetic agonists am-5-HT, 5-MT, 5-CT and 8-OH-DPAT. doi:10.1371/journal.pone.0065052.g003 Figure 3. Effect of various agonists on Trica5-HT1 in CHO-WTA11 cells. (A) Receptor activation after stimulation with 100 mM of agonist, shown as the percentage of activation achieved with 100 mM of 5-HT (set at 100%). (B) Dose-dependent activation of Trica5-HT1 with synthetic 5-HT receptor agonists, shown as the percentage of activation achieved with 1 mM 5-HT (maximum response = 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Data represent the mean 6 SEM of (A) three independent measurements (each performed in duplicate) and (B) seven independent measurements (each performed in duplicate) for 5-HT, or four independent measurements (three performed in triplicate, one performed in duplicate) for the synthetic agonists am-5-HT, 5-MT, 5-CT and 8-OH-DPAT. doi:10.1371/journal.pone.0065052.g003 concluded that Trica5-HT1 does not couple via Gq to the Ca2+ signaling pathway. In HEK293 cells, effects on the cAMP level were examined for 5-HT concentrations ranging from 1 pM to 100 mM. Relative high variation in the data can be explained by assay based variation since cells were not counted when dispensed. Basal levels of cAMP did not significantly change in cells transfected with an empty vector. In Trica5-HT1 expressing cells, a dose-dependent decrease in intracellular, NKH-477 stimulated cAMP levels was registered (Figure 6). Half maximal reduction of cAMP was observed at 82.7 nM 5-HT (logIC50 = 27.0860.30), mean 6 SEM). Trica5-HT1 thus inhibits the cAMP production, probably via the Gi protein. Maximal attenuation of cAMP synthesis (640%) was attained with 5-HT concentrations $10 mM. concentrations up to 100 mM. doi:10.1371/journal.pone.0065052.t001 Discussion Additionaly, Am5- HT1A was found in regions known to be important in visual information processing, such as the optic lobes. Based on behavioral tests 5-HT is assumed to reduce the positive phototactic behavior of honeybees [14]. In P. americana, antibodies directed against Pea5-HT1, stained some large somata in the pars intercerebralis [29]. Thus, expression of Trica5-HT1 in beetle brains may be due to expression in the mushroom bodies and/or cells of the pars intercerebralis. Receptor expression in the gut may indicate a role of 5-HT in the regulation of digestion or gut contraction. In humans, for example, as much as 95% of the total 5-HT content may reside in the intestine [56,57]. It is probable that a substantial release of 5-HT in other animal species, such as Tribolium, takes place in the digestive tract as well. In several insects, 5-HT immunoreactive nerve fibers have been localized in different parts of the intestinal tract [58–60]. 5-HT was also shown to modulate muscle contractions of the gut in several insects [61–66], although it may also act as a paracrine factor and, for example, provoke the release of other factors from neuroendocrine cells. Even then, high 5-HT levels in a given release organ do not necessarily co-incide with the local level of receptor expression, which might explain the relatively low 5-HT1 transcript levels in the gut compared to the brain. Although only low transcript levels were observed in the salivary glands of adult beetles, 5-HT also has been shown to be important in salivation in several insect species [4–7,67]. In the cockroach, 5-HT1 receptor expression was shown in the salivary gland by RT-PCR and Western blotting [29]. Low expression levels of Trica5-HT1 receptors in the salivary gland suggest that possible 5-HT effects on salivation are regulated by other 5-HT receptor subtypes. For example, in the blowfly, Calliphora vicina, salivary glands express 5-HT2 and 5-HT7 receptors [67]. To examine the downstream signaling pathway of Trica5-HT1, the receptor was expressed in CHO-PAM28 and HEK293 cells. Since no Ca2+ response was measured in CHO-PAM28 cells expressing Trica5-HT1, the receptor does not engage Gq and the PLC signaling pathway. In HEK293 cells, the receptor was cotransfected with a CRE-luciferase construct to detect changes in intracellular cAMP levels. 5-HT was found to decrease NKH-477- stimulated cAMP synthesis in a dose-dependent manner. In accordance with other 5-HT1 receptors, Trica5-HT1 couples to Gi/o proteins that impair adenlylate cyclase activity. Discussion CHO-PAM28 and HEK293 cells were used to determine the downstream signaling pathway of Trica5-HT1. No effect of 5-HT was observed in CHO-PAM28 cells transfected with empty vector or in cells transfected with the receptor. Therefore it can be In the present study, we have characterized Trica5-HT1, a 5- HT1 receptor of the red flour beetle, T. castaneum. The obtained sequence has considerable similarity with orthologous receptors from other invertebrates [14,27,29,48–53] and mammals [20,54]. The sequence contains typical characteristics of 5-HT1 receptors, such as a large third intracellular loop, a short C-terminal region, a DRY motif in the second intracellular loop as well as other conserved consensus sequences. Notably, differences between the cloned and annotated nucleotide sequences were uncovered. These included single base mismatches and an intron in the annotated open reading frame (Figure S1). So far, no introns have been reported in the coding regions of vertebrate genes encoding 5-HT1 receptors [55]. However, introns were found in the D. melanogaster genes encoding Dm5-HT1A and Dm5-HT1B [27]. Table 1. EC50 values of agonists for Trica5-HT1 receptor activation in CHO-WTA11 cells. agonist EC50 (mM) logEC50 (mean ± SEM) 5-HT 0.095 27.0160.043 am-5-HT 10.74 24.9760.18 5-CT 24.72 24.6160.17 5-MT 91.84 24.0460.12 8-OH-DPAT 551.0 23.2660.59 doi:10.1371/journal.pone.0065052.t001 Table 1. EC50 values of agonists for Trica5-HT1 receptor activation in CHO-WTA11 cells. When studying the Trica5-HT1 transcript levels with qRT- PCR, highest expression was observed in the brain (without optic lobes), followed by the optic lobes. More detailed localization studies in other insects also showed receptor expression in the optic May 2013 \| Volume 8 \| Issue 5 \| e65052 May 2013 \| Volume 8 \| Issue 5 \| e65052 6 Pharmacology of a Beetle 5-HT1 Receptor Figure 4. Effect of various antagonists on 5-HT mediated activation of Trica5-HT1 in CHO-WTA11 cells. (A) Effect of antagonists (100 mM) on 5-HT (100 nM) mediated receptor activation. Receptor activation is shown as the percentage of activation achieved with 100 nM of 5-HT (, EC50 value) (set at 100%). (B) Dose-dependent effect of 5-HT receptor antagonists on 5-HT (100 nM) mediated receptor activation. Receptor activation is shown as the percentage of activation achieved with 10 nM of antagonist (set at 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Discussion Data represent the mean 6 SEM of (A) three independent measurements (each performed in duplicate) and (B) two (prazosin and methysergide) or three (methiothepin and SB-269970) independent measurements (each performed in triplicate). doi:10.1371/journal.pone.0065052.g004 Figure 4. Effect of various antagonists on 5-HT mediated activation of Trica5-HT1 in CHO-WTA11 cells. (A) Effect of antagonists (100 mM) on 5-HT (100 nM) mediated receptor activation. Receptor activation is shown as the percentage of activation achieved with 100 nM of 5-HT (, EC50 value) (set at 100%). (B) Dose-dependent effect of 5-HT receptor antagonists on 5-HT (100 nM) mediated receptor activation. Receptor activation is shown as the percentage of activation achieved with 10 nM of antagonist (set at 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Data represent the mean 6 SEM of (A) three independent measurements (each performed in duplicate) and (B) two (prazosin and methysergide) or three (methiothepin and SB-269970) independent measurements (each performed in triplicate). doi:10.1371/journal.pone.0065052.g004 lobes as well as other parts of the brain [9,10,14,27,29]. In D. melanogaster, Dm5-HT1A and Dm5-HT1B receptors are expressed in the mushroom bodies [9,10], the region of the brain involved in learning and memory. Also in the mushroom bodies of A. mellifera, high levels of Am5-HT1A were found [14]. Additionaly, Am5- HT1A was found in regions known to be important in visual information processing, such as the optic lobes. Based on behavioral tests 5-HT is assumed to reduce the positive phototactic behavior of honeybees [14]. In P. americana, antibodies directed against Pea5-HT1, stained some large somata in the pars intercerebralis [29]. Thus, expression of Trica5-HT1 in beetle brains may be due to expression in the mushroom bodies and/or cells of the pars intercerebralis. Receptor expression in the gut may indicate a role of 5-HT in the regulation of digestion or gut contraction. In humans, for example, as much as 95% of the total 5-HT content may reside in the intestine [56,57]. It is probable that a substantial release of 5-HT in other animal species, such as Tribolium, takes lobes as well as other parts of the brain [9,10,14,27,29]. In D. melanogaster, Dm5-HT1A and Dm5-HT1B receptors are expressed in the mushroom bodies [9,10], the region of the brain involved in learning and memory. Also in the mushroom bodies of A. mellifera, high levels of Am5-HT1A were found [14]. doi:10.1371/journal.pone.0065052.t002 Discussion am-5-HT, 5-CT and 5-MT. The synthetic agonists were more than a 100-fold less potent than 5-HT and seemed to have a lower efficacy. However it is possible that the agonists did not reach their maximum response although high concentrations (1 mM) were already tested. Similar properties have been reported for other insect 5-HT1 receptors [14,29]. Only a very poor response was observed with 8-OH-DPAT, an agonist for mammalian 5-HT1 and 5-HT7 receptors. Also on A. mellifera and P. americana 5-HT1 receptors, 8-OH-DPAT acted as a poor agonist [14,29]. For the antagonists, no inhibition was measured upon application of butaclamol, ketanserin and mianserin. Although WAY-100635 is known as a potent and selective antagonist of mammalian 5-HT1A receptors, no clear dose- dependent effect was observed. In A. mellifera, however, WAY- 100635 acted as a partial antagonist of Am5-HT1 [14] and an effective inhibition of agonist stimulated Pea5-HT1 was observed [29]. A more effective inhibition was measured for the mammalian 5-HT7 receptor antagonist, SB-269970, but no information regarding the possible effects of this antagonist is known from receptors from other arthropods [14,27,29,52,53]. Besides 5-HT, three additional ligands caused dose-dependent activation of Trica5-HT1, i.e. am-5-HT, 5-CT and 5-MT. The synthetic agonists were more than a 100-fold less potent than 5-HT and seemed to have a lower efficacy. However it is possible that the agonists did not reach their maximum response although high concentrations (1 mM) were already tested. Similar properties have been reported for other insect 5-HT1 receptors [14,29]. Only a very poor response was observed with 8-OH-DPAT, an agonist for mammalian 5-HT1 and 5-HT7 receptors. Also on A. mellifera and P. americana 5-HT1 receptors, 8-OH-DPAT acted as a poor agonist [14,29]. For the antagonists, no inhibition was measured upon application of butaclamol, ketanserin and mianserin. Although WAY-100635 is known as a potent and selective antagonist of mammalian 5-HT1A receptors, no clear dose- dependent effect was observed. In A. mellifera, however, WAY- 100635 acted as a partial antagonist of Am5-HT1 [14] and an effective inhibition of agonist stimulated Pea5-HT1 was observed [29]. A more effective inhibition was measured for the mammalian 5-HT7 receptor antagonist, SB-269970, but no information regarding the possible effects of this antagonist is known from other arthropods. An effective dose-dependent inhibition was measured for prazosin, methiothepin and methysergide, although their IC50 values were in the micromolar range. Discussion However, when interpolating these experimental data to physiological processes in Tribolium, one must be aware of possible discrepancies between effects observed in cultured cell lines and intracellular processes occurring within the in vivo context of the organism. Table 2. IC50 values of antagonists for Trica5-HT1 receptor inhibition in CHO-WTA11 cells. antagonist IC50 (mM) logIC50 (mean ± SEM) prazosin 1.39 25.8960.18 methiothepin 16.38 24.7960.13 methysergide 33.97 24.4760.094 WAY-100635 no clear dose-dependence of inhibition butaclamol 204.1 23.6960.20 SB-269970 205.4 23.6960.25 ketanserin no inhibition yohimbin no inhibition mianserin no inhibition doi:10.1371/journal.pone.0065052.t002 Table 2. IC50 values of antagonists for Trica5-HT1 receptor inhibition in CHO-WTA11 cells. CHO-WTA11 cells were used to investigate the pharmacolog- ical characteristics of Trica5-HT1. Application of 5-HT to these cells resulted in dose-dependent receptor activities. The EC50 value of 95.15 nM is similar to values reported for 5-HT1 May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org May 2013 \| Volume 8 \| Issue 5 \| e65052 7 Pharmacology of a Beetle 5-HT1 Receptor Figure 5. Effect of 5-HT on Trica5-HT1 in presence of different concentrations of antagonist. Dose-dependent activation of Trica5-HT1 was measured in CHO-WTA11 cells with 1 nM –1 mM 5-HT in presence of 10 nM –1 mM antagonist: (A) prazosin, (B) methysergide and (C) methiothepin. Receptor activity is shown as the percentage of activation achieved with 1 mM of 5-HT (set at 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Data represent the mean 6 SEM of three independent measurements (each performed in duplicate) for prazosin and methiothepin and two independent measurements (each performed in duplicate) for methysergide. NA, no antagonist. doi:10.1371/journal.pone.0065052.g005 Figure 5. Effect of 5-HT on Trica5-HT1 in presence of different concentrations of antagonist. Dose-dependent activation of Trica5-HT1 was measured in CHO-WTA11 cells with 1 nM –1 mM 5-HT in presence of 10 nM –1 mM antagonist: (A) prazosin, (B) methysergide and (C) methiothepin. Receptor activity is shown as the percentage of activation achieved with 1 mM of 5-HT (set at 100%). Cells treated with BSA-medium only were used to define the basal level of 0%. Data represent the mean 6 SEM of three independent measurements (each performed in duplicate) for prazosin and methiothepin and two independent measurements (each performed in duplicate) for methysergide. NA, no antagonist. doi:10.1371/journal.pone.0065052.g005 receptors from other arthropods [14,27,29,52,53]. Besides 5-HT, three additional ligands caused dose-dependent activation of Trica5-HT1, i.e. References 15. Sitaraman D, Zars M, Laferriere H, Chen YC, Sable-Smith A, et al. (2008) Serotonin is necessary for place memory in Drosophila. Proc Natl Acad Sci U S A 105: 5579–5584. 1. Jones BJ, Blackburn TP (2002) The medical benefit of 5-HT research. Pharmacol Biochem Behav 71: 555–568. 1. Jones BJ, Blackburn TP (2002) The medical benefit of 5-HT research. Pharmacol Biochem Behav 71: 555–568. 2. Cohen RW, Friedman S, Waldbauer GP (1988) Physiological control of nutrient self selection in Heliothis zea larvae - the role of serotonin. J Insect Physiol 34: 935–940. 16. Sitaraman D, Laferriere H, Birman S, Zars T (2012) Serotonin is critical for rewarded olfactory short-term memory in Drosophila. J Neurogenet 26: 238–244. 3. Chiang RG, Chiang JA, Davey KG (1992) A sensory input inhibiting heart-rate in an insect, Rhodnius prolixus. Experientia 48: 1122–1125. 17. Hoyer D, Clarke DE, Fozard JR, Hartig PR, Martin GR, et al. (1994) International Union of Pharmacology classification of receptors for 5- hydroxytryptamine (Serotonin). Pharmacol Rev 46: 157–203. 4. Berridge MJ, Patel NG (1968) Insect salivary glands - Stimulation of fluid secretion by 5-hydroxytryptamine and adenosine-39,59-monophosphate. Science 162: 462–463. 18. Hoyer D, Hannon JP, Martin GR (2002) Molecular, pharmacological and functional diversity of 5-HT receptors. Pharmacol Biochem Behav 71: 533–554. 5. Trimmer BA (1985) Serotonin and the control of salivation in the blowfly Calliphora. J Exp Biol 114: 307–328. 19. Hannon J, Hoyer D (2008) Molecular biology of 5-HT receptors. Behav Brain Res 195: 198–213. 6. Novak MG, Ribeiro JMC, Hildebrand JG (1995) 5-Hydroxytryptamine in the salivary-glands of adult female Aedes aegypti and its role in regulation of salivation. J Exp Biol 198: 167–174. 20. Nichols DE, Nichols CD (2008) Serotonin receptors. Chem Rev 108: 1614– 1641. 21. Peroutka SJ, Howell TA (1994) The molecular evolution of G protein-coupled receptors - Focus on 5-hydroxytryptamine receptors. Neuropharmacol 33: 319– 324. 7. Walz B, Baumann O, Krach C, Baumann A, Blenau W (2006) The aminergic control of cockroach salivary glands. Arch Insect Biochem Physiol 62: 141–152. 22. Walker RJ, Brooks HL, Holden-Dye L (1996) Evolution and overview of classical transmitter molecules and their receptors. Parasitology 113: S3–S33. 8. Colas JF, Launay JM, Kellermann O, Rosay P, Maroteaux L (1995) Drosophila 5- HT2 serotonin receptor: coexpression with fushi-tarazu during segmentation. Proc Natl Acad Sci U S A 92: 5441–5445. 23. Acknowledgments The authors gratefully acknowledge Marc Parmentier (University of Brussels, Belgium) and Michel Detheux (Euroscreen S.A., Belgium) for providing both CHO cell lines. They also thank Dr. Kristel Vuerinckx for her help with dissections and cDNA sample preparations. receptor subtypes with specific pharmacological properties have evolved within the main classes of vertebrate receptors. Pharmacology of a Beetle 5-HT1 Receptor Pharmacology of a Beetle 5-HT1 Receptor Figure 6. Dose-dependent effect of 5-HT on intracellular cAMP levels in HEK293 cells. The effect of 5-HT (1 pM - 100 mM) on the luciferase bioluminescence in HEK293 cells expressing Trica5-HT1 due to changes in intracellular cAMP levels. Receptor activity is shown as the percentage of activation achieved with 10 mM of NKH-477 (set at 100%). Luciferase bioluminescence due to basal intracellular cAMP levels is set at 0%. The data represent the mean 6 SEM of four independent measurements (three performed in triplicate, one performed in duplicate). doi:10.1371/journal.pone.0065052.g006 characterization of all members constituting the invertebrate 5-HT receptor family are needed to establish a reliable classification scheme. Detailed pharmacological information for each 5-HT receptor subtype will also aid in functional in vivo studies and can be very useful for insect pest control. Author Contributions Conceived and designed the experiments: RV JVB AB HV. Performed the experiments: RV CL HV. Analyzed the data: RV CL HV. Contributed reagents/materials/analysis tools: AB JVB. Wrote the paper: RV AB JVB HV. During the last decades, knowledge about 5-HT (and other biogenic amine) receptors has increased. However, comprehensive data on the pharmacology of insect or other invertebrate 5-HT receptors is still missing. Unequivocal identification and extensive Conceived and designed the experiments: RV JVB AB HV. Performed the experiments: RV CL HV. Analyzed the data: RV CL HV. Contributed reagents/materials/analysis tools: AB JVB. Wrote the paper: RV AB JVB HV. (TIF) Table S1 Nucleotide sequences of primers for T. castaneum housekeeping genes. (DOCX) Figure 6. Dose-dependent effect of 5-HT on intracellular cAMP levels in HEK293 cells. The effect of 5-HT (1 pM - 100 mM) on the luciferase bioluminescence in HEK293 cells expressing Trica5-HT1 due to changes in intracellular cAMP levels. Receptor activity is shown as the percentage of activation achieved with 10 mM of NKH-477 (set at 100%). Luciferase bioluminescence due to basal intracellular cAMP levels is set at 0%. The data represent the mean 6 SEM of four independent measurements (three performed in triplicate, one performed in duplicate). d i 10 1371/j l 0065052 006 Discussion All three are known to be non-selective antagonists for mammalian 5-HT receptors, including 5-HT1 receptors. At high concentrations ($1 mM), they completely inhibited activation of Trica5-HT1 by 100 nM of 5-HT. When comparing dose-response curves of 5-HT in the presence of different concentrations of antagonist, all three antagonists behaved as competitive antagonists. Both the IC50 and the pA2 values indicated that prazosin was the most potent antagonist, followed by methiothepin and methysergide. In conclusion, our data support previous findings that primary structures and signaling properties are well conserved between vertebrate and invertebrate receptors, yet pharmacological prop- erties can differ significantly between both phyla, and even between different invertebrate species. The differences in phar- macological profiles of vertebrate and invertebrate receptors may be due to their large evolutionary distance. Selection during evolution most likely was based on receptor properties such as ligand binding and G protein coupling, but not on conservation of recognition sites for man-made, synthetic ligands. Furthermore May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org 8 Supporting Information Figure S1 Nucleotide sequence of the T. castaneum 5- HT1 receptor sequence (Trica5-HT1, accession no. KC196076). Inverted triangles indicate differences resulting in another amino acid between the current sequence derived from cloned cDNA and the annotated sequence from Beetlebase. Diamonds indicate silent mutations. The arrow indicates the splice site where a stretch of 75 residues is present in the Beetlebase sequence. (TIF) Pharmacology of a Beetle 5-HT1 Receptor 47. Kenakin TP (1982) The schild regression in the process of receptor classification. Can J Physiol Pharmacol 60: 249–265. 29. Troppmann B, Balfanz S, Baumann A, Blenau W (2010) Inverse agonist and neutral antagonist actions of synthetic compounds at an insect 5-HT1 receptor. Br J Pharmacol 159: 1450–1462. 48. Sugamori KS, Sunahara RK, Guan HC, Bulloch AGM, Tensen CP, et al. (1993) Serotonin receptor cDNA cloned from Lymnaea stagnalis. Proc Natl Acad Sci U S A 90: 11–15. 30. Tribolium Genome Sequencing Consortium, Richards S, Gibbs RA, Weinstock GM, Brown SJ, et al. (2008) The genome of the model beetle and pest Tribolium castaneum. Nature 452: 949–955. 49. Olde B, McCombie WR (1997) Molecular cloning and functional expression of a serotonin receptor from Caenorhabditis elegans. J Mol Neurosci 8: 53–62. 31. Hauser F, Cazzamali G, Williamson M, Park Y, Li B, et al. (2008) A genome- wide inventory of neurohormone GPCRs in the red flour beetle Tribolium castaneum. Frontiers in Neuroendocrinology 29: 142–165. 50. Angers A, Storozhuk MY, Duchaine T, Castellucci VF, DesGroseillers L (1998) Cloning and functional expression of an Aplysia 5-HT receptor negatively coupled to adenylate cyclase. J Neurosci 18: 5586–5593. 32. Beeman RW, Haas S, Friesen K (2009) Beetle wrangling tips. (An introduction to the care and handling of Tribolium castaneum). URL: http://bru.gmprc.ksu. edu/proj/tribolium/wrangle.asp (Last modified on 21.08.2012). 51. Barbas D, Zappulla JP, Angers S, Bouvier M, Castellucci VF, et al. (2002) Functional characterization of a novel serotonin receptor (5-HTap2) expressed in the CNS of Aplysia californica. J Neurochem 80: 335–345. 33. Kim HS, Murphy T, Xia J, Caragea D, Park Y, et al. (2010) BeetleBase in 2010: revisions to provide comprehensive genomic information for Tribolium castaneum. Nucleic Acids Research 38: D437–D442. 52. Chen A, Holmes SP, Pietrantonio PV (2004) Molecular cloning and functional expression of a serotonin receptor from the Southern cattle tick, Boophilus microplus (Acari : Ixodidae). Insect Mol Biol 13: 45–54. 34. Lord JC, Hartzer K, Toutges M, Oppert B (2010) Evaluation of quantitative PCR reference genes for gene expression studies in Tribolium castaneum after fungal challenge. J Microbiol Methods 80: 219–221. 53. Spitzer N, Edwards DH, Baro DJ (2008) Conservation of structure, signaling and pharmacology between two serotonin receptor subtypes from decapod crustaceans, Panulirus interruptus and Procambarus clarkii. J Exp Biol 211: 92–105. 35. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. Pharmacology of a Beetle 5-HT1 Receptor (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3: RESEARCH0034. , p J p 54. Kroeze WK, Kristiansen K, Roth BL (2002) Molecular biology of serotonin receptors structure and function at the molecular level. Curr Top Med Chem 2: 507–528. 36. Torfs H, Detheux M, Oonk HB, Akerman KE, Poels J, et al. (2002) Analysis of C-terminally substituted tachykinin-like peptide agonists by means of aequorin- based luminescent assays for human and insect neurokinin receptors. Biochem Pharmacol 63: 1675–1682. 55. Gerhardt CC, vanHeerikhuizen H (1997) Functional characteristics of heterologously expressed 5-HT receptors. Eur J Pharmacol 334: 1–23. g y p p J 56. Gershon MD, Tack J (2007) The serotonin signaling system: From basic understanding to drug development-for functional GI disorders. Gastroenterol- ogy 132: 397–414. 37. Blanpain C, Migeotte I, Lee B, Vakili J, Doranz BJ, et al. (1999) CCR5 binds multiple CC-chemokines: MCP-3 acts as a natural antagonist. Blood 94: 1899– 1905. gy 57. Berger M, Gray JA, Roth BL (2009) The expanded biology of serotonin. Annu Rev Med 60: 355–366. 38. George SE, Bungay PJ, Naylor LH (1997) Evaluation of a CRE-directed luciferase reporter gene assay as an alternative to measuring cAMP accumulation. J Biomol Screen 2: 235–240. 58. Klemm N, Hustert R, Cantera R, Na¨ssel DR (1986) Neurons reactive to antibodies against serotonin in the stomatogastric nervous sytem and in the alimentary canal of locust and crickets (Orthoptera, Insecta). Neurosci 17: 247– 261. 39. Hearn MG, Ren Y, McBride EW, Reveillaud I, Beinborn M, et al. (2002) A Drosophila dopamine 2-like receptor: Molecular characterization and identifica- tion of multiple alternatively spliced variants. Proc Natl Acad Sci U S A 99: 14554–14559. 59. Pietrantonio PV, Jagge C, McDowell C (2001) Cloning and expression analysis of a 5HT7-like serotonin receptor cDNA from mosquito Aedes aegypti female excretory and respiratory systems. Insect Mol Biol 10: 357–369. 40. Ballesteros JA, Weinstein H (1995) Integrated methods for the construction of three-dimensional models and computational probing of structure-function relations in G protein-coupled receptors. Methods in Neurosciences 25: 366– 428. 60. Molaei G, Lange AB (2003) The association of serotonin with the alimentary canal of the African migratory locust, Locusta migratoria: distribution, physiology and pharmacological profile. J Insect Physiol 49: 1073–1082. 61. Davey KG (1962) The mode of action of the corpus cardiacum on the hind gut in Periplaneta americana. J Exp Biol 39: 319–324. References Ayala FJ, Rzhetsky A, Ayala FJ (1998) Origin of the metazoan phyla: Molecular clocks confirm paleontological estimates. Proc Natl Acad Sci U S A 95: 606–611. 9. Yuan Q, Lin F, Zheng X, Sehgal A (2005) Serotonin modulates circadian entrainment in Drosophila. Neuron 47: 115–127. 24. Hauser F, Cazzamali G, Williamson M, Blenau W, Grimmelikhuijzen CJ (2006) A review of neurohormone GPCRs present in the fruitfly Drosophila melanogaster and the honey bee Apis mellifera. Progress in Neurobiology 80: 1–19. 10. Yuan Q, Joiner WJ, Sehgal A (2006) A sleep-promoting role for the Drosophila serotonin receptor 1A. Curr Biol 16: 1051–1062. 11. Dierick HA, Greenspan RJ (2007) Serotonin and neuropeptide F have opposite modulatory effects on fly aggression. Nature Gen 39: 678–682. 25. Tierney AJ (2001) Structure and function of invertebrate 5-HT receptors: a review. Comp Biochem Physiol A Mol Integr Physiol 128: 791–804. 12. Anstey ML, Rogers SM, Ott SR, Burrows M, Simpson SJ (2009) Serotonin mediates behavioral gregarization underlying swarm formation in desert locusts. Science 323: 627–630. 26. Witz P, Amlaiky N, Plassat JL, Maroteaux L, Borrelli E, et al. (1990) Cloning and characterization of a Drosophila serotonin receptor that activates adenylate cyclase. Proc Natl Acad Sci U S A 87: 8940–8944. 13. Ott SR, Verlinden H, Rogers SM, Brighton CH, Quah PS, et al. (2012) Critical role for protein kinase A in the acquisition of gregarious behavior in the desert locust. Proc Natl Acad Sci U S A 109: E381–E387. 27. Saudou F, Boschert U, Amlaiky N, Plassat JL, Hen R (1992) A family of Drosophila serotonin receptors with distinct intracellular signalling properties and expression patterns. The EMBO Journal 11: 7–17. 14. Thamm M, Balfanz S, Scheiner R, Baumann A, Blenau W (2010) Characterization of the 5-HT1A receptor of the honeybee (Apis mellifera) and involvement of serotonin in phototactic behavior. Cell Mol Life Sci 67: 2467– 2479. 28. Schlenstedt J, Balfanz S, Baumann A, Blenau W (2006) Am5-HT7: molecular and pharmacological characterization of the first serotonin receptor of the honeybee (Apis mellifera). J Neurochem 98: 1985–1998. May 2013 \| Volume 8 \| Issue 5 \| e65052 May 2013 \| Volume 8 \| Issue 5 \| e65052 PLOS ONE \| www.plosone.org 9 Pharmacology of a Beetle 5-HT1 Receptor 41. Gether U (2013) Uncovering molecular mechanisms involved in activation of G protein-coupled receptors. Endocr Rev 21: 90–113. p J p 62. Freeman MA (1966) The effect of drugs on the alimentary canal of the African migratory locust, Locusta migratoria. Comp Biochem Physiol 17: 755–764. 42. Kristiansen K, Kroeze WK, Willins DL, Gelber EI, Savage JE, et al. (2000) A highly conserved aspartic acid (Asp-155) anchors the terminal amine moiety of tryptamines and is involved in membrane targeting of the 5-HT2A serotonin receptor but does not participate in activation via a ‘‘salt-bridge disruption’’ mechanism. J Pharmacol Exp Ther 293: 735–746. 63. Cook BJ, Holman GM (2013) Comparative pharmacological properties of muscle function in the foregut and the hindgut of the cockroach Leucophaea maderae. Comp Biochem Physiol C 61: 291–295. 64. Huddart H, Oldfield AC (1982) Spontaneous activity of foregut and hindgut visceral muscle of the locust, Locusta migratoria - II. The effect of biogenic amines. Comp Biochem Physiol C 73: 303–311. 43. Strader CD, Fong TM, Graziano MP, Tota MR (1995) The family of G- protein-coupled receptors. Faseb J 9: 745–754. 44. Noda K, Saad Y, Graham RM, Karnik SS (1994) The high-affinity state of the beta 2-adrenergic receptor requires unique interaction between conserved and nonconserved extracellular loop cysteine. J Biol Chem 269: 6743–6752. 65. Banner SE, Osborne RH, Catell KJ (1987) The pharmacology of the isolated foregut of the locust, Schistocerca gregaia - I. The effect of range of putative transmitters. Comp Biochem Physiol C 88: 131–138. 45. Roth BL, Shoham M, Choudhary MS, Khan N (1997) Identification of conserved aromatic residues essential for agonist binding and second messenger production at 5-hydroxytryptamine(2A) receptors. Mol Pharmacol 52: 259–266. p y 66. Luffy D, Dorn A (1991) Serotoningeric elements in the stomatogastric nervous system of the stick insect, Carausius morosus, demonstrated by immunohistochem- istry. J Insect Physiol 37: 269–278. 46. Barak LS, Tiberi M, Freedman NJ, Kwatra MM, Lefkowitz RJ, et al. (1994) A highly conserved tyrosine residue in G protein-coupled receptors is required for agonist-mediated beta(2)-adrenergic receptor sequestration. J Biol Chem 269: 2790–2795. 67. Ro¨ser C, Jordan N, Balfanz S, Baumann A, Walz B, et al. (2012) Molecular and pharmacological characterization of serotonin 5-HT2alpha and 5-HT7 receptors in the salivary glands of the blowfly Calliphora vicina. PLoS One 7: e49459. PLOS ONE \| www.plosone.org May 2013 \| Volume 8 \| Issue 5 \| e65052 10
https://openalex.org/W2094904305	https://bmcclinpharma.biomedcentral.com/counter/pdf/10.1186/1472-6904-10-1	English	null	Quantitative relationship between the octanol/water partition coefficient and the diffusion limitation of the exchange between adipose and blood	BMC clinical pharmacology	2,010	cc-by	8,994	Quantitative relationship between the octanol/ water partition coefficient and the diffusion limitation of the exchange between adipose and blood David G Levitt David G Levitt © 2010 Levitt; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Correspondence: levit001@umn.edu Department of Integrative Biology and Physiology, University of Minnesota, 6-125 Jackson Hall, 321 Church St. S. E., Minneapolis, MN 55455, USA Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Open Access Abstract Background: The goal of physiologically based pharmacokinetics (PBPK) is to predict drug kinetics from an understanding of the organ/blood exchange. The standard approach is to assume that the organ is “flow limited” which means that the venous blood leaving the organ equilibrates with the well-stirred tissue compartment. Although this assumption is valid for most solutes, it has been shown to be incorrect for several very highly fat soluble compounds which appear to be “diffusion limited”. This paper describes the physical basis of this adipose diffusion limitation and its quantitative dependence on the blood/water (Kbld-wat) and octanol/water (Kow) partition coefficient. Methods: Experimental measurements of the time dependent rat blood and adipose concentration following either intravenous or oral input were used to estimate the “apparent” adipose perfusion rate (FA) assuming that the tissue is flow limited. It is shown that the ratio of FA to the anatomic perfusion rate (F) provides a measure of the diffusion limitation. A quantitative relationship between this diffusion limitation and Kbld-wat and Kow is derived. This analysis was applied to previously published data, including the Oberg et. al. measurements of the rat plasma and adipose tissue concentration following an oral dose of a mixture of 13 different polychlorinated biphenyls. Results: Solutes become diffusion limited at values of log Kow greater than about 5.6, with the adipose-blood exchange rate reduced by a factor of about 30 for a solute with a log Kow of 7.36. Quantitatively, a plot of FA/F versus Kow is well described assuming an adipose permeability-surface area product (PS) of 750/min. This PS corresponds to a 0.14 micron aqueous layer separating the well-stirred blood from the adipose lipid. This is approximately equal to the thickness of the rat adipose capillary endothelium. Conclusions: These results can be used to quantitate the adipose-blood diffusion limitation as a function of Kow. This is especially important for the highly fat soluble persistent organic chemicals (e.g. polychlorinated biphenyls, dioxins) whose pharmacokinetics are primarily determined by the adipose-blood exchange kinetics. © 2010 Levitt; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction any medium, provided the original work is properly cited. Methods Dependence of the adipose-blood exchange rate (Clr) on the intrinsic capillary permeability-surface area product (PS) and the blood-water partition coefficient (Kbld-wat) Figure 1 shows a schematic diagram of the factors involved in the solute exchange between the tissue and blood. The upper case letters in Figure 1 indicate the absolute blood (CB, CA, CV) or tissue (CT) concentration while the lower case indicates the free aqueous blood (cB) or tissue (cT) concentration. It is this free aqueous concentration that determines the diffusional exchange rate between the blood and tissue. The capillary concen- tration varies as a function of the distance (x) from the arterial end of the capillary. It is assumed that the tissue compartment is well mixed and can be represented by an average value that does not depend on x. It is assumed that the tissue consists of N capillaries/ cm3, all with exactly the same geometry, blood flow, permeability, etc. It is also assumed that the relation between the capillary and tissue concentration is in a pseudo steady state. The steady state differential equa- tion for the concentration in the capillary as a function of position is However, there are some notable exceptions to this general rule where the adipose-blood exchange of sev- eral highly lipid soluble molecules seems to be diffusion limited [5-7]. The purpose of this paper is to provide a detailed mechanistic analysis of the origin of the diffu- sion limitation of these highly lipid soluble molecules. It will be shown that the magnitude of the diffusion limita- tion can be directly related to and predicted by the lipid/water partition coefficient. This analysis provides general criteria for predicting the degree of adipose tis- sue diffusion limitation just from knowledge of the lipid/water (or octanol/water, see below) partition coeffi- cient. This is the first detailed discussion of the quanti- tative relationship between lipid/water partition and diffusion limitation that I am aware of.    F dCB dx aPN c x c B T 2 [ ( ) ] (1) (1) where F is the tissue perfusion rate (kg/min/kg), P is the intrinsic capillary permeability defined in terms of the free water concentration (cm/min), N is the number of capillaries per cm3 tissue, and a is the capillary radius (cm). The total blood concentration CB can be related to the free water concentration (cB) using the blood/ water partition coefficient (CB = Kbld-wat cB). Background is “flow limited” - that is, the venous blood leaving the capillary has equilibrated with the well mixed tissue space. For the flow limited model this tissue-blood exchange depends on only two parameters: a) the tissue perfusion rate (kg/min/kg); and b) the blood/tissues par- tition coefficient. This model has the major advantage that if one has determined the perfusion rate using one solute, the tissue-blood kinetic exchange can be pre- dicted for any other solute for which the blood/tissue partition is known. Physiologically Based Pharmacokinetics (PBPK) refers to the approach of modeling drug kinetics using a realistic physiological description of the animal [1,2]. A central feature of this approach is the quantitative description of the tissue-blood exchange. The most basic approach (and the standard one) is to assume that this exchange Page 2 of 13 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 There are a number of solutes for which this flow limited model is clearly not valid and for which a “dif- fusion limited” model must be used. One such class are the large (e.g. inulin) or highly protein bound (e.g. dicloxacillin and ceftriaxone) extracellular solutes which have a significant capillary permeability limita- tion [3]. Similarly, a number of water soluble mole- cules such as actinomycin-D and methotrexate have been shown to have cell membrane limited uptake [4]. It is not surprising that these highly water soluble solutes with their low cell membrane permeability might be diffusion limited. These solutes represent relatively rare exceptions and the great majority of solutes that are used in PBPK modeling are highly lipid soluble with high (nearly infinite) cell membrane permeability and the predictions of the flow limited PBPK model generally provides good agreement with the experimental pharmacokinetics. Relationship between the anatomic adipose perfusion rate (F) and the “apparent” perfusion rate (FA) dCT dt F C t K C t K F C t C t K A A bld wat T ad wat A A T ad b        [ ( ) ( ) / ] [ ( ) ( ) / ld] (7) The general, diffusion limited, differential equation describing the change in the well mixed tissue concen- tration (CT(t)) produced by a time dependent arterial input (CA(t)) is: (7) The adipose tissue concentration (CT) depends on just two parameters - the apparent perfusion rate (FA) and the adipose/blood partition coefficient (Kad-bld). dCT dt F C t C t FK c t c t A V bld wat A V      [ ( ) ( )] [ ( ) ( )] (4) (4) Methods Figure 1 Schematic diagram of the concentrations in capillary blood and adipose tissue and the factors involved in the tissue-blood solute exchange. The upper case letters indicate the absolute blood (CB, CA, CV) or tissue (CT) concentration while the lower case (cB, cT) indicates the free aqueous blood (cB) or tissue (cT) concentration. permeability (i.e. diffusion) limited if a large fraction in the blood is solute bound (i.e. large Kbld-wat). of the true anatomic perfusion rate (F). Using eq [6], the general differential equation that describes the time dependent adipose tissue concentration (CT(t)) as a function of the arterial concentration (CA(t)) is then: Methods Integrating eq [1] over the length of the capillary (L) and solving for the venous concentration leaving the capillary: This analysis is especially important for the persis- tent organic pollutants (e.g. dioxins and polychlori- nated biphenyls) whose pharmacokinetics are dominated by the kinetics of adipose-blood exchange [8]. The use of PBPK model predictions is essential for this solute class because it is not possible to accurately measure their pharmacokinetics in humans. As the fol- lowing analysis shows, the adipose-blood exchange rate for the most highly lipid soluble solutes can be as much as 30 times slower than is predicted assuming flow limited kinetics. c c c c PS K F v T A T bld wat      [ ]exp( ( / ) / ) (2) (2) where PS (min-1) is the permeability-surface area pro- duct per tissue weight (S = 2πaNL). Equation [2] can be related to the fractional clearance or equilibration that occurs in one pass through the capillary: Clr c c c c PS K F A V A T bld wat         ( ) / ( ) exp( ) ( / ) / 1   (3) (3) (3) The approach described here is based on an analysis of experimental measurements of the “apparent” adipose perfusion rate (FA). The “apparent” rate is the perfusion rate that would be predicted assuming that the exchange is blood flow limited. It will be shown that the ratio of this “apparent” rate and the true anatomic adipose per- fusion rate (FA/F) can be used to quantitate the degree of diffusion limitation. For >>1, the venous concentration leaving the tissue (cv) is nearly equal to the tissue concentration (cT), the clearance (Clr) approaches 1 and the solute is flow lim- ited. The clearance depends on both the intrinsic per- meability (PS) and the blood-water partition coefficient. Solutes that have a high intrinsic permeability may be Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 3 of 13 Figure 1 Schematic diagram of the concentrations in capillary blood and adipose tissue and the factors involved in the tissue-blood solute exchange. The upper case letters indicate the absolute blood (CB, CA, CV) or tissue (CT) concentration while the lower case (cB, cT) indicates the free aqueous blood (cB) or tissue (cT) concentration. Relationship between Kbld-wat and octanol/water partition coefficient (Kow) where CV(t) is the time dependent venous concentra- tion leaving the tissue and cA and cV are the unbound free blood water concentration. By definition, if the tis- sue exchange is flow limited, then the venous concentra- tion equilibrates with the tissue concentration (cv(t) = cT (t)): For the solutes used here the experimental value of Kbld- wat is not available and the following procedure was used to estimate it. It is well recognized that the tissue/blood partition for the highly lipid soluble molecules is roughly equal to the tissue/blood lipid concentration ratio [9-11], indicating that, as a first approximation, solutes in the blood are bound as if they were binding to an equivalent blood lipid (i.e. oil) fraction. Thus, the blood/ water concentration can be approximated by: dCT flow dt F K c t c t bld wat A T lim [ ( ) ( )]    (5) (5) Using the definition of clearance (eq. [6]), the general diffusion limited equation (eq. [5]) can be written in the form: K f K f K bld water L oil wat L ow     (8) (8) where Koil-wat is the oil (e.g. olive oil)/water partition coefficient and fL is the equivalent blood lipid concen- tration. The last equality in eq. [8] results from the observation that, for non-polar solutes, Koil-wat is approximately equal to the octanol/water partition coef- ficient (Kow). A detailed analysis of the dependence of the relationship between Koil-wat and Kow on chemical structure is described in section 2 of the supplemental file (Additional File 1). Equation [8] is only an approxi- mate estimate of Kbld-wat since there is clearly some dCT dt Clr FK c t c t F K c t c t F bld wat A T A bld wat A T A       [ ( ) ( )] [ ( ) ( )] Clr F (6) (6) where FA is defined as the “apparent” perfusion rate. Equation [6] is identical to the flow limited expression (eq. [5]), the only difference is that FA is used in place where FA is defined as the “apparent” perfusion rate. Equation [6] is identical to the flow limited expression (eq. [5]), the only difference is that FA is used in place Page 4 of 13 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 a function of time. Relationship between Kbld-wat and octanol/water partition coefficient (Kow) Given the time dependence of the blood concentration entering the tissue and assuming that the tissue is “apparently” flow limited, the adipose tissue concentration depends on only two parameters: FA and Kad-bld (eq. [7]). A continuous smooth plasma concentration curve was generated from the experimen- tal plasma data points (see additional file 1: section I for details) and used as the arterial input (CA(t)). Using this time dependent arterial input, the standard flow limited organ equations (eq. [7]) were solved for the adipose concentration as a function of time. The two adipose parameters FA and Kad-bld were then adjusted to give the optimal fit to the experimental adipose concentra- tion measurements. In most cases the experiments were carried out for long enough times (up to 132 days) that the adipose tissue had come close to equilibrating with the plasma and Kad-bld could be determined directly from this equilibrium value. specific binding to the cellular and protein blood com- ponents [12,13]. The analysis described below relies on using compounds with a very large range of Kow (103) so that small errors in the absolute value of Kbld-wat are relatively insignificant. The value of fL (= Kbld-wat/Koil-wat = Kbld-oil ≈1/Kad-bld) can be estimated from the equilibrium partition between blood and adipose tissue (i.e.,oil). In humans, this ratio is about 0.005 for a large series of polychlorinated biphenyls [14]. In addition, in a recent large compilation of Ktis-bld by deBruyn and Gobas [15] the value of 1/Kad- bld is in the range of 0.005 for most of the very highly fat soluble solutes. This value of 0.005 is close to the directly measured values of blood lipid fraction in humans [9] and rats [10] and a value of fL = 0.005 will be assumed here. With the exception of PCB 2,2’,5,5’ (Table 1), all the Oberg et. al. [16] solutes had values of fL (= 1/Kad-bld) of about 0.005. The numerical calculations were obtained using PKQuest_Java [17], a freely distributed PBPK software routine that can be downloaded from http://www. pkquest.com. In addition, the free file “Persistent organ- ics” contains the complete experimental data sets for For each solute that was modeled, the value of the experimental octanol/water partition coefficient (log Kow) and the model values of the “apparent” rat adipose perfusion rate (FA) and adipose/blood partition coefficient (Kad-bld) are listed. The last column describes the experimental details of the measurements. For “chemical” measurements, the parent solute was directly measured. Tracer (14C, 3H) measurements of the total (parent and metabolite) concentration were used only if there was independent evidence that the metabolite concentration was negligible. parent solute was directly measured. Tracer (14C, 3H) measurements of the total (parent and metabolite) concentration were used evidence that the metabolite concentration was negligible. eled, the value of the experimental octanol/water partition coefficient (log Kow) and the model values of the “apparent” rat adipose e/blood partition coefficient (Kad-bld) are listed. The last column describes the experimental details of the measurements. For Results The three panels in Figures 2 and 3 show the model fits to the experimental data of Oberg et. al. [16] for the solutes with the lowest (2,4,4’ PCB, log Kow = 5.67), intermediate (2,3,3’,4,4’ PCB, log Kow = 6.65) and high- est (2,2’,3,4,4’,5,5’, log Kow = 7.36) value of Kow. Figure 2 shows the absolute plot and the insets shows the early time data. Figure 3 shows the semi-log plot of the same data. The red line is the smoothed fit to the plasma data which is used as the arterial input concentration to the organ. For all the solutes investigated, this smoothed curve provided a nearly perfect fit to the experimental blood concentration data points. In addition to the Oberg et. al. data set, the literature was screened and a number of other solutes were also modeled. One important experimental limitation for many solutes is that only the total C14 or H3 labeled equivalent was measured and the parent and metabolite compounds were not distinguished. Some of these tracer measurements were used in this analysis (see table 1) if there was supporting information that the labeled meta- bolite concentration in the plasma and tissue is rela- tively low. All of the experimental data that was used, along with comments about the experimental limita- tions, are summarized in table 1. The model fit to the adipose concentration data points is a function of two parameters: FA, and Kad-bld. These two parameters were adjusted to give the optimal fit to the adipose data, and these are the results that are shown in Figures 2 and 3 and are listed in Table 1. The adipose tissue fits are clearly not perfect. However, because of the large qualitative difference in the kinetics for the 3 solutes shown in Figure 2, there is no question that there are significant differences in the correspond- ing values of FA. This is dramatically illustrated in Fig- ure 4 where the plots of the optimal values of FA are compared with the best fits that can be obtained using the value of FA for the neighboring solute in Figure 2. Results For example, in the middle panel of Figure 4, the red line shows the optimal fit (FA = 0.04) to the experimen- tal adipose concentration data points for PCB 2,3,3’,4,4’ For published data that was available only in graphical form, the values were read off the graph using the pro- gram UN-SCAN-IT (version 6.0, Silk Scientific Corporation). Value of the anatomic adipose perfusion rate (F) p p ( ) The determination of PS from eq. [9] requires an esti- mate of the anatomic adipose perfusion rate (F). The reported values for F in rat adipose tissue vary over a large range depending on methodologies, age, strain and condition (i.e. conscious or anesthetized). In addition, there are large differences in blood flow at different ana- tomic locations, varying from about 0.15 for epididymal to 0.55 kg/min/kg for mesenteric fat [18,19]. Most recent studies using labeled microspheres in conscious rats report values in the range of 0.18 to 0.25 kg/min/kg [20-22]. A value for F of 0.2 kg/min/kg was assumed in the following calculations. F F Clr c c c c Kow PS fLF A A V A T / ( ) / ( ) exp( ) /          1     (9) (9) As a first approximation, the value of the intrinsic PS (and b) for a given organ and species can be regarded as a constant, independent of the solute. This is because PS is proportional to the aqueous diffusion coefficient which is roughly proportional to 1/radius (Stokes-Ein- stein relation). For all the solutes considered here the molecular radius varies be less than a factor of 2, a var- iation that is negligible compared to the variations in Kow (≈103). Thus, assuming that b is constant, mea- surements of for a series of solutes with a wide range of values of Kow can be used to estimate the value of PS using the known values of fL (= 0.005) and F (0.2 kg/ min/kg). Experimental data The experimental data was obtained from the literature. The central results are those of Oberg et. al. [16] who simultaneously measured the rat plasma and adipose tis- sue concentration following an oral dose of a mixture of 13 different polychlorinated biphenyls, varying from 3 to 7 substituted chlorines with log Kow varying from 5.67 to 7.36 (table 1). These simultaneous chemical measurements represent the ideal data set for this analysis and this is the only measurement of this type that I am aware of. Experimental measurement of “apparent” perfusion rate (FA) FA was determined from experimental measurements in the rat of the blood and adipose tissue concentration as Table 1 Summary of the experimental data and the model analysis. Reference Solute log Kow (ref) FA (kg/min/kg) Kad-bld Comments Oberg [16] PCB 2,4,4’ 5.67 [25] 0.15 314 Simultaneous chemical measurements PCB 2,2’,5,5’ 5.84 [25] 0.05 89 PCB 3,3’,4,4’ 6.36 [25] 0.08 136 PCB 2,2’,3,4,5’ 6.29 [25] 0.08 136 PCB 2,2’,4,5,5’ 6.38 [25] 0.035 132 PCB 2,3,3’,4,4’ 6.65 [25] 0.04 263 PCB 2,3’,4,4’,5 6.74 [25] 0.03 226 PCB 2,2’,3,4,4’,5 6.83 [25] 0.012 226 PCB 2,2’,4,4’,5’ 6.92 [25] 0.013 226 PCB 2,3,3’,4,4’,5 7.18 [25] 0.015 292 PCB 2,3,3’,4,4’,5’ 7.18 [25] 0.012 263 PCB 2,2’,3,3,4,4’,5 7.27 [25] 0.007 263 PCB 2,2’,3,4,4’,5,5’ 7.36 [25] 0.005 263 Muhlebach [23] PCB 2,2’,4,4’,5,5’ 6.92 [25] 0.02 319 14C, Chemical, not metabolized Ebling [36] Thiopental 2.85 [37] 0.18 6.5 Chemical Dallas [38] Perchloroethylene 3.13 [39] 0.25 124 Chemical Parham [40] Dichlorodiphenylsulfone 3.9 (est) 0.12 89 14C, low metabolite Yamaguchi [41] Hexachlorobenzene 5.46 [42] 0.12 80 Chemical measurements Hexabromobenzene 6.07 [43] 0.03 27 Wang [5] TCDD 2,3,7,8 5.95 [44] 0.02 80 3H, low metabolite Kedderis [35] TBDD 2,3,7,8 6.5 [44] 0.0024 89 3H, low metabolite Komsta [45] PCDE 2,2’,4,4’,5 6.38 [46] 0.015 40 Chemical For each solute that was modeled, the value of the experimental octanol/water partition coefficient (log Kow) and the model values of the “apparent” rat adipose perfusion rate (FA) and adipose/blood partition coefficient (Kad-bld) are listed. The last column describes the experimental details of the measurements. For “chemical” measurements, the parent solute was directly measured. Tracer (14C, 3H) measurements of the total (parent and metabolite) concentration were used only if there was independent evidence that the metabolite concentration was negligible. Table 1 Summary of the experimental data and the model analysis. Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 5 of 13 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 ratio of FA/F, the value of the adipose clearance (Clr) and the parameter for a given solute can be obtained from eqs. [3] and [6]. Using the relationship between Kbld-wat and Kow (eq. [8], can be expressed in terms of Kow: most of the solutes discussed here and a detailed tutor- ial for reproducing these calculations. Summary of procedure used to estimate the adipose- blood diffusion limitation (Clr) and the intrinsic adipose capillary permeability-surface area product (PS) Experimental measurements of the rat arterial blood (CA (t)) and adipose tissue (CT(t)) concentrations as a func- tion of time were obtained from the literature. From a numerical solution of the differential eq. [7], the values of the two parameters FA and Kad-bld that gave the best agreement between the predicted and experimental adi- pose tissue concentration is then obtained. From the Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 6 of 13 Figure 2 Absolute plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line). The three panels show the results of Oberg et. al. [16] for a PCB with low (log Kow = 5.67), moderate (log Kow = 6.65) and high (log Kow = 7.36) octanol/water partition coefficient. The insets show the results at short times. Figure 2 Absolute plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) Figure 2 Absolute plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line). The three panels show the results of Oberg et. al. [16] for a PCB with low (log Kow = 5.67), moderate (log Kow = 6.65) and high (log Kow = 7.36) octanol/water partition coefficient. The insets show the results at short times. (solid circles). The green line is the predicted fit if the FA were equal to that of PCB 2,4,4’ (FA = 0.15) and the blue line is the optimal fit if the FA was that of PCB 2,2’,3,4,4’,5,5’ (FA = 0.005). The values of Kad-bld for the green and blue lines have been adjusted so that the curves fit the long time (i.e. equilibrium) data point. al. data is that the first data point is at 6 hours and the early time adipose kinetics are missed. To address this, results are shown in Figure 5 for the very slowly meta- bolized PCB 2,2’,4,4’,5,5’ which has been carefully inves- tigated by Muhlebach et. al.[23] and has adipose data points at early times (5, 15, 30, 60 and 240 minutes). This data (including the early times) is well described by an FA of 0.02, which is in good agreement with the Oberg. et. al. data for PCBs with a similar Kow. Summary of procedure used to estimate the adipose- blood diffusion limitation (Clr) and the intrinsic adipose capillary permeability-surface area product (PS) Three of Table 1 lists the model values for FA and Kad-bld for the 13 PCBs studied by Oberg et. al. [16] along with a number of other solutes. One limitation of the Oberg et. Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 7 of 13 Figure 3 Semi-log plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line). The three panels show the results of Oberg et. al. [16] for a PCB with low (log Kow = 5.67), moderate (log Kow = 6.65) and high (log Kow = 7.36) octanol/water partition coefficient. The insets show the results at short times. plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond he model predictions for adipose (blue line) and blood (red line). The three panels show the results of Oberg et. al. w (log Kow = 5.67), moderate (log Kow = 6.65) and high (log Kow = 7.36) octanol/water partition coefficient. The insets show mes Figure 3 Semi-log plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line). The three panels show the results of Oberg et. al. [16] for a PCB with low (log Kow = 5.67), moderate (log Kow = 6.65) and high (log Kow = 7.36) octanol/water partition coefficient. The insets show the results at short times. the solutes in Table 1 (thiopental, perchloroethylene and dichlorodiphenylsulfone) have low values of Kow (and Kbld-wat) and would be expected to be flow limited. Also listed in Table 1 are two pairs of solutes that are in the same chemical class and used similar experimental mea- surements. The first pair isTCDD and TBDD, for which both the pharmacokinetic and Kow measurements were made by the same laboratory. The second is the the solutes in Table 1 (thiopental, perchloroethylene and dichlorodiphenylsulfone) have low values of Kow (and Kbld-wat) and would be expected to be flow limited. Also listed in Table 1 are two pairs of solutes that are in the same chemical class and used similar experimental mea- surements. The first pair isTCDD and TBDD, for which both the pharmacokinetic and Kow measurements were made by the same laboratory. Summary of procedure used to estimate the adipose- blood diffusion limitation (Clr) and the intrinsic adipose capillary permeability-surface area product (PS) The second is the hexachlorobenzene and hexabromobenzene pair whose kinetics were described in the same publication. For both of these pairs, there is a significant decrease in the values for FA with an increase in Kow. Figure 6 compares the model FA dependence for TCDD and TBDD, similar to the plots in Figure 4. The crucial test of this model of diffusion limitation is to determine if the fractional clearance (= Clr = FA/F) is Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 8 of 13 Figure 4 Comparison of the model fits to the adipose concentrations as a function of the “apparent” perfusion rate (FA) for the three solutes described in Figures 1 and 2. For each solute, the optimal value of FA for that solute (red line) is compared with the optimal fit that can be obtained using the value of FA for the other solutes (green and blue line). For the green and blue fits, the value of the adipose/blood partition (K) was adjusted to approximately fit the last time point. The value of FA and K for each line is listed at the bottom of each panel. Figure 4 Comparison of the model fits to the adipose concentrations as a function of the “apparent” perfusion rate (FA) for the three solutes described in Figures 1 and 2. For each solute, the optimal value of FA for that solute (red line) is compared with the optimal fit that can be obtained using the value of FA for the other solutes (green and blue line). For the green and blue fits, the value of the adipose/blood partition (K) was adjusted to approximately fit the last time point. The value of FA and K for each line is listed at the bottom of each panel. described by eq. [9] Figure 7 shows the plot of the Oberg et. al. [16] values of FA/F (solid circles) as a func- tion of Kow. The red line is the plot of 1-exp(-b/Kow), with b (eq. [9]) adjusted to provide the best fit to the Oberg et. al. data. For the value of b (= 0.75 × 106) that provides the best fit to the data in Figure 6, PS = bfLF = 750 min-1 (assuming fL = 0.005 and F = 0.2). tissue-solute exchange is diffusion limited. Summary of procedure used to estimate the adipose- blood diffusion limitation (Clr) and the intrinsic adipose capillary permeability-surface area product (PS) The capillary permeability limitation produced by this blood lipid solute binding is directly analogous to the carbon monoxide (CO) pulmonary diffusion limitation that results from the high affinity of hemoglobin for CO [24]. This same effect was also invoked to quantitate the capillary diffusion lim- itation of albumin bound b-lactam antibiotics [3]. As seen in eq. [3], given a finite intrinsic permeability (PS), the solute must become diffusion limited if Kbld-wat is large enough. The only question is at what value of Kbld-wat does the solute become diffusion limited? Discussion As described by eq. [3], the blood-water partition coeffi- cient (Kbld-wat) has a critical role in determining whether Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 9 of 13 Figure 5 Absolute plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line) for the PCB 2,2’,4,4’,5,5’ data of Muhlebach et. al. [23]. The insets show the results at short times. Figure 5 Absolute plot of the time dependence of the experimental adipose (green triangle) and blood (yellow diamond) concentration and the model predictions for adipose (blue line) and blood (red line) for the PCB 2,2’,4,4’,5,5’ data of Muhlebach et. al. [23]. The insets show the results at short times. To answer this question, the degree of diffusion lim- itation was determined for a series of solutes with vary- ing values of octanol/water partition coefficient (Kow). The “diffusion limitation” was quantitated in rat adipose tissue from experimental measurements of the ratio of the “apparent” perfusion rate (FA) relative to the ana- tomic perfusion rate (F). Assuming that Kbld-wat is roughly proportional to Kow (eq. [8]), one should observe an increase in diffusion limitation as Kow increases. One would expect that this relationship would only be roughly satisfied because the exact relationship between Kow and Kbld-wat should depend on the specific blood protein binding and the detailed chemical struc- ture [12,13]. This is the reason that it is especially important to have the Oberg et. al. [16] experimental data set for a large series of solutes with similar chemi- cal structures and with Kow varying over a 50 fold range. Another factor that introduces variations in the relation between diffusion limitation and Kow is that there are large variations in the experimental measure- ments of Kow, depending on the experimental technique that is used. An advantage of the Oberg et. al. data set is that the values of Kow for all the solutes was deter- mined using the same approach [25]. As shown in Table 1, for the Oberg et. al. data there is a qualitative increase in the degree of diffusion limitation (i.e. reduction in “apparent” perfusion rate FA) as Kow increases. Quantita- tively, this diffusion limitation is quite will described by eq. [9] (Figure 7). small values of Kow and one would expect them to be flow limited. Discussion Consistent with this prediction, their values of FA are close to the assumed anatomic value (F) of 0.2 kg/min/kg. Results are also shown in table 1 for hexachloroben- zene, hexabromobenzene, TCDD, TBDD and PCDE. The hexachlorobenzene and hexabromobenzene pair are directly comparable since their kinetics were described in the same publication using identical procedures. Similarly, the TCDD and TBDD pair results were car- ried out by the same lab using similar procedures. For both of these solute pairs with very similar structures, there is a significant increase in diffusion limitation (i.e. decrease in FA) with an increase in Kow (see also Figure 6 for a sensitivity analysis of the TCDD and TBDD results). These five solutes have values of Kow in the same range as the flow limited solutes studied by Oberg et. al. [16] and, qualitatively, have a similar increase in diffusion limitation (decrease in FA) with increasing Kow. Quantitatively, these solutes do not fall on exactly the same FA/F versus Kow curve (Figure 7) that was obtained for the Oberg solutes. As discussed above, this is not surprising because one would predict that the propor- tionality between Kow and Kbld-wat should have some solute dependence and should also depend on the speci- fic details of, for example, the Kow measurements. From the plot of FA/F versus Kow for the Oberg et. al. data (Figure 7), one can estimate that the corresponding value of the “intrinsic” PS for rat adipose tissue is about 750/min (see Results). To put this very large value in perspective, the highest capillary permeability that has been directly measured is a PS of about 1/min for Na+ Also listed in Table 1 are the values of FA for some other solutes. Three of these solutes (thiopental, per- chloroethylene, and dichlorodiphenylsulfone) have very Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 10 of 13 Figure 6 Perfusion rate dependence (FA) of the model fits to the experimental adipose concentration measurements (solid circles) for tetrachlorodibenzo-p-dioxin (TCDD, Wang et. al. [5]) and tetrabromodibenzo-p-dioxin (TBDD, Kedderis et. al. [35]). The red line is the predicted fit using the optimal value of FA for that solute and the green line is the fit using the FA of the other solute. For the green lines, the value of the adipose/blood partition (K) was adjusted to approximately fit the last time point. Discussion Figure 6 Perfusion rate dependence (FA) of the model fits to the experimental adipose concentration measurements (solid circles) for tetrachlorodibenzo-p-dioxin (TCDD, Wang et. al. [5]) and tetrabromodibenzo-p-dioxin (TBDD, Kedderis et. al. [35]). The red line is the predicted fit using the optimal value of FA for that solute and the green line is the fit using the FA of the other solute. For the green lines, the value of the adipose/blood partition (K) was adjusted to approximately fit the last time point. in heart capillaries [26]. Solutes with a PS of 750/min only become diffusion limited because they have a very large value of Kbld-wat (≈5000). capillary wall and that the diffusion in the capillary blood and adipose tissue is not limiting. Diffusion in the blood should not be rate limiting because it is well stir- red and the solute is bound at a high concentration rela- tive to the water. The relative rates of diffusion in the aqueous and fat tissue are described by: One can use this value of PS to estimate the “equiva- lent” thickness of the rat blood-adipose diffusion limit- ing barrier. The permeability is equal to: Aqueous Diffusion D c Lipid Diffusion D c K D c w w L L oil wat L w     (12) P D W W  / (10) P D W W  / (10) (12) where DW is the aqueous diffusion coefficient and W is the thickness. Thus, relating W to PS: and the ratio of lipid to aqueous diffusion is: and the ratio of lipid to aqueous diffusion is: Lipid Aqueous Diffusion K D D oil wat L w / /   (13) W D S PS W  / (11) (11) (11) W D S PS W  / (13) where is S is the rat adipose capillary surface area. Using values of 5 × 10-6 cm2/sec [27] for PCB DW and 35 cm2/cm3 [28] for rat S, the value of W is about 1.4 × 10-5 cm for a PS of 750/min. This value for W (0.14 μm) is approximately equal to the thickness of the rat adipose capillary endothelium (0.25 μm) [29] and seems a reason- able estimate of the aqueous diffusive barrier between the blood (where the solutes are well-stirred and bound to albumin and the lipoproteins) and the adipose lipid. Since the lipid (i.e. olive oil) viscosity is about 85 times greater than water, the Stokes-Einstein relation predicts that the lipid diffusion coefficient is about 85 times smaller than in water. However, this relation is valid only if the solute is much larger than the solvent, which is not true for diffusion in olive oil. For small solutes, such as O2 or N2, the diffusion coefficient in olive oil is only about 3 times smaller than in water [30]. Roga- cheva et. al. [31] measured diffusion coefficients of 2- nonanone and benzaldehyde in oil that were about 10 times less than the water value. As a rough estimate, it An inherent assumption in this analysis is that the rate limiting step is the aqueous diffusion across the Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 Page 11 of 13 Figure 7 Plot of the degree of diffusion limitation (FA/F) as a function of the octanol/water partition coefficient for the 13 PCBs studied by Oberg et. al. [16] (see Table 1 for numerical values). The red line is a plot of eq. [9] with a b (eq. [9]) of 0.75 × 106, adjusted to provide the best fit to the data. Figure 7 Plot of the degree of diffusion limitation (FA/F) as a function of the octanol/water partition coefficient for the 13 PCBs studied by Oberg et. al. [16] (see Table 1 for numerical values). The red line is a plot of eq. [9] with a b (eq. [9]) of 0.75 × 106, adjusted to provide the best fit to the data. rats [32] and humans [33] have similar adipose capillary densities, they should have similar capillary surface areas (S). Also, the blood lipid fraction (fL) is similar for rats and humans. (11) One factor that does differ between rats and humans is the adipose blood flow (F), with the aver- age human value of about 0.05 kg/kg/min [34] about 1/4 the rat value (0.2 kg/kg/min). Thus, the human value for b is about 4 times larger than the rat value and there should be corresponding less diffusion limitation for the same value of Kow. is assumed that the diffusion coefficient in olive oil for the solutes studied here is about 20 times less than in water (DL/DW = 0.05). Since Koil-wat for the solutes con- sidered here is 105 or greater, aqueous diffusion (eq. 13) is clearly the rate limiting step. A novel feature of this analysis is the approach that was used to determine the “apparent” adipose perfusion. In the previous publications from which this experimen- tal data was obtained, the blood and adipose concentra- tions were simultaneously modeled using the complete PBPK multi-tissue model. In many cases the model fit to the blood concentration is only poorly approximated by this PBPK model, presumably because of errors in the model assumptions (e.g. non-linearity). In contrast, in this analysis the exact experimental blood concentra- tion is fit by a smooth curve so that the model input to the adipose tissue is identical to the experimental arter- ial concentration. The adipose perfusion rate is then adjusted to give the best fit to the experimental adipose concentration. This provides a significantly more accu- rate estimate of adipose perfusion then the total PBPK model fit with the incorrect blood concentration. List of abbreviations 7. Kramer HJ, Drenth H, vandenBerg M, Seinen W, DeJongh J: Physiologically based pharmacokinetic model for tetrachlorobenzyltoluenes in rat: comparison of in vitro and in vivo metabolic rates. Toxicol Sci 2001, 63:22-28. PBPK: Physiologically based pharmacokinetics; F: ana- tomic organ perfusion rate (kg/min/kg); FA: “apparent” perfusion rate assuming the organ kinetics are flow lim- ited; a: capillary radius (cm); L: capillary length (cm); N: capillary density (#/cm3); S: 2πaNL: capillary surface area surface area per cm3 tissue (1/cm); P: intrinsic per- meability (cm/min); PS: intrinsic organ permeability-sur- face area product (min-1); Clr: fraction of solute that equilibrates across capillary in one pass; Ci: total con- centration (mole/kg) in organ i. i: B(blood), A(artery), V(vein), T(adipose); ci: free concentration in water (moles/cm3); fL: “equivalent” lipid fraction of blood; Kbld-wat: blood/water partition coefficient; Kad-wat: adi- pose/water partition coefficient; Kad-bld: adipose/blood partition coefficient; Koil-wat: oil/water partition coeffi- cient; Kow: octanol/water partition coefficient; Dw and DL: Diffusion coefficient in water and lipid (cm2/sec); W: thickness (cm) of equivalent aqueous layer corre- sponding to intrinsic permeability; PCB: polychlorinated biphenyl; PCDE: pentachlorodiphenyl ether; TCDD: tet- rachlorodibenzo-p-dioxin; TBDD: tetrabromodibenzo- p-dioxin. 8. Persistent organic pollutants. Berlin Heidelberg: Springer-Verlag 2003. g g g g 9. Molen van der GW, Kooijman SA, Slob W: A generic toxicokinetic model for persistent lipophilic compounds in humans: an application to TCDD. Fundam Appl Toxicol 1996, 31:83-94. 10. Emond C, Charbonneau M, Krishnan K: Physiologically based modeling of the accumulation in plasma and tissue lipids of a mixture of PCB congeners in female Sprague-Dawley rats. J Toxicol Environ Health A 2005, 68:1393-1412. 11. Brown JF Jr, Lawton RW: Polychlorinated biphenyl (PCB) partitioning between adipose tissue and serum. Bull Environ Contam Toxicol 1984, 33:277-280. 12. Parham FM, Kohn MC, Matthews HB, DeRosa C, Portier CJ: Using structural information to create physiologically based pharmacokinetic models for all polychlorinated biphenyls. Toxicol Appl Pharmacol 1997, 144:340-347. 13. Noren K, Weistrand C, Karpe F: Distribution of PCB congeners, DDE, hexachlorobenzene, and methylsulfonyl metabolites of PCB and DDE among various fractions of human blood plasma. Arch Environ Contam Toxicol 1999, 37:408-414. 14. Wolff MS, Thornton J, Fischbein A, Lilis R, Selikoff IJ: Disposition of polychlorinated biphenyl congeners in occupationally exposed persons. Toxicol Appl Pharmacol 1982, 62:294-306. 15. deBruyn AM, Gobas FA: The sorptive capacity of animal protein. Environ Toxicol Chem 2007, 26:1803-1808. 16. Conclusions Although it has been previously recognized that some highly fat soluble persistent organic chemicals are diffu- sion limited [5-7], this analysis provides the first physical explanation of this diffusion limitation along with its quantitative dependence on Kow. This diffusion limita- tion follows directly from the basic physiology of the blood tissue exchange. There must be some finite aqu- eous unstirred layer between the blood and the adipose lipid and this layer will become rate limiting if Kbld-wat is large enough. The results described here show that this diffusion limitation reduces the apparent rat adipose perfusion rate from the anatomic value of about 0.2 kg/ min/kg for the flow limited solutes (log Kow < 5) to a value of about 0.005 kg/min/kg for a PCB with a log Kow of 7.36. The thickness of the limiting layer It is of interest to try to extrapolate these rat results to humans. The degree of diffusion limitation for a given Kow is determined by the parameter b = PS/(fLF) (eq. [9]). P corresponds to the diffusion limiting aqueous layer and should be similar for rat and human. Since Page 12 of 13 Page 12 of 13 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 estimated from this analysis is about 0.14 μm, approxi- mately equal to the thickness of the adipose capillary epithelial cell. pharmacokinetic behavior of TCDD in female Sprague-Dawley rats. Toxicol Appl Pharmacol 1997, 147:151-168. pharmacokinetic behavior of TCDD in female Sprague-Dawley rats. Toxicol Appl Pharmacol 1997, 147:151-168. 6. 6. Kedderis LB, Mills JJ, Andersen ME, Birnbaum LS: A physiologically based pharmacokinetic model for 2,3,7,8-tetrabromodibenzo-p-dioxin (TBDD) in the rat: tissue distribution and CYP1A induction. Toxicol Appl Pharmacol 1993, 121:87-98. Authors’ contributions DGL is entirely responsible for the contents of this paper and has read and approved the final manuscript. 21. Kuwahira I, Gonzalez NC, Heisler N, Piiper J: Regional blood flow in conscious resting rats determined by microsphere distribution. J Appl Physiol 1993, 74:203-210. List of abbreviations Oberg M, Sjodin A, Casabona H, Nordgren I, Klasson-Wehler E, Hakansson H: Tissue distribution and half-lives of individual polychlorinated biphenyls and serum levels of 4-hydroxy-2,3,3’,4’,5-pentachlorobiphenyl in the rat. Toxicol Sci 2002, 70:171-182. 17. Levitt DG: PKQuest_Java: free, interactive physiologically based pharmacokinetic software package and tutorial. BMC Res Notes 2009, 2:158. Additional file 1: I. Mathematical representation of the time dependent plasma concentration.II. Prediction of oil/water partition coefficient using octanol/water coefficient. Click here for file 18. West DB, Prinz WA, Greenwood MR: Regional changes in adipose tissue blood flow and metabolism in rats after a meal. Am J Physiol 1989, 257: R711-716. [ http://www.biomedcentral.com/content/supplementary/1472-6904-10-1- S1.DOC ] 19. Crandall DL, Goldstein BM, Huggins F, Cervoni P: Adipocyte blood flow: influence of age, anatomic location, and dietary manipulation. Am J Physiol 1984, 247:R46-51. y 20. Wada DR, Harashima H, Ebling W, Osaki EW, Stanski DR: Effects of thiopental on regional blood flows in the rat. Anesthesiology 1996, 84:596-604. Competing interests p g The author declares that they have no competing interests. The author declares that they have no competing interests. 22. Delp MD, Evans MV, Duan C: Effects of aging on cardiac output, regional blood flow, and body composition in Fischer-344 rats. J Appl Physiol 1998, 85:1813-1822. Received: 15 September 2009 Accepted: 7 January 2010 Published: 7 January 2010 Received: 15 September 2009 Accepted: 7 January 2010 Published: 7 January 2010 23. Muhlebach S, Wyss PA, Bickel MH: Comparative adipose tissue kinetics of thiopental, DDE and 2,4,5,2’,4’,5’-hexachlorobiphenyl in the rat. Xenobiotica 1985, 15:485-491. Additional file 1: I. Mathematical representation of the time dependent plasma concentration.II. Prediction of oil/water partition coefficient using octanol/water coefficient. Click here for file [ http://www.biomedcentral.com/content/supplementary/1472-6904-10-1- S1.DOC ] References 24. Forster RE: Exchange of gases between alveolar air and pulmonary capillary blood: pulmonary diffusing capacity. Physiol Rev 1957, 37:391-452. 1. Reddy MB, Yang RSH, Clewell HJ, Andersen ME: Physiologically Based Pharmacokinetic Modeling. Hoboken, New Jersey: Wiley-Interscience 2005. 2. Edginton AN, Theil FP, Schmitt W, Willmann S: Whole body physiologically-based pharmacokinetic models: their use in clinical drug development. Expert Opin Drug Metab Toxicol 2008, 4:1143-1152. 3. Levitt DG: PKQuest: capillary permeability limitation and plasma protein binding - application to human inulin, dicloxacillin and ceftriaxone pharmacokinetics. BMC Clin Pharmacol 2002, 2:7. 4. Lutz RJ, Dedrick RL, Zaharko DS: Physiological pharmacokinetics: an in vivo approach to membrane transport. Pharmacol Ther 1980, 11:559-592. 5. Wang X, Santostefano MJ, Evans MV, Richardson VM, Diliberto JJ, Birnbaum LS: Determination of parameters responsible for 1. Reddy MB, Yang RSH, Clewell HJ, Andersen ME: Physiologically Based Pharmacokinetic Modeling. Hoboken, New Jersey: Wiley-Interscience 2005 2. Edginton AN, Theil FP, Schmitt W, Willmann S: Whole body physiologically-based pharmacokinetic models: their use in clinical drug development. Expert Opin Drug Metab Toxicol 2008, 4:1143-1152. 25. Hawker DW, Connell DW: Octanol-water partition coefficients of polychlorinated biphenyl congeners. Environ Sci Technol 1988, 22:382-387. 26. Crone C, Levitt DG: Capillary permeability to small solutes. Bethesda, Maryland: American Physiological Society 1984. 3. Levitt DG: PKQuest: capillary permeability limitation and plasma protein binding - application to human inulin, dicloxacillin and ceftriaxone pharmacokinetics. BMC Clin Pharmacol 2002, 2:7. 27. Werner D, Ghosh U, Luthy RG: Modeling polychlorinated biphenyl mass transfer after amendment of contaminated sediment with activated carbon. Environ Sci Technol 2006, 40:4211-4218. p 4. Lutz RJ, Dedrick RL, Zaharko DS: Physiological pharmacokinetics: an in vivo approach to membrane transport. Pharmacol Ther 1980, 11:559-592. 28. Paaske WP: Absence of restricted diffusion in adipose tissue capillaries. Acta Physiol Scand 1977, 100:430-436. pp p 5. Wang X, Santostefano MJ, Evans MV, Richardson VM, Diliberto JJ, Birnbaum LS: Determination of parameters responsible for Page 13 of 13 Page 13 of 13 Levitt BMC Clinical Pharmacology 2010, 10:1 http://www.biomedcentral.com/1472-6904/10/1 29. Fraselle-Jacobs A, Jeanjean M, Hauser N, Remacle C: Effect of aging on the morphology of epididymal adipose tissue in the rat. Exp Gerontol 1987, 22:389-408. Davidson D, Eggleton P, Foggie P: The diffusion of atmospheric ga 30. Davidson D, Eggleton P, Foggie P: The diffusion of atmospheric gases through fats and oils. Q J Exp Physiol Cogn Med Sci 1952, 37:91-105. 31. References Jackson JA, Diliberto JJ, Birnbaum LS: Estimation of octanol-water partition coefficients and correlation with dermal absorption for several polyhalogenated aromatic hydrocarbons. Fundam Appl Toxicol 1993, 21:334-344. 45. Komsta E, Chu I, Villeneuve DC, Benoit FM, Murdoch D: Tissue distribution metabolism and excretion of 2,2’,4,4’,5-pentachlorodiphenyl ether in the rat. Arch Toxicol 1988, 62:258-262. 46. Kurz J, Ballschmiter K: Vapour pressures, aqueous solubilities, Henry’s law constants, partition coefficients between gas/water (Kgw), n-octanol/ water (Kow) and gas/n-octanol (Kgo) of 106 polychlorinated diphenyl ethers (PCDE). Chemosphere 1999, 38:573-586. 46. Kurz J, Ballschmiter K: Vapour pressures, aqueous solubilities, Henry’s law constants, partition coefficients between gas/water (Kgw), n-octanol/ water (Kow) and gas/n-octanol (Kgo) of 106 polychlorinated diphenyl ethers (PCDE). Chemosphere 1999, 38:573-586. References Rogacheva S, Espinosa-Diaz MA, Voilley A: Transfer of aroma compounds in water-lipid systems: binding tendency of beta-lactoglobulin. J Agric Food Chem 1999, 47:259-263. 32. Kampf C, Bodin B, Kallskog O, Carlsson C, Jansson L: Marked increase in white adipose tissue blood perfusion in the type 2 diabetic GK rat. Diabetes 2005, 54:2620-2627. 33. Pasarica M, Sereda OR, Redman LM, Albarado DC, Hymel DT, Roan LE, Rood JC, Burk DH, Smith SR: Reduced adipose tissue oxygenation in human obesity: evidence for rarefaction, macrophage chemotaxis, and inflammation without an angiogenic response. Diabetes 2009, 58:718-725 34. Levitt DG: Heterogeneity of human adipose blood flow. BMC Clin Pharmacol 2007, 7:1. 35. Kedderis LB, Diliberto JJ, Birnbaum LS: Disposition and excretion of intravenous 2,3,7,8-tetrabromodibenzo-p-dioxin (TBDD) in rats. Toxicol Appl Pharmacol 1991, 108:397-406. 36. Ebling WF, Wada DR, Stanski DR: From piecewise to full physiologic pharmacokinetic modeling: applied to thiopental disposition in the rat. J Pharmacokinet Biopharm 1994, 22:259-292. 37. Freese E, Levin BC, Pearce R, Sreevalsan T, Kaufman JJ, Koski WS, Semo NM: Correlation between the growth inhibitory effects, partition coefficients and teratogenic effects of lipophilic acids. Teratology 1979, 20:413-440. 38. Dallas CE, Muralidhara S, Chen XM, Ramanathan R, Varkonyi P, Gallo JM, Bruckner JV: Use of a physiologically based model to predict systemic uptake and respiratory elimination of perchloroethylene. Toxicol Appl Pharmacol 1994, 128:60-68. 39. Leo A, Hansch C, Elkins D: Partition coefficients and their uses. Chemical Reviews 1971, 71:525-616. 40. Parham FM, Matthews HB, Portier CJ: A physiologically based pharmacokinetic model of p,p’-dichlorodiphenylsulfone. Toxicol Appl Pharmacol 2002, 181:153-163. 41. Yamaguchi Y, Kaswano M, Tatsukawa R: Tissue distribution and excretion of hexabromobenzene and hexachlorobenzene administered to rats. Chemosphere 1986, 15:453-459. 42. Bahudur NP, Shiu W, Boocock DGB, Mackay D: Temperature dependence of octanol-water partition coefficient for selected chlorobenzenes. J Chem Eng Data 1997, 42:685-688. 43. Kuramochi H, Maeda K, Kawamoto K: Measurements of water solubilities and 1-octanol/water partition coefficients and estimations of Henry’s lay constants for brominated benzenes. J Chem Eng Data 2004, 49:720-724. 43. Kuramochi H, Maeda K, Kawamoto K: Measurements of water solubilities and 1-octanol/water partition coefficients and estimations of Henry’s lay constants for brominated benzenes. J Chem Eng Data 2004, 49:720-724. 44. Jackson JA, Diliberto JJ, Birnbaum LS: Estimation of octanol-water partition coefficients and correlation with dermal absorption for several polyhalogenated aromatic hydrocarbons. Fundam Appl Toxicol 1993, 21:334-344. 44. Pre-publication history Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit The pre-publication history for this paper can be accessed here: http://www. biomedcentral.com/1472-6904/10/1/prepub The pre-publication history for this paper can be accessed here: http://www. biomedcentral.com/1472-6904/10/1/prepub Submit your next manuscript to BioMed Central and take full advantage of: Submit your next manuscript to BioMed Central and take full advantage of: doi:10.1186/1472-6904-10-1 Cite this article as: Levitt: Quantitative relationship between the octanol/water partition coefficient and the diffusion limitation of the exchange between adipose and blood. BMC Clinical Pharmacology 2010 10:1. doi:10.1186/1472-6904-10-1 Cite this article as: Levitt: Quantitative relationship between the octanol/water partition coefficient and the diffusion limitation of the exchange between adipose and blood. BMC Clinical Pharmacology 2010 10:1.

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.